**Preface to "Nursing and Society"**

Papers in this book from the Special Issue of the International Journal of Environmental Research and Public Health will explore the contribution of nurses to society. The year 2020 is considered by the World Health Organization as the International Year of the Nurse and Midwife. This book serves as an incentive to support the visibility of the contribution of nurses to society, emphasizing the crucial role of nurses in achieving greater attention to community health and the social and environmental determinants that intervene in the health of the population. Nurses are the heart of health teams. They play an essential role in developing new community care models, health promotion, disease prevention, and treatment.

> **Elena Fern´andez-Mart´ınez, Lisa Alves Gomes, Cristina Li´ebana-Presa** *Editors*

### *Article* **Heart Disease, Now What? Improving Quality of Life through Education**

**Lisa Gomes <sup>1</sup> , Cristina Liébana-Presa 2,\* , Beatriz Araújo <sup>3</sup> , Fátima Marques <sup>4</sup> and Elena Fernández-Martínez <sup>5</sup>**


**Abstract:** Introduction: The management of chronic illness assumes a level of demand for permanent care and reaches a priority dimension in the health context. Given the importance of nursing care to post-acute coronary syndrome patients, the objective of this study is to evaluate the impact of an educational intervention program on quality of life in patients after acute coronary syndrome. Method: Quasi-experimental study with two groups: an experimental group exposed to the educational intervention program and the control group without exposure to the educational intervention program. Results: The results showed statistically significant differences between both groups (*p* < 0.001). Although only valid for the specific group of subjects studied, the educational intervention program enabled significant gains in quality of life. Conclusions: According to the findings of the study, a systematized and structured educational program, integrated into the care organization and based on transition processes, is effective in developing self-care skills and improves the quality of life in patients after acute coronary syndrome.

**Keywords:** coronary disease; cardiac rehabilitation; health education; quality of life; self-care

### **1. Introduction**

Cardiovascular diseases (CVD) are the main cause of death [1,2] decade after decade and self-care is not at the top of the pyramid of best practices for chronic disease management.

Changes in the population's health/disease patterns pressure worldwide health care systems' sustainability. These changes include an aging population, an increasing number of individuals with CVD risk factors, and the prevalence of multimorbidity [3]. The burden of heart disease in Europe indicates that we are presently far from achieving the expected success [4]. Health care systems believe that people who seek and need health care will obey the recommendations proposed by professionals.

Patients with heart disease have needs in various dimensions of their lives that require continuous health care, which the current systems have difficulties in satisfying. In the health–disease transition, rehabilitation depends largely on health literacy, adherence to rehabilitation programs, and the patient's active participation in the management of their therapeutic regime [5]. Skills and ability development, behavior, and lifestyle changing is a challenge given the complex treatment regimes.

In the specific case of acute coronary syndrome (ACS), disease control requires a rigorous and long-lasting therapeutic plan. There is a need for patient and family involvement

**Citation:** Gomes, L.; Liébana-Presa, C.; Araújo, B.; Marques, F.; Fernández-Martínez, E. Heart Disease, Now What? Improving Quality of Life through Education. *Int. J. Environ. Res. Public Health* **2021**, *18*, 3077. https://doi.org/10.3390/ ijerph18063077

Academic Editor: Paul B. Tchounwou

Received: 5 February 2021 Accepted: 15 March 2021 Published: 17 March 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/).

and compliance with the set of proposed recommendations [6]. Non-compliance with therapeutic adherence reaches such proportions that it can be considered a new cardiovascular risk factor [7].

Therefore, it is essential to develop strategies that promote greater adherence to the prescribed treatment. These educational programs can represent an important strategy in the fight against heart disease because they enable the development of self-care skills necessary for patients to reach their potential and quality of life (QoL) [8].

However, cardiac patients have difficulties in identifying and managing signs and symptoms related to heart disease, in adhering to therapy, in performing daily life activities, and even in interpersonal management [9]. For these reasons, self-care education is fundamental because it is about influencing behavior change by increasing knowledge, changing attitudes, and developing skills. Education goals include cardiac patients' participation in decision making for lifelong continuing care, awareness-raising, and functional performance [10].

The educational process is gradual, systematized, and personalized. In clinical practice, nurses find it difficult to implement these educational programs due to the high cost of hospitalizations and scientific advances in treatment that have shortened the length of hospital stays for patients with heart disease. This implies that in the rehabilitation plan the educational process begins on the first day of hospitalization [11].

Cardiac rehabilitation programs (CRP) have a strong educational component. Recognized in recent decades as an intervention with a cost-effective impact, CRP assume an integral part of the multidimensional treatment. They improve prognosis, reduce the number of readmissions and health expenses, as well as increase QoL [12]. However, in Portugal, as in other countries, the rate of admission to cardiac rehabilitation programs is very low [13]. The uncertainty about what will happen after discharge and how the patient can adapt to their new health condition justifies a care practice that facilitates the process of making informed decisions. Given that not everyone has access to CRP, the need for implementing educational interventions during hospitalization becomes even more important to improve patient outcomes.

Aware of the importance of nursing care for the post-ACS patient, the objective of this research is to design, implement, and evaluate the impact of an educational intervention program during the hospitalization period on the development of self-care skills and QoL. This article also adds data to the study previously published by Gomes and Reis (2019), and contributes to a better understanding of the importance of nurse-led educational intervention [14].

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

### *2.1. Study Design and Setting and Procedure*

The authors conducted quasi-experimental research, with the establishment of two groups—an experimental group, and a control group. The control group received standardized nursing care, and the experimental group, in addition to the standardized nursing care, had access to the educational intervention program. The 67 participants were adult/elderly patients diagnosed with ACS, hospitalized at an intensive care coronary unit (ICCU) in the northern region of Portugal, who spoke and understood Portuguese, had preserved cognitive and verbal ability, and agreed to participate in the study [14].

However, we encountered the impossibility of randomizing the groups due to the organization of the ICCU and the cardiology unit. If randomization were the choice, the participants of the control group would attend the educational intervention program since the ICCU is an open space and the cardiology unit does not have single rooms. To solve this problem, we opted for the consecutive series method, which consisted of testing the educational intervention in a non-parallel way [15]. Throughout 2016, during two-week periods, patients were included in the intervention group and, the following week new patients were included in the control group and so on, until reaching the required number of participants for our study. Despite the non-randomization, this method allows

the same probability of the patient being included in one of the groups [16]. Table 1 shows the distribution of the groups by weeks/months.


**Table 1.** Distribution of participants by the experimental and control groups.

Each phase of the program has specific objectives, as well as different strategies for action. Table 2 summarizes these phases. The four sessions provided moments for intrinsic motivation, learning, and reflection. Nurse–patient interviews conducted during hospitalization and one month after discharge was the strategy adopted for data collection. Due to the geographical dispersion of the hospital's affluence area, authors opted to contact patients by telephone.

**Table 2.** Sessions developed during the educational intervention program for patients with acute coronary syndrome.


The ethics committee and administration council of the institution and the medical director of the ICCU approved the study protocol. This study did not entail any cost to the participant or institution. It did not involve therapeutic actions; invasive, radiological procedures; or the collection of biological products. Each patient filled out a declaration of informed consent.

### *2.2. Educational Intervention Program*

To obtain significant learning for the development of self-care skills, the program had a structured and systematic approach with four sessions [14,17] and involved three educational areas. At the end of the sessions, the patient should be able to: (i) know the risks of heart disease (cognitive area); (ii) know how to control the signs and symptoms of heart disease and know how to perform their most instrumental self-care (psychomotor area); (iii) understand that changing their behaviors increases their QoL (affective area). Figure 1 illustrates the educational intervention program framework and the outcome indicators.

Given that hospitalization is increasingly shorter and that there is a need to repeat information for processing, the authors used the theoretical model of social learning of Bandura [18]. Based on the statement of Bandura's theory of self-efficacy, that expectations of self-efficacy are dependent on contexts and situations of achievement, patients are encouraged to imagine success, anticipate potential outcomes, and so respond confidently with more adaptive strategies to overcome barriers. The patients need to trust their abilities to achieve the proposed goals. The more confident and motivated the patient, the greater the success and opportunity to enhance a new health behavior.

**Figure 1.** Educational intervention program framework and outcome indicators.

**Figure 1.** Educational intervention program framework and outcome indicators. Given that hospitalization is increasingly shorter and that there is a need to repeat information for processing, the authors used the theoretical model of social learning of Bandura [18]. Based on the statement of Bandura's theory of self-efficacy, that expectations of self-efficacy are dependent on contexts and situations of achievement, patients are encouraged to imagine success, anticipate potential outcomes, and so respond confidently The educational intervention program had the following pedagogical strategies: (i) interviews with the nurse rehabilitation specialist; (ii) an educational video; (iii) checklist/pamphlet; (iv) telephone follow-up interview. The first session began in the ICCU, 12 to 24 h after hospitalization and depending on the hemodynamic stability of the patient. The nurse rehabilitation specialist conducted an interview with didactic resources; the authors opted for the visualization of a video with the following themes and contents:

	- The educational intervention program had the following pedagogical strategies: (i) • Therapeutic management;

to 24 h after hospitalization and depending on the hemodynamic stability of the patient. The nurse rehabilitation specialist conducted an interview with didactic resources; the authors opted for the visualization of a video with the following themes and contents: How does the heart function? What is coronary heart disease? Diagnosis of coronary disease; Therapeutic management; Modifiable risk factors; Healthy lifestyle. The second session was a group session and held the following day, aimed at identifying the patient's knowledge regarding the contents exposed in the video, and clarifying doubts and misassumptions. The authors used this strategy since the patients' beliefs about the disease and the personal experience also result from the social experience. Loring and Holman explain that social persuasion can improve the perception of self-efficacy [19]. During the third session held on the day of hospital discharge, the nurse used a checklist with the contents of the video. The checklist had the following purpose: (i) organize and systematize the information so that the nurse can identify difficulties, and (ii) serve as a leaflet for patient guidance. The fourth and last session occurred one month after discharge, with the aim of monitoring the patient and reinforcing information.

#### The second session was a group session and held the following day, aimed at identifying the patient's knowledge regarding the contents exposed in the video, and clarifying *2.3. Instruments Used to Collect Data*

doubts and misassumptions. The authors used this strategy since the patients' beliefs about the disease and the personal experience also result from the social experience. Lor-For data collection, authors used a sociodemographic and clinical context questionnaire, the therapeutic self-care scale and the MacNew heart disease health-related QoL questionnaire.

ing and Holman explain that social persuasion can improve the perception of self-efficacy [19]. During the third session held on the day of hospital discharge, the nurse used a checklist with the contents of the video. The checklist had the following purpose: (i) organize and systematize the information so that the nurse can identify difficulties, and (ii) serve as a leaflet for patient guidance. The fourth and last session occurred one month after discharge, with the aim of monitoring the patient and reinforcing information. Doran et al. [20] developed the therapeutic self-care scale (TSCS) with the objective of assessing the capacity for self-care in acute care contexts. This scale assesses the person's ability to perform four categories of self-care activities: taking medication as prescribed by the doctor; identifying and managing symptoms; performing activities of daily living; and managing changes in health status. The maximum score is 60 points and corresponds to a high level of performance in therapeutic self-care. The original version of the instrument by Doran et al. was translated, validated, and adapted for the Portuguese population by Cardoso, Queirós, Fontes Ribeiro, and Amaral [21].

The MacNew heart disease questionnaire is a self-administered modification of the original quality of life after myocardial infarction (QLMI) instrument [22]. It assesses the feelings of patients who have suffered an acute myocardial infarction and are attending

a CRP. Höfer et al. [23] states that the MacNew QLMI scale is a specific QoL assessment instrument, but it is used also in other cardiac pathologies, such as angina. To calculate the scores of the questionnaires, the authors used the average of all items and for the total QoL. For presentation and interpretation of results, the variable QoL has the following intervals: 1 to 3, 3 to 5, and 5 to 7. Patients with averages in the range 1 to 3 have worse QoL, patients with averages between 3 and 5 are considered to have moderate QoL, and high QoL is assumed for patients with an average between 5 and 7 points. Leal et al. [24] validated the MacNew QoL version for the Portuguese population.

