*Protocol* **Knowledge, Attitudes, Behavior, Acceptance, and Hesitancy in Relation to the COVID-19 Vaccine among Pregnant and Breastfeeding Women: A Systematic Review Protocol**

**Vincenza Gianfredi 1,\*, Alessandro Berti 1, Marilena D'Amico 1, Viola De Lorenzo <sup>1</sup> and Silvana Castaldi 1,2**


**\*** Correspondence: vincenza.gianfredi@unimi.it

**Abstract:** A new coronavirus, SARS-CoV-2, was identified at the end of 2019. It swiftly spread all over the world, affecting more than 600 million people and causing over 6 million deaths worldwide. Different COVID-19 vaccines became available by the end of 2020. Healthcare workers and more vulnerable people (such as the elderly and those with comorbidities) were initially prioritized, followed by the entire population, including pregnant and breastfeeding women. Despite the safety and efficacy of COVID-19 vaccines, a certain level of skepticism was expressed, including among pregnant and breastfeeding women. There were several reasons for this reluctancy, among them, fear of side-effects for both women and fetuses. Nevertheless, acceptance, as well as hesitancy, were time, country and vaccine specific. This review will collect available evidence assessing knowledge, attitudes, behaviour, practice and acceptance/hesitancy of pregnant/breastfeeding women in relation to the COVID-19 vaccination. The PubMed/MEDLINE, Scopus and EMBASE databases will be consulted. A predefined search strategy that combines both free text and MESH terms will be used. The systematic review will adhere to the PRISMA guidelines and the results will be reported in both narrative and summary tables. A meta-analysis will be conducted if data are available.

**Keywords:** pregnant women; lactating; breastfeeding; COVID-19 vaccine; acceptance

#### **1. Introduction**

Vaccine hesitancy is a complex phenomenon that is listed as one of the ten threats to global health by the World Health Organization (WHO) due to the consequent decrease in vaccination coverage [1]. Several definitions of vaccine hesitancy have been proposed. The WHO defines vaccine hesitancy as a delay in the acceptance or refusal of vaccines despite their availability. Dubè et al. considered vaccine hesitancy to comprise a spectrum, with active demand for vaccines or their complete refusal at the two ends and vaccine-hesitant individuals in between [2]. However, heterogeneities in attitudes/practices/knowledge exist, with some people refusing some vaccines but agreeing to take others, or some others delaying or accepting vaccines but being unsure to do so. In addition, Peretti-Watel et al. defined vaccine hesitancy as a complex decision-making process influenced by several contextual factors [3]. On this basis, in 2015, the Strategic Advisory Group of Experts on Immunization (SAGE) Working Group on vaccine hesitancy developed a theoretical model named the 3Cs to explain the complexity of vaccine hesitancy and its determinants [4]. The 3Cs refers to the three main factors: *complacency*, *convenience* and *confidence*. From 2015 to the present, the 3Cs model was revised based on the outcome of literature reviews and theoretical considerations [5]. The new 5Cs model is based on five psychological antecedents of vaccination: (1) *confidence* in the safety and efficacy of vaccines and trust in the system and providers that deliver them, (2) *complacency*, reflecting a low-perception of the risks linked to the vaccine-preventable disease and the consequent belief that the vaccination is not necessary; (3) *constraints*, physical and psychological barriers that cause vaccination to be perceived as inconvenient, threatening the

**Citation:** Gianfredi, V.; Berti, A.; D'Amico, M.; De Lorenzo, V.; Castaldi, S. Knowledge, Attitudes, Behavior, Acceptance, and Hesitancy in Relation to the COVID-19 Vaccine among Pregnant and Breastfeeding Women: A Systematic Review Protocol. *Women* **2023**, *3*, 73–81. https://doi.org/10.3390/ women3010006

Academic Editors: Claudio Costantino and Maiorana Antonio

Received: 28 December 2022 Revised: 16 January 2023 Accepted: 16 January 2023 Published: 20 January 2023

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

conversion of vaccination intention into actual behavior; (4) *calculation*, the active effort in searching for information about risks and benefits of vaccination, though this commitment is not always associated with the ability to understand studies and data; and (5) *collective* responsibility, defined as "the willingness to protect others by one's own vaccination by means of herd immunity. The flipside is the willingness to have a free ride when a sufficient number of other people are vaccinated".

