**Personalized Nursing and Health Care**

Editors

**Riitta Suhonen Minna Stolt David Edvardsson**

MDPI • Basel • Beijing • Wuhan • Barcelona • Belgrade • Manchester • Tokyo • Cluj • Tianjin

*Editors* Riitta Suhonen Department of Nursing Science University of Turku Turku Finland

Minna Stolt Department of Nursing Science University of Turku Turku Finland

David Edvardsson School of Nursing & Midwifery La Trobe University Melbourne Australia

*Editorial Office* MDPI St. Alban-Anlage 66 4052 Basel, Switzerland

This is a reprint of articles from the Special Issue published online in the open access journal *Journal of Personalized Medicine* (ISSN 2075-4426) (available at: www.mdpi.com/journal/jpm/ special issues/personalized nursing).

For citation purposes, cite each article independently as indicated on the article page online and as indicated below:

LastName, A.A.; LastName, B.B.; LastName, C.C. Article Title. *Journal Name* **Year**, *Volume Number*, Page Range.

**ISBN 978-3-0365-7459-2 (Hbk) ISBN 978-3-0365-7458-5 (PDF)**

© 2023 by the authors. Articles in this book are Open Access and distributed under the Creative Commons Attribution (CC BY) license, which allows users to download, copy and build upon published articles, as long as the author and publisher are properly credited, which ensures maximum dissemination and a wider impact of our publications.

The book as a whole is distributed by MDPI under the terms and conditions of the Creative Commons license CC BY-NC-ND.

### **Contents**


Reprinted from: *J. Pers. Med.* **2021**, *11*, 945, doi:10.3390/jpm11100945 . . . . . . . . . . . . . . . . **99**


Personalized Healthcare: The Importance of Patients' Rights in Clinical Practice from the Perspective of Nursing Students in Poland, Spain and Slovakia—A Cross-Sectional Study Reprinted from: *J. Pers. Med.* **2021**, *11*, 191, doi:10.3390/jpm11030191 . . . . . . . . . . . . . . . . **225**

### **David Ramiro-Cortijo, Maria de la Calle, Vanesa Benitez, Andrea Gila-Diaz, Bernardo Moreno-Jim´enez and Silvia M. Arribas et al.**

Maternal Psychological and Biological Factors Associated to Gestational Complications Reprinted from: *J. Pers. Med.* **2021**, *11*, 183, doi:10.3390/jpm11030183 . . . . . . . . . . . . . . . . **239**

### **About the Editors**

### **Riitta Suhonen**

Ph.D., RN Riitta Suhonen is Professor in Nursing Science and has a speciality Gerontological Nursing Science since 2011. RS is the Director of the Doctoral Programme in Nursing Science (DPNurs). She is the fellow (FEANS) of the European Academy of Nursing Science (board and scientific committee member) and Member of the Academia Europaea. RS has a part-time position of Nursing Director in the Turku University Hospital and City of Turku, Welfare Services Division since 2016. RS has worked in the Department of Nursing Science since 2008 and formerly as a quality and development manager in health care district understanding of the healthcare systems, nurse leader and registered nurse. She has served as the former associate editor of the Nursing Open (Wiley) and Nursing Ethics (Sage).

### **Minna Stolt**

My research, function and health, from foot to head, focuses on health from functional perspective. I am interested in evaluating patients' functional health, quality of care, and rehabilitation in different levels of health care; patients' own strategies to care for and promote their own functional health; and to evaluate the competence of health care professionals to provide care and rehabilitation. The basis for this research lies in foot healh research in different age groups and contexts. My research area belongs to research strategy in the Department of Nursing Science in the Clinical Quality research program.

My Ph.D. study (2013) focused on foot health in older people and I developed the Foot Health Assessment Instrument to be used in clinical nursing care. My post doctoral study concentrated on occupational health care, namely in foot health in nurses. My current research deals with functional health and rehabilitation in patients with long-term conditions, such as rheumatoid arthritis.

### **David Edvardsson**

Professor Edvardsson is a recognized international authority in research across care sciences, health promotion and population health, and ageing and dementia. His ongoing scientific research is focused across five themes, health-promotion, care and support for older people; global and population health, ill-health, and burden of disease; developing positive health outcome measures and health-promoting interventions; person-centred care, health and quality of life; and health-promoting environments.

### *Editorial* **Personalized Nursing and Health Care: Advancing Positive Patient Outcomes in Complex and Multilevel Care Environments**

**Riitta Suhonen 1,2,\* , Minna Stolt <sup>1</sup> and David Edvardsson <sup>3</sup>**


This Special Issue of the *Journal of Personalised Medicine* invited manuscripts that further establish the current state of science relating to personalized nursing and health care. We welcomed manuscripts that highlight and further the knowledge base conceptually, instrumentally, observationally and experimentally, with sound theoretical and methodological underpinnings and implications for research, theory and clinical work in the disciplines of nursing, medicine, allied health and beyond. As there has been a rapid development in the academic literature over the last ten years in terms of papers relating to individualization, personalization, patient-, client-, consumer- and person-centredness, with work on conceptual, instrumental, observational and experimental levels, this theme seemed timely and relevant [1,2].

The individuality of care and services is essential for the realisation of healthcare quality, ethical obligations and the development of a deeper understanding of user perspectives necessary for health care, health policy development and increasing patient choice [3]. Healthcare systems in countries should be based on the comprehensive need assessment of individual clients [4,5] and patients to provide individualised, personalised or tailored care [6]. Personalized medical care should not only improve the patient's situation by providing the right diagnosis, prevention or treatment; it also needs to be tailored according to individual characteristics, situation, context, and environment to support people's health power, health careers and thus, their self-management and independent living [7,8]. This is highly important as health care systems have taken responsibility for care that also warrants the increasing responsibility of people's own self-management [9,10]. In order to support self-care, an individualised assessment of care needs is needed; furthermore, individual client's and patient's active participation in determining care and co-designing services is necessary [11]. Self-care is of vital importance for sustainable healthcare; however, it is not sufficiently emphasised. Knowing the clients and patients, assessing their individual needs and responding to these needs in an individualised manner has been found effective, and even cost-effective. Such initiative requires individualised or tailored interventions that are effective in care delivery. However, the complex multilevel interplay between care and service networks for patients, especially for older people, is rarely studied.

Individualised treatment and nursing care is an activity carried out by professionals and provides the perception of personalised care; it can act as an indicator of personcenteredness, requiring person-centred behaviour and other forms of competence [12,13]. Such care does not appear in and of itself, and health professionals need to support new-comers to in providing individualised or person-centred care for citizens [14]. Individualised care and patient-centeredness can be seen as a process or specific set of nursing or other care activities which produce positive patient outcomes [15]. However, there is a need to change the shift from system and professional-centred activities to patientcenteredness and orientation to build usable and effective care options for different groups

**Citation:** Suhonen, R.; Stolt, M.; Edvardsson, D. Personalized Nursing and Health Care: Advancing Positive Patient Outcomes in Complex and Multilevel Care Environments. *J. Pers. Med.* **2022**, *12*, 1801. https://doi.org/ 10.3390/jpm12111801

Received: 8 October 2022 Accepted: 17 October 2022 Published: 1 November 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

**<sup>\*</sup>** Correspondence: riisuh@utu.fi

of people, including individualised environments that support self-management and independence [16,17]. However, health system reforms have not necessarily taken into account the processes from the client's point of view and there is also a lack of studies focusing on the care environment of older people. There is also a need to increase the multidisciplinarity of care assessment, planning, delivery and evaluation, where different professions and disciplines work with the person at the centre of care and treatment; there has never been a better time to do this than now due to developments in precision medicine, personalized treatment and care, digitalisation [18] and consumer/community participation. All these contemporary trends and support systems can facilitate increased agency on behalf of patients/persons in need of care and can facilitate the health literacy and agency needed to individualise care directed by the person with health needs [19]. However, some further developments are needed to improve the 'individual/person literacy' on behalf of health care professionals and how this information is collected, shared and implemented in care decisions [20]. Standards and structures can also restrict the sharing of agency, power, expertise and accountability of healthcare decisions and actions, particularly in the specialized healthcare space. However, as this Special issue demonstrates, there are rapid developments in the space of personalised medicine and care, which indicate a promising future ahead.

We hope to provide an interesting and comprehensive reading experience with this Special issue. We thank all the authors and editorial office professionals for their contributions to this Special Issue; we also thank the journal and publication platform for providing us with the opportunity to collect multi-disciplinary work covering various important approaches. This Special Issue highlights the scientific advancements in the field as well as meaningful results that can be used to advance healthcare systems.

**Author Contributions:** Conceptualization R.S., M.S. and D.E.; Writing—original draft preparation, R.S.; Writing—review and editing, M.S. and D.E. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


### *Review* **How to Measure Foot Self-Care? A Methodological Review of Instruments**

**Jenni Sipilä <sup>1</sup> , Anne-Marie Mäkelä <sup>2</sup> , Sasu Hyytiä <sup>3</sup> and Minna Stolt 1,4,\***


**Abstract:** Foot self-care is an important element of caring for and promoting foot health. However, little is known about the validity and reliability of existing foot self-care instruments. The purpose of this review is to describe and analyze the focus, content, and psychometric evidence of existing instruments for measuring foot self-care. A methodological review of three international scientific databases—Medline (PubMed), CINAHL (Ebsco), and Embase—was conducted in May 2022. The search produced 3520 hits, of which 53 studies were included in the final analysis based on a twophase selection process. A total of 31 instruments were identified, of which six were observed to have been used more than once. Subsequently, the methodological quality of these six instruments was evaluated. It is noted that although a considerable variety of instruments are used in measuring foot self-care, only a small proportion are used consistently. In general, the psychometric testing instruments seem to primarily focus on analyzing content validity and homogeneity. In the future, comprehensive testing of instrument psychometrics could enhance the cumulative evidence of the methodological quality of these instruments. Furthermore, researchers and clinicians can use the information in this review to make informed choices when selecting an instrument for their purposes.

**Keywords:** foot; foot self-care; instrument; measurement; methodological review

### **1. Introduction**

Foot self-care is an important element of caring for and promoting foot health. Active and proper foot self-care help maintain foot health and prevent foot problems. Notably, foot problems are particularly prevalent in older people [1,2] and people with long-term health problems [3–6], such as diabetes mellitus and rheumatic conditions, thus emphasizing the urgent need for implementing preventive foot self-care. In this context, it is necessary to have valid instruments for evaluating patients' foot self-care. However, a systematic summary of existing foot self-care instruments is lacking in the literature.

Self-care, in general, has become a central component of health care and patients' own resources in terms of caring for, maintaining, and promoting their own health [7]. Foot selfcare is a demanding daily task that entails maintaining fine motor skills, general mobility, and upper limb dexterity with sufficient muscle strength [8]. Although a universally agreed-upon definition of foot self-care is lacking, this review defines it as individuals' own activities directed at caring for their own feet, including skin and nail care, foot pain, and use of proper footwear [9]. Foot self-care is important in caring for different foot disorders, such as dry skin, flatfoot, hallux valgus, and metatarsalgia. To conduct foot self-care, it requires competence [10], including personal knowledge, skills, motivation, and physical ability. In this context, proper knowledge relates to evidence and good practice guidelines about foot self-care. Furthermore, skills refer to individuals' abilities to conduct activities, such as skin moisturisation or nail cutting, according to evidence-based

**Citation:** Sipilä, J.; Mäkelä, A.-M.; Hyytiä, S.; Stolt, M. How to Measure Foot Self-Care? A Methodological Review of Instruments. *J. Pers. Med.* **2023**, *13*, 434. https://doi.org/ 10.3390/jpm13030434

Academic Editors: Marijn Speeckaert and José Carmelo Adsuar Sala

Received: 27 January 2023 Revised: 20 February 2023 Accepted: 27 February 2023 Published: 28 February 2023

**Copyright:** © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

guidelines and recommendations. In addition, motivation is also significant for regularly caring for one's feet.

Adherence to foot self-care among people/patients is diverse. Although foot self-care is considered important by patients, it is generally conducted unsystematically [8]. In fact, the importance of preventive foot self-care is often recognized when foot problems have already occurred [11]. Poor or improper foot self-care can, in turn, negatively affect one's foot health status. For example, the foot self-care activities of patients with diabetes who suffer from foot ulcers have been demonstrated to be improper [12,13]. Therefore, a systematic measurement of patients' foot self-care is required to identify potential gaps in their competence.

The evaluation of the validity and reliability of foot self-care instruments is a constant process, as they are used in various contexts and populations. Therefore, cumulative evidence is needed to prove the validity and reliability of a particular instrument. A single study can only provide evidence of the psychometrics in a certain sample and, therefore, cannot be regarded as the only source of evidence. This indicates that robust reporting of the instruments' development process and the results of their psychometric testing are needed to gather relevant evidence of their validity and reliability. Moreover, instruments used as self-reported outcome tools are useful only if there is evidence to support the interpretation of their obtained scores [14]. Therefore, accurate interpretations of reliability and validity considering different settings and samples can be made only when such kinds of necessary information are available. Furthermore, this information aids researchers in understanding the instrument development process, testing, and its results. From a clinical point of view, such information provides an opportunity to use instruments that have been consistently assessed as reliable.

The validity and reliability of foot- and ankle-related instruments have increasingly been under investigation. However, the target of most studies has been the assessment of foot and ankle symptoms and functions rather than foot self-care. For example, recent reviews have prominently focused on patient-reported outcome measures (PROMs) in foot and ankle orthopaedics to reveal variability and deficits in methodological quality and instrumentation [15,16] and highlighted many studies that have used non-validated instruments [17] with limited evidence of their psychometric properties [18]. Similarly, a substantial variability in the measurement properties of instruments assessing footrelated disabilities was identified in the studies focusing on patients with rheumatoid arthritis [19,20]. In contrast, foot- and ankle-related studies conducted on people with diabetes mellitus seemed to maintain a sufficient level of methodological rigour and used valid instruments to measure issues, such as diabetic neuropathy [21]. However, a review focusing on foot self-care instruments still seems to be lacking.

Previous reviews on the impact of foot care education on foot self-care have criticized the heterogeneity of the available assessment tools [22,23]. These instruments focused on measuring changes in knowledge levels, foot care behaviour, and foot health [23]. Moreover, they were developed for the purposes of a single study and measured the technical competence to carry out a certain skill or regularity of desired foot care behaviour rather narrowly [23]. Furthermore, the content areas that the instruments focus on and their validity remain unclear. Furthermore, the lack of validated outcome tools hampers the precision of the instruments' measurements and their ability to measure changes in behaviour, thus revealing flaws in their evaluation of the impact of foot care education.

Poorly conducted foot self-care can increase the risk of foot complications [24] and seriously decrease the level of foot health [25]. Thus, the proper evaluation of foot selfcare requires valid and reliable instruments. To address this challenge, a methodological psychometric review of existing foot self-care instruments could promote and facilitate their use in clinical practice and research. Therefore, the purpose of this review is to describe and analyse the focus, content, and psychometric evidence of existing instruments that measure foot self-care. The ultimate goal is to gather and provide information regarding the accurate assessment of foot self-care and the measurement properties of individual instruments.

Therefore, the primary questions that this review seeks to answer are as follows:


### **2. Materials and Methods**

### *2.1. Design*

This study was conducted by applying a methodological review design. The reporting process was carried out in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [26]. Moreover, a review protocol was planned, but not published, prior to conducting the review.

### *2.2. Eligibility Criteria*

The identified studies were included in the sample for this investigation if they: (1) Were an empirical primary study with a focus on foot self-care; (2) Included an instrument that measures foot self-care (subjective or objective); (3) Provided evidence of psychometric properties of the foot self-care instrument; and (4) Written in English. The exclusion criteria for the studies included (1) Theoretical discussion papers or (2) The use of general self-care instruments.

### *2.3. Information Sources and Search Strategy*

A methodological review was conducted across three international scientific databases (Medline (PubMed), CINAHL (Ebsco), and Embase) in May 2022. The search sentence was ((foot[Title/Abstract]) AND (self[Title/Abstract])) AND (care[Title/Abstract] OR caring[Title/Abstract] OR manage[Title/Abstract] OR management[Title/Abstract] OR efficacy[Title/Abstract]). This search was limited to the title and abstract levels and to studies written in English. Furthermore, no time limit was applied. Moreover, although the review protocol was planned a priori, it was neither published nor registered.

### *2.4. Selection Process*

The process of selecting the relevant studies included screening the records and evaluating their eligibility against the inclusion and exclusion criteria (Figure 1). All duplicate records were excluded in the screening phase. Following this, the titles and abstracts of the studies were inspected by two independent researchers (A-MM, MS). After reaching a consensus, the full texts of the included studies were read and evaluated. After each step, the researchers discussed their selections to ultimately reach a consensus. In cases of disagreement, a third researcher was consulted.

**Figure 1.** Flowchart of the study selection process. **Figure 1.** Flowchart of the study selection process.

### *2.5. Data Collection Process and Data Items*

*2.5. Data Collection Process and Data Items*  A spreadsheet was developed particularly for the purposes of this review. It included the following information: author, year of publication, country of origin, aim of the study, name of the instrument, measurement focus, number of items, response options, and a list of studies using the particular instrument. While retrieving the data, the study authors' original expressions for the instruments were used without inducing any additional interpretation. A spreadsheet was developed particularly for the purposes of this review. It included the following information: author, year of publication, country of origin, aim of the study, name of the instrument, measurement focus, number of items, response options, and a list of studies using the particular instrument. While retrieving the data, the study authors' original expressions for the instruments were used without inducing any additional interpretation.

### *2.6. Quality Appraisal*

*2.6. Quality Appraisal*  The Mixed Methods Appraisal Tool (MMAT, version 2018) [27] was used to evaluate the methodological quality of the selected studies. The MMAT consists of seven items: two general and five design-related items. The response to each item was registered in The Mixed Methods Appraisal Tool (MMAT, version 2018) [27] was used to evaluate the methodological quality of the selected studies. The MMAT consists of seven items: two general and five design-related items. The response to each item was registered in terms of a three-point scale (yes, no, and can't tell).

### terms of a three-point scale (yes, no, and can't tell). *2.7. Synthesis of Results*

The instruments used by patients to measure their performed activities related to foot self-care were first identified from the original articles. After identification, they were listed

and classified into groups according to their names. Notably, some of these instruments were used in multiple studies, while the revised versions of some were also reported.

Subsequently, the articles were categorized into studies that (a) Reported original instrument development research; (b) Reported further validation of a certain instrument; and (c) Used instruments without considering any information regarding their validation or psychometric testing. Based on this categorization, instruments that were used in multiple studies were identified, and further analysis was conducted on them.

Descriptive information of each instrument on the item level was gathered to a separate table. Items were grouped based on their characteristics to show which foot self-care content areas were present in these instruments. To have an understanding of item coverage, the number of items describing a certain aspect of foot self-care was summed up.

The psychometric properties of each instrument were analyzed using the framework proposed by Zwakhalen and colleagues [28], which includes 10 items that cover the most crucial aspects of validity and reliability: (1) Known origin of the items; (2) Sufficient sample for testing (number of participants); (3) Analysis of, and justification for, content validity; (4) Level of criterion validity achieved using correlation; (5) Construct validity in relation to other appropriate knowledge tests; (6) Construct validity of differentiation; (7) Homogeneity; (8) Inter-rater reliability (confirmed through observation or noted in activity); (9) Intra-rater or test–retest reliability; and (10) Feasibility. Each item was scored as either 0, 1, or 2 by the two researchers according to the relevant scoring criteria [28]. To gather an overall level of psychometric evidence, the scores were summed up. On summing up the scores, the maximum score stood at 20, with a higher score representing a higher level of the analyzed psychometric property. Originally, these criteria were created to evaluate a pain assessment tool for people with memory disorders [28]. Although the psychometric properties assessed in the original study are universal, the assessment criteria of this study were modified to correspond to foot self-care content. Moreover, the criteria suggested by Zwakhalen and her colleagues [28] have been previously used to assess the psychometric properties of instruments [29]. Notably, this framework was developed based on the methodological literature as a quality judgement criterion for instruments in nursing and health research.

### **3. Results**

### *3.1. Study Selection*

The search produced a total of 3520 hits (*n* = 1449 for Medline and PubMed, *n* = 466 for CINAHL and *n* = 1605 for Embase). After removing duplicates, 2417 hits were included in the study selection phase. Subsequently, 119 studies were selected after screening the titles and abstracts of the studies. Following this, based on the full texts of the selected studies, 53 that met the eligibility criteria were included in the final analysis.

### *3.2. Study Characteristics*

A total of 31 instruments used in 53 studies were identified for further analysis (Supplementary Table S1). These instruments were observed to predominantly measure self-reported foot self-care behaviours or activities. The number of items in the instruments ranged from 4 to 29, while a five-point response scale was primarily used to indicate the frequency of activities related to foot self-care. Out of these 31 instruments, 25 were used only once or were developed for the purpose of a single study. In addition, six instruments were unnamed. Therefore, the analysis of psychometric properties targeted the instruments (*n* = 6) that were used more than once.

### *3.3. Description of the Analysed Instruments*

This section provides a detailed description of the six instruments that were employed in multiple studies considered in the sample selected for this investigation.

The Diabetes Foot Self-care Behavior Scale (DFSBS) [30] measures the frequency of foot self-care behaviour. It entails seven items: checking the bottom of the feet and between toes, washing between toes, drying between toes after washing, applying lotion, inspecting the insides of shoes, and breaking in new shoes. Furthermore, this scale has two parts. Part 1 assesses the number of days that a respondent performs a certain behaviour during a 1-week period using the five-point scale (0 days, 1–2 days, 3–4 days, 5–6 days, 7 days). Meanwhile, Part 2 evaluates the frequency at which a respondent performs a certain foot self-care behaviour (5-point scale from 1 = never to 5 = always). These ratings are summed up to arrive at a score, with higher scores indicating a better performance of foot self-care behaviour [30]. This section may be divided by subheadings. It should provide a concise and precise description of the experimental results, their interpretation, as well as the experimental conclusions that can be drawn.

The Summary of Diabetes Self-Care Activities (SDSCA) [31] is a self-report instrument dealing with diabetes self-management. It covers all the self-care areas related to diabetes: general diet, specific diet, exercise, blood-glucose testing, foot care, and smoking. Since specific parts of this instrument can be used separately, many studies investigated in this review were observed to implement the foot care section of the SDSCA, which consists of five items that help to identify the number of days in a week that a person has performed diabetes foot self-care activities: feet washing, feet soaking, drying between the toes after washing, foot checks, and footwear inspection. The response to each item is registered on a scale from 0 to 7, based on the number of days that the person performed the activity (the higher the mean, the better the care) [31].

The Nottingham Assessment of Functional Footcare (NAFF) [32] is a self-report instrument for assessing the foot care behaviour of people with diabetes. This tool accounts for a total of 29 items to measure the extent to which people comply with recommended foot care behaviours. Response options for this instrument are provided on a four-point scale ranging from "rarely" to "most of the time." The sum of the items' scores is then calculated, with a higher score indicating better foot self-care [32].

The Diabetic Foot Self-Care Questionnaire of the University of Malaga, Spain (DFSQ-UMA) [33] was formulated to evaluate foot self-care among patients with diabetes. It consists of 16 questions that are divided into three domains: personal self-care, podiatric care, and shoes and socks. Each item is scored on a five-point response scale (1 = very inadequate; 5 = very adequate), while some items explore the frequency of a determined self-care activity (1 = never; 5 = always) [33].

The Foot Self-Care Behaviour Questionnaire [34] measures the frequency of performing foot care behaviour based on 17 items divided into two subscales: preventive foot selfcare (nine items) and potentially foot-damaging behaviour (eight items). Similarly, two different response scales are used: a six-point scale (i.e., twice a day, daily, every other day, twice a week, once a week, or never) and a four-point scale (i.e., always, most of the time, occasionally, or never). The responses are summed up, wherein a higher score indicates more preventive and potentially damaging behaviours.

The Foot Self-Care Observation Guide (FSCOG) [35] is an objective observation measurement for detecting foot self-exam components. It consists of 16 items divided into three categories: foot care (five items), foot check (three items), and foot safety (eight items). The responses are provided on a five-point scale (1 = never, 2 = occasionally, 3 = sometimes, 4 = frequently, and 5 = always), and then the responses are summarized (range 15–75) with a higher score indicating better foot self-care behaviour.

### *3.4. Methodological Quality of the Included Studies*

The methodological quality of the selected studies, as assessed by the MMAT [27], was found to be acceptable (Supplementary Table S2). However, the main deficit observed in the descriptive quantitative studies was related to non-response bias. Furthermore, the blinding of the assessors was seldom achieved in the randomized controlled trials. For the non-randomized studies, the main shortcoming was recognized as accounting for the confounders of the analysis.

### *3.5. Focus and Content of the Instruments*

The selected instruments were implemented for primarily measuring the frequency of foot self-care (DFSBS [30]; SDSCA [31]; DFSQ-UMA [33], foot self-care activities (DFSQ-UMA [33]) and compliance with recommended foot care (NAFF [32]).

At the item level (Table 1), the focus was predominantly on the selection of the type of footwear (*n* = 15), followed by questions related to socks (*n* = 8), footwear assessment (*n* = 7), foot inspection (*n* = 7), drying the feet after washing (*n* = 6), and skin care (*n* = 6). A few other items that were targeted for evaluation include walking barefoot (*n* = 4), skin moisturization (*n* = 4), foot washing (*n* = 3), nail cutting (*n* = 3), attitude towards foot self-care (*n* = 2), and foot soaking (*n* = 1).

**Table 1.** Content and number of items in the instruments measuring foot self-care.


### *3.6. Psychometric Evidence of the Instruments*

The evidence on the psychometric properties of the instruments varied (Table 2, Supplementary Table S3). First, it should be noted that all the assessed foot self-care instruments were developed based on comprehensive literature reviews. In addition, some authors (NAFF [32] and FSCOG [35]) incorporated foot care recommendations or guidelines to strengthen the theoretical basis of their instruments.




*J. Pers. Med.* **2023**, *13*, 434



*J. Pers. Med.* **2023**, *13*, 434

**Table 2.** *Cont.*


The DFSBS [30] was originally tested comprehensively on a sample of 295 patients with diabetes. Evidence of its content validity, construct validity, differentiation, and internal consistency was also provided. In addition, its feasibility was tested using a pilot test [30]. Over time, the DFSBS has been translated into Arabic [38], Iranian [36], Malay [39,41], and Turkish [40] languages. Moreover, several studies that provide evidence of the instrument's construct validity and internal consistency [36,38–41] have been conducted. However, no evidence of its criterion or construct validity was provided, while its reliability testing focused only on homogeneity. Notably, its feasibility was evaluated by a few studies in pilot testing [30,38,39].

Limited reports could be acquired on the psychometric evidence for the SDSCA [31]. Only one study reported evaluating the feasibility of the instrument through a pilot study [45]. Meanwhile, it was noted that the NAFF [47] was tested for content validity and homogeneity [49].

The DFSQ–UMA [33] was found to be based on a comprehensive literature review followed by careful item operationalization. In addition, extensive testing of its psychometric properties was conducted. The original study thoroughly reported the instrument development process and provided evidence of its validity and reliability in a sample of patients with diabetes [33]. However, its intra-rater reliability was not tested.

The Foot Self-Care Behaviour Questionnaire [34] demonstrated both content validity and homogeneity. Furthermore, the Foot Self-Care Observation Guide [35] reported sufficient evidence of its content validity, internal consistency, test-retest reliability, and feasibility.

### *3.7. Synthesis of the Results*

There are several instruments available for measuring foot self-care, all of which especially focus on foot self-care among people with diabetes. However, among the investigated studies, only six instruments were used more than once, thus providing scattered psychometric evidence. In addition, although the instruments were tested for validity and reliability, the testing process for the different instruments varied, with their predominant focus being on content validity and homogeneity.

### **4. Discussion**

The current study identified six instruments that measure foot self-care, all of which focus primarily on patients with diabetes. Although each instrument had sufficient evidence supporting its usefulness in evaluating foot self-care, they focused mostly on content validity and homogeneity. This indicates the need for systematic and comprehensive psychometric testing of these instruments. In addition, a wealth of instruments used in single studies was found to have limited evidence of their development processes and psychometrics. Moreover, with the increasing volume of tools used for this purpose, duplication and variability were some of the challenges faced in choosing specific instruments among the available ones.

The selected instruments varied in terms of their complexity regarding the items and factors covered. A clear definition of the construct is necessary to search for the most accurate instrument in terms of a given context [54]. In this context, foot self-care is a complex construct that encompasses knowledge (to know what to do and how), skills (how to care for the feet in real life by implementing the correct procedures), and attitude (motivation for carrying out foot self-care). However, since there is currently no standardized definition of foot self-care, the selected instruments were free to measure different kinds of factors related to foot self-care. The main content of the instruments was related to foot inspection, the type of footwear, socks, and the warming of the feet. This is probably because these areas are theoretically relevant and fundamental when it comes to feet care for patients with diabetes. However, in the future, a concept analysis of foot self-care could be beneficial for improving the measurement focus of foot self-care instruments. The NAFF [32] was found to be the most comprehensive instrument covering

a wide number of foot self-care activities. Moreover, the target population in the studies was predominantly patients with diabetes. Therefore, constructing an instrument for measuring basic foot self-care activities performed by the general population could be relevant for population-based health promotion programmes.

Although the construct of interest is generally clearly defined when searching for an appropriate instrument, one should be aware of whether development study and psychometric testing of the specific instrument on the target population were performed [54].

In the future, a strong emphasis should be placed on testing construct validity and reliability in terms of an instrument's intra-rater and test-retest (stability) reliability. A particularly significant need is to conduct further psychometric evaluation studies on existing instruments and adapt them accordingly, focusing especially on the psychometric properties that are rarely evaluated, such as reliability, measurement error, and responsiveness [14,16]. From a clinical perspective, more information on the clinical feasibility of the evaluated instruments is necessary. To address this issue, researchers could benefit from planning a detailed testing procedure for their instruments. Since all the elements of validity and reliability are impossible to test in a single study, collecting data from different settings and participants, such as patients under home care or those having long-term health conditions like rheumatoid arthritis, could help cumulate more evidence and, in turn, improve the methodological quality of the instruments. Moreover, researchers should not rely solely on internal consistency as an indicator of reliability [14].

Given the variety of instruments that are currently available, understanding the quality of evidence about an instrument for evaluating its measurement properties is essential to make an informed selection of the most appropriate tool and properly assess foot self-care in the population of interest. The consistent use of instruments that have been assessed as valid and reliable allows for the systematic and credible monitoring and comparison of measured results [16]. This highlights the need for a clear definition of foot self-care to help focus future research accordingly.