### *2.4. Data Analysis*

For data treatment, the authors used descriptive and inferential statistical techniques and used the Statistical Package for Social Science (SPSS) software, version 23 for statistical analysis. The statistical techniques applied were frequency (absolute and relative), measures of central tendency (arithmetic mean and median), measures of dispersion or variability (minimum value, maximum value, and standard deviation), and tests (Chisquare test, Fisher's exact test, Mann–Whitney U test, Wilcoxon test, Spearman's correlation coefficient significance test, and the Shapiro–Wilk test as a normality test). For all tests, the value of 0.05 was set as the limit of significance, that is, the null hypothesis was rejected when the probability of type I error (probability of rejection of the null hypothesis when it was true) was lower than the set value, when *p* < 0.05, that is, *p* < 5%.

### **3. Results**

The results obtained for sociodemographic characteristics are set out in Table 3. In both groups, the majority of the patients were male, with the percentages being 90.6% in the experimental group and 71.4% in the control group. The ages of the patients in the experimental group extended between 40 and 80 years, with an average age of 57.72 ± 10.55 years. In this group, 28.1% of patients were under 50 years old; the same percentage was between 50 and 60 years old and between 60 and 70 years old. Half of the patients are older than 56.50 years old and the frequency distribution departed significantly from the characteristics of a normal or Gaussian curve (*p* = 0.022). We found that most patients of both groups reported being married or cohabiting, with percentages of 81.3% and 60.0%, respectively, in the experimental group and in the control group. In the experimental group, 43.8% of the patients had primary-level education, and in the control group, this percentage was even higher at 74.3%. No participant was illiterate.


**Table 3.** Sociodemographic characteristics.

Note: M; Mean. SD; Standard Deviation. Min–Max; Minimum–Maximum.

The results illustrated in Table 4 allow us to know the clinical characteristics of the patients. The authors found that most elements of both groups had acute myocardial infarction diagnosis, with the percentages being 90.6% and 88.6%, respectively in the experimental and control group. This was followed by unstable angina (9.4% and 11.4%). The Chi-square test revealed that the observed difference is not statistically significant (*p* = 0.784). As for risk factors, it appears that in both groups, dyslipidemia predominates,

with percentages of 78.1% and 71.4%, followed by arterial hypertension, with percentages of 53.1% and 54.3%, followed by smoking—37.5% and 34.3%, and diabetes mellitus—34.4% and 31.4%, respectively. It appears that the most common risk factors are present in both groups. The Chi-square test revealed that the differences observed between the two groups are not statistically significant (*p* = 0.999). Regarding the weight status, 37.5% of the patients in the experimental group were revealed to be in the pre-obesity situation, followed by 34.4% classified as Stage I obese. In the control group, 45.7% of the patients are pre-obese and 22.9% classified as Stage I obese. Through the application of the Mann–Whitney U test, to compare the body mass index, authors found that the differences observed between the two groups were not statistically significant (*p* = 0.292). Above-normal weight predominates in both groups.



The results also suggest improvements in the development of self-care skills and QoL in patients undergoing an educational intervention. Applying the Wilcoxon test to compare the two groups between each evaluation moment, the authors verified the existence of statistically significant differences in both cases (*p* < 0.001). The comparison of the mean and median values suggests that, between the first and in the second evaluation, the elements of the experimental group tended to improve their therapeutic self-care, while in the elements of the control group the trend was the opposite, that is, there was a deterioration in therapeutic self-care (Table 5). Authors also observed that all patients of the experimental group obtained results equal to 60 points (maximum value of the therapeutic self-care assessment scale). Consequently, the average value was 60.00 ± 0.00 points, the median had the same value, and the frequency distribution cannot be considered normal (*p* = 0.000). In the control group, the results ranged between 21 and 58 points, with an average value of 38.77 ± 7.74 points. It appears that half of the patients in this group had values equal to or greater than 39 points and the frequency distribution reveals characteristics similar to that of a normal distribution (*p* = 0.883). The application of the Mann–Whitney U test revealed the existence of significant differences between the two groups (*p* < 0.001) and a comparison of the values of the measures of central tendency reveals that the participants in the experimental group indicated better therapeutic self-care than those in the control group.

**Table 5.** Pre-intervention and post-intervention therapeutic self-care assessment scale results.


Note: M; Mean. SD; Standard Deviation. Me; Median; Min–Max; Minimum-Maximum. \* *p*-values for normal distribution.

The results illustrated in Table 6 regarding the MacNew questionnaire for the patients in the experimental group showed minimum values and maximum values between 3.22 and 6.26 points, with an average value of 4.74 ± 0.65 points. This means that patients on average perceived their QoL positively. Half of the elements in this group showed results equal to or greater than 4.76 points, and the frequency distribution departed significantly from the characteristics of a normal distribution (*p* = 0.114). In the control group, we observed results between 1.74 and 4.96 points. The average value was 3.78 ± 0.66 points, with a median of 3.74 points. The frequency distribution can be considered normal (*p* = 0.249).


**Table 6.** Quality of life assessment scale results.

Note: M; Mean. SD; Standard Deviation. Me; Median; Min–Max; Minimum-Maximum. \* *p*-values for normal distribution.

For the physical dimension of QoL, the authors observed, in the experimental group, results that ranged between 3.20 and 5.60 points, with an average value of 4.34 ± 0.63 points. Half of the respondents in this group had values equal to or greater than 4.40 points, and the frequency distribution can be considered normal (*p* = 0.375). In the control group, the authors observed values between 2.40 and 5.00 points, with an average of 3.43 ± 0.59 points. Half of the elements in this group obtained results above 3.40 points, which shows that the perception of the QoL is close to the value of 3.50 points that represents the average value of the scale. The values are lower than those of the experimental group. The frequency distribution departed significantly from a normal distribution (*p* = 0.182).

In the emotional dimension for the experimental group, values between 3.21 and 6.57 points were observed, with a mean value of 4.80 ± 0.77 points and a median of 4.93 points. The frequency distribution showed characteristics close to a normal curve (*p* = 0.687). In the control group, the observed values were between 1.43 and 5.00 points, with an average of 3.84 ± 0.74 points. Half of the individuals that constituted this group obtained results above 3.93 points and the frequency distribution departed significantly from the characteristics of a normal curve (*p* = 0.029).

In the social-life dimension, the authors found in the experimental group values between 3.00 and 6.67 points, with the mean value being 5.29 ± 0.79 points. Half of the elements in this group showed results greater than 5.33 points. The frequency distribution cannot be considered normal (*p* = 0.042). For the control group, authors found values between 1.83 and 6.17 points, with an average of 4.17 ± 0.91 points. Half of the sample elements showed values equal to or greater than 4.17 points. The frequency distribution shows characteristics close to a normal curve (*p* = 0.543).

Applying the Mann–Whitney U test, the authors can conclude that between both groups there is a statistically significant difference (*p* < 0.001) in global, physical, emotional, and social QoL. The comparison of the values of the measures of central tendency reveals that the elements of the experimental group showed significantly better QoL than those of the control group.

### **4. Discussion**

As for the sociodemographic characterization, the authors found that in both groups, the patients are mostly male, with low literacy. The average age of the participants in the experimental group is 57.72, while the average age of the patients in the control group is 65.09. These results reflect the profile of coronary patients in the Portuguese population, according to governmental reports [25], in which the prevalence of coronary disease is higher in males, with low literacy and incidence between 50 and 70 years old. Regarding the level of education, the findings of this study revealed that for individuals with a low level of education, risk factors were more prevalent. These results were consistent with the findings of the study conducted by Marques da Silva et al. (2019). They studied the prevalence of cardiovascular risk factors and other comorbidities in patients with hypertension in Portuguese primary health care populations. However, despite the higher prevalence of individuals with low levels of education, it is worth highlighting the trend of a gradual increase in risk factors among individuals with higher education levels [26,27].

There was also a higher prevalence in both groups of married/cohabiting patients. In the Okwose et al. [28] study, patients reported that family members are a strong motivating factor. Having a partner or relative to support them and participate in new activity routines is important. Family can promote patients' emotional support and enable therapeutic regimen adherence.

Data suggests that the educational program promotes improvements in the cardiac patient's self-care skills and in global, physical, emotional, and social QoL. The program made patients more proactive in decision-making and assuming responsibilities for their health conditions. In the Riegel et al. [29] study, the behavior-change factors address habits, motivation, decision-making, and the challenges of persistence. However, it is important to consider that illness-related factors address specific issues that make self-care exceedingly difficult—multimorbidity, symptoms, and stressful life events.

The results of the Goodman, et al. [30] study carried out in patients with decompensated heart failure (HF), suggest that low scores in self-care skills can be related to difficulties in decision-making and lack of motivation, and not so much to learning difficulties. Therefore, in the Rice et al. [31] study, results suggest that nurse-led patient education for adults with HF improves QoL and reduces hospital admissions and readmissions—a major cause of health-care costs [32]. However, for this study the authors did not considered readmissions as an outcome. The randomized controlled trial study conducted in Iran with 60 patients that had HF also concluded that a self-management education program is an appropriate strategy for improving QoL in patients with HF [33].

Educational programs implemented during hospitalization aim to enable patients to deal more effectively with the chronic disease after hospital discharge. Improving patient's health outcomes by increasing knowledge and behaviors related to the cause of the disease, treatment, and coping strategies can prevent complications and better QoL [34]. In our opinion, this educational program can be adapted and implemented among other chronic disease patients, since teaching skills can be generalized.

The results indicate that behavioral changes and self-care skills depend on educational programs for patients to learn and practice. The educational sessions are not just for transmitting information but also should involve moments of elucidation, reflection, and discussion accompanied by didactic material such as pamphlets and audiovisual media for better understanding. For practice implications, Rice et al. [31] considered diverse pedagogical methods, such as nurse-led education strategies.

The study design should be considered as a limitation; it would have been interesting to compare two intervention groups with each other. Furthermore, it would be desirable to carry out longitudinal studies to confirm the differences between the groups in the long term after the intervention.

### **5. Conclusions**

This study revealed a positive relationship between the development of self-care skills and the QoL of the post-ACS patient. However, the period of one month between the two assessment moments (hospitalization and one month after discharge) may not have been enough to observe permanent changes in self-care skills and QoL. In future studies, it will be necessary to evaluate the effectiveness of the educational intervention program in a longer follow-up period.

Knowledge alone does not influence self-care behaviors. It is a complex decisionmaking process. The patient uses values and experiences in decision-making and these experiences emerge from situational awareness, their perceptions, and the meanings given to the cardiac event. Each option and response create a set of standards.

In making complex decisions, the nurse supports and helps the post-ACS patient to recognize decision-making situations and provide strategies to facilitate effective responses.

The nurse's educational intention alone is not enough. This intentionality involves implementing educational interventions guidelines, so that patients and families are informed (knowledge development), trained (skills development), and involved in their health care. Nurses being patients' partners in decision-making can indicate the path towards adaptation, autonomy, and QoL, because these decisions will be compatible with personal life goals.

**Author Contributions:** Conceptualization, L.G. and B.A.; methodology, L.G., B.A., and F.M.; software, L.G., E.F.-M., and C.L.-P.; validation, L.G., E.F.-M., and C.L.-P.; formal analysis, L.G., E.F.-M., C.L.-P., and B.A.; investigation, L.G., B.A., and F.M.; resources, L.G. and F.M.; data curation, L.G., E.F.-M., C.L.-P.; writing—original draft preparation, L.G.; writing—review and editing, L.G., E.F.-M., and C.L.-P.; visualization, L.G., E.F.-M., C.L.-P., B.A., and F.M. supervision, L.G., E.F.-M., C.L.-P., B.A., and F.M.; project administration, L.G., E.F.-M., C.L.-P., B.A., and F.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 Ethics Committee of Centro Hospitalar de Trás-os-Montes e Alto Douro, E.P.E, date of approval 27 April 2016.