The 3C or 5C models provide a framework within which vaccine hesitancy/acceptance can be analyzed. In this respect, many studies have been conducted to date to better understand factors associated with vaccine hesitancy/acceptance. These factors include sociodemographic factors, such as ethnicity, age, sex, education, and employment; factors that depend on geographic or social context, such as accessibility and cost; the safety and efficacy of a new vaccine, lack of information or vaccine misinformation; and more personal factors, such as individual responsibility and risk perceptions and trust in health authorities and vaccines [6].

Among sociodemographic factors, pregnancy or breastfeeding represents a crucial period in which women seek information that will guide health decisions in relation to themselves and the unborn child. The safety of the unborn infant has been suggested to be the primary driver of decision-making during pregnancy [7]. In this context, exploring the knowledge, behavior and practice of vaccine acceptance/hesitancy among pregnant/breastfeeding women is important to plan counseling/education actions. This is particularly apposite considering the COVID-19 pandemic and the ensuing availability of new vaccines. A growing body of evidence suggests that COVID-19-related morbidity and mortality has been higher among pregnant women compared with age-matched non-pregnant individuals [8]. However, the novelty of the type of vaccines used, as well as the paucity of data about long-term effects (safety and efficacy) of the COVID-19 vaccine among the general population and pregnant/breastfeeding women, might impact on vaccination acceptance [9]. Emerging data from the Center for Disease Control and Prevention suggests that there has been no increase in side-effects or complications among pregnant women vaccinated against COVID-19 [10]. Considering data on the safety and efficacy of COVID-19 vaccines, public health programs have prioritized pregnant women as a high-risk group for COVID-19 infection and its complications. However, a certain level of vaccine hesitancy is commonly observed among pregnant women [11]. Moreover, pregnancy represents a time during which women, and the related family, are looking for information that can guide their own health choices and those of the unborn child [12]. Considering the above, the novelty of the COVID-19 vaccines, and recognising that vaccine hesitancy is vaccine-specific and depends on the socio-cultural background, we designed a systematic review and meta-analysis protocol with the purpose of better understanding the mechanisms underlying COVID-19 vaccine hesitancy among pregnant/breastfeeding women. Specifically, we aimed to assess the knowledge, beliefs, attitudes, barriers and facilitators relating to acceptance/refusal of the COVID-19 vaccine. Understanding the mechanisms underlying vaccination hesitancy is key to the design, testing and implementation of interventions that can improve vaccine acceptance and coverage in routine and outbreak settings.

#### **2. Experimental Design**

This systematic review protocol was developed based on The Preferred Reporting Items for Systematic Reviews and Meta-Analyses 2015 guidelines, as extended for systematic review protocols (PRISMA-P) [13]. The review protocol was developed, shared among the authors, and submitted to the journal in advance, prior to commencing the review. The review will be conducted in accordance with the Cochrane Collaboration [14] and the results will be reported based on the PRISMA 2020 guidelines [15]. Moreover, if the necessary data are available, we will proceed with statistical pooling and a meta-analysis will be performed. The latter will be carried out and documented in accordance with the Meta-analysis Of Observational Studies in Epidemiology (MOOSE) guidelines [16].

#### *Research Question*

The purpose of the systematic review (potentially with meta-analysis) is to answer the following questions: (1) What is the level of knowledge regarding COVID-19 vaccination among pregnant/breastfeeding women? (2) What are the facilitating/barrier factors associated with pregnant/breast-feeding women's acceptance/hesitancy to receive COVID-19 vaccine?

#### **3. Materials and Equipment**

#### *3.1. Information Sources*

A comprehensive, structured electronic search will be developed based on the research questions and conducted by checking three different scientific databases: PubMed/MEDLINE, Excerpta Medica Database (EMBASE) and Scopus. The electronic search will be conducted in the three databases during the same day by two different authors. If any discrepancy in records identification occurs between the two authors, it will be solved through discussion with a third (senior author) involved in retracing all the steps taken and checking for errors. The literature search will be supplemented by a review of the reference lists of included articles and, if available, by screening the reference lists of relevant similar reviews published previously in international scientific journals, consistent with previous research [17]. In addition, experts in the field will be contacted to potentially retrieve any additional relevant articles, as previously performed [18].