Authors should discuss the results and how they can be interpreted from the perspective of previous studies and of the working hypotheses. The findings and their implications should be discussed in the broadest context possible. Future research directions may also be highlighted.

### *Limitations*

The results of this review need to be interpreted while also considering some limitations. First, although the literature search on the selected databases was comprehensive, adding more databases would have provided more hits. In addition, the systematic search conducted across the three databases provided several duplicates, indicating overlapping content. To ensure accurate and comprehensive search terms, pilot searches were conducted, and the search terms were modified and approved by the research team. One researcher (MS) conducted the search, while the study selection was handled by two independent researchers (A-MM, MS). No discrepancies that needed intervention by a third researcher evaluation were encountered. Moreover, data retrieval and analysis were conducted by two researchers (JS, MS) to ensure the transparency of the analysis.

This review was limited to studies published in English. As a result, studies published in other languages that deal with the development of good-quality instruments or the measurement properties of the selected studies were omitted. Therefore, further research can focus on studies conducted in other languages. Moreover, the selected framework developed by Zwakhalen and colleagues [28] represents a general structure for evaluating the psychometric qualities of health measurement scales.

### **5. Conclusions**

Although many instruments were identified as potentially suitable for evaluating foot self-care, deficits in demonstrating adequate measurement properties were recognized across all the domains of reliability and validity. Particularly, the evidence of the instruments' sensitivity to detecting changes in foot self-care is required before they are used as outcome instruments, such as for interventions.

The number of items evaluated in the selected studies ranged from 4 to 29. This supports the need for a concise instrument to minimize patient burden, maximize patient engagement, and ensure the collection of meaningful data [16].

A considerable variety of instruments are used to measure foot self-care, with a small proportion being used consistently. Moreover, substantial variability exists in their level of methodological rigour. Foot self-care instruments are important indicators of patients' competence in promoting and maintaining their foot health. With precise, welltargeted, and sensitive instruments, health care professionals may be able to monitor the progress of their patients' foot self-care and evaluate the impact of educational foot health interventions [14]. In addition, in terms of their clinical utility, foot self-care instruments are important for enhancing patients' engagement, outcome evaluation, and the evaluation of their motivation for carrying out foot health care. Future research should focus on testing the psychometric properties of the instruments as it could provide the benefit of incorporating tests from modern test theory, such as Rasch analysis. Most importantly, researchers and clinicians can take recourse to the information provided in this review to make informed choices when selecting an instrument for their purposes.

**Supplementary Materials:** The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/jpm13030434/s1, Table S1: Descriptive information about the instruments included in the review. Table S2: Methodological quality of included studies (*n* = 53) using MMAT tool. Table S3: Detailed analysis of psychometric evidence of six instruments measuring foot self-care, analysed against the criteria proposed by Zwakhalen and colleagues [28,30–53,55–84].

**Author Contributions:** Conceptualization, M.S. and A.-M.M.; methodology, J.S., A.-M.M., S.H. and M.S.; formal analysis, J.S. and M.S.; resources, M.S.; writing—original draft preparation, J.S., A.-M.M., S.H. and M.S..; writing—review and editing, M.S. and S.H.; supervision, M.S.; project administration, M.S.; funding acquisition, M.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by Turku University Hospital, grant number 13240.

**Institutional Review Board Statement:** Not applicable.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


**Disclaimer/Publisher's Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

### *Article* **Women's Experiences with Compliance with Pelvic Floor Home Exercise Therapy and Lifestyle Changes for Pelvic Organ Prolapse Symptoms: A Qualitative Study**

**María Torres-Lacomba , Beatriz Navarro-Brazález \* , María José Yuste-Sánchez , Beatriz Sánchez-Sánchez , Virginia Prieto-Gómez and Fernando Vergara-Pérez**

> Physiotherapy in Women's Health (FPSM) Research Group, Physiotherapy Department, Faculty of Medicine and Health Sciences, University of Alcalá, 28805 Madrid, Spain; maria.torres@uah.es (M.T.-L.); marijo.yuste@uah.es (M.J.Y.-S.); beatriz.sanchez@uah.es (B.S.-S.); v.prieto@uah.es (V.P.-G.); fernando.vergara@uah.es (F.V.-P.)

**\*** Correspondence: b.navarro@uah.es

**Abstract:** In this study, we aimed to investigate women's experiences with compliance with prescribed pelvic floor muscle exercises (PFMEs) and lifestyle changes 6–12 months after completing an individual pelvic floor physiotherapy program. This study was targeted to understanding factors affecting adherence to PFMEs and lifestyle changes to deal with pelvic organ prolapse (POP) symptoms. We designed this research as a descriptive qualitative study. We conducted this study from December 2016 to September 2017 in Madrid, Spain. Twenty-six women with symptomatic POP selected using a purposive sampling method participated in six focus groups and three one-to-one semi-structured interviews. Three authors coded and inductively analyzed transcript contents with iterative theme development. A thematic analysis revealed three main themes: (1) symptoms change; (2) PFMEs and lifestyle changes performance; and (3) a health practitioner–patient relationship. Women identified as adherent reported improvement in physical symptoms and emotional and general state as a result of the new knowledge achieved. Fear also promoted compliance with performing PFMEs and adopting lifestyle changes. Likewise, PFMEs preference and routine, integration of PFMEs and lifestyle changes into activities of daily living, support guides, therapeutic alliance, individual supervision, follow-up, and feedback were also identified as adherence facilitators. One of the biggest barriers that we identified was responsibility. Compliance with prescribed PFMEs and lifestyle changes can be improved with effective individual, women-centered, and supervised physiotherapy programs reducing symptoms, including exercises aligned with women's preferences that are easy to integrate in daily living, promoting knowledge and awareness of their condition, providing written or electronic guidelines, with routine follow-up visits offering both positive feedback and clear and consistent messages, and enhancing therapeutic alliance.

**Keywords:** pelvic organ prolapse; pelvic floor muscle exercises; lifestyle changes; therapeutic adherence; women's experience

### **1. Introduction**

Pelvic organ prolapse (POP) is a common condition affecting mostly postmenopausal women, with a peak age prevalence at 60–69 years [1,2]. The symptoms significantly impair women's daily activities and quality of life [1,3]. POP involves the descent of one or more perineal organs (uterus, bladder, and/or rectum) from the normal anatomic position into the vagina [4]. Widespread symptoms include feeling vaginal building or pressure in the pelvis, discomfort in the perineum, low back and pelvic pain, which are commonly associated with other urinary, bowel, and sexual symptoms, such as urinary incontinence, fecal incontinence, constipation, and sexual dysfunction that may or may not be related to POP [4–7]. Moreover, women with POP have reduced pelvic floor muscle (PFM) strength

**Citation:** Torres-Lacomba, M.; Navarro-Brazález, B.; Yuste-Sánchez, M.J.; Sánchez-Sánchez, B.; Prieto-Gómez, V.; Vergara-Pérez, F. Women's Experiences with Compliance with Pelvic Floor Home Exercise Therapy and Lifestyle Changes for Pelvic Organ Prolapse Symptoms: A Qualitative Study. *J. Pers. Med.* **2022**, *12*, 498. https:// doi.org/10.3390/jpm12030498

Academic Editors: Riitta Suhonen, Christian Singer, Minna Stolt and David Edvardsson

Received: 19 November 2021 Accepted: 17 March 2022 Published: 19 March 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

and greater PFM dysfunction, with more severe POP and urinary symptoms [8], and there are weaker PFM involuntary contractions during increases in intra-abdominal pressure such as coughing in women with POP stage I and II than in women without POP [9].

POP can be managed by surgery, conservative management (lifestyle advice, PFM training, and vaginal pessaries), or a combination of these [1,7,10,11]. Conservative treatments are often recommended if the POP is mild or surgery is not indicated. Additionally, surgical treatments are commonly associated with increased risk of postoperative complications and POP recurrence [11–13]. According to the latest meta-analysis, pelvic floor muscle exercises (PFMEs) can effectively improve POP symptoms and stage compared to controls [14]. Voluntary PFM contraction may improve support of the pelvic organs as well as their support in the normal anatomic position by contracting PFM before and during any increase in abdominal pressure such as abdominal straining, cough, etc. [14]. PFMEs are more strongly recommended, but their symptom reduction success rates decrease in the medium and long term as adherence to the program deteriorates [15,16]. Although adherence to PFMEs in the short and medium term has been investigated in women with urinary incontinence, women with POP likely also adhere less to PFME programs over time.

To the best of our knowledge, only one study explored adherence to a PFME program in five women with POP in New Zealand with a single data collection method: a one-to-one interview with a single researcher [17].

Thus, in this study, we aimed to investigate the experience with compliance with PFM home exercises and lifestyle changes in a group of women with POP who completed an 8 week supervised pelvic floor physiotherapy program combining PFM physiotherapy, PFM home exercises, and therapeutic education (which included lifestyle changes) in order to understand the factors that can facilitate or inhibit women's compliance with PFM home exercises and lifestyle changes for improving adherence and person-centered care.

### **2. Materials and Methods**

### *2.1. Study Design*

We conducted this qualitative research between December 2016 and September 2017 at the Research Unit of the Physiotherapy in Women's Health Research Group of the University of Alcalá (Madrid, Spain). We selected a descriptive qualitative design [18], and we conducted the study using focus groups and one-to-one interviews [19]. We used focus groups rather than individual interviews mainly because (a) sensitive and personal disclosures are more likely in a focus-group setting, (b) respondents are more likely to be candid when other similar people are present, (c) there is less individual pressure than in an in-depth interview, and (d) the moderator can more easily reintroduce a topic not sufficiently covered than in a one-to-one interview [20]. The study protocol was approved by the Ethics Committee for Clinical Research of the Príncipe de Asturias Hospital (OE10/2010) in Alcalá de Henares, Madrid, Spain. The study reporting followed the Standards for Reporting Qualitative Research (SRQR) guidelines [21], as well as the COREQ checklist [22]. We followed the ethical principles of the Declaration of Helsinki. All participants provided informed written consent.

### *2.2. Participants*

Selected using a purposive sampling method [23], we considered 26 women diagnosed by their gynecologist, according to the POP-Quantification Scheme [4], with symptomatic POP, recruited from the Hospital Príncipe de Asturias (Madrid, Spain), and referred to a specialized women's health unit to receive an individual woman-centeredness program of pelvic floor physiotherapy to manage symptoms of POP for inclusion in the study. Women over 70 years of age, who gave birth within the six months prior to referral, with stage IV POP, with previous surgery for POP, with psychiatric disease, and who were not able to understand Spanish were excluded. Women who completed the physiotherapy program between 6 and 12 months prior to the start of the study were invited to participate.

The pelvic floor physiotherapy program consisted of 16 individual sessions over 8 weeks (two weekly sessions) with a women's health physiotherapist. The intervention included PFMEs with manual guidance and biofeedback progressing according to the PERFECT scheme from supine to standing and toward functional activities. Women also performed the PFMEs at home, prescribed one to three times per day during the eight-week intervention period. The program also included therapeutic education consisting of instruction with printed and audiovisual materials about the pelvic floor, pelvic floor dysfunctions, the identification of possible precipitating factors, weight loss, constipation, heavy lifting, coughing, high-impact exercise, knack, etc., together with individual strategies for implementing these measures [24]. Upon completion of the physiotherapy program, women were encouraged to continue PFMEs at home (once per day, at least three days per week) and to continue to comply with lifestyle changes.

### *2.3. Data Collection*

Women who completed the pelvic floor physiotherapy program 6–12 months ago with an interest in participation were approached by the research team (M.J.Y.-S.) to determine their availability and convey arrangements with the researcher (M.T.-L.) for both conversational-style focus groups and one-to-one interviews. The six focus groups lasted 60 to 70 min each and involved 4, 3, 5, 4, 4, and 3 women, respectively. The three one-to-one interviews lasted 40 to 50 min each. All the focus groups and one-to one interviews were conducted by a facilitator (a physiotherapist experienced in focus groups and one-to one interviewing, M.T.-L.) in private, closed rooms at the Research Unit of the Physiotherapy in Women's Health Research Group of the University of Alcalá (Madrid, Spain) using a semi-structured approach. In addition, during the focus groups, the facilitator was supported by a qualified pelvic floor physiotherapist (B.N.-B.) who acted as an observer, taking hand-written field notes on nonverbal communications, other observations, etc. We generated the interview questions in the semi-structured guide (Appendix A) with regard to the literature on PFMEs adherence [17,25]. The questions were discussed by the study team: four qualified physiotherapists, two experienced in qualitative research (F.V.-P. and B.S.-S.) and the other two in women's health (B.N.-B. and M.T.-L.); an experienced gynecologist; and a midwife. All interviews were digitally audio-recorded and later transcribed verbatim by two researchers (B.N.-B. and V.P.-G.), with all women's names anonymized in the transcripts. We assigned different codes according to the type of interview (focus group or one-to-one interview) and to the women's intervention order.

### *2.4. Data Analysis*

The analysis was conducted by three members of the research team (B.N.-B., F.V.-P., and M.T.-L.). A triangulation process (investigator and data collection) was conducted [26] to ensure rigor in research. We repeatedly read each transcript. Next, in an iterative and consensus process between three researchers, initial codes were generated and described, codes were grouped into higher-order categories, and then the categories were arranged under potential themes. We performed the transcription coding using ATLAS.ti version 6.1 software (Scientific Software Development GMBH, Berlin, Germany).

When the data being collected were repetitive and no new issues were emerging, we considered that data saturation was achieved, so we ceased collecting data [27]. This arose after six focus group interviews and three one-to-one interviews.

### **3. Results**

Twenty-six women with symptomatic POP who completed a pelvic floor physiotherapy program 6–12 months ago participated in six focus groups and three one-to one interviews. Two women were Latin American, and the remaining were Caucasian. All women spoke and understood Spanish. The women's demographics and POP status are shown in Table 1.


**Table 1.** Women's demographics and prolapse status.

BMI: Body mass index; PFD: Pelvic floor dysfunction; SUI: stress urinary incontinence; UUI: urgency urinary incontinence; MUI: mixed urinary incontinence; FI: Fecal incontinence; P-QoL: Prolapse Quality of Life Questionnaire (\* an improvement of 14.5 points is considered clinically relevant [28]); PFDI-20: Pelvic Floor Distress Inventory Short Form (\*\* an improvement of 13.5 points is considered clinically relevant [29]); \*\*\* an improvement of 9 cmH2O is considered clinically relevant [30]; PFM: Pelvic floor muscles; X (SD): Mean (Standard Deviation).

Women's experiences (facilitators and barriers) in complying with home PFMEs and lifestyle changes were identified in three themes: (1) symptom changes; (2) performance of PFMEs and lifestyle changes; (3) the health practitioner–patient relationship (Figure 1).

**Figure 1.** Summary of themes and codes. PFMEs: pelvic floor muscle exercises. **Figure 1.** Summary of themes and codes. PFMEs: pelvic floor muscle exercises.

### *3.1. Theme 1: Symptom Changes*

*3.1. Theme 1: Symptom Changes*  We identified improvement in physical symptoms as one of the main factors motivating compliance with home PFMEs and lifestyle changes, such as voluntary PFM contractions before and during activities that increase abdominal pressure such as weight bearing, coughing, etc. Women reported that having less of a bulging sensation in the vagina or a reduction in the episodes of urine leakage encouraged them to continue with We identified improvement in physical symptoms as one of the main factors motivating compliance with home PFMEs and lifestyle changes, such as voluntary PFM contractions before and during activities that increase abdominal pressure such as weight bearing, coughing, etc. Women reported that having less of a bulging sensation in the vagina or a reduction in the episodes of urine leakage encouraged them to continue with the exercises.

the exercises. *I am happy… I no longer leak urine…because before I had frequent leaks. It's what bothered and I disliked it the most... Now, yes, I feel calm because the exercises that have I am happy* . . . *I no longer leak urine* . . . *because before I had frequent leaks. It's what bothered and I disliked it the most* . . . *Now, yes, I feel calm because the exercises that have taught me* . . . *I'm fine. (FG3-POP-P5)*

*taught me ... I'm fine. (FG3-POP-P5) At the moment I continue to do them [PFMEs and lifestyle changes], mainly contractions before weightbearing and coughing [knack], they have been very good for me, they have been very good for me, I feel much better because before, it seemed that I was walking with a ball there [she points to her vagina] all day, but now I don't feel anything [an At the moment I continue to do them [PFMEs and lifestyle changes], mainly contractions before weightbearing and coughing [knack], they have been very good for me, they have been very good for me, I feel much better because before, it seemed that I was walking with a ball there [she points to her vagina] all day, but now I don't feel anything [an expression of relief spread across her face]. Sometimes when I get constipated, I feel it a bit [the bulge* *into the vagina], although the position I learned to defecate is better for me, and of course, it is not the same as before coming here to do the exercises and learn those positions to urine, defecate, or contract before coughing, and those things that we learned here...at least I have felt that. (FG2-POP-P4)*

We also found that the fear of worsening symptoms, or feeling the symptoms that they learned to control again, or of surgery as a reminder or trigger to resume the PFMEs and lifestyle changes were facilitating factors.

*As soon as I neglect myself and do the exercises less, or I have an allergy episode that I sneeze many times and I notice heaviness there [pelvis and vagina] or I start again with small leaks [urine leaks], I begin to do the exercise again, to contract [PFM] before sneezing, coughing, etc. Right away I think, I don't want to go back to be like before* . . . *(FG3-POP-P5)*

*I am afraid, afraid that everything will crumble and that is why I force myself to do it [PFMEs and lifestyle changes], because I am afraid. (FG6-POP-P2)*

*Let's see, my neighbor has been operated and she has not been well. I think about the operation. So, the fear, right? I fear the operation, I know women who have put that surgical mesh on. Very bad. And after three months they have put it back on again, and... I don't want to go through that. (FG6-POP-P2)*

*I'm afraid that it will fall more, that everything will collapse, and then that will stay dry, and I will get injured and have problems, that is, I am afraid that it will get worse. (Int-POP-P3)*

This ability to control their symptoms produced an increase in confidence, security, and satisfaction.

*I was lost, emotionally lost, and everything has improved for me, because I did not know who to turn to and then I already met you one day. Well, my physical condition improved* . . . *physically and mentally, I feel more secure. (FG2-POP-P3)*

*I do it [PFMEs and lifestyle changes] to feel better, so that it doesn't get worse [POP] and improve my quality of life. (Int-POP-P3)*

*You feel satisfied, you know that you can retain it [POP and urine leaks], that you can control it, and that you feel freer. (FG4-POP-P3)*

. . . *for me it has been a great benefit. I'm very happy" (Int-POP-P1)*

They also reported perceiving a global improvement, an improvement in their sexual activity, and even in their quality of life.

*I have learned to control my muscles* . . . *. And to also facilitate many things in my body, to know how to control it as well. Sexually it also favors me, that is, it also stimulates many things. And to know my entire body. (FG1-POP-P1)*

Women also associated these positive changes with their newly acquired knowledge and shared it with other women. New knowledge about the anatomy and physiology of the pelvic floor, pelvic floor dysfunctions, risk factors, management, how to correctly perform the PFMEs, etc., seemed give them a sense of control, facilitating adherence. The valuable knowledge gained was key to understanding the importance of incorporating PFMEs and lifestyle changes into their routines and living activities to improve their symptoms and thus prevent both POP and other pelvic floor dysfunctions from progressing.

. . . *everything I have learned here has been surprising for me, and a satisfaction, I had no idea about so many things! Now, I know how important it is for me [PFMEs and lifestyle changes] and I even share it [new knowledge] with my daughters, my friends* . . . *(Int-POP-P2)*

*Knowing what you can do to improve, why you have to do it, understanding it, is fundamental* . . . *Yes, yes, I already contract instinctively, especially when coughing and* *sneezing, before gaining weight too, I control the stimulating drinks, I no hold myself for so long to go to the bathroom to urine, I defecate in a more adequate posture that helps me not to push* . . . *(FG3-POP-P3)*

*And I think that this pelvic floor program should be done for every woman from now on. To adolescents, that they learn everything that we have learned from now on... so that they learn from young, from before getting pregnant to take care of their pelvic floor and that what happens to us does not happen to them. (FG1-POP-P4)*

### *3.2. Theme 2: Performance of PFMEs and Lifestyle Changes*

In the theme of performing PFMEs and lifestyle changes, the experience of each woman with PFMEs was included. Women taught to perform the PFMEs felt they were performing them better and that, therefore, they had control (self-efficacy). In addition, the association of PFMEs with some everyday gestures (e.g., voluntary PFM contractions (Knack) before and during activities that increase abdominal pressure such as weight bearing, coughing, etc.,) act as a reminder to perform PFMEs. Therefore, these activities supported compliance.

*But always the same, the one to maintain* . . . *I like and am good at. Although, I do PFMEs less than before, really, but everything I learned [lifestyle changes] I have fully integrated, and I no longer have to think about it, I just do it. (FG1-POP-P4)*

*I, for example, the one on "the ladder", well, the one on "the ladder" and the one "going down and up fast". I like them and I feel it very well. Also, the one to contract [Knack] before laughing or coughing, I like it a lot because it prevents me from leaks. (FG4-POP-P2)*

*Stand up, especially when I carry weight and sneeze or cough, I feel how it [knack] holds it [POP] inside [into the vagina]. (FG6-POP-P1)*

*That is the one I do best, of course* . . . *"the holding" and the knack, drinking, bathroom, etc. [lifestyle changes], that are totally integrated into my life. (Int-POP-P2)*

Most of the women stated that integrating PFMEs and lifestyle changes into their daily life, the possibility of performing PFMEs in any place, and a PFME regular routine (routine of place, routine of time of day) made it easier to continue these changes. Most women claimed to be much more consistent with lifestyle changes, which they considered fully incorporated into their routines.

*...what is easier to do, I see that I can do it [PFMEs] at any time. When I am standing, I am sitting, I am even washing the dishes, I am with the children, because I hold my baby in my arms, and of course, that forces me, because I know that I am going to force myself and that I am going to do it [Knack] at the same time....I take advantage of those moments. (FG3-POP-P5)*

*I take advantage of it the most, with the iron, you are standing up, you relax a little and come on, let's do it [PFMEs]. I really like that. (Int-POP-P2)*

*Yes, I am also* . . . *At work doing it [PFMEs] at all times* . . . *or waiting for the bus. (Int-POP-P3)*

. . . *well on Monday, Wednesday and Friday I do these [PFMEs], and on Tuesday and Thursday these others [PFMEs]. Always when I go to bed. (FG6-POP-P3)*

*I have incorporated them [PFMEs] into my gym routine, I do them three times a week that I go to the gym. And I also contract [Knack] when she taught me B. [Pelvic floor program Physiotherapist] during Pilates exercises. (FG4-POP-P4)*

*I always do them [PFMEs] after I shower in the morning, it is already routine and I always do them, just like contracting [Knack] before heavy load, I have already automated it. (Int-POP-P1)*

However, several women reported a lack of criteria regarding the time of day of exercises, repetitions, and progression that made it difficult to perform PFMEs.

*And then when I suddenly remember and say to myself: come on, I'm going to do them at noon, even if it's on a carpet. I will always win something* . . . *. And then at night when I go to bed I say: Uffff! And again I have to do this* . . . *.. how many do I have to do, how I continue* . . . *? (FG2-POP-P2)*

*And especially when I look bad, I see that I need it and then I do them a few days, oh my God! Why have I stopped it? Pum pum* . . . *(FG4-POP-P3)*

In addition, individualized ad hoc guides (written, spoken, or apps) providing reminders regarding PFME types, repetitions, progression, and lifestyle changes were described as support material that acted as a reminder facilitating adherence to PFMEs and lifestyle changes.

*What B. [pelvic floor program Physiotherapist] put in the notebook is what I exactly do, and so I don't forget. (FG3-POP-P1)*

*I do what B. [pelvic floor program Physiotherapist] taught me, I have it here written down on a paper [written information], and very well. (Int-POP-P1)*

. . . *I always have a small piece of paper, if I have a doubt...for example: I don't remember the progression of an exercise, so I look at the paper that I have on the bedside table, and I check the paper. (FG2-POP-P1)*

*Well, I preferred to record it on my mobile phone, and that is how I always carry it with me, and if I don't remember something, I listen to it quickly, I do it on the subway when I go to work, and it suits me very well. (Int-POP-P3)*

*The apps she [pelvic floor program Physiotherapist] taught me to do the exercises is fantastic, I carry it on my mobile and I do them with the app. I like it. (FG3-POP-P2)*

Finally, the feeling of responsibility and the need to take care of themselves as they take care of others emerged from the sessions, mostly related as a barrier to integrating PFMEs in their daily living. Their roles as caregivers in the family negatively influenced their adherence to PFMEs, since they take care of the family first, and they do not pay enough attention to their self-care.

*It is our responsibility, the disease [POP], but we are not consistent, and we also think before of others, children, family. (FG5-POP-P4)*

*We are not in the habit of spending time with ourselves. Because it is a bad habit and a bad practice of ourselves. So, we always make the mistake thinking "it doesn't matter". The last one, us. And we are not interested in ourselves. (FG5-POP-P3)*

*Well, let's say that our will fails us, the truth is a bit for me that I think I had to be more responsible, and I haven't done them [PFMEs]. (Int-POP-P1)*

*And we don't worry about ourselves. Always. And even if we don't have anything to do, we don't think about ourselves. (Int-POP-P3)*

*3.3. Theme 3: Healthcare Practitioner–Patient Relationship*

The individualized pelvic floor physiotherapy program including follow-up sessions was identified as essential for performing PFMEs and for compliance with lifestyle changes.

*It is not the same as going to the gynecologist, and they tell you: look, do these exercises that will help you. And then, of course, you say: okay, I do them, but of course, you don't know if you do it well or do it poorly and of course, here they teach you how to do it. Also, how to urinate or defecate, which before you did not do as well as you should, and other things you do not know. So, it's different. (FG3-POP-P3)*

. . . *and B. [pelvic floor program physiotherapist] has been giving me tricks, strategies, helping me to integrate them [PFMEs and lifestyle changes] into my life, to look for the moments* . . . *(FG4-POP-P3)*

*While we are here, we do it, we go, at least I do. And I know how to do it and if I do it well is because B. [pelvic floor program physiotherapist] tell me. And then we continue with the reviews. (Int-POP-P2)*

*I thank B. [pelvic floor program physiotherapist] who helped me to improve, because maybe if I did not come here* . . . *well, uh* . . . *when I also come here for follow-up, I remember to do it, it reinforces me. (Int-POP-P3)*

In addition, the positive feedback of physiotherapists and other health professionals (mainly gynecologists) involved in the therapeutic approach also influenced adherence to PFMEs and lifestyle changes.

*The gynecologist is happy, because he says: we have seen that many people have progressed, and we have seen that I have improved, and POP has been stabilized. (FG3-POP-P3)*

*And it also motivates me that the gynecologist has told me to continue exercising, that I am better and to continue. (FG5-POP-P1)*

The physiotherapists of the program were identified as having a better approach to care than doctors and nurses. The women responded that the aspect of the physiotherapy program that they most valued was their patient, loving, listening, and helpful attitude, reflecting a person-centered attitude (therapeutic alliance).

. . . *.and B. [pelvic floor program physiotherapist] has already taught me with love and patience, and we are learning, because I never knew, and in my country* . . . *Well, I am not from here, but now, thank God, with the exercises that B. has taught me, I can. I am very happy and grateful. (FG2-POP-P4)*

*And here [pelvic floor program], I have felt very good, they [pelvic floor program physiotherapists] have treated me very well, with a lot of patience, they have listened to me, they have helped me, and I have learned a lot and I have improved. (FG6-POP-P2)*

*... for the learning we have had, that you can already carry it out to practice it in your daily life with a program that has been made for each of us, and well, if we do not come to the program, we would not have learned all this [PFMEs and lifestyle changes], me at least. You read four things, you know two or three, you know them in your own way, you don't know if it is well-done as well* . . . *. Well, for me it was the best, finding you [pelvic floor program physiotherapists]. (Int-POP-P3)*

### **4. Discussion**

In this study, we explored women's experiences with compliance with home PFMEs and lifestyles changes for POP symptoms and their facilitators and barriers. Our findings showed that most women, over time, do not perform PFMEs regularly (weekly); however, they comply with lifestyle changes. As shown in prior studies, decline in PFME compliance over time is normal [25,31]. Several researchers have investigated the factors that influence adherence to PFMEs, mainly in women with urinary incontinence, although none specifically addressed adherence to lifestyle changes [15,16,32–35]. Regarding POP, a single qualitative study analyzed adherence to PFME, but it also did not consider adherence to lifestyle changes [17]. So, to the best of our knowledge, this is the first study including factors affecting adherence to lifestyle changes (i.e., strategies for bladder control; contributing factors; knack application to manage everyday pelvic floor challenges; daily PFMEs for at least 15 days after a period of allergy, cold, weight bearing, etc.; performing PFMEs at the end of the sports activities involving weightbearing or impact, in addition to knack; etc.) [24,36–38].

Regarding the methodology, we used a descriptive design in order to understand and determine the experience of women who completed a supervised and individualized pelvic floor physiotherapy program including training for the long-term maintenance of PFMEs and lifestyle changes. This approach allowed us to explore women's experiences with adherence to PFMEs and lifestyle changes for POP from the their perspective and to understand the meanings that the women attach to their behavior [18]. In contrast to the phenomenological study by Hyland et al. [17], a key strength of our descriptive study is the data collection triangulation, combining one-to-one individual interviews and focus groups. The use of investigator triangulation further strengthened this study, resulting in a valid and comprehensive understanding of these phenomena [39].

Our approach produced three themes explaining the factors modifying adherence to PFMEs and lifestyle changes: (1) symptom changes; (2) performing PFMEs and lifestyle changes; and (3) the health practitioner–patient relationship.

### *4.1. Symptom Changes*

Changes in symptoms were identified as an important factor modifying adherence. We found that both the improvement in and symptom relapse in those women who stopped performing the PFMEs or relaxed their changes in lifestyle, mainly knack, were powerful triggers of therapeutic adherence. Previous research in women with incontinence reported that women with more frequent urine leaks before and after PFMEs plus an education physiotherapy program were more adherent to PFMEs one year after physiotherapy than women with less frequent urine leaks [15]. Similarly, Abhyankarome et al. reported that women with POP ask for help based on the symptoms experienced. They explored women's experiences with seeking diagnosis and treatment for POP. They interviewed 22 women receiving POP care through U.K. NHS urogynecology services regarding experiences with seeking diagnosis and treatment for POP and their needs and priorities [40]. Lack of awareness of POP symptoms [40,41] due to mild symptoms resulted in women seeking less help and making less of a sustained effort to perform PFMEs, similar to women with urinary incontinence [15,34,42,43]. However, when women are aware of POP symptoms, when they feel their fear of symptoms worsening, and they feel fear of the condition progressing to the extent that surgery is the only achievable option [40], that is when PFME compliance increases. A wish to avoid surgery as long as possible is a motivator for seeking help and for performing PFMEs [40].