**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* **Translation and First Pilot Validation Study of the "Undergraduate Nursing Student Academic Satisfaction Scale" Questionnaire to the Spanish Context**

**María Dolores Guerra-Martín 1,\* , Alejandro Cano-Orihuela <sup>2</sup> , Raúl Martos-García <sup>3</sup> and José Antonio Ponce-Blandón 3,\***


**Abstract:** Satisfaction helps nursing students to develop skills and improve their academic performance, hence the importance of assessing it by means of a reliable instrument. The objective was to translate and culturally adapt the "Undergraduate Nursing Student Academic Satisfaction Scale" (UNSASS) instrument to the Spanish context. A cross-sectional study was conducted with a representative sample of 354 fourth-year nursing students from University of Seville, Seville, Spain. The validation process was carried out in five phases as follows: direct translation, synthesis of the translations, back translation, consolidation by a panel of experts, and pilot test with nursing students. After two rounds among two expert committees, the Content Validity Index (CVI) varied from 0.85 to 1, obtaining a CVI above 0.8 with the global questionnaire. A scale composed of 48 items and 4 subscales was obtained, resulting in a Cronbach's α coefficient of 0.96. Within the subscales, this coefficient varied between 0.92 and 0.94. No statistically significant differences were found between the total satisfaction of the scale and gender and teaching unit. An inversely proportional relationship was found between the age and the "Support & Resources" scale. The "Escala de Satisfacción Académica del Estudiante de Enfermería" (ESAEE) scale was obtained, translated, and adapted to the Spanish context from the UNSASS scale, with satisfactory consistency and validity.

**Keywords:** personal satisfaction; students; nursing; surveys and questionnaires; validation studies

### **1. Introduction**

The new focus of universities is to become facilitating centers for teaching, where the students are able to acquire knowledge through friendly methods and where the institutions base their efforts on promoting a healthy environment of well-being [1,2].

Critical and creative thinking is demanded from health professionals, and nursing education has evolved to adapt to these new requirements. This process has gradually taken place thanks to efficient professional competences aimed at obtaining a high level of satisfaction for the user of the educational system [3].

Students' satisfaction helps build self-confidence and is a source of support to develop skills and acquire knowledge, thus helping to improve academic performance [4,5].

The current generation of students is flooded with a great variety of stimuli, with an element that reinforces the motivation to learn in the students being important; in this sense, satisfaction is very much connected to the increase in motivation, through which they learn more and better, which is useful for their future professional practice [6]. However, satisfaction in relation to obtaining any degree is complemented by stressors such as suffering, diseases, disabilities, or death of patients [7].

The non-intellective competences involved in academic performance can be important resources for promoting academic adjustment and satisfaction, favoring the retention and

**Citation:** Guerra-Martín, M.D.; Cano-Orihuela, A.; Martos-García, R.; Ponce-Blandón, J.A. Translation and First Pilot Validation Study of the "Undergraduate Nursing Student Academic Satisfaction Scale" Questionnaire to the Spanish Context. *Int. J. Environ. Res. Public Health* **2021**, *18*, 423. https://doi.org/10.3390/ ijerph18020423

Received: 15 November 2020 Accepted: 4 January 2021 Published: 7 January 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/).

persistence processes in college. In the same way, working on the non-cognitive skills closely linked with study satisfaction could also improve performance and academic success, thereby increasing domain-specific satisfaction. In this case, for example, individual work on intrinsic motivation or group work that enhances the ability to relate to one's fellow students could improve satisfaction with one's study habits [8].

Furthermore, some studies support the finding that the environment at university can be perceived as stressful and that developing the ability to handle negative situations effectively increases perceived support from peer relationships and can create a virtuous cycle that helps students improve this competence [9]. Thus, some authors suggest the importance of developing one's own scale to measure the satisfaction of Nursing Degree students [7], carrying out experiences in which it is clear that these students complement the stressors that young people commonly have with those additional derived from the practical environment, in which day by day they find stress-generating stimuli. Moreover, there is the need that these students have to establish relationships with other health professionals, and the fact of having to play a new role for which they have not yet been fully trained.

Thus, there are various studies whose objective was to validate standardized scales targeted at evaluating nursing students' satisfaction, by dealing with either one or more of its aspects or factors, and directed towards a specific idiomatic and/or cultural context. Currently, there are validated scales that evaluate satisfaction and effectiveness in the clinical learning setting, as well as satisfaction in clinical practices [10–13]. In the case of Asadizaker et al. [4], with their "Satisfaction with First Clinical Practical Education" (SFCPE) scale, they evaluated in a more comprehensive manner this realm of nursing education, through the "Instructor performance" (α = 0.92), "Integrated plan" (α = 0.82)", "Feelings and perceptions" (α = 0.78), "Learning atmosphere" (α = 0.73), "Scheduling" (α = 0.70), "Facilities" (α = 0,65), and "Access to professionals" (α = 0.60) subscales. Furthermore, given the importance that the first experience in a clinical setting, the main factor assessed by this scale, is the cornerstone of the future in professional nursing for all nursing students, the authors defend the use of this instrument, concluding that it is valid and reliable.

In the scope of the European Space of Higher Education, Lepiani et al. [14] conducted a study focused on the satisfaction of first-year nursing students when addressing Basic knowledge, the Organization of teaching, the Skills developed, the Teaching–learning process, Access and attention in the students, and the Curriculum and its structure as study factors. Satisfaction assessment scales have also been described in the university context, not only in the nursing field, which have shown to be useful and could also be applied in contexts of assessment of the satisfaction of undergraduate nursing students [15].

In contrast, among the scales of multidimensional appreciation are studies such as the one focused on determining nursing students' satisfaction [16], where a level of satisfaction in 11 factors was determined. Additionally, the study developed by Gloria and Ortiz [1] established the quality of life level and the factors related to it in such students, when applying the "Calidad de Vida y Satisfacción" (Q-LES-Q) questionnaire.

Despite the usefulness of all these instruments, a multidimensional, validated, and reliable "Undergraduate Nursing Student Academic Satisfaction Scale" (UNSASS) was found focusing on the measurement of satisfaction of nursing students with a Nursing Degree and all the integral aspects that make up the student's learning period. This scale was developed by Dennison and El-Masri [17] and validated in the Canadian context, whose function is to evaluate the level of satisfaction over four dimensions, namely "In-Class Teaching", "Clinical Teaching", "The Program", and "Support & Resources", where each scale obtained Cronbach's alpha coefficients of 0.92, 0.91, 0.91, and 0.74, respectively, and of 0.96 for the entire scale, which denoted an excellent overall internal consistency.

It is true that other instruments have been developed to evaluate the satisfaction of nursing students, in different cultural contexts, such as the work of Chen and Lo [18], validating the "Nursing Student Satisfaction Scale" (NSSS) instrument for its use in the USA context with high internal consistency (α = 0.96), where satisfaction was evaluated in three

factors, namely Curriculum and teaching, Environment, and Professional social interaction. Subsequently, Domingues et al. [19] once again validated this tool, the NSSS scale, although in this case for the Brazilian context. It was composed of three factors, namely "Curricular dimension and teaching", "Environmental dimension", and "Social/Professional interactions". Despite the recognition of the usefulness of this scale in different contexts, it clearly covers a smaller set of scenarios in its studied dimensions, excluding aspects as relevant as the support and necessary resources and the differentiation between clinical learning environments and academic environments ("in-class" learning). These aspects are however effectively addressed by the UNSASS scale, proving to be one of the most complete due to its properties of internal consistency, reliability, measurement of errors, construct validity, structural validity, and criterion validity. Furthermore, the scale follows the standardized quality assessment criteria methodology of studies on measurement properties of health status measurement instruments [20].

Thus, a translation and validation research of the aforementioned scale to the Spanish context were justified. We therefore proposed the research as this study's objective to translate and culturally adapt the multidimensional, validated, and reliable "Undergraduate Nursing Student Academic Satisfaction Scale" instrument to the Spanish context. The study also aimed to verify the level of satisfaction of fourth-year nursing students and the existing relationships between the level of satisfaction and gender, age, or the teaching units to which they belonged.

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

### *2.1. Design and Participants*

A validation study of the "Undergraduate Nursing Student Academic Satisfaction Scale" (UNSASS) questionnaire adapted to the Spanish context was conducted, which included a non-experimental, quantitative, descriptive, and cross-sectional pilot test, between December 2018 and July 2019, where data collection took place in April 2019.

The study population were the 354 fourth-year nursing students belonging to the Virgen Macarena, Virgen del Rocío, or Virgen de Valme teaching units, from the School of Nursing, Physiotherapy and Podiatry of the University of Seville, Seville, Spain, where a stratified sampling was conducted respecting the proportion of students of the target population. The students were recruited by means of e-mail messages, and the instrument was provided online using the Google Forms tool. From the study population divided by teaching unit composed of 74 from Virgen del Rocío, 226 from Virgen Macarena, and 54 from Virgen de Valme, a final sample of 32 students (9.04%) of the entire school was obtained, of which 7 (9.46%), 19 (8.41%), and 6 (11.11%) were from each of the teaching units, respectively, numbers that fall within the 30–40 students estimated in the scientific literature [21].

In addition to asking about gender, age, and teaching unit, the instrument was applied to fourth-year students since they were immersed in the context of undergraduate nursing studies for a longer period of time and, therefore, possessed more knowledge, excluding those who did not understand Spanish and who did not attend the previous years in that school.

### *2.2. Translation and Cultural Adaptation Process*

The instrument used for validation in the Spanish context was the "Undergraduate Nursing Student Academic Satisfaction Scale" (UNSASS) [17]. To this end, the guidelines for cultural adaptation and validation of health questionnaires by Ramada, Serra, and Declós [22] and by Romero-Martín, Gómez-Salgado, De la Fuente-Ginés, Macías-Seda, García-Díaz, and Ponce-Blandón [23] were followed, consisting of five phases, namely direct translation, synthesis, back translation, consolidation by expert committees, and pilot test of the questionnaire.

Initially, two independent and bilingual nurses with work experience and knowledge of the Spanish university system in force conducted a linguistic, cultural, and conceptual

translation of the instrument [24]. Subsequently, the researchers synthesized both translated versions comparing by pairs, where the existing discrepancies between the two versions were identified and discussed, reaching consensus. The consensual version in Spanish was back-translated into English by two independent translators, who did not have access to the original version. Subsequently, pair corrections were made, and the prototype of the scale called "Escala de Satisfacción Académica del Estudiante de Enfermería (ESAEE)" was created. For logical appearance validity, two expert committees were created with a total of 13 participants—one was made up of 7 professors from diverse university settings and the other was constituted by 6 fourth-year nursing students [21]. The objective of both expert committees was to assess whether each item was understandable, with sufficient clarity in each of them.

### *2.3. Applicability and Feasibility*

In the pilot phase, each of the participants was asked to complete the version of the ESAEE instrument translated and adapted to the Spanish context and, later on, they were given the choice to express if they found any difficulty in understanding any of the items, as well as to leave comments on them. In this sense, if at least 15% of the students had difficulties with any item, it should be reviewed; but none of the participants found any difficulty in answering the questions. We also asked them to indicate the time they needed to answer the survey, the result being slightly less than 15 min.

Likewise, it was considered that, in adaptations, it is relevant not only to show evidence of a possible linguistic equivalence between the original and the adapted instruments, but also to state that the adaptations are equivalent from a conceptual point of view [24].

When calculating the relevance or validity of the questionnaire, the "Content Validity Index" (CVI) was used, in which, for each item to be considered acceptable in the final questionnaire, its relevance had to be assessed with a value of 3 or 4 out of 4 on a Likert scale by more than 78% of the experts [25,26], that is, a CVI equal to or higher than 0.78, and that it was understandable for at least 80% of the experts [21,27].

Those items that did not meet the aforementioned requirements were reviewed and reassessed, contacts being made with the primary translators and with those in charge of the back translation, in order to verify that the changes we were going to implement faithfully reflected the original scale. With the translators' approval, a series of changes were introduced in some items; the others were kept unchanged following the translators' suggestions.