#### *3.2. Search Strategy*

A comprehensive and specific search strategy will be developed, combining medical subject headings (MeSH) and free text words. The search strategy will be defined according to the population, exposure, and outcome (PEO), as suggested by the Cochrane Collaboration [14]. We will first develop a search strategy for PubMed/MEDLINE. Then keywords and search terms will be adapted for use in the other two bibliographic databases. The Boolean operators AND and OR will be appropriately and logically combined in order to build the search strategy. The search strategy will be based on the following terms: (P) pregnant and breast-feeding women (and synonyms); (E) COVID-19 vaccination (and synonyms); and (O) knowledge, attitude, and practice (including factors associated with acceptance/hesitancy) regarding the COVID-19 vaccination (and synonyms). The specific search strategy will initially be created by a health specialist with extensive experience in conducting systematic reviews. Susequently, the search strategy will be adjusted based on input from the project team. The search strategy for PubMed/MEDLINE that will be adopted in the review is reported in Table 1.

**Table 1.** Search strategy developed in PubMed/MEDLINE.


#### **4. Detailed Procedure**

#### *4.1. Inclusion/Exclusion Criteria*

Studies will be selected based on the inclusion/exclusion criteria described below, defined according to the PEO strategy, along with additional information related to the study design, language, and time-span. Original population-based observational studies assessing the knowledge, attitudes and practice of pregnant or breastfeeding women in taking/refusing COVID-19 vaccination will be included in the review. By observational studies is implied all cross-sectional, case-control or cohort (prospective and retrospective) studies. Only English language, peer-reviewed articles published in international scientific journals will be considered. All articles published between 2019 and the date of the review's conclusion will be considered eligible for inclusion. The systematic review's exclusion criteria include the following: studies not performed among humans or that were conducted on a different population (for instance, the general public, women in general, parents or only mothers of children older than one year, and children's caregivers in general); studies combining data with different and multiple outcomes, or assessing different outcomes not listed in our inclusion criteria (for instance, articles assessing the efficacy, serology, immunology, safety, and development of the COVID-19 vaccine in pregnant or breastfeeding women); articles assessing acceptance/hesitancy/refusal against vaccines other than COVID-19; articles not written in the English language and those not published in peer-reviewed international journals; non-observational studies, e.g., trials (randomized or non-randomized controlled trials); and, lastly, non-original research papers, including reviews or meta-analyses, articles with no quantitative information or details, and non-full-text papers (e.g., letters to the editor, conference papers, commentary notes, expert opinions, abstracts). There will be no restrictions based on the type of setting, such as community-based or hospital-based populations.

#### *4.2. Selection Process*

All the retrieved studies will subsequently be downloaded to the EndNote software (EndNote® for Microsoft, Redmond, WA, USA, 2020). Duplicates will be removed using an automatic function in the EndNote software, followed by a manual check by one of the authors. The remaining articles will then be assessed for eligibility, firstly based on the title and abstract, followed by their full text. Two authors will independently undertake the two-step screening process by applying the inclusion/exclusion criteria detailed above. If any doubt or disagreement should arise during the two screening steps, this will be solved through a direct comparison between the views of the two authors. If divergences still persist, a final arbitrator will settle any disagreements over inclusion. Reviewer authors will be blind to the journal title, authors, and their institutions/affiliations. However, to increase agreement between the two reviewer authors, a pilot assessment will be conducted on 20 randomly selected retrieved articles [19]. Repeated articles and multiple publications from the same study will be excluded and all the reasons for exclusion will be reported. The results of the selection process will be detailed at each stage and reported using the PRISMA flow diagram.