These symptoms can affect women's self-esteem, body image, and quality of life [40] and can have social, psychological, and sexual impacts. The more the symptoms progress, the more they affect quality of life, social well-being, and sexual health [44–46]. Ghetti et al. found, in their qualitative study describing the emotional burden experienced by forty-four women seeking treatment for POP, that women's psychological well-being is intimately related with their POP symptoms [47]. Salovey et al. reported that physical health and emotional states influence one another. Positive emotional states, such as the satisfaction expressed by the women in this study, and their empowerment regarding the control of their symptoms may promote healthy perceptions, beliefs, and physical well-being [48]. Women also linked satisfaction and sense of control to embodied knowledge. Women stated that gaps in their pelvic health knowledge have been addressed. Pintos-Díaz et al., in a qualitative study exploring the reasons Spanish women with urinary incontinence sought help, stated that women' have complaints about the overall lack of information. They also found that the knowledge of many women about pelvic floor and pelvic floor dysfunctions was based on beliefs or myths [43]. Pelvic health education is effective in dealing with myths, false beliefs, and misinformation about the pelvic floor and pelvic floor dysfunction [43]. Moreover, all women reported using their new knowledge, mainly lifestyle changes, and PFMEs. They even shared this information with others. Adherence was linked with knowledge. The more knowledgeable and insightful the women were about their pelvic floor and pelvic health, and the more they experienced an effect on their symptoms, the more they adhered to PFMEs [49]. These findings agree with those of other studies conducted in women with urinary incontinence [15,50,51].

This theme did not emerge the study by Hyland et al. [17]. This could be due to the clinical characteristics of the women in each study. Although women were homogeneous regarding the stage of POP (II and III) and they could also have been homogeneous in relation to the POP symptoms, whether the five women interviewed by Hyland et al. suf-

fered from other pelvic floor dysfunctions is unknown. In our study, women also reported urinary incontinence (76.9%) and anal incontinence (46.4%) symptoms. POP often co-exists with other pelvic floor dysfunctions, mainly those involving urinary incontinence [5,12,52].

### *4.2. Performing PFMEs and Lifestyle Changes*

Compliance with PFMEs and lifestyle changes was associated with regular routines and their integration into their daily lives, PFME movability, and PFME preference regarding control and self-efficacy. Most women performed the PFMEs, which allowed them to have more control and a greater sense of performing PFMEs effectively, which could be related to self-efficacy with a PFME-specific task. A woman's belief in her own ability to perform PFMEs seems to influence adherence to PFME as well as to predict the intention to adhere to PFMEs [15,36,53,54]. These findings agree with those of prior studies in women with urinary incontinence [34] as well as with the findings reported by Hyland et al. in women with POP symptoms [17].

Women also valued the individualized ad hoc guidelines because they explained PFME types, repetitions, progression, and some lifestyle changes. This support material acted as a reminder that facilitated adherence to PFMEs and lifestyle changes. Several studies have evaluated strategies to improve adherence to PFMEs [15,55–57] in women with urinary incontinence. Some researchers tested the efficacy of electronic reminders or exercise diaries, finding that these are effective in increasing adherence to PFMEs in an unsupervised approach [55–57]. As barriers to seeking treatment, Abhyankarome et al. reported that women with POP expressed forgetting to perform the exercises after some time. To recall the exercises, some of them used telephone apps or alarms as reminders and some paired the exercises with daily activities [40].

Women identified the largest barrier to PFME adherence as the feeling of responsibility and the need to take care of others. In the current study, not only did women prioritize others over themselves, but they also expressed their poor commitment to their own healthcare. This may be due to the role of women in society as family planners and caregivers [58,59], probably because of their feeling of moral and affective obligation [60,61]. Despite the social changes over the last decades and the increased involvement of men in caring for the family, women continue to be the main caregivers, at least in Spanish society [62]. These findings agree with those of Alewijse [15] and Hyland [17].

In studies on PFME adherence, adherence is associated with performing PFMEs, seldom with bladder training or with the functional use of pelvic floor muscles in activities of daily living. In our study, women stated that they were more compliant with the functional use of pelvic floor muscles in their day-to-day activities (i.e., knack) than with PFMEs. Moreover, the functional use of the pelvic floor muscles acts as a reminder to perform PFMEs, and therefore supports compliance. This is probably because knack is an effective tool in reducing urine loss, as well as a valid tool for reducing bladder neck movement, which can also minimize POP symptoms [63,64]. To the best of our knowledge, this is the first study exploring adherence to lifestyle changes (including functional use of the pelvic floor muscles) in women with POP symptoms.

### *4.3. Health Practitioner–Patient Relationship*

Women in our study highlighted the importance of their relationship with healthcare professionals, mainly with the gynecologist and with physiotherapists. This could because these women underwent a previous pelvic floor physiotherapy program, and in this eightweek program, the physiotherapists had more opportunities to explain the program and have discussions with the women. The most valued aspect, as in Abhyankarome et al., was the physiotherapists' person-centered attitude. This issue has not emerged in other studies on adherence to PFMEs. This could be because those studies were less supervised or the interventions were unsupervised. The therapeutic relationship or therapeutic alliance seems to improve adherence, satisfaction, and quality of life [65].

This therapeutic alliance may have also been reinforced by the follow-ups and the univocal positive feedback from all of the health professionals involved in the diagnosis and treatment of the women in the study. The follow-up and feedback were described as a demand from women with symptoms of POP, expressing "a strong need for longerterm monitoring and periodic follow-up of prolapse symptoms following the appropriate training in PFME" [sic] [40]. These follow-ups and univocal positive feedback may have simultaneously reinforced self-efficacy [35].

Our findings of the experiences with long-term PFME adherence confirm those reported in the literature on women with urinary incontinence [15,16,25,32–34] and with POP symptoms [17]. Our findings also provide new evidence of the experiences of women with POP symptoms and adherence to lifestyle changes.

### *4.4. Limitations*

The qualitative approach employed in this study also has some limitations. By nature, this approach prevents the generalization of the findings to further ethnically diverse groups. Although the purpose of qualitative research is not to generalize, our results can be extended mainly to Spanish-culture Caucasian women (although two Latin American women also participated) who have previously attended an individualized and supervised woman-centered pelvic floor physiotherapy program in an urban and outpatient academic and research women's health physiotherapy unit. Therefore, our findings cannot be extended to women with POP symptoms who have not visited a women's health clinic or specialist.

Although the interviewer (facilitator, M.T.-L.) was an independent researcher to the pelvic floor physiotherapy program that all women previously completed, the observer (B.N.-B.) in the focus groups interviews was one of the pelvic floor physiotherapy program physiotherapists. Therefore, women may also have said what they thought the physiotherapist would have wanted to hear. Nevertheless, the fact that the women and one of the researchers had a relationship may have contributed to a safe, confidential, and conducive environment for women to talk about their intimate stories.

### **5. Conclusions**

To conclude, long-term compliance with prescribed PFMEs and lifestyle changes can improve with effective, individual, women-centered, and supervised pelvic floor physiotherapy programs, thereby reducing symptoms. The program should include exercises agreeing with women's preferences that are easy to integrate into their activities of daily living, promoting knowledge and awareness of their condition, providing written or electronic guidelines adapted to the specific needs of women, with routine follow-up visits offering both positive feedback and clear and consistent messages from all of the healthcare professionals involved, and enhancing the therapeutic alliance.

An increased understanding of the facilitators of and barriers to performing PFMEs and lifestyle changes may help providers better understand women's needs and provide individualized care.

**Author Contributions:** Conceptualization, M.T.-L. and F.V.-P.; methodology, M.T.-L., F.V.-P. and B.N.-B.; software, V.P.-G.; validation, M.J.Y.-S., V.P-G. and B.S.-S..; formal analysis, M.T.-L., F.V.-P. and B.N.-B.; investigation, M.T.-L., F.V.-P. and B.N.-B.; resources, M.T.-L.; writing—original draft preparation, M.T.-L.; writing—review and editing, M.T.-L., F.V.-P. and B.N.-B.; visualization, B.S.-S.; supervision, M.T.-L.; project administration, M.J.Y.-S.; funding acquisition, M.T.-L. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research is part of a study funded by the Health Institute Carlos III (Protocol PI10/01756) of the Spanish Health Ministry.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Ethics Committee of at the Hospital Príncipe de Asturias (protocol number: OE10/2010) in Alcalá de Henares (Madrid), Spain, and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

**Informed Consent Statement:** Informed consent was obtained from all the subjects involved in the study.

**Data Availability Statement:** Data are held securely by the research team and may be available upon reasonable request and with relevant approvals in place.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **Appendix A**


### **References**


### *Article* **Extracting New Temporal Features to Improve the Interpretability of Undiagnosed Type 2 Diabetes Mellitus Prediction Models**

**Simon Kocbek 1,\*, Primož Kocbek <sup>2</sup> , Lucija Gosak <sup>2</sup> , Nino Fijaˇcko <sup>2</sup> and Gregor Štiglic 1,2,3**


**Abstract:** Type 2 diabetes mellitus (T2DM) often results in high morbidity and mortality. In addition, T2DM presents a substantial financial burden for individuals and their families, health systems, and societies. According to studies and reports, globally, the incidence and prevalence of T2DM are increasing rapidly. Several models have been built to predict T2DM onset in the future or detect undiagnosed T2DM in patients. Additional to the performance of such models, their interpretability is crucial for health experts, especially in personalized clinical prediction models. Data collected over 42 months from health check-up examinations and prescribed drugs data repositories of four primary healthcare providers were used in this study. We propose a framework consisting of LogicRegression based feature extraction and Least Absolute Shrinkage and Selection operator based prediction modeling for undiagnosed T2DM prediction. Performance of the models was measured using Area under the ROC curve (AUC) with corresponding confidence intervals. Results show that using LogicRegression based feature extraction resulted in simpler models, which are easier for healthcare experts to interpret, especially in cases with many binary features. Models developed using the proposed framework resulted in an AUC of 0.818 (95% Confidence Interval (CI): 0.812−0.823) that was comparable to more complex models (i.e., models with a larger number of features), where all features were included in prediction model development with the AUC of 0.816 (95% CI: 0.810−0.822). However, the difference in the number of used features was significant. This study proposes a framework for building interpretable models in healthcare that can contribute to higher trust in prediction models from healthcare experts.

**Keywords:** diabetes mellitus type 2; prediction model; LogicRegression; interpretability

### **1. Introduction**

Morbidity and mortality are often results of Type 2 diabetes mellitus (T2DM). In addition, T2DM presents a substantial financial drain for individuals and families, health systems, and societies. Globally, the incidence and prevalence of T2DM are increasing rapidly [1]. In 2017, it was estimated that 425 million people had any diabetes (approx. 5.5% of the worldwide population), of which 90% had T2DM. According to projection estimations, the prevalence is going to increase substantially in the coming years; by 2045, for example, a 48% increase of prevalence from the above numbers is expected, or in absolute numbers, an estimated 629 million people (approx. 6.6% of the worldwide population) are expected to be suffering from any diabetes [2]. T2DM can also lead to a substantially increased risk of macrovascular and microvascular disease, especially in inadequate glycemic control [3]. Impaired fasting glucose typically leads to slow progression of T2DM and, more importantly, its symptoms may remain undetected for many years.

**Citation:** Kocbek, S.; Kocbek, P.; Gosak, L.; Fijaˇcko, N.; Štiglic, G. Extracting New Temporal Features to Improve the Interpretability of Undiagnosed Type 2 Diabetes Mellitus Prediction Models. *J. Pers. Med.* **2022**, *12*, 368. https://doi.org/ 10.3390/jpm12030368

Academic Editors: Riitta Suhonen, Minna Stolt and David Edvardsson

Received: 6 December 2021 Accepted: 25 February 2022 Published: 28 February 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

Electronic Health Records (EHR) enable researchers to perform predictive modeling by providing a large amount of data [4] and many links have been found between patient health, the environment, and clinical decisions [5]. Nowadays, data mining techniques are applied to various fields of science, including healthcare and medicine [6]. Usually, techniques such as pattern recognition, disease prediction, and classification are used. Although multiple methods are available to build prediction models, prediction accuracy and data validity are often not realistic for model application in practice. Models usually perform well in specific datasets used to build the prediction models but are frequently not adapted sufficiently well when used on other datasets [7].

There is growing interest in clinical prediction, but models' interpretation is rarely based on end-user needs [8], and there is a lack of model interpretability techniques [9]. Interpretability of results based on predictive models is crucial in critical areas such as healthcare and is essential for adopting models. People often do not understand predictive models and therefore do not trust them [10]. LogicRegression can be used to improve the interpretability of predictive models.

LogicRegression is an adaptive classification and regression procedure which searches for Boolean (logic) combinations of binary variables that best explain the variability in the outcome [11,12]. LogicRegression looks for logical combinations of binary features. We can explain the variability of the outcome feature and thus reveal the features and interactions related to the response and whether they have predictive capabilities [11].

The purpose of this paper is to use LogicRegression to make final models less complex (i.e., with less features) and the features that appear in the interpretation of predictive models much more understandable. This is also important for health professionals, as they do not have the necessary knowledge to apply prediction models or interpret the results obtained. This is also important from the patient's point of view and the provision of personalized healthcare. Simple interpretation will make it easier for the patient to understand the operation of the predictive model and outcome. The paper presents an example of using extracted features using Logic Regression to improve the personalized interpretability of the prediction models to the end-users.

### **2. Materials and Methods**

### *2.1. Data*

EHR data consisted of health check-ups and prescribed drugs data from four Slovenian primary healthcare providers for a period of approximately 3.5 years from 12 December 2014 to 27 July 2018. Data for 21,138 medical records and 114 potential useful features were exported from the healthcare information systems after the on-site anonymization process. Our first step was the removal of features with more than 20% of missing data (73 potential features remain). Since our focus when building prediction models was on the fasting plasma glucose level (FPGL) measurement (mmol/L) and results of Finnish Diabetes Risk Score (FINDRISC) features, which included Age, Gender, BMI, Waist circumference, Active\_30\_min, Medication, High\_BS, Grocer, and Diab\_fam we selected cases with all those values present (4086 such cases remained). We next removed (a) cases with more than 50% of the features were not available (4067 cases left), (b) removed all duplicate entries (in cases of multiple patient visits only the most recent visit was included) (3535 cases left), (c) cases not having a previous diabetes diagnosis (3176 cases left) and entries where: (d) FPGL was not reported giving us a total of 3120 records of patient visits were left for development of a prediction model to estimate the risk of undiagnosed T2DM. Data included demographics, questionnaire answers for lifestyle choices, physiological measurements, and prescribed medications for two time periods.

Binary features were created for prescribed drugs and questionnaire responses, which resulted in nine numeric and 161 binary features where specific drug related feature was coded as positive in cases where a patient was prescribed with the specific drug during the last 4 months prior to the visit. The target feature was binary, where positive cases were defined as having FPGL higher than 6.1 mmol/L consisting of 24.71% (*n* = 771) of patient visits.

We imputed the remaining missing values using the MissForest based approach [7], which on average meant features with 12.25% missing values as we initially already removed features with 20% or more missing values. MissForest is used to impute missing values particularly in the case of mixed-type data. It can be used to impute continuous and/or categorical data including complex interactions and nonlinear relations. The summary information of the basic predictive and target features can be seen in Table 1. Please see Table A1 for list of all features used in the experiments.


**Table 1.** Summary table basic predictive and target features for healthcare centers.

### *2.2. Experimental Setup*

The data were split into 80% to derive five extracted features using Logic Regression [13] and 20% to build and evaluate the final prediction models.

Finally, we created three datasets with the following features: all numeric and binary (170), all numeric and logic (14), and all numeric, binary, and logic features (175). On each dataset, we built a predictive model separately using the same training data.

The Least Absolute Shrinkage and Selection Operator (LASSO) [13] was used to build prediction models. We repeated each 10-fold cross-validation ten times to estimate the variance in Area under the ROC curve (AUC) that was used as our classification performance metric.

### **3. Results**

We split the results in this section into two parts. First, we present selected logic attributes extracted from the dataset for the undiagnosed T2DM prediction use case. Next, we present the performance evaluation of the model.

### *3.1. Feature Extraction Using LogicRegression Approach*

To demonstrate the practical example of using LogicRegression based extraction of new features to improve interpretability of the prediction models, we provide the results of the first cross-validation run.

The selected use case resulted in five logic features (Table 2) extracted from the complete set of features.

In Table 3, we list all features that were selected in at least 50% of runs in our experiments with LASSO on the dataset with numeric and logic features, while Table 4 lists features for the dataset with numeric, binary and logic features. Frequency (freq) shows in how many experiment runs each feature appeared in the final set of features.


**Table 2.** Extracted logic features with corresponding LogicRegression rules and descriptions.

**Table 3.** Selected features with the Least Absolute Shrinkage and Selection Operator (LASSO) on the dataset with numeric and logic features.


**Table 4.** Selected features with LASSO on the dataset with binary, numeric, and logic features.


It can be observed that L1, L2, and L3 were used by prediction models derived from the data in all folds of all evaluation runs. Thus, confirming a high contribution of extracted logic features.

In the case of results from a much wider set of features (Table 4), we can see a higher variance in selection by the final prediction models. Four (L2, L3, L4, L5) logic features can be found among the varaibles that were selected in at least 50% of evaluation runs. (SD)) = 0.03) in no\_logic to 0.819 (SD = 0.03) in all\_logic and finally 0.829 (SD = 0.03) in the num\_logic dataset. When looking at the number of selected feature averages and its variation, we can observe that it slowly increases from 21.7 (SD = 11.18) in no \_logic to 23.7 (SD = 10.09) in all\_logic but it then almost halves to 13.35 (SD = 0.63) in the num\_logic

+Heart\_beat 100 Heart\_beat +Age 100 Age in years

−Gender 100 Gender

+Body\_weight 87 Body weight −Non\_smoker 87 Non-smoker +L2 79 Logic attribute 2 <sup>−</sup>Q321 78 Most often used oil is vegetable

−Non\_drinker 77 No alcohol consumption −Q583 75 Handle stress with hardship +Q62 74 Parent, brother, or sister have

+BMI 69 Body Mass Index −Q161 63 2 meals per day on average −Q301 51 No habit of using salt at the ta-

In Figure 1, we summarize AUC and a selected number of features for all three datasets: no\_logic (numeric and binary features), all\_logic (numeric, binary, and logic fea-

We can observe a slowly increasing average AUC from 0.816 (Standard Deviation

+Q45 100 Ever measured high blood

+Q32 93 Using drug(s) for lowering

sugar? Yes

blood pressure

oil

diabetes

ble

#### *3.2. Performance Evaluation* dataset. The SD is steadily increasing in the first two cases, but then it decreases sharply

*3.2. Performance Evaluation* 

tures), and num\_logic (numeric and logic features).

In Figure 1, we summarize AUC and a selected number of features for all three datasets: no\_logic (numeric and binary features), all\_logic (numeric, binary, and logic features), and num\_logic (numeric and logic features). to below 1 (SD = 0.63), which means that out of the 100 repetitions in 92 cases 13 or 14 features were selected in the num\_logic dataset. This indicates a very stable final prediction models when comparing num\_logic based solutions to no\_logic or all\_logic.

*J. Pers. Med.* **2021**, *11*, x FOR PEER REVIEW 5 of 10

**Figure 1.** Selected features with the Least Absolute Shrinkage and Selection Operator (LASSO) on the dataset with numeric and logic features. **Figure 1.** Selected features with the Least Absolute Shrinkage and Selection Operator (LASSO) on the dataset with numeric and logic features.

We can observe a slowly increasing average AUC from 0.816 (Standard Deviation (SD)) = 0.03) in no\_logic to 0.819 (SD = 0.03) in all\_logic and finally 0.829 (SD = 0.03) in the num\_logic dataset. When looking at the number of selected feature averages and its variation, we can observe that it slowly increases from 21.7 (SD = 11.18) in no \_logic to 23.7 (SD = 10.09) in all\_logic but it then almost halves to 13.35 (SD = 0.63) in the num\_logic dataset. The SD is steadily increasing in the first two cases, but then it decreases sharply to below 1 (SD = 0.63), which means that out of the 100 repetitions in 92 cases 13 or 14 features were selected in the num\_logic dataset. This indicates a very stable final prediction models when comparing num\_logic based solutions to no\_logic or all\_logic.

### **4. Discussion and Conclusions**

In this paper, we compared three dimensionality reduction approaches to improve the interpretability of undiagnosed T2DM prediction models (Please note that the calibration of a prediction models was not the scope of this paper and presents a limitation). A simple LASSO regression approach is compared to two variants where a pre-selection of predictive features is conducted on the training set using LogicRegression to consequently simplify a final set of features obtained by the LASSO regression. We kept all original features with added logic features in the first variant, while in the second variant, we kept only numeric and logic features.

Results showed that logic features resulted in simpler models with lower number of features, which are potentially easier to interpret by healthcare experts. This is especially important in the field of personalized medicine. Measured AUC was similar to more complex models, where all features were included. It should be noted that although our method resulted in a lower number of features, some of the logic features may not be straightforward to interpret (e.g., the feature L3 in this paper). To address this issue, we plan to include an interactive system in our future work, where the user would specify the maximum number of original features included in generated logic features in cases where the final model would include many complex logic features. As a result of the current work, in cases when some of the final features are hard to interpret, we recommend that the user uses LogicRegression settings to adjust the complexity of final logic features for achieving satisfactory results.

When healthcare professionals and patients know which features are important in obtaining the outcome of a prediction model and how they can be combined, it helps to understand and increase the level of trust in the decision-making systems [10]. With greater interpretability of the model, we better understand and interpret the forecast for end-users and improve the support in decision-making for health professionals based on data [14]. More complex models such as deep neural networks [15] allow high accuracy but are difficult to explain. Simple models (e.g., decision trees) are less accurate but allow for more straightforward explanations [16]. Therefore, sophisticated machine learning models usually offer better performance than traditional simple models but are difficult for health professionals to understand. However, in many cases simple models also provide good classification performance, which is not significantly different from more complex models [17]. Our results confirm this hypothesis. Comprehensible models are known for their contribution to higher trust in prediction models from the end-users in healthcare.

Interpretability techniques are often categorized according to the time period used to develop the machine learning model [14]. Pre-model approaches are independent of the model and may be employed prior to making a choice on which model to use. Our approach presented in this study belongs in this group of interpretability approaches along with techniques such as Principal Component Analysis (PCA), t-Distributed Stochastic Neighbor Embedding (t-SNE), and some clustering techniques. While Molnar [18] classifies PCA, t-SNE, and clustering methods as interpretable methods, it is worth noting that the interpretability of attributes transformed using PCA, embeddings, or clusters cannot provide comprehensible medical interpretation, but can be used to visualize the results and highlight patterns of interest from an interpretability standpoint. The proposed approach is much more interpretable, despite the possible complex combinations of features that might occur as a result of LogicRegression.

During the experiments, we also observed the unstable behavior of logic regression, where different logic features were selected with each run of the cross-validation. Although this did not influence the average number of selected features it resulted in instability of the interpretability of the model. Another limitation are the combinations of the features used in extracted features. For example, the first extracted feature (L1) suggested that checking whether a person did not experience elevated blood sugar in the past should be accompanied by checking for sulfadiazine and trimethoprim use in the last 4 months – this extracted feature works as a protective factor as seen from Table 3. We see this as a disadvantage of logic regression since different conclusions can be made based on selected features. This could be resolved to some extent by using exhaustive search methods to extract logic features resulting in extremely long running times, presenting another drawback, especially in cases where personalized models would be built. To personalize the solution even further, it would be worth exploring the prediction model development for each specific patient at the time of the examination using the subset of the data where patients similar to the examined patient would be assigned a higher weight in comparison to other patients (boosting principle).

Although our work is the field of healthcare, we believe that our results can also be applied in other emerging fields of applied prediction modeling where interpretability of results is important such as security [19] or ecology [20]. In future work, we will explore effectiveness of our methods in the broader field of security, specifically, to help us understand how misinformation (e.g., intentionally misleading information) is being spread.

**Author Contributions:** Conceptualization, S.K., P.K. and G.Š.; methodology, S.K., P.K. and G.Š.; software, S.K., P.K. and G.Š.; validation, S.K., P.K. and G.Š.; formal analysis, S.K. and P.K.; investigation, S.K., P.K., L.G. and G.Š.; resources, P.K. and G.Š.; data curation, S.K. and P.K.; writing—original draft preparation, S.K., P.K., L.G. and G.Š.; writing—review and editing, S.K., P.K., L.G., N.F. and G.Š.; visualization, P.K.; supervision, G.Š.; project administration, G.Š.; funding acquisition, G.Š. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by Slovenian Research Agency (grants number ARRS N2-0101 and ARRS P2-0057) and the European Union's Horizon 2020 Research and Innovation Program under the Cybersecurity CONCORDIA project (GA No. 830927).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Restrictions apply to the availability of these data. Data was obtained from Nova vizija d.d. and are available from the authors with the permission of Nova vizija d.d.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **Appendix A**

**Table A1.** List of original features with description and their possible values. Please note that Nominal features were processed in such a way that for each possible value a new feature was generated. For example, the feature Q43 resulted in three features for each possible value (new features were named Q431, Q432 and Q433). Drug features are marked with the Anatomical Therapeutic Chemical (ATC) classification. The final set contained 170 features.






### **References**


### *Article* **What Influences Women to Adhere to Pelvic Floor Exercises after Physiotherapy Treatment? A Qualitative Study for Individualized Pelvic Health Care**

**Beatriz Navarro-Brazález , Fernando Vergara-Pérez , Virginia Prieto-Gómez \* , Beatriz Sánchez-Sánchez , María José Yuste-Sánchez and María Torres-Lacomba**

> Physiotherapy in Women's Health (FPSM) Research Group, Physiotherapy Department, Faculty of Medicine and Health Sciences, University of Alcalá, 28871 Alcalá de Henares, Spain; b.navarro@uah.es (B.N.-B.); fernando.vergara@uah.es (F.V.-P.); beatriz.sanchez@uah.es (B.S.-S.); marijo.yuste@uah.es (M.J.Y.-S.); maria.torres@uah.es (M.T.-L.)

**\*** Correspondence: v.prieto@uah.es; Tel.: +34-91-885-4828

**Abstract:** Conservative treatment of pelvic floor dysfunction (PFD) includes therapeutic exercise for pelvic floor muscle (PFM) training or other complementary exercise modalities, such as hypopressive exercises. However, the long-term effectiveness of the conservative treatment depends on a patient's adherence to the exercises and the integration of professional health advice into their daily life. The objective of this study was to establish the adherence experience of women with diagnosed PFD in home-based exercises after an intensive face-to-face physiotherapy treatment. A qualitative study from an interpretive paradigm was developed. Semi-structured individual and group interviews were performed 6 months after finishing individual physiotherapy treatment. The interviews were recorded, fully transcribed and analyzed thematically by creating categories. Thirty-one women were interviewed. The women reported that their adherence to home PFM exercises depended on the exercise program itself, its efficacy, their personal experiences with the exercises, intrinsic factors such as self-awareness or beliefs, and extrinsic factors, such as professional or instrumental feedback. Thus, therapeutic adherence could be more likely with effective physiotherapy programs that include mutually agreed home exercises and simple movements women can build into their daily lives. Improving awareness and knowledge of the pelvic region and the importance of PFM treatment as well as consideration for potential worsening of PFD will also encourage women to adhere to the exercises.

**Keywords:** pelvic floor muscle exercises; pelvic floor dysfunction; qualitative research; therapeutic exercise; therapeutic adherence; women's health physiotherapy

### **1. Introduction**

The pelvic floor muscles (PFM) play an important role in the preservation of urinary and anal continence, in pelvic organ support and in sexual function, among others [1]. The weakness and loss of PFM properties are associated with the development and maintenance of pelvic floor dysfunctions (PFD) [2,3]. Prevalence studies estimate that PFD affect up to 40% of women [4], including symptoms of urinary incontinence (UI), pelvic organ prolapse (POP), and anal incontinence (AI), which are the most frequent PFD. The first line of treatment for mild PFD is focused on specific PFM exercises [5,6], which can involve a therapeutic education program [7] to provide knowledge and self-management strategies to women. In addition, global exercise modalities, such as hypopressive exercises, have recently begun to be included in PFD treatment [8–10], to train the PFM in coordination with posture and adjacent muscles. However, the long-term effectiveness of conservative treatment does not seem to depend exclusively on the exercise method. Patient adherence to a prescribed exercise program and to health professional advice seems to be one of the determining variables to ensure the success of the treatment in the short and long term [11].

**Citation:** Navarro-Brazález, B.; Vergara-Pérez, F.; Prieto-Gómez, V.; Sánchez-Sánchez, B.; Yuste-Sánchez, M.J.; Torres-Lacomba, M. What Influences Women to Adhere to Pelvic Floor Exercises after Physiotherapy Treatment? A Qualitative Study for Individualized Pelvic Health Care. *J. Pers. Med.* **2021**, *11*, 1368. https:// doi.org/10.3390/jpm11121368

Academic Editors: Riitta Suhonen, Minna Stolt and David Edvardsson

Received: 18 November 2021 Accepted: 10 December 2021 Published: 14 December 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

The determining factors of therapeutic adherence to PFM exercises have been analyzed and classified based on patient personal parameters or factors dependent on physiotherapeutic performance. Motivation, perceived self-efficacy and benefit expectations are identified as personal facilitating factors, while misconceptions, false beliefs or low body awareness are presented as personal adherence barriers. Regarding physiotherapy treatment, a structured program and close supervision are positive adherence factors, and a high number of sessions and abstract concepts are associated with poor adherence [12–14].

Hay-Smith et al. [15] inquired about the experiences of women in performing PFM exercises at home and observed that a high degree of self-efficacy was needed, which could be improved with individualized educational programs. Recent qualitative studies in women with PFD point to the patient's demand for [16] and the perceived usefulness [17] of a structured educational program on the functions and dysfunctions of the pelvic floor, on risk factors and on coping strategies. However, further studies to understand and delve into the patient's experiences are required to develop more long-term effective and patient-centered programs [18,19].

In a previously published randomized clinical trial [8], women who experienced stress or mixed urinary incontinence, anal incontinence, or mild pelvic organ prolapse received physiotherapy treatment. Participants were randomly allocated to either PFM training and an educational strategy (PFMT group), or to hypopressive exercises and an educational strategy (HE group), or to PFM training plus hypopressive exercises and an educational strategy (PFMT+HE group). The three programs were conducted individually and faceto-face by the same physiotherapist specialized in women's health. The physiotherapy treatment lasted 8 weeks, with two visits of 45 min each week. Moreover, women were advised to carry out 15 min of daily home exercises based on the specifications of the physiotherapist depending on the intervention group. Despite the difference between the study exercises, the three groups showed similar adherence data, not being able to understand or deepen the reasons for continuing or abandoning the home guidelines.