### *2.4. Statistical Analysis*

When assessing the internal consistency or reliability of the questionnaire, its Cronbach's alpha coefficient was calculated, where a value above 0.7 was considered acceptable, which reveals a strong relationship among the questions of the test, either in each dimension or subscale or in the entire questionnaire [28]. For the construct validity, a confirmatory factor analysis (CFA) was used, and a previous verification of the suitability was checked with the Bartlett sphericity test and the Kaiser–Meyer–Olkin (KMO) coefficient. A significance of *p* < 0.05 for the Bartlett test and a value of KMO > 0.60 were considered acceptable as recommended in the literature [29]. The extraction method used to perform the CFA was the principal component analysis, and the rotation method used was the Varimax method with Kaiser normalization.

When describing the gender, age, and teaching unit variables, the subscales and items of the questionnaire, means, standard deviations and percentages were used.

The Shapiro–Wilk test and the Levene's test were used to observe normality distribution, as well as to determine whether there was homogeneity in the variances, in addition to parametric tests such as the Student's *t*-test and the ANOVA test or non-parametric ones, like Kruskal-Wallis or Mann-Whitney U. For the relationship of two continuous quantitative variables, Pearson's correlation test was performed.

In order to perform the statistical data analysis, the SPSS© statistical software, version 21.0 (IBM Corp., Armonk, NY, USA) was used.

### *2.5. Ethical Considerations*

Prior to the translation and validation process, authorization was asked from the authors of the UNSASS questionnaire, and they gave permission to use the UNSASS scale for its adaptation and use in the present study [17].

The participants of both expert committees (professors and students) and the students who participated in the pilot test were previously informed about the purpose of the study and they provided informed voluntary written consent. Their information was registered anonymously so it would not be possible to identify participants´ answers, guaranteeing anonymity and data confidentiality at all times. The study was approved by the Research Ethics Committee from the Spanish Red Cross Nursing College, University of Seville, with reference number 10/2018. According to this report, the study meets the requirements for research with human beings and complies with current regulations in Spain and the European Union regarding research issues.

### **3. Results**

### *3.1. Translation and Adaptation Phase*

After the described translation process and assessment, the initial version of the instrument, after going through the expert panel, can be found as supplementary information to this article (Table S1, Supplementary Materials), with its variables (type and operational definition) summarized in Table 1.


**Table 1.** Variables, types, and operational definitions.

### *3.2. Item Creation Process and Content Validity*

When presenting a CVI above 0.78 or when being considered clear by at least 80% of the experts in both the first and second rounds, the reevaluated items were accepted as relevant and were included in the final version of the scale.

In the first round of item appreciation by the expert committees, seven items (3, 15, 28, 30, 31, 42, and 44) did not reach the clarity appreciation cutoff point in at least 80% of the experts, and there were also six items (4, 15, 28, 31, 44, and 45) with a CVI below 0.78. Those items were subjected to a second evaluation round with the improvements proposed by the experts and a reassessment of the translation process. In any case, in this first round of the expert panel of professors, some of the suggestions for improvement that the experts proposed for each of the items that did not meet the validity or relevance criteria were taken into account.

Unlike in the panel of professors, the panel of students did not find objections in the items when assessing their clarity and relevance.

Therefore, the items in which professors had discrepancies were reevaluated, rechecking each of the steps previously performed, again comparing the previous versions of the scale, and contacting the primary and native translators who carried out the translations, to verify that the modifications that we were going to make could be a reflection of the original scale, in order to maintain fidelity in the process, since it was not possible to move away from its essence. With the approval of the translators, some modifications were

introduced in the items, and in others, they were kept the same following the suggestions of the translators.

Once the items had been modified, a second analysis was made by the panel of professors in order to verify that the changes made were the ideal ones to achieve optimal clarity and relevance of each of the items. In this second round of the validation process, all the reevaluated items obtained a CVI above 0.78 (Table 2), the reason why they were accepted as relevant and were included in the final version of the scale. They observed that, in its integrity, the scale was expressed with sufficient clarity, thereby concluding this phase of the validation process, including all items for the final version of the scale, even the reassessed items.

**Table 2.** Content Validity Index (CVI) of the reevaluated items after the last round with the expert committees.


CVI: Content Validity Index.

### *3.3. Results of the Pilot Study*

The ESAEE scale was administered to a sample of 32 students. A total of 78.1% of participants (*n* = 25) were women. The mean age of the sample was 22.2 years (21–27), SD = 1.62). The distribution of participants among the different teaching units was 21.8% (*n* = 7), 59.3% (*n* = 19), and 18.7% (*n* = 6) for "Virgen del Rocío", "Virgen Macarena", and "Virgen de Valme" units, respectively.

Table 3 shows the descriptive of the score obtained (mean, standard deviation, and the percentage of the mean score related to the maximum value of each of the subscales and the ESAEE scale as a whole). The subscale 4 "Support & Resources" was the best valued subscale by nursing students. No comprehension or legibility problems were identified for any of the items.

**Table 3.** Mean, standard deviation, and percentage of the mean score in relation to the possible maximum of each of the subscales and of the entire "Escala de Satisfacción Académica del Estudiante de Enfermería" (ESAEE) scale.


Table 4 presents the mean and standard deviation for each of the items and for each of the subscales.

*Int. J. Environ. Res. Public Health* **2021**, *18*, 423


**Table 4.** Mean values and standard deviation for items, subscales, and total ESAEE scale.

*Int. J. Environ. Res. Public Health* **2021**, *18*, 423


### *3.4. Internal Consistency and Reliability*

The four subscales that constitute the ESAEE scale obtained Cronbach's alpha coefficients of 0.94 (In-Class Teaching subscale), 0.94 (Clinical Teaching subscale), 0.92 (Program Design and Delivery subscale), and 0.92 (Support & Resources subscale). The ESAEE scale obtained the highest Cronbach's alpha coefficient, i.e., 0.96.

### *3.5. Internal Validity of the Scale*

Regarding the confirmatory factor analysis (CFA), before the CFA, KMO = 0.93 and *p* < 0.001 was achieved in the Bartlett test. The CFA results revealed and confirmed the four original factors that accounted for 67.3% of the total variance and a factor loading above 0.4 in all the items. Table 5 presents, through the rotated component matrix, the values of the CFA model and the standardized coefficients of the model items that best fit (only factor loading values above 0.4 are expressed).

### *3.6. Hypothesis Contrast Analysis*

The relationship of gender and teaching unit with the satisfaction of the surveyed students was analyzed, not only in each of the factors or dimensions but also in the entire scale, and no relationship was observed. However, in the case of the relationship between age and satisfaction there was no relation between them, both in the entire scale and in all the subscales, except for the "Support & Resources" subscale (*p* = 0.003; Pearson's correlation coefficient of −0.513), which indicated an inversely proportional relationship, with moderate intensity.





**Table 5.** *Cont*.

Only factor loading values above 0.4 are expressed.

### **4. Discussion**

The final version of the ESAEE scale is composed of 48 items divided into 4 subscales (In-Class Teaching, Clinical Teaching, Program Design and Delivery, and Support & Resources), identically to the original UNSASS scale [17]. All the items, as Dennison and El-Masri [17] did in their development of the UNSASS scale, were evaluated both in terms of clarity and of relevance, this last characteristic by means of the CVI. Similarly, in order to maintain fidelity with the original in the adaptation of the instrument, the same Likert-type scale scoring was followed, and no strata were proposed for the classification of satisfaction.

By means of the expert committees (professors and students), it was verified that the ESAEE scale presents logical appearance validity. The reliability of the UNSASS scale was assessed using Cronbach's alpha, and showed excellent reliability for the entire scale and for the subscales, with the ESAEE scale presenting reliability levels even higher than the original. However, it is necessary to highlight that the calculation of Cronbach's alpha was not performed from the perspective of a scale made up of 5-point ordinal variables, as suggested by some in the literature [30,31] but rather as nominal polycotomic variables. However, in any case, this same literature suggests that with this type of analysis the Cronbach's alpha would probably have been even higher.

With respect to the "Satisfaction with First Clinical Practical Education" (SFCPE) questionnaire developed by Asadizaker et al. [4] to assess satisfaction in the nursing students with their first clinical setting, it consists of seven factors, and there are similarities when analyzing the nursing student's satisfaction between the two scales, although the ESAEE scale is more compact for having fewer factors, four specifically.

In the first subscale of ESAEE, mostly devoted to theoretical teaching in classrooms, there is no correlation with the SFCPE scale, since the latter is more focused on the clinical scope. However, the second subscale, "Clinical Teaching", is in fact widely adopted and, at the same time, subdivided into several factors that can relate between the subscales. In this sense, the SFCPE adopts factors like "Feelings and perceptions" (Factor 3) (related to items 18, 19, and 29 from ESAEE), "Clinical atmosphere" (Factor 4) (related to items 21, 23, and 25 from ESAEE), "Instructor performance" (Factor 1) (related to items 22, 24, 26, 30, and 31 from ESAEE), or "Access to professionals" (Factor 7) (related to items 17, 20, and 27 from ESAEE). The third subscale in ESAEE, "Program Design and Delivery", is related to the SFCPE scale in the following factors: 2. "Integrated plan" and 5. "Scheduling". Finally, the "Support & Resources" subscale from ESAEE is related to Factor 6. "Facilities" of the SFCPE scale.

In relation to gender, no statistically significant difference was found between satisfaction and this variable. This coincides with the study by Salamonson et al. [32], in which no relationship was found between satisfaction and the age of the nursing students, coinciding with the results obtained in the ESAEE scale, with the exception of the "Support & Resources" subscale, where the relationship with satisfaction was in fact statistically significant. Studies such as the ones by Milton-Wildey et al. [33] and Domingues et al. [19] set forth that younger students present greater satisfaction levels than older students, which is in agreement with the results obtained in the "Support & Resources" subscale from ESAEE.

In the ESAEE scale, total satisfaction was high (75%). Regarding the decreasing order of the students' satisfaction levels with the different subscales, we found that the "Support & Resources" subscale (83.8%) was the best valued, followed by "Clinical Teaching" (78.4%), "Program Design and Delivery" (72.6%), and, in fourth and last place, "In-Class Teaching" (70.8%).

The "In-Class Teaching" subscale from ESAEE obtained a higher result than that obtained in the "Cuestionario de Satisfacción del Estudiante" by Jiménez et al. [34], where the "Desempeño del profesor" factor (related to the "In-Class Teaching" subscale from ESAEE), was in second place, but with a percentage of 64%. The results obtained in our study (mean of 3.54) are in agreement with those obtained by Domingues et al. [19] with a mean of 3.57 for the "Curricular dimension and teaching" factor.

The results obtained for the "Clinical Teaching" subscale (78.4%) from ESAEE were consistent with those obtained in the study by Espeland and Indrehus [10], where 70% of the students were satisfied with the "Clinical practice".

The "Program Design and Delivery" from ESAEE is related in the study by Lepiani et al. [18] with the "Organization of teaching" factor, which obtained a mean score of 3.81 in the students' satisfaction levels, similarly to our study (3.63). In the study by Domingues et al. [19], this subscale is related to the "Curricular dimension and teaching" factor, which obtained a mean score of 3.57, also similar to the result obtained in the "Program Design and Delivery" subscale.

The results of the "Support & Resources" subscale from ESAEE, with items related to the infrastructures and to the administration and services staff, are in agreement with those obtained in the study by Pecina [3], where the areas best evaluated were "IT services" and "Infrastructures". In the study by Lepiani et al. [14], the "Facilities" factor was the worst evaluated (mean of 2.87); however, this result does not coincide with the higher satisfaction level obtained in subscale 4 (mean of 4.19).

A very high reliability was obtained in the ESAEE scale (α: 0.96), coinciding with the values obtained in the studies by Asadizaker et al. [4] and Baykal et al. [14]. In addition, Domingues et al. [19] and Dennison and El-Masri [17] reached Cronbach's α coefficients of 0.92, 0.97, 0.93, and 0.96, respectively; however, lower results, though reliable, were obtained in Pecina [3] and in Salamonson et al. [29], namely 0.83 and 0.80, respectively.