#### *4.3. Data Extraction*

The data extraction process will be performed in duplicate by two reviewer authors. A standardized and pre-defined Excel (Microsoft Excel® for Microsoft 365 MSO, USA, 2019) spreadsheet will be used to extract data from the included studies [20]. The spreadsheet will initially be piloted on one-third (or no more than five, depending on the total number) of included articles to increase consistency between the two reviewer authors [21]. The following information will be extracted from each article included: author name, study period, country where the study was conducted, study settings, main characteristics and the study population's number, study completion rates (attrition), tool(s) used to assess the outcomes, number of items, whether the tool(s) was/were validated or not, manner in which the questionnaire was administered, recruitment methods, outcomes of interest, outcomes definition, main results, and funds and conflicts of interests, if any. Vaccine coverage will also be recorded, if available.

Nevertheless, recognising that outcomes of interest are composite measures, we will extract data directly as reported in the original articles even if unvalidated instruments are used to assess the outcomes. Furthermore, despite a general consensus on the definition of vaccine hesitancy/acceptance, no unequivocal tool or operationalization method is in place to evaluate it [22]. Rather, several instruments and statistical methods are commonly used. In light of this, we anticipate substantial variation in the methods used to report results. For this reason, when available, we will also extract methodological information, such as whether the tool was validated or not and the statistical analysis undertaken. Lastly, if studies report data using risk estimates, for instance, odds ratio (OR), risk ratio (RR) or hazard ratio (HR), we will collect the maximally adjusted data, along with the list of variables used for the adjustment [23].

The data collected will support the assessment of the study quality and will be used for data synthesis.

#### *4.4. Quality Assessment*

The risk of bias of the included studies will be independently assessed by two reviewers. Disagreements between the reviewers will be discussed and resolved by consensus. However, insights from a third reviewer will be sought if necessary. The Joanna Briggs Institute (JBI) quality assessment tools will be used to assess the potential risk of bias in each included article [24]. We opted to use the JBI tools due to the availability of separate checklists for each study design (e.g., cross-sectional, case-control, and cohort studies) [25]. The JBI tools are based on eight items that explore seven different domains: (i) participant selection, (ii) setting definition, (iii) ascertainment of the exposure; (iv) validity of condition measurement; (v) identification of confounders and dealing strategy, (vi) ascertainment of the outcome, and (vii) appropriateness of the statistical analysis [24].

Assessed papers will be categorized based on their methodological qualities by applying a scoring system available in [11]. Specifically, for each of the eight items, four options are allowed: Yes, No, Unclear and Not applicable. We will assign 2 points for yes, −2 points for no, −1 point for unclear and 0 points for not applicable. The total score could range between −16 and 16. Articles scoring from −16 to 4 will be classified as low quality, articles scoring from 5 to 9 as moderate, and articles scoring equal to or more than 10 (and up to 16) as high quality.

#### **5. Expected Results**

The quantitative and qualitative results of the literature will be presented using descriptive tables. As previously performed [26], a narrative description of the main characteristics of the study (for instance, the study design, study period, country where the study took place), the population characteristics (for instance, the age of the women and their status), the methodology (for instance, the manner in which the survey was administered, if the tools were validated), and the outcome, will be obtained from the included studies. This description will help to identify similarities and differences among the studies. The main results will be presented with reference to the 5C model and synthesized using a narrative approach [5].

A pilot exploration of how many results will result was conducted on PubMed/MEDLI-NE (13 January 2023). A total of 184 records were identified.

#### *5.1. Quantitative Analysis*

If at least two studies report data for the same outcome using OR, RR, or HR and their 95% confidence interval (CI), then we will proceed to pool data through meta-analysis. When two or more studies report estimated risks (OR, RR or HR) for a specific factor, both random and fixed effects models will be used to calculate the pooled effect size. The pooled effect size will be reported as the OR with a 95% CI. We will assess the heterogeneity of

the studies using both the chi-square test and the I2 statistic, as previously performed [27]. Heterogeneity will be classified into four categories based on the I2 value (higher: I2 > 75%, moderate = I<sup>2</sup> ranging between 75% and 50%, low = I<sup>2</sup> ranging between 50% and 25%, and low = I2 < 25%). Publication bias will be assessed via visual inspection of the funnel plot and by means of the Egger regression asymmetry test, with statistical significance set at *p* < 0.10 [28]. If there is any publication bias, the trim and fill method will be performed [29]. All analyses will be conducted using Prometa3® software (Internovi, Cesena, Italy).