Therefore, the aim of the present study was to establish the experience of women to adhere to home exercises recommended by their physiotherapist after completing an intensive face-to-face treatment, which included PFM exercises and/or hypopressive exercises and an individually tailored educational program.

### **2. Materials and Methods**

### *2.1. Research Team and Reflexibility*

From October 2014 to June 2016, semi-structured individual interviews and focus groups were conducted. The personal interviews were carried out by the physiotherapist who performed the 2 months of individual treatments. This assumed that the participants knew the physiotherapist, and a prior relationship of trust could have been established. The physiotherapist (B.N.-B.) was a woman with more than 5 years of professional experience in women's health and with previous experience in qualitative interviews. The focus groups were led by another woman physiotherapist (B.S.-S.), a university professor, who also specialized in women's health and was experienced in qualitative research interviews. She was not connected with the study in order to achieve the appearance of a new perspective [20], and the participants did not know her. All the interviews began with the presentation of the objective of the study, the guarantee of the confidentiality of the contributions and the request for permission to record the audio of the interviews.

### *2.2. Study Design*

A qualitative descriptive study was developed from an interpretative paradigm design to present the women's perspectives on the phenomenon of adherence [21] to the pelvic floor exercise program after a randomized clinical trial [8]. The Clinical Research Committee of the Principe de Asturias Hospital (OE20/2013) approved the study, which was registered at ClinicalTrials.gov (NCT02259712). The consolidated criteria for reporting qualitative research (COREQ) were followed for reporting [22].

### *2.3. Participant Selection and Setting*

Purposive sampling was used to select women from the three intervention groups. Women who were included in the randomized clinical trial and were in the follow-up period after the physiotherapy treatment were invited to participate in the qualitative study. The inclusion criteria were to have completed the 2 months of physiotherapy treatment in any of the three study groups, for a period of between 3 and 6 months to have elapsed after its completion, to have availability to attend the interview, speak Spanish fluently and to freely agree to participate. Since no statistically significant differences were found in the adherence of the three study groups, the qualitative study was proposed and aimed at all participants, regardless of the assigned treatment group. The capture of the individual interviews was carried out in person during the physiotherapy reviews, inviting participants to carry out the interview at a time that best suited them. Group interviews were scheduled, and the participants were contacted by phone to give them a face-to-face appointment.

A total of 20 individual semi-structured face-to-face interviews were completed: nine interviews of women allocated to the PFMT group, six interviews of women in the HE group, and five interviews of participants who were in the PFMT+HE group. Furthermore, three group interviews were conducted, one for each intervention group with the participation of three to five women in each group. Three participants rejected the interview due to time incompatibility. The appointments were developed in a private room at the FPSM-RG laboratory at the University of Alcalá (Madrid, Spain). During the interviews, only the participants and the interviewer were present.

Since the age of the participants [23], the PFD symptoms, or the improvement found with the physiotherapy treatment could influence the results of the interviews [14], a descriptive analysis of the sample was carried out. This description only provides a descriptive character of the participant and not a criterion for interpretive analysis. To verify normal distribution of the data, the Kolmogorov–Smirnov statistical test was used. The normal quantitative variables were described with means and standard deviations. The categorical variables were described with absolute frequencies and percentages. For this statistical analysis, the IMB SPSS Statistics version 20 software was used.

The mean age of the participants was 49.74 (10.78) years, with a BMI of 26.01 (4.88) kg/m<sup>2</sup> , 16 (51.6%) of them had experienced menopause, and in relation to the history of pregnancy and childbirth had an average of 2 (1) vaginal deliveries. Annual household income indicated that the study population belonged to the middle class; 35.5% of the women had a university education, 51.6% had a basic education, and one participant indicated that she had never attended school. All the participants suffered from some PFD, since it was an inclusion criterion, presenting UI in 87.1%, IA in 48.4%, and POP in 51.6% of the women. Comparison of PFM strength and quality of life survey results before and after the physiotherapy treatment indicated improvements in all participants. Table 1 shows the characteristics of the sample in more detail.


**Table 1.** Demographic and clinical data of the participating women.


**Table 1.** *Cont.*

*n*: Number; SD: Standard deviation; BMI: Body mass index; UI: Urinary incontinence; AI: Anal incontinence; POP: Pelvic organ prolapse; PFDI-20: Pelvic Floor Distress Inventory Short Form; PFIQ-7: Pelvic Floor Impact Questionnaire Short Form.

### *2.4. Data Collection*

Prior to the start of the interviews, a group of three physiotherapists, specializing in women's health (M.T-L. and M.Y-S.) and qualitative research (F.V.-P.), and a gynecologist developed referral questions (Table 2). After the guide questions were prepared, they were presented to the interviewers to consider if they needed any changes. A pilot test was carried out with a participant. This interview was excluded because it was not recorded and was only conducted to ensure the understandability and relevance of the questions.

**Table 2.** Guide questions for the individual and focal group interviews.


All the interviews included in this study were audio recorded with participant permission. No field notes were taken, the interviews were conducted only once, and the interview transcriptions were not returned to participants.

The mean duration was 24 min for individual interviews and 65 min for focal group interviews. Data saturation was considered when the information collected was consistent with the previous interviews. This fact was shared and discussed with the physiotherapists who participated in transcribing the interviews.

### *2.5. Data Analysis*

Five physiotherapist members of the FPSM-RG (B.N.-B., F.V.-P, V.P.-G., M.T.-L. and B.S.-S.) manually transcribed the recorded interviews. Different codes were established as part of the intervention group allocation, to sort the interview (individual or in group) and the participant intervention order, with the purpose of securing the women's anonymity. Researchers' triangulation process was performed, where three physiotherapist members analyzed the resultant texts (B.N.-B., F.V.P and M.T.-L.). In an iterative consensus process, the researchers conducted an open and axial coding since a previous theoretical framework was not used, as well as related codes and categories. The transcription encoding was performed using the ATLAS.ti version 6.1 software (Scientific Software Development GMBH, Berlin, Germany). The interviews and the categorization of the themes were carried out in Spanish. Quotes and categories were later translated into English.

### **3. Results**

Five themes emerged from the literal transcription analyses: the exercise program, the program efficacy, personal experiences with exercises, intrinsic factors, and extrinsic factors (Table 3).


**Table 3.** Main themes and codes extracted from participants' interviews.


**Table 3.** *Cont.*

### *3.1. The Exercise Program*

Inclusion in the program itself was identified as a motivating factor. Many women expressed the need for several appointments with health professionals about their PFD to be informed about a physiotherapy treatment. Moreover, the program was free, costly at a private level and almost non-existent in public services.

HEG4: "The only thing they told me about was a possible operation".

PFMTG3: "Nor the gynecologist, nor the midwifery; no one tell you anything, there is no information".

On a social level, PFD did not seem to be recognized either, as the following participant pointed out:

PFMTI3: "It is a pity that I am not so heard, as for example if you have a bone problem you go to trauma, but I do not know about you if it is not for the last doctor."

The moment when women should access the program was also mentioned because many considered that conservative treatments should be implied before the symptoms showed up or when they were minimally aware.

PFMT+HEG4: "As a birth preparation is stablished, it should be a repaired postlabor treatment".

HEG1: "and in menopause..."

Furthermore, participants were generally satisfied with the exercise program because of the pleasant relationship with the physiotherapist, the knowledge gained, and the physical and mental well-being acquired.

PFMTI1: "It was worth it, I am delighted, and I would repeat it, for me it has been a very good and very positive experience."

### *3.2. Program Efficacy*

Close to the program satisfaction were the positive results achieved with the physiotherapy treatment. The improvement in perceived symptoms was identified as a key factor in treatment adherence:

HEG1: "Since I perform the exercises, I feel better, more secure. Before that, I avoid myself to do things, because it was uncomfortable . . . But not anymore! I feel safer".

Many women showed signs of general well-being, indicating that they felt happy, good about themselves, with better self-esteem and with the ability to return to a normal life, in which they could control their symptoms. That feeling of well-being helped them continue with the treatment:

PFMTI1: "I keep my self-esteem high, my relationships normal and my life normal, everything positive that entails, I have achieved everything I wanted, I am interested in keeping it."

These benefits provided by the program were also specified in the physical symptoms of the pelvic floor. In the first place, the participants' previous experience with their physical symptoms played a significant role, since each of these symptoms diminished an aspect of their daily life:

PFMT+HEI5: "Every two steps I was afraid of cough, okay, not much, droplets, but today one, later another, another, another..."

Additionally, the fact that, after the exercise program, they had managed to minimize these symptoms, was also an adherent factor:

PFMT+HEI3: "Before I was very overwhelmed, it was a small cough and now, I cough calmly so to speak."

However, in some participants in which the symptoms were not totally cured, the limited efficacy of the treatment was a barrier to continuation at home.

PFMTI5: "I don't have a 100% of improvement. [ . . . ] Doing big sacrifices, but at the end, I know that always some drops will escape, so . . . ".

### *3.3. Personal Experiences with Exercises*

Participants revealed that they required a quiet environment and concentration to perform the exercises correctly, which was impracticable due to their family and work duties.

PFMTI4: "I would need more order in my life [ . . . ] I have three kids, the work, the housework, every day, I'm always nervous".

Moreover, women expressed other health issues that made it difficult to continue the exercises, such as colds, allergies, low-back pain, or vaginal infections. Integrating the PFM exercises with other daily activities was essential, and this was closely connected with the perceived difficulty of the exercises. Exercises were said to be performed while walking, driving, doing the housework, or going to the gym. Furthermore, the personal preferences

for the exercises (the type of exercise, posture, the demanded conditions of the exercise) and the perceived efficacy of one over another, influenced the adherence.

PFMTI4: "I only practice the maintenance exercise, because I think that to gain force it's ok".

### *3.4. Intrinsic Factors*

The self-perceived symptoms, the self-perceived treatment efficacy, and the ability to reproduce exercises efficiently as well as to correct themselves were important in adherence. Moreover, the participants' beliefs, their immediate environment and the normalization of symptoms of urinary leakage in society affected adherence.

PFMTG3: "Some of my friends do hypopressive exercises, and I'd like to be in the hypopressive group, because I think that they're really good".

The newly acquired awareness of responsibility changed the women from passive agents to active agents in the improvement of their condition.

HEI3: "Now the responsibility is mine, if I don't do anything . . . there is no magic".

### *3.5. Extrinsic Factors*

Regular contact with the PT, the use of biofeedback devices and mirrors, and the posterior assessments were named as motivating factors.

PFMTI8: "The gynecology gave me a paper with some exercises. But I didn't understand them, nor fancy doing them. When I came here, I saw the importance they have".

Nevertheless, finishing the continuous physiotherapy treatment required some women to withdraw from the weekly PFM training. The evaluation and feedback provided by other health professionals also was a condition of adherence to the exercises.

### **4. Discussion**

This study shows the adherence of women with PFD to the exercises recommended by their physiotherapist 3 to 6 months after an intensive treatment. To our knowledge, this is the first qualitative research that interviewed women with different mild forms of PFD, who underwent an individualized treatment based on PFM guided exercises with biofeedback and/or hypopressive exercises. Five categories emerged from the thematic analysis that explained the adherence-modifying factors: the exercise program, the program efficacy, the personal experience with exercises, patient's intrinsic aspects and extrinsic factors. Each theme contained related codes, from which positive and negative aspects for adherence were extracted.

Other qualitative studies emerged from randomized clinical trials. Our study is consistent with the study by Hyland et al. [24]. They presented the interviews of five women with POP between 5 and 24 months after participating in prescribed PFM exercises and selected positive adherence codes related to the ease of performing the exercises anywhere and their association with everyday moments. Family duties were emphasized as a determining aspect of adherence and maintenance of PFM exercise over time, a complicated challenge influenced by the low self-efficacy perception. The perception of self-efficacy has been identified as one of the main predictors of adherence to the maintenance of home PFM exercises [25]. To reinforce self-efficacy, the use of motivational interviewing has been proposed [15,26] and agreement reached on individualized goals based on the motivational stage of the patient [27], aspects that were included in the three intervention groups.

Adherence rates in our three study groups were similar. In the PFMT group, 71.9% continued exercising weekly at home, 61.3% in the HE group, and 67.7% in the PFMT+HE group [8]. Hay-Smith et al. [16] focused on PFM exercise comprehension and found that participants had trouble distinguishing the different exercise interventions, which could alter the outcome of their study. In our study, we did not identify confusion by the participants in identifying the different exercises, probably because the PFM exercises and the hypopressive exercises are different in their execution [8]. Our qualitative research showed that women found it difficult to continue a daily home exercise routine, especially if they

perceived the exercise to be difficult, exhausting, or incompatible with a simultaneous activity; thus, we were surprised by the high adherence data found in relation to hypopressive exercises. In Spain, hypopressive exercises were popular among women; in fact, a participant from the PFM group revealed in her interview that she would have liked to join the HE group. This social support for an exercise modality could be a significant factor of adherence [28]. Furthermore, the treatment effectiveness perceived by the participants appeared as a positive factor in maintaining its practice at home [29], and in the three intervention groups, the women achieved clinically relevant improvements in their quality of life, which could also explain the similar adherence results.

Women valued the need for a treatment that produced improvements in relation to the symptoms of the pelvic floor, provided them with a feeling of well-being and generated secondary improvements, such as the minimization of low back pain. These findings are in line with the promotion adherence results found by Alewejnse et al. [7], who showed that a well-designed physiotherapy protocol may downplay the addition of a health education program. However, the lack of information and false beliefs have been identified in women with UI [16], assuming educational programs are an essential tool for the health education of these women. In fact, the participants in our study highlighted the knowledge acquired as a positive factor of adherence, finding in other studies how women perceive that it was a useful learning experience for life. The feedback provided by the physiotherapist and the periodic evaluation visits also appeared as facilitating aspects of adherence [30].

Several barriers to adherence were identified based on the experiences of the women in our study. They included difficulty in accessing a pelvic floor physiotherapy service in relation to ignorance of its existence and its high price, as it is not generally integrated into public health services [29–31]; whether treatment is started with advanced symptoms; and failure to achieve complete resolution of the symptoms. Regarding the time when the physiotherapy program should start, the women agreed that around postpartum/first delivery could be a good time. Despite the opinions of women, qualitative studies show that postpartum women are unaware of PFD that may appear after childbirth, and that they believe it is an inevitable consequence and do not know about the treatment alternatives available [31,32]. Grant et al. [33] interviewed 31 women who had given birth in the last 5 years and found that women needed more information and support for proper performance of the MSP exercises, which could also be facilitated by the use of new technologies, such as the creation of an app. Other identified barriers to adherence were related to beliefs about normalization or taboo subjects [31,33], a lack of self-care [34], prioritizing family needs [24], or job obligations [30].

Pakbaz et al. [34] identified feeling ignored by healthcare professionals as an obstacle of attention demand; an assumption in our study was the feeling of gratitude for care as a positive factor for adherence. This appreciation could enable the patients to have good experiences since the physiotherapist who guided the treatment and prescribed the exercises was also the interviewer, which was a possible limitation of the study. To minimize this fact, group interviews were performed by an external physiotherapist who also specialized in women's health. However, to enlist the same intervening physiotherapist for interviews was an advantage because she provided confidence to the interviewed participants, and her knowledge about women was used to interpret the experiences and categorize the information [24]. A strength of our study is that the questions were agreed upon by different health professionals and the process of triangulation and analysis of the information was carried out by five physiotherapist experts in women's health and qualitative research.

The practical implications derived from the present study can be directed first to the planning of effective physiotherapy programs that promote the empowerment of women, and second to the evaluation and management of contextual factors that may be positive or negative. Knowing and exploring both the beneficial factors and the adherence barriers present in each woman would allow the physiotherapist and patient to agree on realistic home exercises [30] and in meaningful environments. For future research, it is considered

essential to explore and deepen the experience of women regarding their needs related to improvement in their PFD, their opinions on current management and the possibilities for them to take an active role in their own recovery. Designing mixed-method studies or qualitative studies after intervention research could delve into the personal reasons for adhering or not to a treatment plan, especially important in chronic or long-terms disorders such as PFD.

### **5. Conclusions**

In the present qualitative study on the experience of maintaining home exercises after intensive physiotherapy treatment in women with PFD, five central themes were identified: the exercise program, program efficacy, personal experience with the exercises, intrinsic factors, and extrinsic factors. Interventions with perceived effectiveness and easy exercises suitable for integration into daily life would enhance therapeutic adherence. Providing knowledge so that women recognize their pelvic floor and the importance of PFM exercises as well as take an active role in their self-care process should be a goal of health professionals, also reinforced by professionals.

**Author Contributions:** Conceptualization, B.N.-B. and M.T.-L.; methodology, F.V.-P. and M.J.Y.-S.; software, B.N.-B. and F.V.-P.; formal analysis, B.N.-B., V.P.-G., F.V.-P., B.S.-S. and M.T.-L.; investigation, B.N.-B. and B.S.-S.; resources, M.T.-L.; data curation, M.J.Y.-S.; writing—original draft preparation, B.N.-B. and F.V.-P.; writing—review and editing, M.T.-L.; visualization, V.P.-G.; supervision, B.S.-S. and M.T.-L. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Clinical Research Committee of the Principe de Asturias Hospital (OE20/2013).

**Informed Consent Statement:** Written informed consent has been obtained from the patients to publish this paper.

**Data Availability Statement:** More information on the data presented in this study can be requested from the corresponding author.

**Acknowledgments:** The authors would like to thank all the women who participated in the study.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


### *Article* **Functional Profile of Older Adults Hospitalized in Convalescence Units of the National Network of Integrated Continuous Care of Portugal: A Longitudinal Study**

**Ana Ramos <sup>1</sup> , César Fonseca 2,3 , Lara Pinho 2,3,\* , Manuel Lopes 2,3 , Henrique Oliveira <sup>4</sup> and Adriana Henriques 1,5**


4

**Abstract:** Aim: To evaluate the evolution of the functional profile of older adults admitted to a health unit in Portugal; to relate the functional profile of these individuals with age, sex, education level and emotional state; and to evaluate the probability of the degree of dependence as a function of age and sex. Methods: longitudinal, retrospective study with a sample of 59,013 older adults admitted to convalescence units of the National Network of Integrated Continuous Care of Portugal. Results: In the first 75 days of hospitalization, activities of daily living, mobility and cognitive state improved, but there was a decline after 75 days of hospitalization. The ability to perform instrumental activities of daily living improved in the first 15 days of hospitalization, stabilized until 45 days and then began to worsen. Women had a higher probability of having a severe/complete dependence three years earlier than men (88 years to 91 years). A higher education level and stable emotional state were protective factors against functional decline. Conclusions: The functional profile of older adults improved during the length of stay recommended for hospitalization in convalescence units (30 days). It is critical for health systems to adopt strategies to prevent declines in the emotional state of frail individuals.

**Keywords:** older adults; functional status; health care; hospitalization; activities of daily living

### **1. Introduction**

With the increasing aging of the population and the increase in life expectancy, it is necessary to pay greater attention to the health of older adults. The European Pathway Association states that for care pathways to be successful, they must obey a set of principles, among which the following stand out: (a) the definition of clear care goals, based on scientific evidence, the best clinical practices and the expectations and characteristics of the person being cared for; (b) the facilitation of communication among all those involved; (c) effective coordination; (d) the correct monitoring and evaluation of results; and (e) the identification of the appropriate resources for the individual and the clinical situation [1]. In this sense, a care pathway should be based on both the integration of care and the continuity of care [2].

The integration of care is considered by the World Health Organization (WHO) to be the result of multifaceted efforts made to promote integration, with benefits for people [3]. It has been considered an international priority in health policy and health management

**Citation:** Ramos, A.; Fonseca, C.; Pinho, L.; Lopes, M.; Oliveira, H.; Henriques, A. Functional Profile of Older Adults Hospitalized in Convalescence Units of the National Network of Integrated Continuous Care of Portugal: A Longitudinal Study. *J. Pers. Med.* **2021**, *11*, 1350. https://doi.org/10.3390/ jpm11121350

Academic Editors: Riitta Suhonen, Minna Stolt and David Edvardsson

Received: 7 October 2021 Accepted: 7 December 2021 Published: 11 December 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

research [4]. Recent data confirm that integrated care models have had benefits in improving the health-related quality of life and functionality of people with multimorbidity and frailty [5]; in reducing hospitalization and readmission rates [6]; in reducing polypharmacy [7]; and in improving patient satisfaction, perceived quality of care and access to services [8].

In Portugal, the National Network of Integrated Continuous Care (RNCCI, acronym in Portuguese) was created in 2006 through Decree Law no. 101/2006, which is intended for people who, regardless of age, are functionally dependent. The RNCCI was developed as an integrated Health and Social Security response that mobilizes the public, private and social sectors. Its intervention objectives are the rehabilitation, readaptation and reintegration of frail individuals who no longer require acute hospital care [9]. The RNCCI is focused on community outreach services and includes hospitals, health centers, district and local social security services, the Solidarity Network and local authorities. There are several types of care provided in the RNCCI, which includes inpatient units (convalescence units; mediumterm and rehabilitation units; long-term and maintenance units; and level 1 pediatric integrated inpatient units) and outpatient units (mental health integrated continuous care units; day and autonomy promotion units; pediatric outpatient units and integrated continuous care teams) [9].

Convalescence Units (CUs) are intended for individuals with a potentially recoverable transitory loss of autonomy, and their purpose is clinical stabilization and functional rehabilitation. These units have their own facilities and are connected to an acute care hospital to ensure health care 24 h a day over a total length of hospitalization of 30 days [9].

As the purpose of CUs is functional rehabilitation, it seems clear to us the importance of assessing the functionality of patients to better evaluate health outcomes. The WHO developed the International Classification of Functioning, Disability and Health (ICF) to standardize the international assessment of functioning and disabilities related to the healthdisease process, taking into account the body's structures and functions and environmental factors [10]. Studies indicate that the evaluation of functionality is crucial in care models for older adults [11,12].

The performance of self-care behavior is considered extremely important for the individual person and for the health system, due to the benefits associated with it [13]. The gains from the development of self-care health behavior are related to reduced risk of complications, healthcare expenditures, hospital readmission rates, increased satisfaction with care, feelings of responsibility, control, independence, and autonomy, adoption of effective coping strategies, improved well-being, functional capacity, quality of life, symptom control, and pain [14–16]. In the care process, we seek to assess dependence in self-care skills (mobility, basic and instrumental activities of daily living, and cognitive status) [17,18], with the purpose of maintaining life, healthy functioning and personal development. The self-care deficit presents itself in degradé, as it can fluctuate in different levels, from mild to complete/severe [15]. The WHO report for the 2021–2030 decade emphasizes the importance of implementing actions that improve the functional ability of older adults, presenting four areas of action for this purpose: a) changing the way we think, feel and act toward age and aging; b) ensuring that communities foster the abilities of older people; c) providing integrated person-centered care and services that meet the needs of older people; and d) providing access to long-term care for older people who need it [19]. One of the strategies defined in the same report to accelerate the implementation of functional ability is to strengthen data, research and innovation [19]. Thus, the aim of the present study is to assess the functional trajectory of older adults hospitalized in CUs of the RNCCI of Portugal. For this purpose, the following objectives were defined: (1) to evaluate the evolution of the functional profile of older adults hospitalized in CUs; (2) to relate the functional profile of these individuals with age, sex, education level and emotional state; and (3) to evaluate the probability of the degree of dependence as a function of age and sex.

### **2. Materials and Methods**

### *2.1. Study Type and Sample*

This was a longitudinal, retrospective study with a sample of 59,013 older adults aged 65 or older hospitalized in health units belonging to CUs of the RNCCI of Portugal.

### *2.2. Instrument*

To evaluate the functional profile, variables of the International Classification of Functioning, Disability and Health (ICF), which contains the ICF components: Body Functions (Mental Functions) and Activities and Participation (Mobility, Self-care, Communication, Domestic Life, Main life areas), were used. The ICF items were transformed into a Likert scale so that they could be analyzed (no problem = 1; mild or moderate problem = 2; severe problem = 3; complete problem = 4). The global Cronbach's α = 0.951 is obtained, which means excellent internal consistency.

### *2.3. Data Collection*

The data were collected from the records made by health workers and entered in the RNCCI portal. To trace the sociodemographic profile, data related to the first evaluation of each hospitalization episode occurring between 1 January 2010, and 27 February 2017, were selected.

Subsequently, the evolution of the functional profile was evaluated through the analysis of biweekly evaluations of older adults hospitalized in CUs who were targets of a set of structured professional interventions (objective 1).

To meet the second objective, which was to relate the functional profile with the variables age, sex, education level and emotional state, principal component analysis of each of the instrument domains (mobility, ADL, IADL, and cognitive state), was combined with cluster analysis. This procedure presented the advantage of reducing the number of input variables in the cluster analysis, thus helping to simplify the characterization of the upstream clusters [20].

To meet the third objective, the probabilities of each class as a function of age and by sex were evaluated.

### *2.4. Statistical Analysis*

To analyze the four components of self-care capacity, over the days of hospitalization, a longitudinal analysis was performed, based on parametric tests (One-way ANOVA and t-Student test) (objective 1).

The exploratory analysis of clusters was performed using the hierarchical method (Analyze Classify Hierarchical Cluster). Since it is big data, it was necessary to perform a random partition of the database, to create a sub-sample with approximately 20% of the data, so that it was possible to process the information by SPSS. When obtaining the agglomeration coefficients (Ward's method), a graphic projection was performed, of the highest (last 30), to visualize their distances, where it was possible to verify that the best solution resided in the retention of 3 clusters. Consecutively, the cluster analysis was carried out using the non-hierarchical optimization method available in IBM SPSS: K-means (objetive 2).

Ordinal regression with the probit link function was used to assess whether age and sex had a significant effect on the probabilities related to the type of dependence (objective 3). The link function was chosen based on the frequency distribution criteria of the classes of the dependent variable "degree of dependence" [21].

### *2.5. Ethical Procedures*

The study was conducted in accordance with the guidelines of the Declaration of Helsinki and was approved by the Ethics Committee of Scientific Research in the Areas of Human Health and Welfare of the University of Évora (report number, 17036; date of approval, 26 April 2017).

### **3. Results**

### *3.1. Sociodemographic and Clinical Characteristics*

The mean age of the sample was 78.93 years (SD = 7.28), with an age range of 65 to 109 years. The majority of older adults admitted to CUs were in the age group of 75 to 84 years (47.2%), followed by 65 to 74 years (29.0%), and last, people aged 85 or older (23.8%). Most of the sample was female (61.5%), had a partner (44.7%), had less than 6 years of schooling (60.4%), and had an unskilled professional level at working age (70.3%). Table 1 provides the sociodemographic characterization of the sample.

**Table 1.** Sociodemographic characterization of older adults hospitalized in convalescence units by sex (2010–2017).


### *3.2. Evolution of the Functional Profile during Hospitalization*

Figure 1 shows the evolution of the functional profile of the sample throughout hospitalization, over time, with the evaluation performed every 15 days since admission. There were significant differences in each of the components: dependence in mobility (F(65.161954) = 143.337; *p* < 0.001), ADLs (F(65.166854) = 479.340; *p* < 0.001), IADLs (F(53.42791) = 9.271; *p* < 0.001) and cognitive state (F(65.161945) = 34.303; *p* < 0.001).

**Figure 1.** Components: mobility, activities of daily living (ADL), instrumental activities of daily living (IADL) and cogni-**Figure 1.** Components: mobility, activities of daily living (ADL), instrumental activities of daily living (IADL) and cognitive status in convalescence units (standardized mean values).

#### tive status in convalescence units (standardized mean values). *3.3. Dependence Clusters*

*3.3. Dependence Clusters* The nonhierarchical exploratory method of cluster grouping was used, yielding the following partition: Cluster 1—29.6% (n = 11,248); Cluster 2—52.1% (n = 19,785); and Cluster 3—18.3% (n = 6932). These clusters differed significantly in the dimensions mobility The nonhierarchical exploratory method of cluster grouping was used, yielding the following partition: Cluster 1—29.6% (n = 11,248); Cluster 2—52.1% (n = 19,785); and Cluster 3—18.3% (n = 6932). These clusters differed significantly in the dimensions mobility (F(2.37962) = 0.162; *p* < 0.001), ADLs (F(2.37962) = 0.734; *p* < 0.001), IADLs (F(2.37962) = 0.905; *p* < 0.001) and cognitive state (F(2.37962) = 0.811; *p* < 0.001).

Figure 1 shows the evolution of the functional profile of the sample throughout hospitalization, over time, with the evaluation performed every 15 days since admission. There were significant differences in each of the components: dependence in mobility (F(65.161954) = 143.337; *p* < 0.001), ADLs (F(65.166854) = 479.340; *p* < 0.001), IADLs (F(53.42791) = 9.271; *p* < 0.001) and cognitive state (F(65.161945) = 34.303; *p* < 0.001).

(F(2.37962) = 0.162; *p* < 0.001), ADLs (F(2.37962) = 0.734; *p* < 0.001), IADLs (F(2.37962) = 0.905; *p* < 0.001) and cognitive state (F(2.37962) = 0.811; *p* < 0.001). In general terms, the three clusters were quite distinct, with the following configuration, presented in Figure 2:


care deficit); (3) Cluster 3: Older adults with a lower degree of dependence (mild self-care deficit). Figure 3 shows the differences among the three clusters and the variables sex, age group, education, sad/depressed emotional state and anxious emotional state.


**Figure 2.** Mean dependence levels for Cluster 1, Cluster 2 and Cluster 3 in the convalescence units. **Figure 2.** Mean dependence levels for Cluster 1, Cluster 2 and Cluster 3 in the convalescence units.

Figure 3 shows the differences among the three clusters and the variables sex, age group, education, sad/depressed emotional state and anxious emotional state. Distribution of older adults across clusters

Cluster 1: Severe/complete dependence Cluster 2: Moderate dependence Cluster 3: Mild dependence

**Figure 3.** Distribution of people aged 65 years or older in the convalescence units by sex, age group, education level, emotional state: sad/depressed and emotional state: anxious, per cluster. **Figure 3.** Distribution of people aged 65 years or older in the convalescence units by sex, age group, education level, emotional state: sad/depressed and emotional state: anxious, per cluster.