As a limitation of this research, we can mention having conducted an exploratory study of the ESAEE questionnaire, without accompanying it with a reliability test (test and re-test), in addition to only selecting students from the last year of the nursing program. This would have allowed us to calculate the interclass correlation coefficients, which would have given much more consistency to the study. Thus, as an improvement proposal, we shall conduct a reliability test, by having the participants complete the ESAEE questionnaire a second time 15 days after its first application, in addition to expanding the selection to students from the third year, who also have clinical practices as well as theoretical classes. Another limitation to highlight is the absence of an external evidence of validity, given that, due to the scarcity of specific scales available of this type, it would also be very difficult to identify a "gold standard". As it is a first pilot validation study of the scale and, therefore, a line of research in which the authors continue to investigate, the inclusion of a comparison with an external evidence of validity could be suggested for the future.

### **5. Conclusions**

Obtaining the satisfaction level of undergraduate nursing students represents an opportunity to know the aspects that can be improved, in order to implement measures leading to better quality in the studies, from the point of view not only of the curricula, but also of their transposition to theoretical classes, care practices, and the organization of the Nursing School. The aforementioned will improve how the students cope with their entry in the near future into the nursing profession and, with that, into the professional health care provided to people. This research could also be useful for a future review of the aforementioned aspects, either by the Nursing Schools or by the universities.

The Escala de Satisfacción Académica del Estudiante de Enfermería (ESAEE) scale was developed through a 5-phase validation process, adapted to the Spanish context. It consists of 48 items encompassed in four appreciation dimensions (In-Class Teaching, Clinical Teaching, Program Design and Delivery, and Support & Resources), with a Content Validity Index and a Cronbach's α sufficiently high and similar to the original version of the validated questionnaire (UNSASS), which signifies sufficiently high internal consistency and validity of the content of the questionnaire for its validation. Only one moderate negative correlation was observed between the "Support & Resources" subscale and age.

For the future, we intend to conduct a multicenter research study with Nursing Schools from other universities, with a larger sample of nursing students and from different years, in addition to conducting not only a quantitative but also a qualitative approach, since in the world of perceptions and feelings, as in the case of the study of satisfaction, it is necessary to know more in depth what those students feel.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/1660-460 1/18/2/423/s1, Table S1: "Escala de Satisfacción Académica del Estudiante de Enfermería" (ESAEE scale) after translation, expert committee modifications, and piloting.

**Author Contributions:** Conceptualization, M.D.G.-M. and A.C.-O.; methodology, M.D.G.-M., A.C.- O., and J.A.P.-B.; software, M.D.G.-M., A.C.-O., and R.M.-G.; validation, formal analysis, and data curation, M.D.G.-M. and A.C.-O.; writing—original draft preparation, M.D.G.-M. and A.C.-O.; writing—review and editing, R.M.-G. and J.A.P.-B.; visualization, A.C.-O. and R.M.-G.; supervision, M.D.G.-M. and J.A.P.-B. 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 Research Ethics Committee from the Spanish Red Cross Nursing College, University of Seville, with reference number 10/2018.

**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* **Study of the Strengths and Weaknesses of Nursing Work Environments in Primary Care in Spain**

**Vicente Gea-Caballero 1,2 , José Ramón Martínez-Riera 3,\* , Pedro García-Martínez 1,2,\* , Jorge Casaña-Mohedo <sup>4</sup> , Isabel Antón-Solanas 5,6 , María Virtudes Verdeguer-Gómez <sup>7</sup> , Iván Santolaya-Arnedo 8,9 and Raúl Juárez-Vela 8,9**


**Abstract:** Background: Nursing work environments are defined as the characteristics of the workplace that promote or hinder the provision of professional care by nurses. Positive work environments lead to better health outcomes. Our study aims to identify the strengths and weaknesses of primary health care settings in Spain. Methods: Cross-sectional study carried out from 2018 to 2019. We used the Practice Environment Scale of the Nursing Work Index and the TOP10 Questionnaire of Assessment of Environments in Primary Health Care for data collection. The associations between sociodemographic and professional variables were analyzed. Results: In total, 702 primary care nurses participated in the study. Responses were obtained from 14 out of the 17 Spanish Autonomous Communities. Nursing foundation for quality of care, management and leadership of head nurse and nurse–physician relationship were identified as strengths, whereas nurse participation in center affairs and adequate human resources to ensure quality of care were identified as weaknesses of the nursing work environment in primary health care. Older nurses and those educated to doctoral level were the most critical in the nursing work environments. Variables Age, Level of Education and Managerial Role showed a significant relation with global score in the questionnaire. Conclusion: Interventions by nurse managers in primary health care should focus on improving identified weaknesses to improve quality of care and health outcomes.

**Keywords:** nursing; primary care; workplace; quality of health care; nurse's role

The nursing workforce plays a crucial role in health systems globally. According to the World Health Organization (WHO) [1], nurses must work to their full potential if countries are to achieve universal health coverage for the population. Nurses are a key element in the sustainability of the health service, enhancing quality of care and promoting patient safety, satisfaction and confidence [2,3]. These reasons have justified the implementation

**Citation:** Gea-Caballero, V.; Martínez-Riera, J.R.; García-Martínez, P.; Casaña-Mohedo, J.; Antón-Solanas, I.; Verdeguer-Gómez, M.V.; Santolaya-Arnedo, I.; Juárez-Vela, R. Study of the Strengths and Weaknesses of Nursing Work Environments in Primary Care in Spain. *Int. J. Environ. Res. Public Health* **2021**, *18*, 434. https://doi.org/ 10.3390/ijerph18020434

Received: 24 November 2020 Accepted: 1 January 2021 Published: 7 January 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/).

of the international campaign "Nursing Now" and the declaration of the year 2020 as the International Year of the Nurse.

In order to achieve the highest possible quality of nursing care, it is essential to analyze the elements contributing to the delivery of such care; the sum of these elements is the nursing work environment (NWE). Thus, NWE is defined as the characteristics of the workplace that promote or hinder the delivery of quality nursing care [4]. Positive NWEs are characterized by lower rates of mortality, morbidity and adverse events [5–7]; reduced administrative costs and absenteeism [8]; a lower level of burnout and increased patient satisfaction with nursing care and the health organization [9]. Improving health outcomes, the sustainability of the health system and user satisfaction are common goals for healthcare managers, which can be addressed from the perspective of NWE with positive conditions for care. The Magnet Hospital program, for example, accredits centers of excellence and quality of care based on evidence-based practice [10–12]. Magnet centers promote the development of a culture of safety and quality care, staff recognition programs, interdisciplinary communication and horizontal management that contribute positively to the work environment [13,14].

NWE studies have been extensive in hospital settings [15]. However, there is a lack of evidence about the impact of positive primary health care (PHC) NWE on patient outcomes. This may be a handicap for optimizing PHC services, where the configuration of the microenvironment should be conducive to the provision of excellent care. Previous studies [4] have pointed to a possible association between positive PHC NWE and patient outcomes, but the evidence is still scarce.

Previous studies [4,16–18] have assessed the PHC NWE in Spain, concluding that management and leadership of the head nurse, nurse–physician relationship and nursing foundation for quality of care are the most highly valued aspects by primary care nurses, whereas adequate human resources to ensure quality of care is frequently among the least valued characteristics of the workplace. A recent scoping review of the literature suggested that the reality of the PHC NWE is different from that of the hospital NWE. This is due to differences in the decision-making and organizational processes and the relationships between team members. Thus, the evidence available on the hospital NWE may not be applicable to the reality of the PHC NWE [19]. According to Lucas and Nunes [19], the work environment is the most influential factor with the greatest impact on nursing outcomes and on the perceptions of the quality of care and client safety. In order to assess and thus improve the quality of the NWE, Poghosyan et al. [20] developed a global model for optimizing the PHC NWE. In their model, they propose the integration of institutional policies, organizational innovation and research.

We argue that positive PHC NWE can increase the quality of nursing care and, subsequently, improve patient outcomes. However, in order to achieve this goal, it is necessary to analyze the strengths and weaknesses of the PHC NWE. Therefore, we aim to analyze the characteristics of NWE in PHC settings in Spain, identifying these environments' strengths and weaknesses. In addition, we aim to analyze the associations between sociodemographic and professional variables in our sample, as well as the nursing professionals' perception of their NWE.

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

### *2.1. Design*

Cross-sectional study of the strengths and weaknesses of NWE in PHC settings in Spain.

### *2.2. Participants and Study Location*

We recruited a non-probabilistic, multi-stage sample of qualified nurses working in PHC settings in Spain. Due to the limited resources for data collection and the geographical dispersion of PHC professionals, the maximum possible number of participants was proposed as a sampling target during the set data collection period. Inclusion criteria for participation in this study were: being qualified as a nurse, having worked in the same PHC setting for at least 3 months and signing the consent form. Data collection was paused during vacation periods in order to limit the possibility of bias arising from staff turnover.

A total of 817 questionnaires were received, of which 115 (14.7%) were excluded. The reasons for their exclusion were: 7 (6.09%) nurses who did not work at PHC, 27 (23.48%) did not complete the questionnaires correctly and 81 (70.43%) did not meet the criteria for participation in the study.

First, the researchers disseminated the study through social networks (Twitter, Facebook) obtaining an unrestricted sample [21]. Simultaneously, direct dissemination among PHC nurses was achieved by contacting key informants via institutional e-mail (on two occasions, one month apart) to encourage and remind them to participate in the study. This was done between June 2018 and June 2019.

A data collection pack containing the consent form, a questionnaire of sociodemographic variables developed ad hoc and the Practice Environment Scale of the Nursing Work Index (PES-NWI) questionnaire was sent via email to the participants. We used Google Forms®, with a limited response via Internet Protocol (IP), to collect the data electronically. A letter of invitation to participate in the study was also attached, as well as a request to contribute to the dissemination of the study with the purpose of reaching a greater number of PHC nurses through a snowball sampling technique.

### *2.3. Instruments of Data Collection*

An ad hoc questionnaire was designed to collect sociodemographic and professional variables including age, gender, level of education, professional specialization, work experience in PHC settings and management role and responsibilities.

There are multiple measurement tools for the study of NWE [22]. The PES-NWI is among the most widely used and has high levels of consistency and reliability (Cronbach's Alpha 0.807–0.916) [9,23,24]. We used the Spanish version of the PES-NWI, which was adapted and validated for use in Spanish PHC settings by de Pedro-Gómez et al. [25] in 2012. The Spanish PES-NWI is divided into 5 dimensions, namely, nurse participation in center affairs (nine items) (D1), nursing foundation for quality of care (10 items) (D2), management and leadership of head nurse (five items) (D3), adequate human resources to ensure quality of care (four items) (D4) and nurse–physician relationship (three items) (D5). This tool includes a total of 31 items measured on a 4-point Likert scale, with scores ranging between 4 to 124. A favorable environment receives scores of >2.6, neutral or controversial environments receive scores between 2.6 and 2.4 and an unfavorable NWE receives scores of >2.4 for each dimension. Total scores are interpreted as follows: values ≥80.6 are interpreted as positive environments for nursing work, values between 74.5 and 80.5 are identified as controversial environments and values ≤74.4 are classed as negative environments for nursing work [26]. For the present study, Cronbach's alpha for the PES-NWI was 0.937, with a reliability range between 0.836 and 0.935 for each dimension.

More recently, an abbreviated version of the PES-NWI tool was developed by Gea-Caballero et al. [26] with the aim of synthetizing and prioritizing the essential elements for improving PHC settings in the Spanish context. The TOP10 Questionnaire of Assessment of Environments in Primary Health Care (hereinafter TOP10) is divided into 3 dimensions, namely, nurse participation in center affairs (D1a), quality of care (D2a) and human resources (D3a). It comprises 10 items identified as the "essential elements of care"; if not positive, these essential elements of care can seriously affect the quality of care in any given NWE. The total score of the TOP10 questionnaire ranges between 10 and 40. The higher the score, the more favorable the NWE. Cronbach's alpha for the TOP10 questionnaire was reported at 0.816 in a previous study [26]; for the present study, Cronbach's alpha was 0.805. The TOP10 questionnaire was completed by the researchers based on the results obtained from the Spanish version of the PES-NWI tool.

The information for the analysis of the TOP10 was extracted from the PES-NWI questionnaire, since the 10 essential items are part of the 31 that make up the PES-NWI.

### *2.4. Statistical Analysis*

We used descriptive statistics to analyze the sociodemographic and professional characteristics of our sample. Mean and standard deviation for quantitative variables and frequencies, and percentages for qualitative variables, were calculated.

We calculated the Cronbach's alpha coefficient of both the PES-NWI and TOP10 questionnaire to assess internal consistency and reliability as a whole, and for each dimension separately. Reliability was considered excellent when Cronbach Alpha was greater than 0.90; good between 0.80 and 0.89; acceptable between 0.70 and 0.79; questionable between 0.60 and 0.69; poor between 0.50 and 0.59 and unacceptable below 0.50 [27].

We carried out a normality test using Shapiro–Wilk and Kolmogorov–Smirnov tests as appropriate for the quantitative variables. Four variables were found to have a nonnormal distribution, namely, dimensions D1, D3, D4 and D5, and were analyzed using non-parametric methods. For the bivariate analyses of variables with a normal distribution (D2 and TOP10), we used a *t*-test for independent samples for dichotomous variables and ANOVA for polytomous variables. If significant differences in the ANOVA were found, the Bonferroni test was applied to determine which pairs of categories presented significant differences between them.

The SPSS v23 statistical package was used for the statistical analysis, and a significance level of *p* = 0.05 was adopted.

### *2.5. Ethical Considerations*

We safeguarded our participants' confidentiality and anonymity according to Spanish/European data protection regulations (Organic Law 3/2018). The participants were informed about the methods and aims of the study and gave their consent to take part in this investigation. This study was approved by the Research Ethics Committee of the Valencian Community (Xàtiva/Ontinyent, Valencia, Spain). The participants did not receive any compensation for completing the questionnaires.

### **3. Results**

The final sample consisted of 702 qualified PHC nurses. The participants were mainly women (71.9%), aged 40 years or older (61.1%) and with more than 10 years of work experience (52.4%). Most of our participants were educated to degree level (64.3%), 5.4% were nurse specialists and 11.8% were nurse managers or coordinators. Responses were obtained from 14 out of the 17 Spanish Autonomous Communities, with a greater representation from the Valencian Community and the Canary Islands (76.3%).

The results of the PES-NWI are shown in Table 1. The NWE in Spanish PHC settings was positive with a total average score of 82.4. Three dimensions were identified as strengths, namely, nursing foundation for quality of care (D2), management and leadership of head nurse (D3) and nurse–physician relationship (D5), and one dimension was identified as a weakness in the PHC settings studied: adequate human resources to ensure quality of care (D4). Nurse participation in center affairs (D1) was considered as neutral or controversial. Specifically, 17 items were identified as strengths, and only 9 were classed as weaknesses of the PHC settings (items 6, 7, 9, 12, 16, 25–28). The reliability of the PSE-NWI in our sample was confirmed with a Cronbach's alpha 0.937 and a range between 0.836 and 0.935 for each of its dimensions.

The average score from the TOP10 questionnaire was 29.7. Two of its dimensions were identified as strengths: participation in center affairs (D1a) and quality of care (D2a), and one was identified as a weakness: human resources (D3a). The reliability of the TOP10 tool in our sample was confirmed with a Cronbach's alpha 0.805.


*Int. J. Environ. Res. Public Health* **2021**, *18*, 434

*Int. J. Environ. Res. Public Health* **2021**, *18*, 434


74.4–80.6; negative environment: PES-NWI score < 74.4; \* TOP10 items.

The bivariate analysis of the results from PES-NWI and the sociodemographic and professional variables are shown in Table 2. Age was statistically significant for the PES-NWI and 3 out of its 5 dimensions (D2, D3 and D5). The participants' level of education was statistically significant for the overall PES-NWI and 4 of its dimensions (D1, D2, D3 and D5), and management role was found to be statistically significant for the PES-NWI and 4 of its dimensions (D1, D2, D3 and D4). We did not find a statistically significant correlation between PES-NWI and the rest of the sociodemographic and professional variables.

**Table 2.** Bivariate analysis of the results from PES-NWI and the sociodemographic and professional variables.


M: mean; SD: standard deviation; D1: participation of nursing staff; D2: nursing foundation in quality of care; D3: capacity, leadership and support of nursing staff by managers; D4: the size of staff and adequacy of human resources; D5: relationships between nursing and medical professionals; \* significant values *p* < 0.05; *p* a : ANOVA; *p* b : Student's *t*-test.

> We investigated gender inequalities in our sample (Table 3). Our results show that most of the nurses were aged ≤50 years (*p* = 0.031). The representation of women at higher education levels (masters or doctorate) is proportionally lower than that of men (*p* = 0.024) but, paradoxically, women achieve a higher percentage of specialist training (*p* = 0.048).

**Table 3.** Distribution and comparison of socio-demographic and occupational data by gender.


*p*: chi square test; \* *p* < 0.05

### **4. Discussion**

The main objective of this study was to identify strengths and weaknesses in the PHC work environment in Spain. Our results suggest that positive NWEs in PHC in Spain are characterized by nursing foundations for quality care (D2), management and leadership of the head nurse (D3) and the nurse–physician relationship (D5). This is consistent with previous studies in our context [16,17]. Nurse participation in center affairs (D1) was identified as neutral or controversial in our study as opposed to a previous study by Gea-Caballero et al. [16], where it was identified as a strength. Overall, our results differ from those obtained by de Pedro-Gómez et al. [28], who classified the NWE in PHC settings in the Balearic Islands as controversial (80.4 points).

This study fits well with the improvement model proposed by Poghosyan et al. [20]. Political decision-making and organizational innovation in PHC settings are key to improve identified weaknesses. Furthermore, research in healthcare settings is essential to not only increase knowledge of, and improve, both processes and procedures, but also to create an organizational culture that promotes the integration of the best available evidence [18], thus improving patient outcomes and increasing service user satisfaction.

Given the difficulty in finding other studies in the PHC setting, and due to their conceptual proximity, we compared our results with those reported in studies about magnet hospitals (as described in the introduction). Our results show that dimensions D2, D3 and D5 are associated with a positive NWE. This is in agreement with the results from previous studies carried out in "non-magnet hospitals". These are encouraging findings, but they are still far from those obtained in "magnet hospitals", where every single dimension of the PES-NWI was identified in historical studies as a strength [29,30]. This is an encouraging finding as it demonstrates that the transformation of weak or controversial dimensions into strengths is possible, as evidenced by the results obtained in magnet and excellent work environments.

In Spanish hospitals, the same three dimensions, namely, D2, D3 and D5, were shown to be neutral or controversial for the NWE [31]. This diversity in the results suggests that the quality of the NWE in the hospital context depends on external as well as internal characteristics of the healthcare service. Therefore, interventions to improve the NWE in the PHC context should be individualized and based on the results obtained from each separate healthcare institution (microenvironment). The same reflection is applicable to PHC work environments. However, a study by de Pedro et al. [28] identified D2 and D3 only as strengths in hospitals with 300–500 beds. Paradoxically, in international studies about the characteristics of the NWE, we find a greater diversity of scenarios; some identify all the dimensions of the PES-NWI as strengths [32,33], others show management and leadership of the head nurse and nurse–physician relationship as strengths (D3 and D5) [34], and some identify management and leadership of the head nurse as the only strength (D3) [35].

In the PHC NWE in Spain, the size of the workforce or human resources (D4) is identified as a clear weakness, coinciding with national studies in both PHC [16,17,28] and hospital [28,31,36] work environments. These results coincide with those portrayed in international studies about NWE in the hospital setting [34,35]. The comparative studies between "magnet and non-magnet hospitals" reported similar results in historical studies, with human resources (D4) being identified as a weakness in "non-magnet hospitals"; it was not identified as a weakness in "magnet hospitals", but it was the worst valued dimension [29,30]. This same situation was also observed in international studies, both European [32] and Asian [33], with human resources usually being the worst valued dimension. The problem with human resources is particularly serious in Spain, where the nurse–patient ratio is 567 per 100,000 inhabitants, well below the European average (811/100,000) and far from the more industrialized countries, such as Finland, Denmark or Belgium (1500/100,000) [37]. Despite the efforts made in recent years to increase the nursing workforce, and the commitment to nurses as health agents, it is still a limitation that compromises patient safety and quality of care. Furthermore, the nurse–physician ratio in Spain is severely unbalanced. According to the Organisation for Economic Co-operation

and Development (OCED) [38], the number of physicians per inhabitant in Spain is above average (7th place and above countries such as Italy, Australia, France and Finland), but the number of nurses is well below the average worldwide (23rd place out of 26 countries). Finland and Germany triple the number of nurses in Spain, and Norway quadruples it. It should not be forgotten that there is a direct correlation between the ratio of nurses and patient mortality, as well as other unwanted events and health outcomes [5–9].

These facts, framed in a global SARS-COV-2 pandemic, reveal and exacerbate existing problems within the healthcare service. For example, the COVID-19 pandemic has added undue pressure to the health services in Spain, thus highlighting the lack of qualified nurses. As suggested by Seccia Ruggero [39], the replenishment of material resources can be achieved relatively easily, but reinforcement with qualified nurses is difficult to achieve and cannot be done over a short period of time. An adequate nursing provision could contribute to improved outcomes in health crises, such as at the peak of the COVID pandemic, which has led the WHO in April 2020 to call for more investment in nurses [40]. Key stakeholders and those responsible for decision making on healthcare service planning should consider the need to increase the Spanish nursing workforce and draw a plan accordingly in the years to come.

The results from the TOP10 scale [26] were fully consistent with the PES-NWI results, identifying participation of nurses in the affairs of the center (D1a) and the nursing foundations for quality of care (D2a) as strengths, and human resources (D3a) as a weakness. We argue that TOP10 is a simpler way of identifying the strengths and weaknesses associated with the NWE, making it easier and simpler for nurse managers to identify weakness or areas for improvement within their PHC work environments. In addition, the results from the TOP10 scale are valid and reliable as supported by a recent study by Martínez-Riera et al. [41], where a group of community care experts considered that 9 out of the 10 items of the TOP10 scale were essential elements to the PHC NWE.

The comparative study of the sociodemographic and professional variables and the perception of the PHC NWE shows significant differences associated with age, level of education and the level of management in which the professionals were involved. Older nurses (50+) were the most critical with their work environments. In addition, significant differences were found for dimensions D2, D3 and D5 separately. The agerelated differences found in Spanish studies should be assessed with caution due to the average age difference between nurses employed in public and private services, with greater representation in privately managed centers of the age range under 40 years [16].

Nurses educated to doctoral level identified the Spanish PHC NWE as a negative environment for nursing care and also pointed to the dimensions of nursing participation in center affairs (D1), nursing foundations for quality of care (D2) and human resources (D4) as weaknesses. Interestingly, the dimension nursing foundations for quality of care (D2) was identified as a weakness by doctoral nurses and as a strength by the rest of the nursing professionals. The same was observed in a previous study [17] in the Community of Madrid. This may suggest a lower level of job satisfaction among the most the nursing professionals with a highest level of education [33], or perhaps it may reflect a greater capacity for critical thinking. This situation is paradoxical. PHC nurses look after an ageing population with highly complex and chronic conditions. Thus, it would seem reasonable to integrate nurses with high levels of training and those in advanced practice roles in PHC settings [42]. The International Council of Nurses [43] defines advanced nurse practitioners as professionals who have acquired the theoretical knowledge, complex decision-making skills and clinical competencies for extended practice in the country and context for which they are accredited. Advanced training, such as a master's or doctoral degree, is recommended for an advanced nursing practice qualification [44]. Our results show that highly qualified nurses (doctoral level) value their work environment the least, reflecting the fact that the work environment may not be adapted to the academic level of these professionals. We argue that it is necessary to ensure that highly qualified nurses and those in advanced practice roles [45] are able to work to their full potential within PHC

settings in Spain, and recommend that aspects such as the nurses' level of training and expertise, and not simply their seniority and years of experience, are taken into account when designing nursing career pathways. Advanced practice nursing.