The assumption of the slope homogeneity model was validated (߯ଶሺ2ሻ = 4.531;

The model was considered highly significant (߯ଶሺ2ሻ = 274.822; > 0.001(, although

<sup>ଶ</sup> = 0.132; ܴௌ

0.101 × ܵ݁ݔிሻ, where Φ is the standard normal distribution ࣨሺ0,1ሻ), because this function is recommended when the latent variable presents a normal distribution. The coefficients and statistical significance of the adjusted ordinal model are shown in Table

<sup>ଶ</sup> = 0.129). In the ordinal regression

**Interval** 

**df** *p* **Value 95% Confidence** 

*3.4. Degree of Dependence as a Function of Age and Sex*

*p* = 0.104).

2.

Thresholds

Localization

the effect size was small (ܴெி

**Parameters Estimate Standard Error** 

**Table 2.** Estimates and significance of the adjusted "Probit" model.

<sup>ଶ</sup> = 0.096; ܴே

Mild αMild = 2.439 0.065 1421.020 1 <0.001 [2.312; 2.566]

Moderate αModerate = 3.946 0.066 3559.323 1 <0.001 [3.816; 4.075]

Sex (Female) *b*Sex(Female) = 0.101 0.012 69.659 1 <0.001 [0.078; 0.125]

Age *bAge* = 0.043 0.001 2684.676 1 <0.001 [0.041; 0.044]

### *3.4. Degree of Dependence as a Function of Age and Sex*

The assumption of the slope homogeneity model was validated (*χ* 2 (2) = 4.531; *p* = 0.104).

The model was considered highly significant (*χ* 2 (2) = 274.822; *p* < 0.001), although the effect size was small (*R* 2 *MF* = 0.096; *R* 2 *<sup>N</sup>* = 0.132; *R* 2 *CS* = 0.129). In the ordinal regression model, the link function "Probit" was adopted (Φ −1 (*P*[*Y* ≤ *k*]) = *α<sup>k</sup>* − (0.043 × *Age* + 0.101 × *SexFemale*), where Φ is the standard normal distribution N (0, 1)), because this function is recommended when the latent variable presents a normal distribution. The coefficients and statistical significance of the adjusted ordinal model are shown in Table 2.


**Table 2.** Estimates and significance of the adjusted "Probit" model.

The results obtained suggest that with advancing age, the probability of observing higher-order classes, i.e., the probability of observing a higher degree of dependence, increases (bAge = 0.043; *p* < 0.001). Regarding sex, the results obtained suggest a higher probability of observing a higher dependence type in women than in men (bSex(Female) = 0.101; *p* < 0.001); however, this difference is moderate.

The evaluation of the probabilities of each of the classes as a function of age and by sex is shown in Figure 4. The analysis revealed that a) for both women and men, the probability of a dependence profile of "mild" is always inferior to any of the other two profile types; b) for women, starting at 88 years of age, the dependence profile that most likely occurs is "severe/complete"; for men, the same occurs, but starting at 91 years of age (three years later); c) by year of age, the ratio of the probability of observing profiles of lower dependence, compared to the probability of observing profiles of greater dependence, decreases by 4.2% (1-exp(-0.043)); and d) the odds ratio of lower dependence (mild) relative to higher dependence (severe/complete) decreases by 9.6% (1-exp(-0.101)) from males to females.

females.

The results obtained suggest that with advancing age, the probability of observing higher-order classes, i.e., the probability of observing a higher degree of dependence, increases (bAge = 0.043; *p* < 0.001). Regarding sex, the results obtained suggest a higher probability of observing a higher dependence type in women than in men (bSex(Female) = 0.101; *p*

The evaluation of the probabilities of each of the classes as a function of age and by sex is shown in Figure 4. The analysis revealed that a) for both women and men, the probability of a dependence profile of "mild" is always inferior to any of the other two profile types; b) for women, starting at 88 years of age, the dependence profile that most likely occurs is "severe/complete"; for men, the same occurs, but starting at 91 years of age (three years later); c) by year of age, the ratio of the probability of observing profiles of lower dependence, compared to the probability of observing profiles of greater dependence, de-

to higher dependence (severe/complete) decreases by 9.6% (1-exp(-0.101)) from males to

**Figure 4.** Evaluation of the probabilities for mild, moderate and severe dependence for women (a) and men (b) (߯ଶሺ2ሻ = 274.822; > 0.001 ;ܴெி <sup>ଶ</sup> = 0.096; ܴே <sup>ଶ</sup> = 0.132; ܴௌ <sup>ଶ</sup> = 0.129). **Figure 4.** Evaluation of the probabilities for mild, moderate and severe dependence for women (**a**) and men (**b**) (*χ* 2 (2) = 274.822; *p* < 0.001; *R* 2 *MF* = 0.096; *R* 2 *<sup>N</sup>* = 0.132; *R* 2 *CS* = 0.129).

### **4. Discussion**

**4. Discussion**

< 0.001); however, this difference is moderate.

This study evaluated the evolution of the functional profile of older adults hospitalized in CUs of the RNCCI, related it to the variables discussed below and evaluated the probability of the degree of dependence as a function of age and sex. This study evaluated the evolution of the functional profile of older adults hospitalized in CUs of the RNCCI, related it to the variables discussed below and evaluated the probability of the degree of dependence as a function of age and sex.

Regarding the evolution of self-care components throughout hospitalization, in the first 75 days of hospitalization, ADLs, mobility and cognitive state improved, but there was a decline after 75 days of hospitalization. In contrast, the ability to perform IADLs improved in the first 15 days of hospitalization, stabilized up to 45 days, and worsened thereafter. Considering that IADLs include tasks such as preparing meals, washing clothes, and housework, activities that are not performed by patients during hospitalization, it would be expected that with the lack of practice, the ability to perform these tasks Regarding the evolution of self-care components throughout hospitalization, in the first 75 days of hospitalization, ADLs, mobility and cognitive state improved, but there was a decline after 75 days of hospitalization. In contrast, the ability to perform IADLs improved in the first 15 days of hospitalization, stabilized up to 45 days, and worsened thereafter. Considering that IADLs include tasks such as preparing meals, washing clothes, and housework, activities that are not performed by patients during hospitalization, it would be expected that with the lack of practice, the ability to perform these tasks would begin to decrease earlier than would ADLs, mobility or cognitive state because the tasks related to these components are maintained throughout hospitalization. However, at 75 days of hospitalization, the ability to perform these activities also begins to decline. One study reports that 30 to 60% of older adults experience a functional decline in ADL during hospitalization [22]. Low mobility during hospitalization and functional and physical changes can lead to functional deficits in ADLs and cognitive impairments [23]. However, 80% of cases are preventable with effective health care [24,25]. Importantly, individuals who are hospitalized for more than 30 days in CUs are those with less potential for rehabilitation or without family support to provide continuity of rehabilitation at home; the length of hospitalization recommended for individuals referred to CUs is 30 days, which may explain these results. Thus, it is concluded that most individuals stay for the recommended time and experience reductions in acquired deficits in all dimensions.

> Regarding age, as would be expected, most individuals older than 85 years are in the severe/complete dependence cluster, the age group of 65 and 84 years is mostly in the moderate dependence cluster, and most individuals in the age group of 65 to 74 years old

are in the mild dependence cluster. Other authors have also concluded that the older the age, the worse the functionality [26].

When analyzing the clusters by sex, we found that there was a higher percentage of males in the severe/complete dependence cluster and in the mild dependence cluster and a higher percentage of females in the moderate dependence cluster. However, when analyzing the probability of the degree of dependence as a function of age and sex, for women, starting at 88 years of age, the type of dependence profile that most likely occurs is "severe/complete", while for men, the same occurs only after 91 years of age (three years later). This finding should be interpreted with caution because studies are conflicting, as some report that women have a higher degree of dependence than do men and others report that there are no differences. A literature review revealed that the incidence of functional disability was identical between sexes [27]. However, the results from two Portuguese studies with older adults indicated that the functional profile is worse in women than in men [26], regardless of age [28]. Another study that analyzed data from Spain, Portugal and Italy found that compared with men, women after 65–70 years of age had a higher risk of suffering from severe functional limitations. The same study added that in the age group of 75–80 years of age, women were 3.3% more likely than men to have severe functional limitations and that in the age group of 80 years or older, this probability increased to 15.5% [29]. The results from another study indicated that there are significant differences between sexes regarding the probability of occurrence of disability, with a higher probability for women in the following countries: United States, Korea, Southern Europe, Mexico and China. However, these differences do not exist in Northern, Central and Eastern Europe, England, and Israel. The authors conclude that gender inequality in society at the macro level is significantly associated with the probability of women developing disabilities [30]. A European study concluded that there are differences between sexes in relation to some important health indicators and that these differences are higher in Southern and Eastern Europe than in Western and Northern Europe. The same study warns that the presence of a sex difference in health cannot be considered a universal factor because, depending on the indicator and the country, the difference tended to increase, decrease or even reverse with age [31]. It should be noted that neither of the two previous studies included Portugal.

With regard to education, several studies are consistent in indicating that the lower the educational level, the worse the functional profile and the greater the degree of dependence [26,28,32] and that a lower education level is associated with worse physical and mental health outcomes [33–35]. In addition, a study showed that the impact of multimorbidity on ADLs was three times higher at the lowest education level than at the highest education level [36]. Another study added that after 65 years of age, the average probability of severe functional limitations for individuals with a low education level increased to more than 40% at age 80 years and older, while it was 26% in the higher education category [29]. The results of the present study are in line with those described in the literature, with a predominance of individuals who did not attend school in the severe/complete dependence cluster; those who studied for 1 to 6 years were predominant in the moderate dependence cluster, and those with a higher education level (7 or more years) were mostly in the mild dependence cluster. Thus, it is confirmed that in individuals admitted to CUs, the higher the educational level is, the lower the likelihood of dependence; additionally, education is extremely important for health and for active and healthy aging.

Regarding the analysis of emotional state, in the severe/complete dependence cluster, there was a predominance of individuals who felt depressed and anxious for a long time, and in the moderate dependence and mild dependence clusters, there was a predominance of older adults who felt sad or anxious for a short time. In fact, in a literature review, depression was considered one of the main risk factors for functional disability in older adults [27]. Another recent study that analyzed several combinations of multimorbidity concluded that those that included depression were the only consistent predictors of disability [37]. Another study followed the same direction, with the results indicating that

depression and/or cognitive deficit was associated with a substantially greater potential disability than combinations composed exclusively of somatic diseases [38].

The strengths of the present study include the relatively large sample size and the heterogeneity of the participants.

The limitations of the present study include the lack of an analysis of the time period after 2017 and the inclusion of only some of the determinants identified in the literature as influencing functional ability (e.g., emotional state).

### **5. Conclusions**

The recommended length of hospitalization in CUs of the RNCCI of Portugal is 30 days, although some individuals remain longer. During this period, there was a significant improvement in the functional profile of older adults hospitalized in these units, and there was a decline in those who remained beyond 75 days. More studies are needed to understand the reason for the observed decline after this cutoff point.

Women hospitalized in CUs were more likely to have a severe degree of/complete dependence three years earlier than men (88 years compared to 91 years of age, respectively). The higher the education level and the better the emotional state was, the lower the degree of dependence. This study confirms the importance of the education level and emotional state in functional ability in older adults. It is necessary that future studies evaluate the effectiveness of interventions that seek to prevent declines in the emotional state of individuals so that they are applied in clinical practice when a situation occurs that threatens the independence of such individuals, as is the case of people referred to CUs. In addition, the importance of strategies that promote literacy in the population is reinforced.

**Author Contributions:** Conceptualization, M.L. and C.F.; methodology, A.R. and H.O.; validation, M.L., C.F. and A.H.; formal analysis, A.R. and H.O.; investigation, M.L. and C.F.; resources, A.R. and L.P.; data curation, A.R. and H.O.; writing—original draft preparation, A.R. and L.P.; writing—review and editing, all authors; supervision, M.L., C.F. and A.H.; project administration, M.L.; funding acquisition, M.L. and C.F. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by FEDER. Programa Interreg VA España-Portugal (POCTEP), grant number 0499\_4IE\_PLUS\_4\_E. This research received an external funding for the PhD scholarship from the Foundation for Science and Technology, grant number SFRH/BD/140865/2018, Portugal (of the one of the authors Ana Ramos).

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethics Committee of Scientific Research in the Areas of Human Health and Welfare of the University of Évora (report number 17036 and date of approval 26 April 2017).

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Data are available from the authors upon reasonable request and with permission of University of Évora.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


### *Article* **Exploring the Social Networks of Women Bereaved by Stillbirth: A Descriptive Qualitative Study**

**Tosin Popoola \* , Joan Skinner and Martin Woods**

School of Nursing, Midwifery and Health Practice, Victoria University of Wellington, Wellington 6012, New Zealand; Joan.Skinner@vuw.ac.nz (J.S.); Martin.Woods@vuw.ac.nz (M.W.) **\*** Correspondence: Tosin.Popoola@vuw.ac.nz

**Abstract:** The loss of a baby to stillbirth is a traumatic experience and can lead to secondary losses, such as the loss of social relationships. In Nigeria, stillbirths are a common public health problem. However, limited attention has been given to the social ramifications of stillbirths. This study describes the social networks of women who have experienced a stillbirth and the factors influencing their social networks. Interviews and social network diagrams were used to collect data from 20 women about their social networks before and after stillbirth. Findings suggest that the experience of shame, unmet expectation of support, and a lack of trust led to relationship changes after stillbirth. Most participants met bereavement needs with their existing social networks before stillbirth, but many participants also experienced relationship losses (even among family networks). Information from social network analysis can reveal the risks and strengths inherent in social networks, which can be helpful for the provision of tailored/personalized bereavement care.

**Keywords:** drawings; Nigeria; perinatal loss; social networks; stillbirth; stillborn

**Citation:** Popoola, T.; Skinner, J.; Woods, M. Exploring the Social Networks of Women Bereaved by Stillbirth: A Descriptive Qualitative Study. *J. Pers. Med.* **2021**, *11*, 1056. https://doi.org/10.3390/jpm11111056

Academic Editors: Riitta Suhonen, Minna Stolt and David Edvardsson

Received: 28 September 2021 Accepted: 19 October 2021 Published: 21 October 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

### **1. Introduction**

The loss of a baby to stillbirth is a devastating and traumatic experience for women [1,2]. However, women do not just lose a baby when they experience stillbirth. When a child is lost to stillbirth, women often lose their emerging social status as an 'expecting mother' [3,4], leading to shame and low self-esteem [5] The experience of shame after stillbirth often leads to social withdrawal, loneliness, and relationship deterioration [5,6], which may lead to prolonged or complicated grief [4].

Generally, bereaved people experience changes in their social networks [7,8], and this is no different for mothers of stillborn babies [9,10]. However, in stillbirth bereavement, mothers may feel unsupported and isolated [9,11]. As a result of social withdrawal due to stigma, women's social networks may become smaller, disconnected, or underresourced [10–12] and their family may emerge as the primary source of support [1,2]. However, the social ramifications of stillbirths extend beyond the family [1,13]. Even if the family was supportive, the bereaved mother would need others outside the family to successfully reintegrate back into society [4]. However, there is limited focus on the social networks of women bereaved by stillbirths.

Social networks, understood here as the people and institutions through which an individual receives and gives social support, is important in stillbirth bereavement. After a stillbirth, women turn to their social networks to seek and receive social support [4]. However, many factors determine whether social support will be exchanged. For example, in spousal bereavement, Morrigan et al. [8] found that unmet expectations of support led to relationship loss and changes. Similarly, Aoun et al. [7] found that the amount, timing, function, and structure of social support influenced bereaved peoples' perception of the helpfulness or unhelpfulness of their social networks. However, research continues to favor the individual and psychological aspects of stillbirth bereavement over social dimensions of stillbirth loss.

In Nigeria, stillbirths are a significant public health problem. In 2019 alone, 171,428 babies were stillborn in Nigeria [14]. Dated but important studies on perinatal loss in Nigeria suggest that social support protects against depression and anxiety [15,16]. However, social norms have also been reported to prevent new social relationships after stillbirth in Nigeria [17]. For example, research suggests that Nigerian women have little to no opportunities to connect with their social networks after stillbirth due to the absence of rituals and funerals for stillborn babies [17,18]. Despite this, no research has examined the social ramifications of stillbirth loss in Nigeria.

Women's social networks hold social resources such as thew emotional, financial, and psychological support needed for a healthy grieving. As a result, it is crucial to identify the social networks that women bereaved by stillbirth have access to or do not have access to. This research seeks to describe the social networks of a sample of Nigerian women who have experienced the loss of a baby through stillbirth and the factors influencing their social networks.

### **2. Method**

This study utilized a descriptive qualitative design. After obtaining ethical approvals from the Ethical Review Committee of the Saki Baptist Medical Centre and the Human Ethics Committee of the Victoria University of Wellington (#23450), women bereaved by stillbirth were recruited. Twenty women aged 22 to 44 years (mean = 33.5) were recruited through snowballing from Saki, a town predominantly occupied by the Yoruba ethnic group in southwest Nigeria (Table 1). The sociocultural norms and attitudes of the Yoruba people about stillbirth have been described elsewhere (Popoola et al., 2021b). Potential participants who expressed interest in discussing their stillbirth experience were contacted by the first author (TP) who also happens to be a native of Saki town. Since the focus of the research was on the participants' social networks, eligibility was limited to women whose stillbirth was more than six months but less than three years ago. The eligibility criterion was intended to minimize distress during the early stages of grief and allow a timeframe where mothers could still remember those who played a role in their adjustment to loss.

**Table 1.** Demographic characteristics of the study participants.


### *2.1. Data Collection*

All women who participated in this study signed written informed consent before data were collected with face-to-face semi-structured interviews and social network diagrams in 2017 (Table 2). Most of the participants (*n* = 19) were interviewed in their homes based on their preferences. Interviews preceded social network diagrams since we wanted to build rapport and ease the participants into talking about a traumatic loss. The drawings were made on a Livescribe A5 notebook with the Livescribe pen since it allowed digital capturing of the diagrams and recording of the interviews.

**Table 2.** Interviews and diagrams' instructions.


The instructions that guided the creation of the diagrams were adapted from Pienaar et al.'s [19] communication mapping methodology, an approach that is embedded in communicative ecology models. Broad et al. [20] defined communicative ecologies as the network of connections individuals or groups depend upon to achieve their goals. Thus, from a communicative ecology perspective, the different people who are involved in an ecology, their relationships, and the social institutions and structures that connect them are the focus of analysis [21]. Visually representing social networks through diagrams has been found to facilitate meaningful discussions about social networks and the social context in which giving and receiving social support occurs [22]. In this study, elucidating information through social network drawings was important since women bereaved by stillbirth often perceive that their grief is not real and may be reluctant to discuss their experience publicly [4].

To construct the social network diagram, each participant was asked to draw a picture of herself on the Livescribe A5 notebook. After that, the participant was then asked to add her social networks to the image. Each participant was also asked to join herself with each social network using a maximum of three lines to show the quality of relationships. To identify the changes that might have occurred to the social networks after stillbirth, each participant was asked to look at the diagram and draw another one showing her social networks before stillbirth. After drawing the social network diagrams, the participants were invited to reflect on the changes in their social networks. The participants' reflections on their drawings were validated against the interview data and any discrepancies were discussed and clarified. The collection of two drawings per participant means 40 diagrams were produced in total. The interviews averaged 45 min.

### *2.2. Analysis*

The data collection processes included audio-recordings and digital capture of the diagrams, using a Livescribe A5 notebook/pen. The interviews that were conducted in the Yoruba language were translated to English during transcribing. A back-translation to Yoruba was conducted to ensure translation accuracy, and a second native Yoruba speaker verified this. The back-translation had to consider the challenge presented by the use of metaphors by the participants. Consistent with the nature of the data, we adopted both quantitative and qualitative methods of analysis. Raw images from the participants were used to present their social networks. However, a simple statistical analysis was used to analyze the number of people in the social networks and relationships (such as family). A qualitative analysis (thematic) was performed to explain the factors influencing the participants' social networks. Starting with data familiarization, the dataset was read repeatedly to identify statements relevant to the participants' social networks and the factors that determined why a participant turned to someone for help or not. Based on the significant statements identified from the data, the authors jointly developed a codebook and used it to group the significant statements into themes.

### *2.3. Rigor*

The criteria for ensuring rigor in qualitative research by Lincoln and Guba (cited in Connelly [23]) was followed in this study. The first author (TP) collected all of the data, and this enhanced the consistency of the data collection procedure. The first author also had the advantage of cultural insights, which enhanced the quality of the data interpretation. The use of culturally appropriate probes also assisted with the achievement of prolonged engagement with the participants. Participants also had opportunities to confirm the accuracy of interview transcripts/summaries and those who did (*n* = 7) made no changes. A reflective journal was kept during the project's data collection phase, which was used to debrief the co-authors. The findings are grounded in participants' narratives and all the authors agreed with the final thematic schemata. Pseudonyms SK# was used to present quotations.

### **3. Results**

Before stillbirth, the total number of people found in the social network diagrams of the 20 participants was 127. For individual participants, the lowest social network was 2, while the highest social network was 15 (average = 6.35) before stillbirth. After stillbirth, the total number of people found in the social network diagrams of the 20 participants reduced to 99. For individual participants, the lowest social network after stillbirth was 2, while the highest social network was 8 (average = 4.95). The social network analysis revealed six types of social relationships: family, friends, acquaintances, colleagues, neighbors, and healthcare providers. Family networks (consanguineous and affinal kin) accounted for more than half of participants' social networks before and after stillbirth, while healthcare providers were the least represented (Table 3).

The social network of most of the participants was stable, with 11 participants experiencing no loss or gain in social relationships after stillbirth. This suggests that the majority of the participants met the needs of bereavement with their existing social networks. Eight participants experienced relationship losses after stillbirth, with the most remarkable loss being amongst family relationships, where a total of 15 family members were dropped after stillbirth. Compared to the loss of family networks, the relationship loss in other domains was minimal: acquaintance = 4, friends = 4, colleagues = 4, neighbor = 1. Only one participant gained new social networks (from 2 to 6), with the new additional social networks being acquaintances from the place of worship. Table 3 presents the descriptive statistical analysis of the participants' social networks before and after stillbirth.


**Table 3.** Description of the participants' social networks.

### *3.1. Findings Regarding Women's Social Networks and the Factors Influencing It*

The factors that influenced the participants' social networks are presented in three categories: the perception of shame, expectation of support, and trust.

### 3.1.1. The Perception of Shame

"After I put to bed [gave birth], I hid the loss from people because I was ashamed that people might call me a failure . . . " (SK1)

"Whenever I saw someone who knew I was pregnant, my heart would skip [nervous] and I always tried to scurry away as if I had done something wrong . . . I felt like I had some impediments that I needed to be ashamed of". (SK5)

As indicated in the above quotes, participants tried to conceal the loss by avoiding social interactions. All participants experienced self-shame, with some likening the shame of stillbirth to other socially stigmatizing events such as incarceration. For example, one participant who concealed the loss by misleading others about her whereabouts used a metaphor about incarceration to discuss how stillbirth had devalued her. The participant said "kini idunu elewon ton'so ago mo owo [what pride does an ex-convict have to raise shoulders in the community]" (SK10).

The participants concealed the loss by engaging in two types of strategies. The first strategy used to avoid shame was the observance of a protracted period of mourning. Most participants exceeded the culturally stipulated period of mourning (40 days) since they were unsure how people would interpret their demeanor and conduct.

"Grieving as a mother whose child passed away is very tricky and challenging. On the one hand, you cannot move on too quickly because people expect a lengthy and genuine portrayal of soberness from you . . . your conduct should convince people that the loss truly and deeply pained you. On the other hand, you also cannot dwell on it for too long because people expect you to be grateful for your own life . . . So, grieving a stillborn child is like a performance; the timing of your re-entrance into the society, the way you carry yourself and your countenance must genuinely reflect your sadness but also your gratitude . . . Performing this role is hard". (SK13)

The second type of behavior that the participants used to avoid shame was relocating from their familiar environment. One participant even implied that "*suicide was better than facing shame [iku ya ju esin lo] (SK15)*". To avoid facing the shame of stillbirth, some participants relocated from their homes for up to six months, thinking that the passage of time would make people forget about the pregnancy. One participant who relocated for over six months said, "*ki oju ma ri ibi, gbogbo ara lo gun e [in order for my eyes not to see shame, I had to flee with all my body]" (SK18)*.

Both strategies employed to avoid shame resulted in relationship loss or inhibited the formation of new social networks. As illustrated in Figure 1, efforts to conceal the loss from others mean the family was more likely to be left in the social network after stillbirth, making the family a critical support network.


**Figure 1.** Iku ya ju j'esin lo [death is more preferable to shame] (SK15).

Supportive social interactions tend to counteract the negative impact of shame, as one participant described.

"When it happened [stillbirth], it felt like I was alone and I felt like there was nothing good about me anymore. But with the support of my mother, my level of shame started to reduce, and I started to become more comfortable in the presence of others". (SK13)

### 3.1.2. Expectation of Support

After the stillbirth, participants evaluated their relationships based on the perception of received/expected support and the level of interest shown by others in the grief. Using the metaphor below, some participants said they realized that some people in their social networks were fair-weather friends and could not be relied on for support.

"owo epo ni araye n'banila, won ki n'banila t'eje [people want to taste part of the oil (palm oil) in your hands, but not the blood]". (SK12)

All relationships were evaluated for their worth and usefulness, but it seemed that women were more critical of family members and close friends in terms of expected support. Unlike other relationships, participants were more likely to describe the support they expected or received from family members as a form of obligation.

"The person [child's father] who was supposed to help and support me was nowhere to be found. Everything he was supposed to do as a spouse and the father of the child, he did not do". (SK12)

"I would have loved my mother-in-law to be there for me, but she ended up disappointing me". (SK17)

"I did not feel that those around me understood my grief. Even though people surrounded me, I felt alone . . . Everyone was saying the same thing and doing the same thing, but I had other needs that nobody cared about". (SK14)

As seen in Figure 2, family or close social relationships that did not meet the mothers' expectations tend to be removed from the social network after stillbirth. Single women also tend to have the lowest social network after stillbirth. The highest number of people

in single women's social networks was four (ranged from two to four) and the quality of relationships also tend to be fragile.

**Figure 2.** Owo epo ni araye n'banila, won ki n'banila t'eje [people will help you lick your fingers if drenched in palm oil, but not when it is drenched in blood] (SK12).

When social networks met the expectation of support, the quality of the relationship between women and their social networks strengthened. As seen in Figure 3, a participant dropped her sister from her social network after stillbirth. However, the quality of the relationship increased with her mother and husband since they met the participants' emotional needs.

"My mum was very supportive . . . she moved closer to me and listened to my views about how I felt about the whole situation. During that time, I needed someone to talk to, not just people who would tell me it will be alright". (SK9)

**Figure 3.** Ti aja ba ni eni lehin, a pa obo [with the support of others, a person can survive any challenges] (SK9).

### 3.1.3. Trust

A lack of trust between participants and their social relations also prevented the formation of new relationships. Certain behaviors such as gossiping, blaming and insensitive comments created distrust and hindered mothers from engaging with others.

"When it happened, I felt like some people thought I deserved to lose the child because I had a home birth. Whenever I walk past, they gossip about me and I sometimes overhear them saying, 'look at the woman who did not value the life of a child' . . . I came out of the experience as a very introverted and paranoid person". (SK17)

"After the incident, the doctors and the nurses did not even give me the chance to gather myself together before they started asking how we were going to pay the hospital bills . . . they did not show any human feeling and I did not trust them to look after me". (SK10)

While a lack of trust prevented the formation of social relationships, participants bonded with those who were deemed trustworthy in their social network. The majority of the participants received the most support from their spouses and found them trustworthy.

"From my experience, I think the greatest gift you can receive when you lose a child is a true friend that you can trust with your pains . . . My husband was the one who really stood by me. He didn't go to work until two weeks after the incident . . . ". (SK2)

### **4. Discussion**

The findings of this study confirm earlier studies that have reported that bereavement leads to relationship changes [7,8,10]. The factors that influenced relationship changes in this study: shame [5,13], the unmet expectation of support [2] and mistrust [2,10] have been confirmed separately by other studies. Of the three factors that influenced the participants' social networks, shame/stigma has received the most attention in stillbirth research. In response to perceived/anticipated shame, women may try to conceal the loss [5,13], leading to relationship changes [2,10]. Consistent with earlier studies [5,6], we found that our participants employed strategies such as relocating or observing prolonged mourning to avoid shame, but this resulted in relationship loss or stagnancy. While concealment strategies might provide temporary salve for mothers, Pollock et al. [5] note that such strategies might heighten the risk of internalized grief and guilt, which may worsen bereavement outcomes in the long term.

In times of loss and crisis, people commonly turn to their social network for support [4]. However, being embedded within a network of relationships does not guarantee support for bereaved people [7]. Factors that determine whether support will be perceived as helpful or unhelpful have been explored from the size or quality of social networks but with conflicting results. For example, among people who have experienced traumatic events, Platt et al. [24] found that heterogeneous social network, not perception of strong social support, was more protective against post-traumatic stress disorder. In contrast, Ferlander et al. [25] found that a homogenous social network comprising of immediate family was more protective against depression for Russian women. Limited studies on bereavement, including stillbirths, suggest that type, size, and quality of social networks are essential for bereavement outcomes [7,26].

The needs of bereaved people are often complex and multifaceted [7,8]. Common needs during stillbirth bereavement, such as rebuilding identity, planning funerals, resolving psychological trauma, planning future reproductive health, and meeting financial needs, may be best met by a combination of intrapersonal and interpersonal social networks [1,27]. In this study, the participants' social networks did not include stillbirth support groups and there was an under-representation of health personnel. In addition, the participants who were single also tend to have smaller social networks and this echoes earlier studies that have reported that unmarried women have limited support after stillbirth [26,28]. This finding highlights that single women and those whose social networks are devoid of stillbirth support groups and health professionals may have limited support and feel socially isolated.

Similar to earlier studies [2,10], our findings also suggest that women rely more on family networks than any other types of relationships. In stillbirth bereavement, women may have no other choice but to rely on their family and existing social networks due to stigma and lack of social recognition of stillbirth loss. However, while the family may not be able to meet all the bereavement needs, Cacciatore et al. [26] found that family support is the only type of support that predicted anxiety and depression. Our findings replicate previous studies that family support is not a guarantee in stillbirth bereavement [2,6]. Previously, it has been reported separately that incongruent grieving styles, migration and insensitive comments from family members lead to family breakdown after stillbirth [2,6,29]. Our study adds that moral obligations within family relationships can play dual roles in bereavement; it can guarantee support for mothers, but also increase the likelihood of family breakdown. Family crises pose a severe threat to mothers' wellbeing and bereavement outcomes, especially in low and middle-income countries where social support groups, trained counsellors, and psychiatrists may not always be available [27]. Therefore, health personnel must recognize that the family network may be under pressure to meet bereavement needs, especially when external support is limited.

In this study, the mean social network of mothers after stillbirth was 4.95 (range 2–8). This is smaller compared to women in general, where an average number of 10 people has been reported among women's social networks [30–32]. The paucity of evidence on the social network of mothers of stillborn babies means we cannot compare the size of social networks that we found with other studies. However, it is important for those caring for women bereaved by stillbirth to know whether they should focus their interventions on social networks' size/quality, or structure/contents. Future studies are needed to increase our understandings of aspects of social networks that influence bereavement outcomes.

### **5. Strength and Limitations**

This study employed a creative and novel method to facilitate dialogue about a tabooed and challenging subject. The use of drawings permitted metaphors, which unlocked the barriers posed by language and social norms around the issue. The use of dual media of interviews and drawings allowed the participants to express feelings that may have been difficult in words. The limitations in this study are related to the methodological decisions taken. First, the participants' social network may be more extensive than we found if this study was conducted immediately after stillbirth. Second, the network size that we found may differ if the participants were asked to list or estimate their social networks instead of drawings (see Morrigan et al. [8], for instance). Third, time since loss might have resulted in the under-estimation of the social networks, but we did not notice any significant difference between participants based on time since the loss. Fourth, real-life relationships are intertwined and conceptually distinguishing them by the nature of relationships might overlook some nuances in the way people experience their social relationships. Fifth, due to the small sample size and the fact that all the participants belonged to the same Yoruba ethnic group, the transferability/generalizability of the findings to other contexts need to be assessed carefully. However, while the social networks that we found in this study might not represent all possible ties and relationships, it presents the social network that the participants identify as significant for themselves after stillbirth.

### **6. Conclusions and Practice Recommendations**

The loss of social networks and the inability to form new social networks after stillbirth suggests that mothers of stillborn babies may feel unsupported, isolated and miss out on critical bereavement support. Our study adds to the emerging literature on social networks of bereaved people [7,8] by suggesting that bereaved mothers' social networks can be influenced by shame, mistrust and unmet expectation of support.

The use of drawings to capture the social networks of mothers proved to be useful in analyzing the social relationships and the risk factors for social isolation. Since the experience of grief is unique and individual, the social network technique used in this study can help health personnel to provide personalized/person-centered bereavement care. When diagrams are used to elucidate the social relationships of women bereaved by stillbirth, health personnel may find that mothers' social networks are conflictive, broken or fragmented and may be tempted to want to fix such broken relationships. However, Sanicola [33] argued that the goal of care is not to remedy broken ties but to facilitate collective caregiving. This means that healthcare providers can achieve better bereavement outcomes if they locate bereavement care within the context of supportive and reliable relationships.

Despite the loss of family networks, the family was still a reliable source of support for the participants. However, the vulnerabilities of family relationships to breakdown cannot be overlooked in clinical care. Family breakdown can negatively impact the bereavement outcome, especially in Nigeria, where there is no external support for mothers after stillbirth [11,17]. Due to the risk that stillbirth presents to women and their families, healthcare providers need to approach stillbirth bereavement from a family-centered perspective. Healthcare providers can help mothers and their families discuss expectations and link the family with appropriate and supportive resources.

**Author Contributions:** Conceptualization T.P.; data acquisition T.P., data analysis and interpretation T.P., J.S. and M.W.; manuscript drafting T.P.; critical revisions T.P., J.S. and M.W. All authors have read and agreed to the published version of the manuscript.

**Funding:** The research was funded by the Victoria University of Wellington, grant number 400348.

**Institutional Review Board Statement:** The protocol for this study was granted ethical approvals by Victoria University of Wellington Human Ethics Committee and Saki Baptist Medical Centre Ethical Review Committee.

**Informed Consent Statement:** Written informed consent was obtained from all participants involved in the study.

**Data Availability Statement:** The de-identified transcripts of interviews and diagrams are still held by the primary author and can be provided upon reasonable request and sound methodological justification.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


### *Article* **Use of Digital Educational Technologies among Nursing Students and Teachers: An Exploratory Study**

**Fernanda Loureiro 1,\* , Luís Sousa <sup>2</sup> and Vanessa Antunes <sup>1</sup>**


**\*** Correspondence: floureiro@egasmoniz.edu.pt

**Abstract:** The emergence of digital educational technologies (DET) raises questions regarding the personalization of both teaching and care. DET use implies profound changes with consequences in nursing care and in nursing teaching-learning process. With the purpose of contributing to the improvement of the teaching-learning process through the use of DET, an exploratory-descriptive, cross-sectional, and observational study, with a quantitative approach (descriptive and inferential statistics), was developed. Online questionnaires were applied (*n* = 140 students and *n* = 23 teachers) after ethics committee approval. Results point to low cost and access without time/space limits as the main benefits, and decreased interaction, less physical contact, and technical difficulties as constraints. Globally, there was no difference between students and teachers in the use of DET. Still, men report more constraints than women. In this sample, the use of DET is still at an early stage. Both students and teachers are still unfamiliar with the scope and possibilities of these tools, not taking full advantage of the potential they have to offer. The impact of DET used in personalized nursing care is still yet to be understood.

**Keywords:** education; distance; education; learning; digital

### **1. Introduction**

Digital educational technologies (DET) have been used increasingly in recent years due to technological evolution. It implies profound changes in teaching practices with consequences in the teaching-learning process [1]. The use of platforms not only for providing academic content, but also for distance learning, have a great impact on both teachers and students: in the format of the contents that are made available, in the language used, class duration, time management, methods and hours of study, and in communication strategies. The recent public health pandemic emphasized the need for digital technologies use to maintain activity with social distance. In this context, DET represents an added value since they increase the interactivity and promote additional space for knowledge construction [2]. DET can be used in an asynchronous or a synchronous way. When used in its asynchronous form, it offers as added benefits: the possibility of a flexible schedule, the availability of quality material at low cost, and the potential to reach a high number of students without geographical limits [3]. However, issues on humanization of education and knowledge systematization are raised and, additionally, although DETs offer a wide range of possibilities, they do not guarantee learning quality [4]. Still, there is some reluctance in its use, particularly by teachers, as DETs break with traditional education, placing the emphasis on the student as the protagonist of the teaching-learning process [5].

As technology evolves, it has been integrated into different areas where education is included. Having this in mind, DET can be defined as the use of technological resources to improve teaching and learning, promote educational development and access to data and material [1]. It implies the use of virtual scenarios, online platforms, and digital

**Citation:** Loureiro, F.; Sousa, L.; Antunes, V. Use of Digital Educational Technologies among Nursing Students and Teachers: An Exploratory Study. *J. Pers. Med.* **2021**, *11*, 1010. https://doi.org/10.3390/ jpm11101010

Academic Editors: David Edvardsson, Minna Stolt and Riitta Suhonen

Received: 7 September 2021 Accepted: 5 October 2021 Published: 8 October 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

resources, among others. Historically, educational sciences have used DET as a resource to improve teaching since the 80s, but in nursing its use is more recent. Benefits can be pointed out from its use, such as improved interaction and content integration [2,4], easy access, more dynamism and interaction with students outside a physical school space [1]. However, constrains can also be pointed to, namely financial (adopting stateof-the-art technology can be expensive) and personal (teachers may be reticent to change traditional educational methods). It is not integrated in any theory of learning, but it fits a diverse theoretical framework. A major element of this theoretical frameworks is socio-constructivism that underlines the role of social processes in individuals' learning [6]. Social constructivism-based theory understands learning as a process that occurs through supportive collaboration with other people and leads to knowledge and evolution [7].

In nursing, the use of DET is seen as a resource that is complementary to traditional teaching [4] with increasing use and good results such as enhanced nursing students' problem-solving skills [6]. An example of the use of DET with other educational strategies is its use combined with case studies improving its benefits [8]. However, Chavaglia et al. [9] pointed to the need for greater diversification in the use of DET. In their study with undergraduate nursing students, they concluded that digital power point presentations, email and a google search engine were the predominant digital tools used by students.

A link between the use of DET, nursing education and sustainability is also found in the literature. Education in sustainable development is recognized as important particularly in health professions such as nursing. The use of DET offers an opportunity to reinforce this link since DET are sustainable and can be applied with low environmental impact [10]. Other innovative andragogical approaches are also being used, such as augmented reality and virtual simulation. Foronda et al. state that results from the use of these technologies are encouraging, since they suggest efficacy in improving nursing students learning outcomes [11].

However, some caution is required in the implementation of DET, especially by teachers. They have to be conscious of factors that influence the success in DET's application, and a minimum quality requirement has to be assured [10]. Usually, the types of DETs used by teachers are limited to those available in their institution and with easy access [12], which limits the quality of teaching and consequently the quality of the learning process. On the other hand, there has been a greater trend towards personalization of care and patient-centeredness. The use of this type of technology in teaching and its impact on care is still unknown. However, it seems to be consensual that DET may expand nursing education and consequently improve patient outcomes [11].

For the present study non-presential classroom was considered as classes that occur without physical presence in the same room of both students and teachers. The purpose of this study is to contribute to the improvement of the teaching-learning process in the nursing undergraduate degree through the use of DET in non-presential classroom teaching. The objective is to evaluate the use of DET in non-presential classroom teaching, benefits, constraints, and implications for teaching-learning process in nursing students. The following questions were defined: how is the use DET fulfilled in non-presential classroom teaching in nursing? What are the benefits and constrains to the use of DET? What are the implications for teaching-learning process of nursing students with the use of DET?

### **2. Materials and Methods**

### *2.1. Study Design*

An exploratory-descriptive, cross-sectional, and observational study was outlined. The research design fits into the quantitative paradigm with the use of surveys as a data collection technique.

The study was approved by the school board as well as by the school ethics committee. It was an independent study not integrated in any course or curricular unit. The study was applied to 230 inquiries and invitation to participate in the study was sent by email

with a survey link attached. Prior to fill the survey information regarding framework, aim, confidentiality, ethical issues and researchers contacts was provided. Consent was obtained prior to data collection. Participants had to mandatorily give their consent to advance with participation.

### *2.2. Sample and Setting*

The population consists of all nursing students and teachers who lecture the undergraduate nursing course at a private nursing school in Lisbon, Portugal. The sample is therefore of simple non-probabilistic type. Students from all academic years were invited to participate (*n* = 205). Teachers who teach classes of any type (*n* = 25) were selected and teachers involved solely in clinical nursing were excluded.

### *2.3. Instrument*

The online survey designed for this study includes 60 items based on previously published literature [13–16]. It had two parts: first part with sociodemographic data such as sex, age, and type of experience in the use of DET. Additional data collected included academic year for students' and professional experience and type of employment for teachers. The second part of the survey had questions related to the use of DET, selected and adapted from a previous literature review.

Types of DET [13] included it a 5-point Likert scale with 23 items within 5 dimensions (organizational tools, communications tools, presentation tools, learning assessment tools and identity transformation tools). Benefits of using DET include 10 dichotomous items and were retrieved from Kokol et al. [14]. These authors applied a survey to 125 nursing students' study and identified benefits and challenges of online education. Constraints of using DET was adapted from Lloyd et al. [15] and comprise 13 dichotomous items. The original scale included 22 items and authors performed exploratory factor analysis for validity and Cronbach alpha for reliability. Four factors were extracted that had high reliabilities (0.892; 0.806; 0.805; 0.870). In our study, we adapted the items for our context. Finally, implications in the use of DET include 14 dichotomous items adapted from an integrative review on digital technologies in the teaching of nursing skills by Silveira and Cogo [16].

Before its application, the surveys were sent to 5 experts. Content validity index was calculated as described by Polit and Becket [17] to verify the extend of expert agreement. A value of 0.90 was obtained which is considered adequate by the same authors.

Additionally, for instrument validity, to understand how types of DET were related and to favor statistical analysis, exploratory principal components factor analysis with varimax rotation was performed on the 23 items that addressed types of DET. The Kaiser-Meyer-Olkin measure verified the sampling adequacy for the analysis, KMO = 0.759. Additionally, Bartlett's sphericity test indicated that correlations between items were sufficiently large, X<sup>2</sup> = 1146.772, *p* < 0.001.

Initially, for the first-dimension sevens factors were identified that explained 64.9% of variance. However, two item (calendars and electronic portfolio) had correlations <0.3. By removing them and forcing analysis to the initial 5 factors that were identified in the literature, a 53.8% of the variance was obtained as explained on Table 1. Regarding communalities values ≥0.4 were considered. We also included the item: "Instant messages" although h<sup>2</sup> value is 0.386 because it is a marginal value and due to its conceptual importance in the context of this study.


**Table 1.** Exploratory factorial analysis for types of DET.

\*-h<sup>2</sup> : communalities.

For reliability assessment the Cronbach Alpha coefficient as well as the Kuder-Richardson formula were determined. Regarding types of DET, the alpha Cronbach values range between 0.594 in factor 5 and 0.739 in factor 1 as shown in Table 1. For benefits, constraints and implications thar are dichotomous items in the Kuder-Richardson formula indicated the following values: 0.670 (benefits); 0.502 (constrains) and 0.584 (implications). The values obtained are considered satisfactory [18], and additionally, it must be taken into consideration that it is a first attempt to use this instrument.

### *2.4. Procedure*

Participant's recruitment was performed by email. Since this was an independent study, not integrated in any course or discipline, researchers contacted students and professors through email addresses. An email was sent to all the participants inviting them to participate in the study. This email contained a survey link that allowed access to the online instrument. It was available for 1 week, and after that time, a reminder was sent, and the survey was available for one week more. The number of answered surveys were monitored by the researchers at the end of each week. After two weeks had passed the link was removed and data was transferred for analysis. All surveys were fully completed by participants.

### *2.5. Data Analysis*

Descriptive and inferential statistics were used for data analysis. Statistical analyses was performed with SPSS statistical package, 27.0 version (SPSS, Inc., Chicago, IL, USA). Categorical variables are presented as percentages. All variables followed a non-normal distribution so nonparametric test namely Mann-Whitney U test was used for comparisons between groups. To verify association between categorical variables Phi/Cramer V was used. A *p* value of 0.05 or less was considered statistically significant.

### **3. Results**

### *3.1. Sample Characteristics*

Of the study population, 163 subjects (140 students and 23 teachers) completed and returned the surveys, so the response rate was 70.8%. Most respondents were female (85%; *n* = 138) and had no training or formal education on DET, only experience as users (86%; *n* = 140). Sociodemographic characteristics are summarized in Table 2.



### *3.2. Types of DET*

Regarding the type of DET used, they were grouped into the five categories identified in factorial analysis. For presentation data, answers were clustered into two group: very frequently and frequently answers and occasionally, rarely or never answers. Results are summarized in Table 3.


**Table 3.** Results regarding type of DET used by teachers and students.

VF—very frequently; F—frequently; O—occasionally; R—rarely; N—never; \* significant at *p* < 0.05.

Mann-Whitney U test was used to assess if there were association between student's sample and teachers sample and the use of DET. No significative association was found except on the item "tests" (U = 2094.5; *p* = 0.016).

As to technologies used, results show the use of the technologies provided and available in the institution where the survey was applied: Moodle (98.8%; *n* = 161) and Microsoft Teams (98.2%; *n* = 160). Other technologies identified included Zoom (22.1%; *n* = 36), Mentimeter (4.9%; *n* = 8), Kahoot (4.3%; *n* = 7), Skype (2.5%; *n* = 4) and Google classrooms (1.2%; *n* = 2).

As to type of classes DET they were mostly used in theoretical class (93%; *n* = 153). In practical and laboratory classes DET were used less frequently (37% and 28%, respectively).

*3.3. Benefits and Constrains in the Use of DET*

Regarding benefits and constraints responses are summarized in Table 4.


**Table 4.** Results regarding benefits and constraints in the use of DET.

\* significant at *p* < 0.05.

To assess association between type of inquiries (students or teachers) Phi/Cramer's V was calculated. There were only significant statistical differences between teachers and students in two items: "it allows the integrating of multiple learning tools (pdf, links, app, among others)" and "promotes student accountability". Teachers refer to more benefits in the integration of multiple learning tools (V = 0.170; *p* = 0.030), and students attribute more benefits in item: "promotes accountability" (V = 0.170; *p* = 0.030). Regarding constrains, the same procedure was followed; however, no significant statistical differences were found.

### *3.4. Implications in the Use of DET*

The last question of the questionnaire referred to implications in the use of DET. Answers are synthetized in Table 5.


**Table 5.** Results regarding implications in the use of DET.

Regarding implications also Phi/Cramer's V was calculated to assess association between type of inquiries (students or teachers) and implications. Nevertheless, there were no significant statistical differences between the groups.

Globally, there was no difference between students and teachers in the use of DET. Still, when analysis is performed by gender differences were found regarding Factor 3 in types of DET and constrains.

Women got a higher score concerning the types of DET used (Factor 3) compared to men, that is, women use these resources more than men (U = 1531.5; *p* = 0.040). Still, men report more constraints than women in the use of DET (U = 2255; *p* = 0.005).

As stated earlier, all questions had an open space for additional answers; however, there were few answers and those who did respond mentioned aspects already identified.

### **4. Discussion**

This study allowed us to gather evidence related to the use of DET, being a first step to the improvement of the teaching-learning processes in the nursing undergraduate degree. Our results show a scarce use of DET by both teachers and students. Literature points to an increase in the use of DET in nursing teaching [16]. Our results show that there is space for improvement in this area. A possible explanation for this result may be the existence of interpersonal and institutional barriers, training and technological constrains and the lack of cost/benefit analysis [15].

Overall, 86% of our sample had no formal training on DET, which is in line with the literature. However, studies highlight that digital technologies are used both with academic and personal purposes [19]. Probably teachers in this sample do not feel the need for formal education as the day-to-day use of these tools allows them to use them without limitations or constraints. Additionally, digital tools are, naturally intuitive and user friendly. When considering our sample of students, they belong to a generation raised with information technology, internet, and social networks, which make them more confident manipulating new platforms and devices, not really requiring training in the area.

As mentioned in the results chapter, the statistical procedure allowed to group the types of DET in five new factors, categorized as follows: Training and Discussion Tools (factor 1); Communication tools (factor 2); Presentation and assessment tools (factor 3); Organization tools (factor 4); and Complementary learning tools (factor 5).

The training and discussion tools (factor 1) included the following items: discussion forums, blogs, avatars, virtual scenarios, and immersive technology. In this category, results show that avatars and virtual scenarios are poorly used. This types of DET are identified as particularly relevant in nursing education since they may provide a solution to issues such as: faculty shortages, scarcity of clinical placements and limited onsite laboratory space [11]. Additionally, its use allows students to repeatedly simulate procedures and care or even recreate high risk events that they may not contact within clinical context [11,16].

Communication tools category (factor 2) includes: videocalls, web conferences, audio, and video. Videocalls (74%) and audio (74%) are the preferred communication tools for the total sample. Although web conferences can save students time and prevent the inconveniences of traveling, video and audio conferencing provide the benefit of visual aids, allowing participants to make use of multiple senses, improving their concentration levels and increasing their capacity to absorb more information [20,21].

Within the presentation and assessment tools category (factor 3), the following items were identified: slides, tutorials, image sharing, bibliography sharing, activities, and tests. Tutorials got the lowest score (39%). This technology is still underused in our sample, but perhaps the changes in the lifestyle of today's society could lead to a greater need and recognition of its use in this setting. The traditional classroom can restrict daily life, as it requires anticipated planning. Otherwise, the tutorial can be paused, rewound or fastforwarded, and the lecture can be heard as many times as needed. These characteristics fit the current generation of students that crave the digital world and are extreme consumers of technology [22]. On the other hand, slides got the most representative score, as it is used by 89% of our sample. The use of DET to share slides can be framed as a simplistic and traditional use of this type of technologies.

In this category, it is also possible to verify an association between gender and the use of presentation and assessment tools, as women got a higher score compared to men (U = 1531.5; *p* = 0.040). Additionally, men report more constraints than women in the use of DET (U = 2255; *p* = 0.005). These results are quite interesting, as there is the general idea that women are outnumbered in informatics. Nevertheless, literature shows that the very first computer programmers and IT users were women, and that technology are more frequently used in female-dominated areas [23], such as teaching and nursing. Cai et al. [24] found that there are no significant gender differences in the attitudes toward technology. However, these authors argue that in the academic context, women are more prepared for technology use than the general female population.

In types of DET, factor 4 covers the following organization tools: online schedule, mind maps or graphic organizers, and file and management store. This last one is the most used by 67% of our sample. It is a more cost-effective system, as it reduces organization bureaucracy and saves space. Students can easily search, access, and share files. Additionally, support for decision-making and knowledge discovery can be achieved in an effectively way through the use of massive amounts of data that can be easily stored [24].

Instant messages and chats, research, and social network were considered as complementary learning tools (factor 5). Instant messages and chats (84%) are the mostly used by both students and teachers in our sample. They may be more relevant because they are one of the oldest communication tools, and usually a quick option when network failures occur. Within this category, the results also demonstrate that DET are used for research by 72% of the sample. These results are aligned with literature that shows that traditional libraries are being partially replaced by digital search, specially by students who prefer it because they are easy and quick to search [25].

More than half of our sample identified the following benefits of using *DET*: the integration of multiple learning tools (66%), low cost (68%) and access without time or space limits (71%). Pinto and Leite [19] also mention that the use of DET has effects on the interaction time between students and teachers extending it beyond the traditional academic period. This access 24/7 can also be seen as a constraint and reduce teacher's quality of life. However, considering the students' perspective, it can be a major benefit. Männistö et al. [6] reported that the use of DET in nursing teaching enhances motivation for learning. When used in nursing education DET can improve the learning experience particularly in clinical learning settings [26]. Regarding low cost, there is no robust evidence that e-learning is a more cost-effective way to deliver knowledge when compared to traditional methods [27]. Yet, for students and teachers it may prove to be a more economical option as it avoids travel or food expenses.

Still regarding the benefits, no significant associations were identified between the teacher and student groups, except for the items: it allows the integrating of multiple learning tools (pdf, links, app, among others)" and "promotes student accountability". These results are quite surprising because, given the age difference between the two groups in our sample, a more pronounced association would be expected. Literature shows that older adults are less likely to use technology that younger adults [28,29].

Our results show that constraints had greater statistical evidence than benefits. The more relevant were: the decreased interaction between students/teachers (72%), the less physical contact (77%), and the technical difficulties (67%). These aspects are pointed as the main barriers to the use of DET in literature and in fact, digital technologies can be used to enhance education, but they cannot completely replace face-to-face teaching [26]. Regarding technical difficulties, Naveed et al. [30] showed that efficient technology, infrastructure readiness and system reliability are some of the main critical success factors in implementing DET. To overcome these constrains several measures can be taken such as the use of combined methodologies (both presential and non-presential), a balanced and flexible schedule that maintains physical contact between students and teachers and permanent technical support available for users.

Implications identified by more than half of our sample include both the possibility of using software in simulation scenarios (57%) and the possibility to repeat simulations until learning, ensuring patient safety (55%). Overall, 66.9% of our sample consider that DET stimulates students' independence (66%) and stimulate self-learning (71%). In nursing, the possibility to repeat simulations is identified as important and relevant for learning [31]; additionally, it is a significant factor for patient safety. Nevertheless, in our sample, only 14% considered that the use of DET improves performance in the execution of techniques. This could be due to a broader interpretation of the DET concept, as the technology by itself does not improve the practical skills required for nurses. Rather, the use of specific technology tools, such as clinical virtual simulation, can be used as a complementary strategy to DET, improving clinical reasoning skills [31].

The impact of DET on nursing care and patient outcomes is still poorly explored. Although there seems to be a link between improved nursing education and improved patient outcomes, this issue must be further explored. Additionally, there is a trend towards more centered and personalized models of care and the use of massive forms of education may not be suitable to this trend. It should be noted that some of the digital tools used in education are used in the health area, also with the aim of personalizing and, in this case, improving care. With regard to students, DET may be adapted to their educational and personal needs, which translates into a more student-centered teaching. Additionally, due to its positive impact, these technologies are being implemented in continuing education and professional development. A good example is the use of immersive technologies that allow nurses to have dynamic experiences with patients and be more prepared for real-world clinical settings. Nevertheless, the impact of DET on nursing care and patient outcomes is still poorly explored. Although, there seems to be a link between improved nursing education and improved patient outcomes, this issue should be further studied.

It is also worth exploring if more student-centered teaching is later reflected in more people-centered nursing care.

Palvia et al. [32] state that e-education is advancing and is here to stay all around the world, which brings implications such as the need for improvement of telecommunications infrastructure, the acknowledgement of online education as equivalent to traditional face to face education and the globalization of e-education, which is inevitable, similar to the globalization of email or e-commerce. This authors also state that both online (virtual) and offline education must be combined so the virtues of both can be used.

It is necessary to have a more in-depth knowledge on this subject and a greater investment, both by teachers and educational institutions. Additionally, to understand the impact of the use of these technologies on nursing students learning, both in academic and clinical contexts, is needed. Nursing schools must invest in updated information technologies appropriate to the nursing curriculum, as well as providing training on their use to teachers and students. In our study, we verify that both teachers and students use mainly the tools made available by the school. Therefore, one way to improve the use of DET is to make more tools available. Additionally, in our setting the current teacher's generation is more familiar with the first tools that were created and used as DET so those are the ones they mostly use. In our sample, most teachers do not have formal education on the use of DET and this should be included in the annual training program.

As to limitations this study was applied after the first lockdown that occurred in Portugal which triggered an exponential growth in the use of DET. As more periods of lockdown occurred both teachers and students were forced to improve the use of DET. Therefore, we consider that if the application of this study had occurred after these periods the results would have been different. Study design and type of sample are also limitations since our exploratory-descriptive study used a convenience sample and therefore results cannot be generalized.

### **5. Conclusions**

Our results show that the use of DET in this sample is it is still at an early stage. Both students and teachers are still unfamiliar with the scope and possibilities of these tools, not taking full advantage of the potential they have to offer. The integration of multiple learning tools, low cost, and access without time or space limits are considered the main benefits of DET. On the other hand, the decreased interaction between students/teachers, the less physical contact, and the technical difficulties, are seen as the greater constraints. The main implications of DET are the possibility of using software in simulation, and the possibility to repeat simulations until learning, ensuring patient safety. It is also considered as a method that stimulates students' independence and self-learning.

Although distance education through DET is a few years old, as far as nursing education is concerned, this concept is still very recent. Globalization and, more recently, the pandemic context that forced social isolation, have further boosted the introduction of DET in nursing education. Distance education thus becomes an effective strategy to reach people who want or need to be qualified, but who, for different reasons, cannot depart from their context of life and work [2]. Nevertheless, this constitutes a challenge to the nursing traditional teaching-learning methods, which have a predominantly practical and proximity component [27]. The distance between students and teachers, should be used to its full potential, involving students dynamically in the learning process, respecting independence, and autonomy, establishing links between learning and life and professional experience. On the one hand, it is necessary to provide teachers with skills to establish a link with students and stimulate their learning and engagement. It is also recommended that schools select the appropriate DET methods for teaching nursing, namely investing in robust digital platforms and state-of-the-art simulated practice technology [32]. The link between improved nursing education and its implication on nursing care must also be further explored.

**Author Contributions:** Conceptualization, F.L., L.S. and V.A.; methodology, F.L., L.S. and V.A.; software, F.L., L.S. and V.A.; validation, F.L., L.S. and V.A.; formal analysis, F.L., L.S. and V.A.; investigation, F.L. and V.A.; resources, F.L., L.S. and V.A.; data curation, F.L., L.S. and V.A.; writing original draft preparation, F.L.; writing—review and editing, L.S. and V.A.; visualization, F.L., L.S. and V.A.; supervision, F.L., L.S. and V.A.; project administration, F.L. and V.A.; funding acquisition, F.L., L.S. and V.A. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work is financed by national funds through the FCT—Foundation for Science and Technology, I.P., under the project UIDB/04585/2020. The researchers would like to thank the Centro de Investigação Interdisciplinar Egas Moniz (CiiEM) for the support provided for the publication of this article.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Institutional Review Board and Ethics Committee of Escola Superior de Saúde Egas Moniz (reference number 881/2020).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


### *Article* **The Impact of the COVID-19 Pandemic on Nursing Care: A Cross-Sectional Survey-Based Study**

**Marco Clari <sup>1</sup> , Michela Luciani <sup>1</sup> , Alessio Conti 1,\* , Veronica Sciannameo <sup>2</sup> , Paola Berchialla <sup>3</sup> , Paola Di Giulio <sup>1</sup> , Sara Campagna <sup>1</sup> and Valerio Dimonte <sup>1</sup>**


**Abstract:** The COVID-19 pandemic has had a severe impact on nursing care. This cross-sectional survey-based study compared aspects of nursing care and nurses' satisfaction with care provided before and during the first wave of the COVID-19 pandemic. A total of 936 registered nurses (RNs) rated the frequency with which they performed fundamental care, nursing techniques, patient education, symptom management, and nurse–patient relationships before and during the pandemic. A recursive partitioning for ordered multivariate response in a conditional inference framework approach was applied. More frequent fundamental cares were associated with their frequency before the pandemic (*p* < 0.001), caring for COVID-19 patients (*p* < 0.001), and workplace reassignment (*p* = 0.004). Caring for COVID-19 patients (*p* < 0.001), workplace reassignment (*p* = 0.030), and caring for ≤7.4 COVID-19 patients (*p* = 0.014) increased nursing techniques. RNs in high-intensity COVID-19 units (*p* = 0.002) who educated patients before the pandemic, stopped this task. RNs caring for COVID-19 patients reported increased symptom management (*p* < 0.001), as did RNs caring for more non-COVID-19 patients (*p* = 0.037). Less frequent nurse–patient relationships before the pandemic and working in high-intensity COVID-19 units decreased nurse–patient relationships (*p* = 0.002). Despite enormous challenges, nurses continued to provide a high level of care. Ensuring the appropriate deployment and education of nurses is crucial to personalize care and to maintain nurses' satisfaction with the care provided.

**Keywords:** COVID-19; nursing care; patient care planning; quality of health care; personalized care; conditional inference trees

### **1. Introduction**

From 31 December 2019, when the World Health Organization's China Office reported a case of pneumonia of unknown etiology in Wuhan, the Coronavirus Disease 2019 (COVID-19) started to spread across the globe [1]; on 11 March 2020, COVID-19 was declared a pandemic [2]. Italy was one of the first countries outside of China to report cases [1]. As of July 2021, Italy has reported more than 4,200,000 confirmed COVID-19 cases and more than 128,000 COVID-19 deaths [3]. The COVID-19 pandemic has had a severe impact on healthcare systems around the world, affecting the availability of beds in hospitals and intensive care units [4]. The second and following waves of the COVID-19 pandemic are still challenging healthcare systems and professionals [5].