The nurses in a management role identified nursing participation in the affairs of the center (D1), nursing foundations in the quality of care (D2), management and leadership of the head nurse (D3) and human resources (D4) as strengths, with their score being higher than that of their staff nurse colleagues. This was also the case in previous studies carried out in Spain [16,17]. Interestingly, the human resources dimension (D4), which was recognized as a weakness in our study, as well as in previous studies [16], was not identified as such by the nurse managers, who considered it to be a strength. We believe that this phenomenon should be analyzed further through in-depth qualitative interviews with nurse managers, as well as other key stakeholders, in order to fully understand the root cause of this problem. Namely, it is possible that there are specific factors which are affecting the participants' assessment of the impact of the nursing workforce on the NWE. This may include the level of participation of the highest trained professionals and the quality of the relationship between the nurse managers and the rest of the staff.

Finally, from a gender perspective, no significant differences were observed when comparing the NWE with the gender of the participants in our study.

### *Limitations*

We wish to highlight a number of limitations. First, our cross-sectional design does not allow us to infer causality in the relationships between variables. Second, although our sample is larger than that of previous PHC NWE studies, we cannot guarantee the representativeness of the entire nursing population in Spain as some of the Spanish territories are either not represented or under-represented. Third, although precautions were taken to control for duplicate responses, it is possible that some scaped our scrutiny. For these reasons, we recommend that further studies analyzing the NWE in PHC settings with more powerful samples are carried out in order to confirm these data.

### **5. Conclusions**

The NWE in PHC settings in Spain is positive and comparatively better than the NWE in hospital settings. We identified the following strengths: (1) nursing foundation for the delivery of care, (2) management and leadership of the head nurse and (3) nurse– physician relationship, and the following weakness: (1) participation of nurses in the affairs of the center and (2) human resources. We argue that there is room for improvement of the NWE in PHC settings in Spain, and that efforts should be directed towards the neutral and negative aspects identified. Two groups of nurses were particularly critical of their NWE, namely, older nurses and those educated to doctoral level. Nurse managers did not identify human resources as a weakness, contrary to the results from previous national and international investigations. We found no evidence of gender influence on the results obtained.

**Author Contributions:** Conceptualization, V.G.-C., R.J.-V. and J.R.M.-R.; methodology, V.G.-C., R.J.-V., M.V.V.-G., I.A.-S. and P.G.-M.; software, R.J.-V., J.C.-M., I.S.-A. and P.G.-M.; validation, V.G.-C. and J.R.M.-R.; formal analysis, P.G.-M., J.C, I.S.-A. and R.J.-V.; investigation, V.G.-C., R.J.-V., M.V.V.-G., I.A.-S., M.V.V.-G., J.C.-M. and M.V.V.-G.; resources, J.C.-M., I.S.-A. and M.V.V.-G.; data curation, P.G.-M., J.C.-M., I.A.-S., I.S.-A. and R.J.-V.; writing—original draft preparation, V.G.-C., J.R.M.-R., J.C.-M., M.V.V.-G., I.A.-S., I.S.-A. and P.G.-M.; writing—review and editing, V.G.-C., R.J.-V. and P.G.-M.; visualization, J.R.M.-R. and M.V.V.-G.; supervision, R.J.-V., J.R.M.-R. and M.V.V.-G.; project administration, V.G.-C. and J.R.M.-R. 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 Health Department of Xàtiva/Ontinyent (26-2-13), and others Ethics Committes: Health Department of Elx-Crevillent (21-3-14), Health Department of Elx Hospital General (6-11-13), Health Department of Torrevieja (21-3-14), and Health Services of Canary Islands (19-6-17).

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

**Data Availability Statement:** Not applicable.

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

### **References**


### *Article* **Friendship and Consumption Networks in Adolescents and Their Relationship to Stress and Cannabis Use Intention**

**María Cristina Martínez-Fernández <sup>1</sup> , Cristina Liébana-Presa 1,\* , Elena Fernández-Martínez <sup>2</sup> , Lisa Gomes <sup>3</sup> and Isaías García-Rodríguez <sup>4</sup>**


**Abstract:** Background: Cannabis is an illegal psychoactive substance that's use is widespread among adolescents. During adolescence, many changes can cause stress. In this phase, the group of friends becomes increasingly important, being a situation of vulnerability for the beginning of cannabis use, either as an escape mechanism or due to peer's influence. Therefore, the purpose of this study is to describe and analyze the structure of the consumption and friendship network, the intention to use cannabis, and the stress in a secondary school class. Methods: An online platform with validated self-reported questionnaires were used for data collection. Results: The sample consisted of adolescents (*n* = 20) aged 14–16 from a third-year class of compulsory secondary education in Ponferrada (León, Spain). Significant differences were obtained concerning consumption intention and the different network metrics in both the friendship and consumption networks. Subsequently, the representation of these networks was carried out. Conclusions: Social Network Analysis is a very useful tool that provides a picture of the context in which adolescents are located. In the consumption network, there are central actors who have not yet consumed cannabis; this is a crucial moment to implement prevention strategies.

**Keywords:** cannabis; adolescents; stress; social network analysis; network; friendship

### **1. Introduction**

According to the World Health Organization (WHO), cannabis is the most widely consumed drug among young people in 2018; approximately 4.7% of young people aged 15 to 16 years had consumed it at least once [1]. In Spain, the most recent data indicate that cannabis is the illegal psychoactive substance with the highest prevalence of use, with an average age of onset of 14.8 years; in fact, 398,600 students aged 14 to 18 years had consumed cannabis during 2016 [2]. Substance use is related to multiple risk factors, including school failure and problematic behaviors [3]. Additionally, substance use can lead to physical, psychological and social disorders that demand the design of effective prevention policies [4]. Substance use and abuse is, therefore, one of the main risk factors for health, where adolescent users are more likely to manifest social and personal issues, lower psychological adjustment and emotional competence [5].

Adolescence is a period of change. A key characteristic of youth is that it is a relevant phase in the consolidation, gain or loss of previously acquired habits and lifestyles, which has an impact on the future health status of individuals [6]. Emotional and behavioral adjustment problems have been seen as mediators between cannabis use and psychosis risk [7]. Early initiated cannabis use results as a significant marker of mental health and

**Citation:** Martínez-Fernández, M.C.; Liébana-Presa, C.; Fernández-Martínez, E.; Gomes, L.; García-Rodríguez, I. Friendship and Consumption Networks in Adolescents and Their Relationship to Stress and Cannabis Use Intention. *Int. J. Environ. Res. Public Health* **2021**, *18*, 3335. https://doi.org/10.3390/ ijerph18073335

Academic Editor: Paul B. Tchounwou

Received: 26 February 2021 Accepted: 21 March 2021 Published: 24 March 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/).

behavioral risk [8]. Therefore, well-being in adolescents plays a key role in preventing substance use [7]. Moreover, burnout experienced by adolescents in high school can have an influence on cannabis use, being relevant to the consideration of academic stress in the prevention of cannabis use, since its use causes a loss of academic expectations and motivation [9].

As previously mentioned, cannabis use may appear as an escape mechanism from stress [10]; in fact, those adolescents who do not use cannabis cope with stress in a more flexible way [11]. There are cognitive schemes based on beliefs of grandiosity and insufficient self-control that are significantly associated with drug use [12]. The theory of planned behavior provides an explanation for cannabis use and describes the relationship between cognitive characteristics of individuals and the development and maintenance of behavioral patterns [13]. Developmental theories indicate that the transition to adolescence is decided by an increase in the frequency of peer interactions, the adoption of more sophisticated interpersonal behaviors, new social roles and experiences, adolescents' motivation to develop a stable sense of identity, and young people's reliance on peer feedback (and their perceived status by peers) [14].

In this context, the concept of Social Network Analysis (SNA) emerged, is proving an impact on the health. Peers are fundamental in the social organization of adolescents, as they are present in the academic environment and activities. Likewise, during adolescents' development, they increasingly acquire more autonomy and independence from their family environment [15,16]. SNA has been widely applied in public health, social support, social capital, influences on health behaviors, and the social structure of information dissemination [17]. Many studies have been carried out using the SNA in adolescents, indicating how the position in the peer group and popularity have an influence on relevant aspects, such as leisure activities and eating behaviors that spread in the friendship network, influencing overweight [18] and even sleeping habits, where the most popular adolescents sleep less than the rest of the individuals in the network [19].

SNA is a theoretical and methodological paradigm that makes possible to evaluate the relational context empirically and to capture contexts of social interaction, which determine the behavior of the actors that are part of that context [17]. Thus, one of the key components in adolescent well-being are social networks. The literature has found that having friends or being connected to friendship networks that exhibit risky behaviors (smoking or drinking alcohol) implies an increased risk of engaging in these behaviors, both initially and over time [20]. SNA can be helpful in better understanding the mechanisms underlying the connection between friendships and risk behaviors [20]. Adolescent contacts are important for the establishment of adolescent health, as well as for the acquisition and maintenance of risk behaviors [21,22].

Peer influence is defined as a phenomenon characterized by the presence of both selection and socialization. By understanding why adolescents fit in with their peers, it is possible to develop preventive measures that alternatively address the psychological motivations that lead to conformity at present [14]. However, peer influence from adolescence to adulthood has been found to be later much less clear [23].

Following the exposed problem, this study has the following question: what is the relationship between the structural characteristics of friendship and cannabis use networks, an individual's intention to use cannabis and young adolescents' stress? The SNA can become a useful tool that will allow us to know the behavioral pattern of the network in order to plan concrete and effective interventions on the identified risk behaviors (cannabis use intention and stress), in short, to promote healthy and sustainable networks. Therefore, the aim of this study is to describe the structure of the consumption and friendship network, the intention to use cannabis and the stress of young adolescents in a secondary school class, and, furthermore, to analyze the relationships between network structural variables, consumption intention and stress and to represent these relationships in order to identify peer leaders and plan interventions that promote health (less stress and less consumption).

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

A cross-sectional, descriptive and correlational study was conducted. A nonprobabilistic convenience sample was selected. The sample is composed of all students enrolled in a 3rd-year class in a Compulsory Secondary Education (E.S.O.) center in the city of Ponferrada (León, Spain) during the 2019/2020 academic year. The criteria for selecting this class are twofold: (i) the age of the students is at the age of initiation of cannabis use and (ii) previous academic courses that ensure the coexistence of students. The selected class has 20 students.

### *2.1. Variables and Measuring Instruments*

Stress: Stress was measured using the Student Stress Inventory–Stress Manifestations (SSI–SM) [24] validated in Spanish for adolescents [25]. This questionnaire consists of 22 items, with a five-point Likert-type scale (not at all, rarely, sometimes, often and totally). These items are distributed in three factors: emotional (α = 0.79), physiological (α = 0.62) and behavioral (α = 0.66).

Cannabis use: To measure cannabis use, the Spanish Survey on Drug Use in Secondary Education (ESTUDES) [2] is used. This questionnaire is composed of different items, according to the substances, from which we selected those referring to the block of cannabis use. The section is made up of 10 items related to the time that cannabis has been used, the first time it was used and how it is used.

Cannabis Use Intention: The validated Cannabis Use: Intention Questionnaire (CUIQ) was used for the youth population [26]. This questionnaire comprises 12 items. Each one of the items is evaluated by means of a Likert-type scale from 1 to 5 points.

Networks: To determine the classroom consumption and friendship network, a limited census of actors is used that is matched to the list of classroom peers. Students are asked to nominate only those actors/peers in their class with whom they would go out to consume. With the data collected, a sociocentric matrix is obtained. In the same way, to obtain the classroom friendship network, students were asked to nominate from the census of actors those classmates with whom they share their free time. A 4-point Likert-type scale was used, where 0 means "I never share my free time," and 4 means "We are always together."

### *2.2. Procedure*

Data were collected by online questionnaire. Authorization was obtained by the corresponding Compulsory Secondary Education center, and the teacher of the class involved was contacted to make them aware of the procedure. Data collection took place on different days in February 2020. The online questionnaire was carried out in different web programming languages, PHP (Zend, Minneapolis, MN, USA) and MySQL (Oracle, Santa Clara, CA, USA) for its dynamization, together with a front-end based on HTML5 (World Wide Web Consortium (W3C), Cambridge, MA, USA), CSS (World Wide Web Consortium (W3C), Cambridge, MA, USA), JavaScript (Oracle, Santa Clara, CA, USA) (and jQuery (The OpenJS Foundation, San Francisco, CA, USA)), complying with different standards and measures that facilitate its visualization on different devices (responsive design).