Nurses have been recognized as fundamental actors in public health crises and have played a major role in the COVID-19 pandemic; however, the pandemic has had a severe impact on nursing care. This is due to the challenges associated with the preparedness and

**Citation:** Clari, M.; Luciani, M.; Conti, A.; Sciannameo, V.; Berchialla, P.; Di Giulio, P.; Campagna, S.; Dimonte, V. The Impact of the COVID-19 Pandemic on Nursing Care: A Cross-Sectional Survey-Based Study. *J. Pers. Med.* **2021**, *11*, 945. https://doi.org/10.3390/jpm11100945

Academic Editor: Riitta Suhonen

Received: 30 July 2021 Accepted: 22 September 2021 Published: 23 September 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

response to emergencies shown by several healthcare systems in different care settings [6,7]. In the face of a heavy workload, nurses have had to wear personal protective equipment (PPE) [4,8] and have been confronted with a lack of PPE [9,10], staff [10,11], and other resources [9,12], all of which have led to decreased mental health and well-being [13,14], occupational satisfaction [15], and high infection rates among nurses [16]. Nurses also reported being reassigned due to changes in human resource allocations, having to quickly learn new skills and competencies, having to work with newly-graduated nurses, and difficulties in communicating with patients and their families due to PPE and isolation. All of the factors mentioned above could have affected personalized care, an essential aspect of nursing, during the COVID-19 pandemic [17]. The concept of personalized health care in nursing is influenced by the care environment and the ability of general nursing care to meet a patient's needs, which were inevitably affected by the emergency, thereby potentially impacting clinical outcomes and satisfaction with care.

However, to the best of our knowledge, no rigorous study exists on how the pandemic has impacted nursing care. In particular, no studies have examined differences in the care provided by nurses who cared for COVID-19 and non-COVID-19 patients, or between those who were reassigned due to the pandemic and those who continued to work in their unit. Moreover, there is still a limited understanding of the factors associated with nurses' satisfaction with the care provided during the pandemic. This information could help decision makers ensure that nurses in a given unit have the appropriate education, skill mix, and patient-to-nurse ratio, thereby improving clinical practice and care personalization during this pandemic [5] and future health emergencies. Hence, this study aimed to identify changes in nursing care by comparing aspects of nursing care and satisfaction with care provided before and during the first wave of the COVID-19 pandemic, examining differences between nurses who cared for COVID-19 and non-COVID-19 patients, and between those who were reassigned and those who continued to work in their unit.

### **2. Materials and Methods**

### *2.1. Study Design and Participants*

This cross-sectional study included registered nurses (RNs) in Italy who delivered nursing care in the 3 months before study enrollment. No restrictions regarding the type of patient nor the setting were applied.

### *2.2. Procedures*

RNs were invited to complete an online questionnaire, which was available between 12 May and 31 July 2020. Invitations were disseminated through ads on social media (Facebook, Twitter, Instagram), informational links on the websites of the Nursing Councils, and through texts and e-mails sent directly to RNs, using contact lists obtained from nursing schools in each of the Italian regions. Every 2 weeks, ads and informational links were reposted, and texts and e-mails were resent to RNs. The response rate (RR) was calculated as the RR2 [18], i.e., the sum of complete (I) and partial (P) questionnaires divided by the sum of complete, partial, non-questionnaires (NC, defined as respondents who logged on to the questionnaire but did not complete any item), and other (O, defined as respondents who could not fit in any of previous classifications; in this study, this category was not present): RR2 = (I + P)/((I + P) + (NC + O)). For this study, the RR2 was 81.4%.

### *2.3. Instruments*

The online questionnaire was composed of six sections that covered: (I) changes in nursing care due to the COVID-19 pandemic; (II) changes in work organization; (III) ethics choices; (IV) most challenging case; (V) additional education needed to care for COVID-19 patients; (VI) socio-demographic characteristics. In the present analysis, only sections I, II, and VI were considered.

In section I, RNs reported the frequency with which they carried out the following tasks: fundamental care (i.e., personal hygiene, elimination, nutrition, mobility) [19],

nursing techniques (i.e., respiratory support, vascular access, device positioning and management), patient education (i.e., respiratory exercises, medication, education), symptom management (pain, dyspnea, fatigue), and nurse–patient relationships (i.e., personal interactions with patients). These frequencies were reported for two time periods: before the COVID-19 pandemic and during what RNs perceived to be the worst week of the pandemic, using a 5-point Likert-type scale (1 = never; 2 = rarely; 3 = sometimes; 4 = often; 5 = most of the time). For each time period, RNs were also asked to rate their overall satisfaction with the care they provided (1 = very poor; 2 = poor; 3 = fair; 4 = good; 5 = excellent).

In section II, RNs reported the number of patients they cared for before the COVID-19 pandemic, if that number increased, remained stable, or decreased during the pandemic, if they were reassigned to another unit, and, if so, how many times. RNs were also asked if they thought there was enough time to prepare for the pandemic in terms of work organization, education/training, and their personal lives, with responses given on 5-point Likert-type scale (1 = not at all; 2 = not really; 3 = neutral; 4 = somewhat; 5 = very much). Finally, RNs were asked to report the number of patients they personally cared for during their last shift, and those caring for COVID-19 patients were also asked to report the number of patients who required no respiratory support, high-flow oxygen, non-invasive ventilation, and mechanical ventilation. The number of patients was then weighted based on the respiratory support provided, with a higher coefficient for patients with mechanical ventilation and a lower coefficient for those with no support.

### *2.4. Analysis*

Continuous variables were described using medians and interquartile ranges (IQRs), or means and standard deviations (SDs). The Mann–Whitney U test was performed to evaluate differences in quantitative variables, and Chi-square or Fisher's exact test was used for categorical variables as appropriate.

A recursive partitioning for ordered multivariate response in a conditional inference framework approach was applied. Conditional inference trees were constructed to identify the pattern of work organization, nurse education/training, and personal lives associated to the different levels of fundamental care, nursing techniques, patient education, symptom management, and nurse–patient relationships in post COVID-19.

Independent variables were selected and split through multiplicity-adjusted *p*-values following the Bonferroni's criterion. The split of variables determines a set of rules associated to different values of the dependent variable. In more detail, for each variable, the conditional regression tree determines the optimal split and a partitioning is performed, selecting the input variable with the highest multiplicity-adjusted *p*-value. Then, a binary split is performed on the selected input variable, and this process is recursively performed until a stopping criterion is reached. The stopping criterion was based on significant results, i.e., splitting continues until the minimum of the adjusted *p*-values is less than a pre-specified level of significance (0.05) or otherwise stops [20].

In our analysis, a conditional inference tree was constructed for each nursing task as dependent variables (fundamental care, nursing techniques, patient education, symptom management, and nurse–patient relationships). Independent variables included in the trees were: values from before the pandemic for investigated nursing tasks, caring for COVID-19 patients, gender, age, education, geographic area, work experience, working unit, workplace reassignment, preparedness (in terms of work organization, education/training, and personal lives), decrease/increase in number of patients, number of non-COVID-19 patients, a weighted sum of COVID-19 patients with different respiratory support, and type of contract.

We also performed sensitivity analyses using multivariate logistic regression. Likert scores for the investigated nursing tasks before and during the COVID-19 pandemic were dichotomized into never/rarely/sometimes (0) and often/most of the time (1), and odds ratios and 95% confidence intervals were computed. Missing data were deleted listwise. Analyses were performed using R version 3.6.1 [21].

### *2.5. Ethics*

RNs were informed about the study before accessing the online questionnaire, and consent was obtained before they began the questionnaire. RNs were not compensated for their role in the study, and participation was voluntary. All data were collected anonymously, and respondents could leave the questionnaire at any time. The study was approved by the University of Torino Ethics Committee (Approval no. 279061–01/07/2020) and conducted following the Declaration of Helsinki guidelines.

### **3. Results**

A total of 936 RNs completed the online questionnaire (68.2% female); the median age in the sample was 39 years (IQR 30–49), and 40.7% of RNs had a bachelor's degree (Table 1).


**Table 1.** Participants' characteristics.

<sup>1</sup> Presence of missing data.

Most RNs worked in Northern Italy during the pandemic (86.5%). The median work experience was 13 years (IQR 5–25). Most RNs worked in a hospital setting (67.8%), 77.1% cared for COVID-19 patients, 28% worked in a dedicated COVID-19 unit, and 35.9% were reassigned following reorganizations to increase beds for COVID-19 patients. Almost half of RNs reported that they cared for fewer patients during the pandemic (45.8%): the median patient-to-nurse ratio among RNs caring for non-COVID-19 patients was 8 (IQR 3–15), compared to 2.4 (IQR 1–5) among those caring for COVID-19 patients. RNs caring for COVID-19 patients were significantly younger and had less work experience. Furthermore, there was a higher number of male RNs among those caring for COVID-19 patients. Lastly, in areas of high COVID-19 prevalence (i.e., Northern Italy and the Lombardy Region) more RNs were reassigned to COVID-19 hospital units. RNs reported there was little time to prepare for the pandemic in terms of work organization (median 2 (rarely), IQR 1–3), education/training (median 2, IQR 1–2), and their personal lives (median 1 (no time), IQR 1–2) (Table 1).

The highest number of reassigned RNs was observed in the Lombardy Region (19.0% versus 10.1%; *p* = 0.003). Reassigned RNs (*p* = 0.042; median 38, IQR 29–48 versus median 40, IQR 30–50 among those not reassigned) felt poorly prepared in terms of work organization (*p* = 0.004; median 1, IQR 1–2 versus median 2, IQR 1–3) and their personal lives (*p* = 0.009; median 1, IQR 1–2 versus median 1, IQR 1–2) (Table 2).


<sup>1</sup> Presence of missing data; NA Not available.

Reassigned nurses reported caring for a higher number of patients (*p* < 0.001), and they comprised a higher number of self-employed (9.2% versus 3.8% among those not reassigned; *p* = 0.028) and temporary contract workers (public temporary 3.3% versus 2.6%; private temporary 3.9 versus 2.9). RNs were mostly reassigned from medium–low-intensity facilities, and especially from community care to tertiary care hospitals (*p* = 0.003); RNs reassigned to COVID-19 units were most often transferred from medium–low-intensity facilities to tertiary care hospitals (*p* < 0.001). Conversely, RNs who were not reassigned and did not care for COVID-19 patients were significantly older (*p* < 0.001; median 45.5, IQR 36–51 versus median 37.5, IQR 30–48 among reassigned RNs who did care for COVID-19 patients) and had more work experience (*p* < 0.001; median 20, IQR 10–30 versus median 11, IQR 4–23) (Table 2).

### *Nursing Tasks*

The frequency of fundamental care tasks before the pandemic was associated with the frequency during the pandemic (*p* < 0.001; before: median 3, IQR 2–4; during: median 3, IQR 2–5). In Figure 1, the conditional inference tree with the frequency of fundamental care tasks as a dependent variable, reported at the bottom of the figure as a boxplot, is represented. At the top of Figure 1, we can see that the first split of the decision tree was

performed based on the frequency of fundamental care registered in the pre-pandemic period (node 1). RNs who frequently performed fundamental care before the pandemic continued this practice (*p* < 0.001), as we can see in the rightmost branch of the tree. Furthermore, among those who declared to not usually perform fundamental care before the pandemic, caring for COVID-19 patients (*p* < 0.001) or being reassigned (*p* = 0.004) significantly increased the frequency of fundamental care during the COVID-19 pandemic, as expressed in the boxplot on the Likert scale, reported in the bottom of the Figure 1, left part. performed based on the frequency of fundamental care registered in the pre-pandemic period (node 1). RNs who frequently performed fundamental care before the pandemic continued this practice (*p* < 0.001), as we can see in the rightmost branch of the tree. Furthermore, among those who declared to not usually perform fundamental care before the pandemic, caring for COVID-19 patients (*p* < 0.001) or being reassigned (*p* = 0.004) significantly increased the frequency of fundamental care during the COVID-19 pandemic, as expressed in the boxplot on the Likert scale, reported in the bottom of the Figure 1, left part.

0.003); RNs reassigned to COVID-19 units were most often transferred from medium– low-intensity facilities to tertiary care hospitals (*p* < 0.001). Conversely, RNs who were not reassigned and did not care for COVID-19 patients were significantly older (*p* < 0.001; median 45.5, IQR 36–51 versus median 37.5, IQR 30–48 among reassigned RNs who did care for COVID-19 patients) and had more work experience (*p* < 0.001; median 20, IQR 10–

The frequency of fundamental care tasks before the pandemic was associated with the frequency during the pandemic (*p* < 0.001; before: median 3, IQR 2–4; during: median 3, IQR 2–5). In Figure 1, the conditional inference tree with the frequency of fundamental care tasks as a dependent variable, reported at the bottom of the figure as a boxplot, is represented. At the top of Figure 1, we can see that the first split of the decision tree was

*J. Pers. Med.* **2021**, *11*, x FOR PEER REVIEW 6 of 13

30 versus median 11, IQR 4–23) (Table 2).

*Nursing Tasks*

**Figure 1.** The conditional inference tree with the frequency of fundamental care tasks as dependent variable, reported at the bottom of the figure as boxplot. **Figure 1.** The conditional inference tree with the frequency of fundamental care tasks as dependent variable, reported at the bottom of the figure as boxplot.

In Figure 2, the conditional inference tree about fundamental care is represented. RNs caring for COVID-19 patients who frequently performed nursing techniques before the pandemic continued this practice (*p* < 0.001; before: median 4, IQR 3–5; during: median 5, IQR 4–5), as we can see by following the rightmost branch. RNs who cared for COVID-19 patients (*p* < 0.001), those who were reassigned (*p* = 0.030), and those who assisted ≤7.4 patients (*p* = 0.014) significantly increased the frequency of nursing techniques, with a median Likert scale of 5, as shown in the leftmost boxplot (Figure 2). In Figure 2, the conditional inference tree about fundamental care is represented. RNs caring for COVID-19 patients who frequently performed nursing techniques before the pandemic continued this practice (*p* < 0.001; before: median 4, IQR 3–5; during: median 5, IQR 4–5), as we can see by following the rightmost branch. RNs who cared for COVID-19 patients (*p* < 0.001), those who were reassigned (*p* = 0.030), and those who assisted ≤7.4 patients (*p* = 0.014) significantly increased the frequency of nursing techniques, with a median Likert scale of 5, as shown in the leftmost boxplot (Figure 2). *J. Pers. Med.* **2021**, *11*, x FOR PEER REVIEW 7 of 13

**Figure 2.** The conditional inference tree with the frequency of nursing techniques as dependent variable, reported at the bottom of the figure as boxplot. **Figure 2.** The conditional inference tree with the frequency of nursing techniques as dependent variable, reported at the bottom of the figure as boxplot.

RNs who performed patient education often before the pandemic and worked in high-intensity COVID-19 units (*p* = 0.002) stopped performing this task, as shown in the RNs who performed patient education often before the pandemic and worked in high-intensity COVID-19 units (*p* = 0.002) stopped performing this task, as shown in the

rightmost branch in Figure 3, while those in other settings continued their usual practice

In Figure 4, the decision tree with symptom management as the outcome variable is reported. The frequency of symptom management during the COVID-19 pandemic was similar to that before the pandemic (*p* < 0.001; before: median 4, IQR 3–5; during: median 4, IQR 3–5), but it increased for RNs who cared for COVID-19 patients (*p* < 0.001), as shown in the leftmost branch. Caring for a higher number of non-COVID-19 patients (>6) increased the frequency of symptom management (*p* = 0.037), as reported in the eighth node,

Nurse–patient relationships before the pandemic were associated with those during the pandemic (*p* < 0.001; before: median 5, IQR 4–5; during: median 4, IQR 2–5), but nurses working in high-intensity COVID-19 units who reported frequent nurse–patient relationships before the pandemic had no chance to relate with patients during the pandemic (*p* = 0.002), as shown in the rightmost branch of the conditional decision tree reported in Figure

**Figure 3.** The conditional inference tree with the frequency of patient education as dependent variable, reported at the

with a median Likert scale of 4.

5.

bottom of the figure as boxplot.

(*p* < 0.001; before: median 3, IQR 2–4; during: median 3, IQR 1–4).

bottom of the figure as boxplot.

the bottom of the figure as boxplot.

at the bottom of the figure as boxplot.

**Figure 2.** The conditional inference tree with the frequency of nursing techniques as dependent variable, reported at the

*J. Pers. Med.* **2021**, *11*, x FOR PEER REVIEW 7 of 13

rightmost branch in Figure 3, while those in other settings continued their usual practice (*p* < 0.001; before: median 3, IQR 2–4; during: median 3, IQR 1–4). rightmost branch in Figure 3, while those in other settings continued their usual practice (*p* < 0.001; before: median 3, IQR 2–4; during: median 3, IQR 1–4).

RNs who performed patient education often before the pandemic and worked in high-intensity COVID-19 units (*p* = 0.002) stopped performing this task, as shown in the

**Figure 3.** The conditional inference tree with the frequency of patient education as dependent variable, reported at the bottom of the figure as boxplot. **Figure 3.** The conditional inference tree with the frequency of patient education as dependent variable, reported at the bottom of the figure as boxplot.

In Figure 4, the decision tree with symptom management as the outcome variable is reported. The frequency of symptom management during the COVID-19 pandemic was similar to that before the pandemic (*p* < 0.001; before: median 4, IQR 3–5; during: median 4, IQR 3–5), but it increased for RNs who cared for COVID-19 patients (*p* < 0.001), as shown in the leftmost branch. Caring for a higher number of non-COVID-19 patients (>6) increased the frequency of symptom management (*p* = 0.037), as reported in the eighth node, with a median Likert scale of 4. In Figure 4, the decision tree with symptom management as the outcome variable is reported. The frequency of symptom management during the COVID-19 pandemic was similar to that before the pandemic (*p* < 0.001; before: median 4, IQR 3–5; during: median 4, IQR 3–5), but it increased for RNs who cared for COVID-19 patients (*p* < 0.001), as shown in the leftmost branch. Caring for a higher number of non-COVID-19 patients (>6) increased the frequency of symptom management (*p* = 0.037), as reported in the eighth node, with a median Likert scale of 4. *J. Pers. Med.* **2021**, *11*, x FOR PEER REVIEW 8 of 13

**Figure 4.** The conditional inference tree with the frequency of symptom management as dependent variable, reported at **Figure 4.** The conditional inference tree with the frequency of symptom management as dependent variable, reported at the bottom of the figure as boxplot.

Nurse–patient relationships before the pandemic were associated with those during the pandemic (*p* < 0.001; before: median 5, IQR 4–5; during: median 4, IQR 2–5), but nurses working in high-intensity COVID-19 units who reported frequent nurse–patient relationships before the pandemic had no chance to relate with patients during the pandemic (*p* = 0.002), as shown in the rightmost branch of the conditional decision tree reported in Figure 5.

RNs' satisfaction with the care provided remained stable during the pandemic (*p* < 0.001; before: median 4, IQR 3–4; during: median 3, IQR 2–4), but younger nurses (aged ≤27 years) tended to judge their care as poor (*p* = 0.047), with a median Likert scale of 2 (node 12, Figure 6). Similarly, nurses with less work experience (≤13 years) reported a decreased quality of care (*p* = 0.032). RNs who reported high preparedness in terms of education/training showed increased care satisfaction (*p* = 0.039), with a median Likert

**Figure 5.** The conditional inference tree with the frequency of nurse–patient relationships as dependent variable, reported

scale of 4 (node 10, Figure 6).

the bottom of the figure as boxplot.

at the bottom of the figure as boxplot.

**Figure 5.** The conditional inference tree with the frequency of nurse–patient relationships as dependent variable, reported **Figure 5.** The conditional inference tree with the frequency of nurse–patient relationships as dependent variable, reported at the bottom of the figure as boxplot.

RNs' satisfaction with the care provided remained stable during the pandemic (*p* < 0.001; before: median 4, IQR 3–4; during: median 3, IQR 2–4), but younger nurses (aged ≤27 years) tended to judge their care as poor (*p* = 0.047), with a median Likert scale of 2 (node 12, Figure 6). Similarly, nurses with less work experience (≤13 years) reported a decreased quality of care (*p* = 0.032). RNs who reported high preparedness in terms of education/training showed increased care satisfaction (*p* = 0.039), with a median Likert RNs' satisfaction with the care provided remained stable during the pandemic (*p* < 0.001; before: median 4, IQR 3–4; during: median 3, IQR 2–4), but younger nurses (aged ≤27 years) tended to judge their care as poor (*p* = 0.047), with a median Likert scale of 2 (node 12, Figure 6). Similarly, nurses with less work experience (≤13 years) reported a decreased quality of care (*p* = 0.032). RNs who reported high preparedness in terms of education/training showed increased care satisfaction (*p* = 0.039), with a median Likert scale of 4 (node 10, Figure 6). *J. Pers. Med.* **2021**, *11*, x FOR PEER REVIEW 9 of 13

**Figure 6.** The conditional inference tree with the RNs' satisfaction with the care provided as dependent variable, reported at the bottom of the figure as boxplot. **Figure 6.** The conditional inference tree with the RNs' satisfaction with the care provided as dependent variable, reported at the bottom of the figure as boxplot.

identified in the conditional inference trees.

The results of sensitivity analyses, performed through logistic regressions, confirmed those obtained from the decision trees. In the logistic regression models, the variables associated with the care provided during the COVID-19 pandemic were the same as those The results of sensitivity analyses, performed through logistic regressions, confirmed those obtained from the decision trees. In the logistic regression models, the variables associated with the care provided during the COVID-19 pandemic were the same as those identified in the conditional inference trees.

### **4. Discussion**

tainment measures.

**4. Discussion** The present study described how nursing care was affected during the first wave of the COVID-19 pandemic and provided indications as to why nurses changed their practices. The first wave of the COVID-19 pandemic hit healthcare systems hard, affecting all The present study described how nursing care was affected during the first wave of the COVID-19 pandemic and provided indications as to why nurses changed their practices. The first wave of the COVID-19 pandemic hit healthcare systems hard, affecting all aspects of nursing care; more nursing techniques were performed to care for patients,

aspects of nursing care; more nursing techniques were performed to care for patients, but the nurse–patient relationship suffered. These changes in nursing care led nurses to report

The impact of the COVID-19 pandemic on nursing care mainly was felt on a clinical level. COVID-19 is a life-threatening illness that requires complex, clinically-focused, and personalized care; many patients were hospitalized for long periods with unpredictable outcomes. The COVID-19 pandemic was influenced by RNs' previous care habits, indicating the real professional commitment demonstrated by nurses, regardless of the circumstances [22]. Although the pandemic caused high levels of stress among nurses due to the high workload and uncertainty [23], nurses endeavored to guarantee the same level of care they provided before the pandemic to each patient. The prioritization of nursing care provided was affected [24], likely due to the unknown trajectory of the disease, the increase in fundamental care due to high patient dependency, the acquisition of new technical skills, and difficulties in communication and patient education due to infection con-

Nurses who were reassigned to COVID-19 units increased the amount of fundamental care provided. A possible explanation could be that the COVID-19 context often requires many specialized skills, and newly-hired nurses may have preferred to perform more fundamental care while they learned these skills. Conversely, reassigned nurses could have increased the amount of fundamental care simply because COVID-19 patients

The physiological needs of COVID-19 patients also required the performance of more nursing techniques. In fact, nurses who cared for COVID-19 patients performed these techniques far more frequently during than before the pandemic. Nurses had to adapt to the needs of these patients, often learning or refreshing techniques such as non-invasive

require more personalized support for their activities of daily living [25].

but the nurse–patient relationship suffered. These changes in nursing care led nurses to report lower satisfaction with the care provided.

The impact of the COVID-19 pandemic on nursing care mainly was felt on a clinical level. COVID-19 is a life-threatening illness that requires complex, clinically-focused, and personalized care; many patients were hospitalized for long periods with unpredictable outcomes. The COVID-19 pandemic was influenced by RNs' previous care habits, indicating the real professional commitment demonstrated by nurses, regardless of the circumstances [22]. Although the pandemic caused high levels of stress among nurses due to the high workload and uncertainty [23], nurses endeavored to guarantee the same level of care they provided before the pandemic to each patient. The prioritization of nursing care provided was affected [24], likely due to the unknown trajectory of the disease, the increase in fundamental care due to high patient dependency, the acquisition of new technical skills, and difficulties in communication and patient education due to infection containment measures.

Nurses who were reassigned to COVID-19 units increased the amount of fundamental care provided. A possible explanation could be that the COVID-19 context often requires many specialized skills, and newly-hired nurses may have preferred to perform more fundamental care while they learned these skills. Conversely, reassigned nurses could have increased the amount of fundamental care simply because COVID-19 patients require more personalized support for their activities of daily living [25].

The physiological needs of COVID-19 patients also required the performance of more nursing techniques. In fact, nurses who cared for COVID-19 patients performed these techniques far more frequently during than before the pandemic. Nurses had to adapt to the needs of these patients, often learning or refreshing techniques such as noninvasive ventilation support and arterial puncture for blood gas analysis [26]. Symptom management was also performed more frequently given the higher number of patients as healthcare systems reached capacity [27].

A biomedical perspective is often predominant in life-threatening situations, but it should be balanced by the value of caring for others and the individualization of care, something on which technical skills should also focus [28]. This is particularly crucial for COVID-19 patients, who have to face an unknown disease without the support of their loved ones and likely without a close relationship with healthcare providers. Education is the most commonly omitted task when resources are strained [29], and nurse–patient relationships and patient education in our study were reduced due to infection prevention measures [30]. Despite these problems, nurses tried to invent new ways to interact with patients, such as drawing on gowns, printing their pictures to show their faces, and writing their names on face shields [31].

Another relevant finding was the lower satisfaction with care provided among RNs from areas of high COVID-19 prevalence and those who worked in high-intensity units. This may be due to the lack of therapeutic solutions and the patients' reduced chance of recovery, especially the elderly, making nurses feel less confident that the care they provided was adequate [32]. Nurses working in primary care were also less satisfied with the care provided, probably due to the exceptional number of patients that were treated at home and in other community settings [33].

The devastating, rapid impact that the pandemic has had on the Italian healthcare system might explain some of the lack of preparedness reported by our respondents. During the first wave, COVID-19 was an unknown disease, and a trial and error approach was often applied, with frequent changes in therapies, use of ventilation, and supination [34]; thus, nurses had to adapt their daily work to rapidly changing guidelines and protocols, individualizing the care they provided. Moreover, in Italy, newly-graduated nurses were hired to work in new community care units specifically dedicated to COVID-19 patients, which aimed to identify those who required medical assistance and those who could stay isolated at home with telephone follow-ups [35]. More experienced community care nurses were also recruited for these units, which led to a decreased availability of nurses for

established home care services. These factors could have influenced the preparedness of nurses working in community settings.

The pandemic globally highlighted the nursing shortage and the lack of adequate, coordinated management responses to population health crises [36]. However, contextual factors must be considered when evaluating care outcomes, which should not be merely interpreted quantitatively. In this regard, the patient-to-nurse ratio should not be the only index collected to evaluate staffing needs but, perhaps, environmental and organizational factors should also be considered [37]. Healthcare systems should maintain, and be prepared to implement, plans for pandemic events, and hospital managers should draw up specific emergency plans that define the human resources required in case of a long-term pandemic with recurrent waves, based on competencies, skill mix (e.g., of experienced and novice nurses), and job rotations. Nurse staffing should be proportionate to the clinical complexity of patients and to the organizational complexities to individualize care, minimizing the risk of missed care and patient death [38]. These actions would improve the healthcare system's response and alleviate the stress imposed by crises, although maintaining the highest standard of nursing care delivered.

The present study has a number of limitations. The majority of the respondents were from the areas most affected by the pandemic in Italy, and most worked in hospital settings. Moreover, our sample could not completely represent the Italian nursing population, which consists of around 367,000 individuals. This may have produced some response bias as nurses from hospital settings may have felt more implicated in the pandemic and, thus, may have been more interested in the survey. Moreover, considering that data collected were self-reported, findings may be affected by the respondents' emotional or physical condition. High negative and low positive affect have been associated with an emotional autobiographical memory [39]; in the emotionally-charged pandemic, recall may also have influenced our results. Our findings should be interpreted considering that data were collected during the first wave in Italy, the first European country hit by the pandemic, and the first country that had to reorganize its healthcare system to respond to the emergency. Moreover, the use of a cross-sectional design means that causality cannot be proven. However, the decision tree analysis used, and the sensitivity analyses performed, increased the confidence in the inferences.

### **5. Conclusions**

This study highlighted the impact of the COVID-19 pandemic on nursing care and the differences between nurses who were and were not reassigned. Despite all of the difficulties faced by nurses, they were generally satisfied with the care provided, except for younger nurses and those with limited working experience. Furthermore, in spite of the healthcare reorganization, and the need to educate, prioritize, and individualize their activities to meet the needs of patients with complex clinical conditions, nurses continued to provide a high level of care, individualizing their practices and ensuring the highest quality of care. Nurses who felt more prepared in terms of education and training were more satisfied with the care provided, and an increase in the number of patients decreased the frequency of fundamental care and nurse–patient relationships. Nurses caring for COVID-19 patients performed nursing techniques more often, to the detriment of patient education. Ensuring the appropriate deployment and education of nurses is crucial to personalize care, especially during a pandemic, and to maintain nurses' satisfaction with the care provided. Policy makers should consider these results to create structured plans to address long-term pandemics and ensure appropriate nurse staffing in hospitals and primary care settings.

**Author Contributions:** Conceptualization, M.C., M.L. and A.C.; methodology, M.C., M.L. and A.C.; validation, V.S. and P.B.; formal analysis, V.S. and P.B.; investigation, V.D.; data curation, M.C., M.L. and A.C.; writing—original draft preparation, M.C., M.L., A.C. and V.S.; writing—review and editing, P.B., P.D.G., S.C. and V.D.; supervision, P.B., P.D.G., S.C. and V.D.; project administration, M.C. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethics Committee of the University of Torino (Approval no. 279061–01/07/2020).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author. The data are not publicly available due to privacy reason.

**Acknowledgments:** We would like to thank all of the nurses who completed the survey in such difficult times.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


### *Article* **Kinesio Taping vs. Auricular Acupressure for the Personalised Treatment of Primary Dysmenorrhoea: A Pilot Randomized Controlled Trial**

**Elena Mejías-Gil <sup>1</sup> , Elisa María Garrido-Ardila 1,\* , Jesús Montanero-Fernández <sup>2</sup> , María Jiménez-Palomares <sup>1</sup> , Juan Rodríguez-Mansilla 1,\* and María Victoria González López-Arza <sup>1</sup>**

> <sup>1</sup> ADOLOR Research Group, Department of Medical-Surgical Therapy, Faculty of Medicine and Health Sciences, Extremadura University, 06006 Badajoz, Spain; elenamejias92@gmail.com (E.M.-G.); mariajp@unex.es (M.J.-P.); mvglez@unex.es (M.V.G.L.-A.)

**Abstract:** Background: Dysmenorrhoea is the medical term for menstrual pain. The World Health


**Citation:** Mejías-Gil, E.; Garrido-Ardila, E.M.; Montanero-Fernández, J.; Jiménez-Palomares, M.; Rodríguez-Mansilla, J.; González López-Arza, M.V. Kinesio Taping vs. Auricular Acupressure for the Personalised Treatment of Primary Dysmenorrhoea: A Pilot Randomized Controlled Trial. *J. Pers. Med.* **2021**, *11*, 809. https://doi.org/10.3390/ jpm11080809

Academic Editors: Riitta Suhonen, Minna Stolt and David Edvardsson

Received: 26 July 2021 Accepted: 17 August 2021 Published: 19 August 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

Organization estimates that up to 81% of women of childbearing age are affected by this condition, and it is one of the leading causes of absenteeism from work and school among women. Although there are pharmacological treatments available for menstrual-pain relief, they do not respond to all women's needs. Therefore, there is a need to study and develop non-pharmacological alternatives to broaden the individualised treatment options for dysmenorrhea. There are scarce studies published on non-pharmacological treatments, such as kinesio tape and auricular acupressure for the relief of menstrual pain, but the scientific evidence available suggest that these techniques may be beneficial in addressing this problem. The objective of this pilot study was to assess and compare the effectiveness of kinesio tape and auricular acupressure to decrease pain and drug intake in women with primary dysmenorrhoea. Methods: This was a double-blind randomized clinical controlled trial. The period of study was from September 2017 to August 2018. Women enrolled in the University of Extremadura and who had primary dysmenorrhoea were randomized to five groups: control (n = 23), kinesio tape (n = 23), placebo kinesio tape (n = 23), auricular acupressure (n = 23) and placebo auricular acupressure (n = 22). Measures were taken during the pretreatment phase (at four menstrual cycles), during the post-intervention phase (at four menstrual cycles) and during the follow-up phase (at the first and third menstrual cycles after the treatment was completed). The primary outcome measures were mean pain intensity, maximum pain intensity, number of painful days and dose of drug intake during menstruation, measured with the Visual Analogue Scale. The secondary outcome measures were the length of the cycle, the length of menstruation, the drug intake and the type of drug. Results: In all, 108 participants completed the study. The statistical analysis (MANOVA, ANOVA, t-paired and McNemar tests) showed that kinesio tape and auricular acupressure have a beneficial effect on pain relief (mean pain intensity, *p* < 0.001; maximum pain intensity, *p* < 0.001; number of painful days, *p* = 0.021; dose of drug intake, *p* < 0.001). In addition, once the treatments were withdrawn, the auricular-acupressure group maintained lower scores during the first follow-up cycle (*p* < 0.001). Conclusions: Kinesio tape and auricular acupressure decrease pain and drug intake in women with primary dysmenorrhoea. The changes in the auricular-acupressure group seemed to last longer. The results suggest that these techniques could be used as complementary personalised therapies to the pharmacological treatment and not as a substitution.

**Keywords:** dysmenorrhea; kinesio tape; auricular acupressure; pain

### **1. Introduction**

Dysmenorrhoea is the medical term for menstrual pain. It is a cramping sensation felt in the supra pubic area of the abdomen and can be accompanied by muscle pain, headache and nausea [1,2]. It is classified as primary or secondary according to the absence or presence of underlying pathologies that trigger the pain respectively [2,3].

The World Health Organization estimates that 81% of the female population can be affected by this condition. However, the percentage can vary in different countries [4]. This implies that more than half of the female population of childbearing age experiences this type of pain at least once a month. It has been observed that dysmenorrhoea has a negative impact on the academic, work, sport and social life of women, who can also see their quality of life affected [2,4,5].

Although there are pharmacological treatments available for menstrual-pain relief, they do not respond to all women's needs. Allergies to drugs, lack of effectiveness or refusal to use medication (for personal, religious or social reasons) leave part of these women with no effective treatment for their symptoms [2,6–11].

Therefore, non-pharmacological treatments which provide person-centred and individualised care, such as those proposed in Physiotherapy and Traditional Chinese Medicine, can be very useful. Many publications have focused their studies on the relief of various types of pain through the use of non-pharmacological therapies, such as kinesio tape, thermotherapy, electrotherapy, massage, acupuncture and auricular acupressure obtaining encouraging results [12–16]. In particular, kinesio taping and auricular acupressure are treatment techniques that have been used for a long time in the management of pain from different causes. However, there is very little scientific evidence of the effectiveness of these techniques for primary dysmenorrhoea. Four studies that analyzed the use of kinesio taping in primary dysmenorrhoea can be found in the literature [17–20]. Their results suggested an improvement of menstrual pain. Similarly, the three articles that studies the effectiveness of auricular acupressure in primary dysmenorrhoea showed a pain-relief effect [21–23]. These results would point towards the possible benefits of the use of these non-pharmacological treatment for the management of dysmenorrhoea.

The application of kinesio taping is based on its proprioceptive and skin receptor stimulation effects. When applied correctly, it influences muscle tone and can induce muscle relaxation. In addition, part of the analgesic effect of this technique is based on its ability to decrease interstitial pressure [24,25]. This causes a reduction of the stimulus received by nociceptors and normalizes local blood and lymphatic circulation, thus also eliminating mediators of pain and inflammation [24,25]. Moreover, the literature also suggests the use of mechanisms related to the Gate Control theory, whereby a tactile sensory stimulus interferes with the perception of pain intensity [26].

Although the mechanism of action of auricular acupressure is still under study, the research conducted by using functional magnetic resonance imaging (fMRI) and positron emission tomography (PET) has revealed the presence of brain activity in the areas corresponding to the structure represented by the point on which the stimulus is applied [27–29]. This stimulus has also been shown to trigger specific responses in brain regions related to pain inhibition and to influence the release of endorphins, melatonin and serotonin [27,28].

The objective of this pilot study was to assess the effectiveness of kinesio taping and auricular acupressure improving pain and decreasing drug intake in women with primary dysmenorrhoea, comparing both treatment approaches between them.

### **2. Materials and Methods**

### *2.1. Study Design*

This pilot study was a single-blind randomized clinical controlled trial. The study took place within the University of Extremadura (Spain), in an outpatient setting. The period of study was from September 2017 to August 2018. The study protocol was approved by the Bioethical Commission of the University of Extremadura in Spain (registration number: 58/2017). The trial was registered with the ClinicalTrials.gov registry (Study

Identifier: NCT04400968). All participants signed a written informed consent. The data were guaranteed to be protected and anonymous. The CONSORT statements were used to conduct and report the trial.

### *2.2. Participants and Procedures*

The target population was women enrolled in the University of Extremadura who had primary dysmenorrhoea. Participants were recruited in September 2017. The inclusion criteria were women between 18 and 30 years old affected by primary dysmenorrhoea grade 2 and 3 of Andersch and Milsom classification [30], to have attended gynecologist consultation for a general revision in the last 2 years, to have menstrual pain, to have regular menstrual cycles of 21 to 38 days, and to not have an intrauterine device or to be on oral contraceptive treatment. The exclusion criteria were to have been diagnosed with a condition that could influence menstrual-pain perception and to know or have been previously treated with techniques used in the interventions and pregnancy.

In the first place, an interview was conducted at the begging of the academic year (2017/2018) in order to select the sample. Once the participants were recruited, they were assigned an alphanumeric identifier and randomized into the five study groups, using the SPSS statistical program. The program selected the participants to be included in each group equally. Since the sample was not fully divisible among the five groups, the order of preference to complete the groups was allocated by randomly assigning in a ranking from 1 to 5.

Then a pretreatment phase of 4 menstrual cycles started. During this period, the participants completed the questionnaires at their home to collect information regarding the symptoms experienced in each menstrual cycle. The questionnaires were codified by an identifier number assigned to each participant to ensure masking of identity and group allocation and, therefore, to ensure blinding of the data-collection process and analysis.

Once the pretreatment phase was completed, a four menstrual cycles treatment phase commenced. During this phase the participants continued with the same protocol to collect data and received the treatments assigned to their group. After each cycle finished, the treatments were discontinued, and the follow-up phases started. In the first and second follow-up phase, the data corresponding to the first and the third cycle after the treatments were finished was collected respectively.

Due to the nature of the treatments, participants knew whether they belonged to one of the kinesio-taping or auricular-acupressure groups. However, they did not know whether the technique applied was the placebo or the real one. The therapist in charge of applying the treatments could not be blinded in order to apply the treatments correctly.

The primary outcome measures were four in total: Mean pain intensity for the 3 first days of menstruation, maximum pain intensity, number of painful days and dose of drug intake. Pain intensity was measured with the Visual Analogue Scale (VAS). This scale measure pain in a scale from 0 to 10, where 0 means no pain and 10 means maximum and excruciating pain [31]. All outcome measures were registered at the initial interview and every day during the bleeding period in the pretreatment, treatment and follow-up phases. As the pretreatment and the treatment phases consisted of four periods each, there were ten measures in all.

The secondary outcome measures included the length of the cycle, the length of menstruation, the drug intake and the type of drug. The epidemiological data of the sample (age, body height, body weight, age of menarche and age of first pain) were collected at the baseline measurement.

The sample consisted of 114 participants who were randomly allocated to a control group, a kinesio-tape group, a placebo kinesio-tape group, an auricular-acupressure group and a placebo auricular-acupressure group (Figure 1). Details of the intervention following the Template for Intervention Description and Replication (TIDieR) [32] guidelines are provided in Supplemental Materials Table S1.

lines are provided in Supplemental Materials Table S1.

**Figure 1.** CONSORT flow diagram. **Figure 1.** CONSORT flow diagram.

The therapy and the placebo treatments were always placed within 4 h from the beginning of the menstrual cycle and were maintained during 72 h. When the adhesive tapes lost fixation and became detached, the treatment material was replaced as soon as possible (never later than 2 h from detachment). The cases that could not receive the treatments on the scheduled time were excluded. The therapy and the placebo treatments were always placed within 4 h from the beginning of the menstrual cycle and were maintained during 72 h. When the adhesive tapes lost fixation and became detached, the treatment material was replaced as soon as possible (never later than 2 h from detachment). The cases that could not receive the treatments on the scheduled time were excluded.

following the Template for Intervention Description and Replication (TIDieR) [32] guide-

The participants received the necessary information in order to correctly maintain the tapes and adhesives. All groups received the same information, not differentiating between treatment and placebo groups, to ensure blinding. All participants committed to maintain secret their experiences and not to comment with the rest of women during the study period. The participants received the necessary information in order to correctly maintain the tapes and adhesives. All groups received the same information, not differentiating between treatment and placebo groups, to ensure blinding. All participants committed to maintain secret their experiences and not to comment with the rest of women during the study period.

As it was considered unethical, the drug intake was not forbidden if pain relief was necessary. However, all subjects were asked to delay the intake until the appearance of symptoms. In order to control the influence of the medication, the questionnaires included sections where the intake was registered. As it was considered unethical, the drug intake was not forbidden if pain relief was necessary. However, all subjects were asked to delay the intake until the appearance of symptoms. In order to control the influence of the medication, the questionnaires included sections where the intake was registered.

The control group did not receive any treatment. However, the controls completed all the questionnaires to collect the information regarding their symptoms in order to observe their progress with no intervention.

### *2.3. Statistical Analysis*

The obtained data were analyzed through the IBM SPSS Statistics 22.0 version (Statistical Package for the Social Sciences). A descriptive analysis of all the outcome measures was performed. Changes on drug intake were analyzed by the McNemar test. In order to assess the primary outcome measures, we distinguished between two steps. Firstly, for each single outcome we applied a one-way multivariate ANOVA considering the ten measurements and the five groups, so that we could contrast the existence of global influence of the treatment. Moreover, drug intake was included as a second factor in the model in order to detect a possible bias due to this circumstance. In the following step, we focused on analysing just the average between the four periods considered during the pretreatment and during treatment stages. This way, the number of phases got reduced to four (average pretreatment, average treatment, follow-up 1 and follow-up 2) for each primary outcome measures. Then, a more exhaustive analysis was performed for each one. On the one hand, we assessed the inter-group differences at each phase by a one-way ANOVA F test or Kruskal–Wallis's H test (depending on the level of skewness or symmetry of the distribution of the different outcomes). On the other hand, we analyzed separately the within-group changes of the outcome measures during the intervention as compared with the pretreatment scores by a t-paired test or Wilcoxon's W test.

The sample size does not respond to a previous calculation. However, as many participants as possible were first recruited and posteriorly randomly allocated into five equal groups of 25 women. Finally, according to Cohen's d, the groups sample size lead to a statistical power of 80% for a 2-sided level 5% *t*-test to detect an effect size of 0.8.

### **3. Results**

Figure 1 shows the study flow diagram. Except for the auricular-acupressure placebo group, which had no loses, all groups had similar drop-outs rates. This was 5.6% of the total initial sample.

The average duration of the first cycle was 30.26 ± 0.31 days, with 5.63 ± 1.05 days of menstruation. These values hardly changed along the ten menstruations that were followed up. All the participants reported regular drug intake for pain relief. In fact, 88% of women took medication during the first period previous to the application of the treatments. Nevertheless, we observed that this proportion decreased along the treatment phase to 62% during the last menstruation (*p* < 0.001), since 33 participants stopped taking medication. Most of them (namely 29) belonged to the auricular-acupressure and the kinesio-tape groups. This percentage increased progressively after the treatment during the follow-up phases.

Tables 1 and 2 include descriptive statistics of the rest of secondary outcomes and the primary ones, respectively.


**Table 1.** Epidemiological data and secondary outcome measures.

CG, control group; KT, kinesio-tape group; KT-P, kinesio-tape placebo group; AP, auricular-acupressure group; AP-P, auricular-acupressure placebo group; AM, age of menarche; AFP, age of first menstrual pain; DC, duration of the cycles; DM, duration of menstruation.


**Table 2.** Changes (mean ± SD) of the primary outcome measures by groups along the four phases. SD: standard deviation.

P-T, pretreatment phase; T, treatment phase; FU-1, follow-up 1 phase; FU-2, follow-up 2 phase; CG, control group; KT, kinesio-tape group; KT-P, kinesio-tape placebo group; AP, auricular-acupressure group; AP-P, auricularacupressure placebo group. Post-hoc results are expressed by letters as follow: for each column, there are significant differences between groups whose letter is different.

A one-way multivariate ANOVA showed the overall changes and influence of the treatment for all the primary outcome measures (*p* < 0.001 in the four of them). When including drug intake as the secondary factor in the model, neither interaction (*p* > 0.05) nor drug intake (*p* < 0.05) was significant. Therefore, there was no evidence of bias due to drug intake. Table 2 also includes the evolution of the primary outcome measures during the study in the following way: between-group comparisons were performed by a one way ANOVA F test or by Kruskal–Wallis's H test, and their *p*-values are shown in the table. Post hoc results are expressed by letters, as usual. Pretreatment within-group comparisons were performed by t-paired or Wilcoxon test. A mean value is marked in bold if it implies a significant improvement in relation to both pretreatment (within-group) and the control group, the placebo auricular-acupressure group and the placebo kinesio-tape group (between-group).

From the analysis of Table 2, firstly, it can be observed that baseline values of the four primary outcome measures were similar for the five experimental groups (second column of Table 2), as expected. Secondly, the most remarkable is the fact that the auricularacupressure group achieved the best results in mean, both during the treatment and followup one phases, for all the primary outcome measures (mean pain intensity, maximum pain intensity, number of painful days and dose of drug intake). In a deeper analysis and taking into account within-group comparisons, these results showed significant improvements in comparison with the pretreatment phase. Regarding the between-group comparison, we observed that, for the mean pain and maximum pain intensity, the auricular-acupressure group achieved significant improvements during both phases (treatment and follow-up one) in relation to the rest of the groups. In the kinesio-tape group, the results obtained from the treatment phase were similar to those from the auricular-acupressure group. During the first follow-up phase, the auricular-acupressure group performed significantly better.

In addition, we can also say that, regarding the dose of drug intake, the auricularacupressure and kinesio-tape groups performed significantly better than the control group and both placebo groups. With regard to the number of painful days, the auricularacupressure group achieved results significantly better than the control group at the treatment phase and significantly better than the kinesio-tape group at the first followup phase.

### **4. Discussion**

This pilot study contributes to the scientific literature with the evidence of the effectiveness of two non-pharmacological treatment approaches, kinesio tape and auricular acupressure for the individualised and person-centred management of primary dysmenorrhoea. In addition, it shows, for the first time, a comparative assessment of both techniques.

Our results add new data from a sample of over 100 women to the few studies previously conducted that analyze the effectiveness of these techniques. In addition, the pre- and post-treatment comparisons' accuracy of our study was improved by the extended period of time used to observe the symptoms in the pretreatment phase. Another strength that of the present clinical trial is the presence of a control group and a placebo group for each treatment technique. Participants were blind at all times in relation to the group they were included in (treatment or placebo) and during the data collection.

Only four previous studies that applied kinesio tape for menstrual-pain relief were found in the literature [17–20]. In relation to the application of auricular acupressure in this condition, three studies were the result of the literature review [21–23]. In comparison to these studies, it can be observed that our study has a longer period of intervention.

Our results regarding the mean and the maximum pain-intensity levels in the kinesiotape group and the auricular-acupressure group coincide with the improvements found by different authors that applied these techniques [17–23]. We were not able to contrast our results in relation to the number of painful days, days of the menstrual cycle and days of the menstruation, as there were no studies found in the literature that analyzed those variables. Although there were no statistically significant changes in the number of days of the menstrual cycle and menstruation after the treatments, we consider that the observation of these variables is important. This is because these variables are risk factors for dysmenorrhoea and can influence the results of the interventions [4].

The scientific evidence available on other non-pharmacological therapies show a great diversity of techniques used for the management of primary dysmenorrhoea, especially related to the field of Physiotherapy and Traditional Chinese Medicine. These include, for example, thermotherapy, massage therapy, electrotherapy, spinal manipulation, Kegel exercises, acupuncture and moxibustion in all its forms (acupressure, electro-acupuncture, laser acupuncture, etc.) [33,34]. However, as in the specific cases of kinesio taping and auricular acupressure, the number of studies published that analyze the effectiveness of these non-pharmacological techniques applied to primary dysmenorrhoea is low. All of these techniques show encouraging results, but, in general, there has been little research on their application to menstrual pain [2,15,33–39]. A remarkable advantage of kinesio taping and auricular acupressure over the other non-pharmacological techniques mentioned is that they are low-cost techniques, self-applicable after proper training and simple and quick to apply. In addition, the patient can keep the tape or the seeds on and continue with

her daily routine, without having to travel to a clinic or invest too much time and money and, therefore, without affecting her rhythm of life.

Although we found no evidence of bias in the pilot study in relation to drug intake, we are not in a position to assess the opposite. Indeed, since most participants took medication during the study, the sample size was big enough for this ambitious statistical task. Nearly 10% of the participants stopped taking medication during all the treatment phase. The dose of drug intake was significantly reduced in the kinesio-tape group (kinesio tape–control = −0.9; kinesio tape–kinesio-tape placebo = −0.81), as well as in the auricular acupressure group (auricular acupressure–control = −1.19; auricular acupressure–auricular ac-pressure placebo = −1.36). The study conducted by Tomás-Rodríguez et al. in 2015 is the only study found that analyzed the drug intake. They assessed the between-group differences and found a difference of 1,09 drug units between the kinesio tape and the placebo groups. Asher et al., in 2010 [40], also found that there are studies on auricular acupressure that have shown a decrease on drug intake for pain relief in other conditions, such as surgery or chronic pain.

This decrease in drug consumption coincides with the improvement of pain levels and suggests that both treatment approaches could be beneficial for the management of primary dysmenorrhoea.

The comparisons between the kinesio-tape and the auricular-acupressure groups did not show statistically significant differences during the treatment phase. Nevertheless, when comparing these groups during the follow-up phases, the auricular-acupressure group maintained the improvements achieved during a longer period. These results could not be contrasted, as our study is the first one comparing these treatment techniques. When auricular acupressure has been used for the treatment of other conditions, such as menstrual headache, the results have revealed a reduction of plasma arginine vasopressin and prostaglandine F2α [41]. The most recent research suggests that the cause of primary dysmenorrhoea is the excess of endometrial prostaglandins E2 and F2α, which increase uterine contractions and painful sensation [1]. The longer duration of pain relief maintained by the auricular-acupressure group could be justified by the decrease of prostaglandins F2α and the secretion of pain-inhibiting agents that, according to Wu et al. (2007) [22] and Alimi et al. (2002) [28], this technique achieves.

The results of the present pilot study can have important implications in the clinical practice. Our data show that a kinesio taping and auricular acupressure decrease pain and drug intake in women with dysmenorrhoea. They are two techniques that are increasingly used by practitioners these days and can be performed safely with the appropriate training. This condition, and, in particular, the pain that is associated with it, has an important negative impact on the quality of life of women [2,5] which could be minimized with the use of the treatment approaches described in this study, as a complement to their pharmacological treatment.

### *Study Limitations*

The main limitation of our study was the impossibility to blind the therapist that applied the techniques. Due to the nature of the treatment, they could clearly see which group the participant was allocated to.

Furthermore, we consider that it was important to analyze whether the drug intake and the type of medicine used for pain relief could influence these results. In order to assess this possible effect, a two-way multivariate ANOVA, considering as factors drug intake and type of drug, was carried out for each outcome measure. Although there were no significant results, neither for interaction nor for main effects, we cannot dismiss the possible influence. We do not have enough evidence to make a fair decision, since the number of participants that did not take any medicine was so small. In order to analyze the problem at that level, we consider that a bigger sample size would be needed.

### **5. Conclusions**

Based on the results obtained in this pilot study, we can conclude that kinesio taping and auricular acupressure have a beneficial effect on pain relief in women with primary dysmenorrhoea. Although both groups showed similar improvements, the changes on the auricular-acupressure group seemed to last longer. The pain relief obtained by both treatment approaches suggests that these techniques could be used as complementary personalised therapies to the pharmacological treatment and not as a substitution. The participants of the kinesio taping and the auricular-acupressure group experienced a decrease in drug intake.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/10 .3390/jpm11080809/s1. Table S1: Descriptions of the interventions conducted following the Template for Intervention Description and Replication (TIDieR).

**Author Contributions:** Conceptualization, M.V.G.L.-A., E.M.-G. and J.M.-F.; methodology, E.M.-G., J.M.-F. and M.V.G.L.-A.; formal analysis, E.M.-G., M.V.G.L.-A. and J.M.-F.; investigation, E.M.-G. and M.V.G.L.-A.; writing—original draft preparation, M.V.G.L.-A., E.M.-G., E.M.G.-A. and J.R.-M.; writing—review and editing, E.M.-G., J.R.-M., M.V.G.L.-A., E.M.G.-A., J.M.-F. and J.R.-M. visualization, M.J.-P., E.M.-G., M.V.G.L.-A., E.M.G.-A., J.M.-F. and J.R.-M.; supervision E.M.-G., M.V.G.L.-A., E.M.G.-A., J.M.-F., M.J.-P. and J.R.-M. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Institutional Ethics Committee of the University of Extremadura (protocol code 58/2017 and date of approval: 7th of July 2017).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** The data underlying this article cannot be shared publicly in order to maintain the privacy of individuals that participated in the study. The data will be shared upon reasonable request to the corresponding author.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


**Cheryl Lin <sup>1</sup> , Rungting Tu <sup>2</sup> , Brooke Bier <sup>1</sup> and Pikuei Tu 1,\***


**Abstract:** Rheumatoid arthritis (RA) patients face psychological hardship due to physical discomfort, disabilities, and anxieties. Previous research indicated a bidirectional relationship and patient desire for emotional support from providers. This study examined lesser-understood RA experiences across the psychological and social contexts in relation to self-perception through the patients' expression of their struggles with these burdens. We conducted four semistructured focus groups and eleven interviews (total *n* = 31). A codebook was developed and refined through iterative transcript coding via NVivo-12. Four emerging themes were identified by inductive, thematic analysis: (1) the patients' healthy appearances were a myth, with subthemes revealing a conflict between an inclination to hide the disease and a desire for validation, while feeling embarrassed by symptom manifestations and disappointment at withdrawal from social interactions; (2) an identity crisis due to diminished functionality, autonomy, and sense of self; (3) RA constantly occupied the mind, as its unpredictability dictated daily schedules and altered plans; and (4) the disease's chronic nature influenced personal outlook to worry about or accept the uncertainty. Even with effective treatment, the invisibility of the disease, the fear and anticipation of flare-ups, and identity clashes caused emotional distress. The insights offer a different perspective on personalized medicine, complementing clinical treatments based on genetic or biomarker profile. For patient-centered holistic care, education is needed to prompt both patients and providers to discuss psychological issues for more customized, integrated interventions. The findings can help inform healthcare teams and families in recognizing and supporting these physical-psychological intertwined experiences, thereby ameliorating patients' wellbeing.

**Keywords:** patient-centered; personalized care; arthritis; autoimmune disease; sociopsychological factors; emotion; depression; self-identity; qualitative study; observational study

**Citation:** Lin, C.; Tu, R.; Bier, B.; Tu, P. Uncovering the Imprints of Chronic Disease on Patients' Lives and Self-Perceptions. *J. Pers. Med.* **2021**, *11*, 807. https://doi.org/10.3390/ jpm11080807

Academic Editors: Riitta Suhonen, Minna Stolt and David Edvardsson

Received: 8 July 2021 Accepted: 16 August 2021 Published: 18 August 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

### **1. Introduction**

Rheumatoid arthritis (RA) is a chronic, systematic inflammatory condition characterized by persistent and progressive joint and autoimmune dysfunction. It is estimated that up to 2% of the global population and about 1.5 million Americans suffer from RA; most are women with an age of onset between 40 and 60 years old [1,2]. The etiology of RA is relatively unknown. However, the resulting prolonged inflammation typically causes joint deformation, stiffness, excessive fatigue, as well as widespread and intense pain [3]. In addition, most patients face psychological hardship due to RA-imposed functional limitations and anxieties [4–6].

Numerous studies have described a bidirectional relationship between RA disease activities and poor mental functioning. The progression of RA often leads to feelings of helplessness, grief, and uncertainty [6–8]. This psychological distress further contributes to the intensification of pain and thus creates a vicious cycle of physical and emotional suffering [9–11]. RA patients have consistently reported a loss of enjoyable activities, a

struggle to find and adhere to effective treatment, and difficulties in actively dealing with disease manifestations, implicating a 14–62% prevalence rate of depression in the RA community [6,12–15]. Moreover, illness-related shame and loneliness perpetuate social withdrawal, and the resulting lack of social support could lead to aggravated disease outcomes [16–18]. Recent studies have also discussed the impacts of the disease on self-esteem and outlook; patients express the loss of autonomy and low self-efficacy in maintaining a sense of independence due to physical disabilities [4,14,19].

These findings underscore the importance of a deeper understanding of the psychological nuances of RA's personal impacts. Previous research has explored such experiences to help plan treatments, but often focused on improving quality of life and coping strategies [14,15,20]. Other studies have deliberated how patient-centered care can complement clinical disease management for a more holistic approach [21,22]. Patient-centeredness encompasses a bio-psychosocial perspective to provide attentive, individualized care with emotional support and augmented communication for the patients' overall wellbeing [23,24]. While emerging personalized medicine emphasizes the individuals' distinctive genetic or biomarker profile to inform prevention and customize treatment [25,26], the patients' mental and contextual states are also critical in determining their response to medications and health outcomes.

Few existing studies have comprehensively explored the intricacies of these experiences faced by RA sufferers on a daily or long-term basis, or investigated personalized intervention techniques addressing such emotional and social struggles. It is imperative to hear and respect patients' voices and incorporate multidimensional therapies targeting biological and psychological pains concordantly [27–29]. Research in rheumatology and other medical disciplines have predominately evaluated the patients' conditions quantitatively [30,31]. This study utilized qualitative methods to uncover and synthesize lesser-understood psychological burdens, whether from or in combination with physical encumbrances. We captured the patients' narratives in their own words and examined the complex existence of RA in relation to the patients' self-perceptions and outlook in personal and social contexts. By identifying and analyzing novel domains of the impact of RA on lives and individual identities, these experiences and sentiments can be better recognized and legitimized. The findings also provide insights for more effective, personalized psychological interventions and support.

### **2. Methods**

We conducted four focus groups (*n* = 20) and eleven individual interviews to elicit participants' descriptions of RA's presence in and influence on their lives (total *n* = 31 in 15 sessions). Eligibility criteria for participating in the study were age 18 years or older, clinically diagnosed with RA, and have received treatment for RA in the previous year. Focus-group participants were recruited and screened by a market research agency to ensure a diverse sample; two experienced moderators facilitated the four groups. Individual interviewees were recruited through flyers posted at local clinics, then screened and interviewed by the research team.

A semistructured discussion guide was constructed, incorporating inputs from the literature [6,30,32–35] and experts, and was used across all sessions. The discussions explored participants' history of diagnosis, hobbies, what a good day or bad day was like, awareness of and sentiments towards RA's impact or associated changes, sources and specifics of their emotional state, struggles with and strategies for managing the disease, and interactions with physicians and families. Participants were also asked to self-rate the extent to which their lives were impacted by RA on a scale of one to seven, one being "no impact" and seven being "extreme impact".

Group and individual interviews were recorded, transcribed, and de-identified. We uploaded text data into NVivo software version 12 for coding and analysis, utilizing an inductive, thematic approach [36]. A codebook was developed and refined through iterative coding by multiple researchers. Common experiences and sentiments emerging from the data were grouped together for potential theme categorization. Disagreements regarding coding and identification or naming of themes were discussed and resolved through consensus. Selective quotes were extracted to represent the essence of each theme in the participants' own words. The research protocol was approved by Duke University Institutional Review Board. Informed consents were obtained from individual participants before each session.

### **3. Results**

### *3.1. Participant Characteristics*

A total of 31 people were interviewed in groups or individually. Participants were heterogenous in terms of demographics (mean age 47.39, SD = 14.38; 77.42% female; 38.7% minorities; Table 1) and disease-related characteristics (mean disease duration: 11.76 years, SD = 8.71; mean RA impact: 4.29 on a scale of 1 to 7, SD = 1.19). The occurrences and intricacies of RA's impact were obtained via qualitative inquiries, described narratively and illustrated in a diagram below.

**Table 1.** Participant self-reported demographic and disease-related information.