### *2.3. Data Analysis*

Qualitative variables are shown as frequencies and percentages. Quantitative variables were expressed as mean and standard deviation. After verifying that the quantitative variables did not follow a normal distribution, using the Kolmogorov–Smirnov test with Lilliefors correction, nonparametric correlations were performed, and Spearman's rho coefficients were obtained. Statistical analyses were carried out using Statistical Package for the Social Sciences software (SPSS v. 26.0) (IBM, Armonk, NY, USA).

For network analysis, the data obtained were transferred to Excel and processed using the UCINET V 6.0 program and NetDraw [27]. Centrality measures (see Table 1) were calculated for the participants.


**Table 1.** Definitions of the Social Network Analysis (SNA) centrality metrics.

### *2.4. Ethical Considerations*

The anonymity and confidentiality of the study subjects were considered at all times. Being underage minors, prior authorization was received from their parents or legal guardians for participation in the study, as well as the informed consent of the participants. The data obtained from the research will be treated in accordance with both the Constitutional Law 3/2018, of December 5, on the Protection of Personal Data and Guarantee of Digital Rights and the General Data Protection Regulation of the European Union EU 2016/679 (GDPR). In addition, permission was requested from the educational center and the competent body for education in the region (Consejería de Educación de la Junta de Castilla y León). The study was approved by the ethics committee (ETICA-ULE-035-2019) of the University of León (Spain), which ensures compliance with ethical and legal aspects.

### **3. Results**

The sample consisted of a total of 20 adolescents, of which 10% were female (*n* = 2), and 90% male (*n* = 18), with an age measurement of 14.45 ± 0.61 (min = 14; max = 16).

First, regarding the results of cannabis use, the data obtained place the prevalence of cannabis use at 10% (*n* = 2). Of these, one individual's last cannabis use was 40 days ago, while the other shows a more habitual use, being the last use 3 days ago. The age of onset of use was 13 years old. One of the two consumers used cannabis mixed with tobacco. When asked that, if cannabis consumption were legal would they consume it, 20% of the students said that they would consume it, while 15% of the sample had already tried it (including consumers). Finally, it is worth noting that one student consumed cannabis alone quite often, while the other indicated that he had never consumed cannabis alone. Regarding the cannabis use intention, the total values are 1.35 ± 1.13; however, we find a high maximum, that result in the equivalence table indicates that it is above the 90th percentile. This means that 90% of young people of the same age have a lower intention to consume than this one.

Secondly, the descriptive results obtained for the variable stress and are set out in Table 2. Stress obtained total values of 37.65 ± 18.45.


**Table 2.** Descriptive statistics of stress.

Note: SSI–SM; Student Stress Inventory–Stress Manifestations. Min–Max; Minimum–Maximum. M; Mean. SD; Standard Deviation. Me; Median.

Thirdly, Table 3 shows the values of the structure of the consumption and friendship networks, the actors who may be in more central positions in the networks and with greater degree of influence. The following parameters are described: indegree, outdegree, degree of proximity (out/in closeness), betweenness, and influence through the eigenvector. It is found that networks are similar in terms of values and density, which may suggest that adolescents select their friends as the peers with whom they would go out to consume.


**Table 3.** Descriptive indicators of centrality in consumption and friendship networks.

Note: Min–Max; Minimum–Maximum. M; Mean. SD; Standard Deviation.

Figure 1 illustrates the distribution of normalized betweenness in the friendship network, showing that a large number of individuals have a very low intermediation capacity. In fact, 30% of the individuals have no intermediation capacities at all.

**Figure 1.** Normalized Betweenness (nBetweenness) centrality distribution.

A correlational analysis between the different variables, stress, cannabis use intention and friendship and consumption networks is shown in Table 4.


**Table 4.** Correlations between centrality metrics of friendship and consumption networks, stress and cannabis use intention.

Note. CUIQ; Cannabis Use: Intention Questionnaire. SSI–SM; Student Stress Inventory–Stress Manifestations. Rho: Spearman's correlation.

\* Correlation is significant at the 0.05 level. \*\* Correlation is significant at the 0.01 level.

No significant correlations were found for cannabis use with these data, given the small sample size of the consumers. Behavioral manifestations of stress involve behaviors, such as acting defensively, neglecting friendships or showing negative attitudes in different interpersonal relationships. In this regard, it is noteworthy that a statistically significant correlation was found between the behavioral manifestations of stress dimension and the metrics of out closeness (r = 0.541), outdegree (r = 0.530) and the degree of betweenness (r = 0.496). In addition, statistically significant correlations were found in this network for in closeness (r = −0.511), emotional manifestations and physical manifestations and outdegree (r = 0.446). The cannabis use intention appears statistically significantly correlated with different network metrics: for the friendship network, outdegree (r = 0.596), out closeness (r = 0.531) and betweenness (r = 0.598), while, in the consumption network, outdegree (r = 0.252), out closeness (r = 0.500) and betweenness (r = 0.549). On the other hand, it shows a statistically significant correlation with total stress (r = 0.622) and its different manifestations: emotional (r = 0.510), physiological (r = 0.687) and behavioral (r = 0. 0479).

Figure 2 represents the classroom friendship network, where males are represented with a blue color and females with a pink color. Cannabis users are identified as squareshaped nodes and nonusers as circles. The size of the nodes varies according to their total stress. The intensity of the relationships is measured through the strength of the ties (0–4), where a greater strength indicates that these students are united by a bond of friendship by spending a large part of their free time together. The size of the nodes varies as a function of the total stress scores. Thus, we find how individuals who are consumers are close and could be considered friends. These nodes, although they have relationships with those actors who are more central, establish ties of greater intensity with actors who are on the periphery. However, it stands out how the girl who uses cannabis establishes ties of friendship with the central actors, being, in addition, one of the individuals with higher levels of stress. In general, we found a dense network, where relationships of varying intensity were high. We found an isolated node and a node on the periphery that only maintains a relationship with two individuals from the others in the class, being also one of the individuals with the highest level of general stress.

**Figure 2.** Friendship network structure and the study variables; (**a**) total of Student Stress Inventory–Stress Manifestations, (**b**) bbehavioral of Student Stress Inventory–Stress Manifestations.

> Figure 3 shows the classroom consumption network from which students were asked to select, out of all their classmates, those individuals with whom they would consume. As in the previous case, the size of the nodes varies as a function of the total stress scores or the behavioral manifestations of stress. The structure is similar to the friendship network, since the central actors are maintained, in this case, the central actor being a cannabis user. Consequently, it may be an indication that the friendship network influences cannabis consumption, being that this consumption is accepted by those nodes with whom they share more time.

**Figure 3.** Consumption network structure and the study variables; (**a**) total of Student Stress Inventory–Stress Manifestations, (**b**) behavioral of Student Stress Inventory–Stress Manifestations.

> Figure 4 shows the two networks, but, in this case, the size of the node varies depending on the cannabis use intention. Although most of the subjects obtained scores according to the mean, we found nodes of greater size, whose scores were indicative of an intention

higher than normal. These actors are found both in the periphery and in the center of the network and are individuals in whom action must be taken through different intervention strategies when they are in a situation of risk at the beginning of cannabis use.

**Figure 4.** Cannabis use intention (**a**) consumption network; (**b**) friendship network.

### **4. Discussion**

The aim of this article is to describe the consumption and friendship network of adolescents and to relate it to stress in its different manifestations and the cannabis use intention. Subsequently, the representation of the friendship and consumption network is carried out for those variables in which significant results have been obtained: the intention to use cannabis, total stress and behavioral manifestations of stress, which are related to acting defensively, neglecting friendships, talking more about peers and teachers, picking on others, etc.

SNA allows us to locate those individuals with positions of influence within the network. Although multiple factors influence cannabis use, the literature indicates that family stressors have a direct impact on the progression to problematic cannabis use, as well as their consequent indirect effects through the school experience of young people [30]. Furthermore, cannabis use causes difficulties in school performance, creating a lack of motivation and interest that feeds back by increasing cannabis use [9]. In adolescence, there is a change in the relationship system of young people, from focusing on the family to associating with peers; in this sense, the SNA helps to explain personal attachment to the community from a local level in the structure of the friendship network and supports the peer influence theory [31]. Previous experiences have employed the SNA in school settings with adolescents to identify those who are considered leaders among their peers and train them to deliver e-cigarette prevention programs to the rest of the class [32]. In this case, through SNA, we have identified the most central actors, two individuals who are in central positions in the consumption network, who may be in positions of risk of initiation, problematic consumption or ability to transmit these behaviors to the rest of their peers, and through whom to carry out prevention strategies and interventions customized for each setting.

In addition, it is necessary to consider the emotional and behavioral adjustment problems that mediate the relationship between cannabis use and risk of psychosis [7], as this study points out, highlighting the importance of prevention by focusing on the mediating

role of emotional and behavioral problems to train young people in socioemotional competencies in school contexts [7]. The literature highlights the importance of prevention and addressing adolescents before actual cannabis use takes place; as [33] says, adolescents who think more positively about being under the influence of marijuana, those with greater approval from their social environment, and those with less confidence in their ability to abstain from using have a greater intention to initiate marijuana use [33]. In addition, adolescents will have a more favorable attitude toward drug use if their contacts with environments and inciting companies and the contacts with drugs maintained by friends are greater [34]. Thus, SNA can be used as an intervention strategy to promote behavioral changes, since networks influence the health of their members by generating a context with its own behavior and norms where members can influence each other through persuasion, information exchange or support [35]. In this sense, the correlation found in this study between the betweenness centrality metric and the intention to use cannabis could be a potential danger, regarding the spreading of this unhealthy habit, as the individuals with higher betweenness values play an important role in the dissemination of behavior through the networks. This is especially relevant if one takes into account that 30% of the individuals have no intermediation capacities in the network.

This study presents advances in the area of cannabis use, since it is the first study to analyze the friendship and consumption network, stress and cannabis use intention. However, it has some limitations that should be considered: This is a small sample, which is not representative of the population, and the results should be interpreted with caution. In addition, there is a possibility that some participants did not declare their consumption, as can be seen in the classroom consumption network with very central actors not consuming, where it can be seen that those who have consumed are very central, but not all of them. In addition, in future studies, it would be useful to carry out longitudinal designs that can indicate how these variables behave over time and be able to carry out causal explanations, as well as designs where health education interventions are proposed to these population groups.

### **5. Conclusions**

SNA is a useful tool that provides a picture of the context where adolescents are located. It allows identifying those that are more isolated, which is considered a disadvantage, and the most popular, who can be chosen as role models by their peers. After knowing the consumption network of the class, we find central actors who that have not initiated cannabis consumption. This indicates a crucial moment to carry out prevention strategies adapted to each school context. These strategies should be implemented at earlier ages, since we found several cannabis-consuming students who may have been influenced by their peers.

**Author Contributions:** Conceptualization, M.C.M.-F., C.L.-P. and I.G.-R.; methodology, M.C.M.-F., C.L.-P. and I.G.-R.; software, M.C.M.-F., C.L.-P. and I.G.-R.; validation, M.C.M.-F., C.L.-P. and I.G.-R.; formal analysis, M.C.M.-F., C.L.-P., E.F.-M., L.G. and I.G.-R.; investigation, M.C.M.-F., C.L.-P., E.F.-M., L.G. and I.G.-R.; resources, M.C.M.-F., C.L.-P., E.F.-M., L.G. and I.G.-R. data curation, M.C.M.-F., C.L.- P. and I.G.-R.; writing—original draft preparation, M.C.M.-F., C.L.-P. and I.G.-R.; writing—review and editing, E.F.-M. and L.G.; visualization, M.C.M.-F., C.L.-P. and I.G.-R.; supervision, C.L.-P. and I.G.-R. 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 Ethics Committee of University de Leon (ETICA-ULE-035-2019).

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

**Data Availability Statement:** Not Applicable.

**Acknowledgments:** We thank Benítez-Andrades J.A. for his contribution and help in the management of data collection and Marqués-Sánchez P. for her willingness, help and vision throughout the process.

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

### **References**


International Journal of *Environmental Research and Public Health*