#### *5.2. Subgroup and Sensitivity Analysis*

If the necessary data are available, the analysis may be stratified according to the subjects' characteristics (e.g., pregnant or breastfeeding women), the country where the study took place and, lastly, the methodological quality of the studies (e.g., only including moderate/high methodological quality studies).

#### **6. Ethical Considerations**

This is a systematic literature review of the available literature using already published data. No interventions are planned, nor will there be any direct data collection from humans/animals. For these reasons, no ethical approval is required. The results of our review will be disseminated among academia, policymakers, healthcare professionals and the general public. For detailed information dissemination, scientific presentations at national and international congresses and conferences, peer-reviewed scientific publications, and posts on both academic and generalist social network platforms will be used.

#### **7. Discussion**

The current review will offer a comprehensive overview of the existing literature on the knowledge, attitudes, behaviour, acceptance, and hesitancy regarding the COVID-19 vaccine among pregnant and breastfeeding women. Although previous reviews have focused on knowledge, attitudes or behaviour in accepting/refusing vaccines, in general, the current review will focus on specific vaccines for COVID-19, which largely differ from other type of vaccines [30]. The differences include, first and foremost, the availability of several vaccines developed by numerous pharmaceutical companies (for instance, seven different vaccines were approved and authorised to be marketed in Europe) [31] that may increase uncertainty among individuals. Despite the fact that all these vaccines showed satisfactory levels of safety and efficacy in clinical trials [31–34], they were all administered to the general public, albeit with different indications [31]. This could cause people to be uncertain about vaccine preferences, which may raise doubts regarding vaccination [35]. Some vaccines share the same technology, while others were developed using different approaches. The first COVID-19 vaccines approved were those using viral mRNA (for example, those from Pfizer and Moderna). Subsequently, recombinant, adjuvanted vaccines (for example, those from Novavax), inactivated, adjuvanted vaccines (for example, from Valneva), and, finally, those using recombinant DNA technology (for example, from Janssen) were developed and approved [31]. Secondly, COVID-19 vaccines were the first mRNA vaccine administered to humans, which may have contributed to fear of long-term sideeffects [36]. Thirdly, due to the novelty of the virus and its rapid global spread, there was an urgent need for safe and effective vaccines. For this reason, many efforts were made to expedite the testing and licensing of the vaccines. Consequently, the public could also have been affected by fear of poorly executed experimental trials and possible unknown side-effects [37]. Additionally, during the COVID-19 pandemic, the general public and pregnant/breastfeeding women directly experienced fear of the disease itself, fear not commonly perceived for other "old" vaccine-preventable diseases for which the vaccination programmes are known to have prevented millions of cases, dispelling the fear of the disease itself and leaving, instead, room for fear of possible, albeit rare and mostly non-serious, vaccine adverse effects [38]. Last, but not least, the great volume of information

(and even disinformation) readily available, especially on the internet and social networks, the so-called infodemic, has a substantial impact on vaccination acceptance [38–40].

The ultimate aim of the current review is to shed light on this still evolving area of research on the assumption that the results could help in understanding the barriers and facilitators of COVID-19 vaccine acceptance among a specific vulnerable sub-population. Pregnant women have a higher risk of severe complications from COVID-19 compared to non-pregnant women of reproductive age [41]. According to a large study conducted by the Centers for Disease Control and Prevention, pregnant women affected by COVID-19 have a higher risk of intensive care unit admission, invasive ventilation, extracorporeal membrane oxygenation, and death than non-pregnant women of reproductive age [8]. In light of this, and the associated high burden, it is of utmost importance to understand the reasons for hesitancy towards or acceptance of COVID-19 vaccines with the ultimate purpose of raising the vaccination rate among pregnant/breastfeeding women.

Some potential limitations of our work should be acknowledged. First, we will only be including articles written in English. This may exclude potentially relevant articles written in other languages, for instance, Chinese, the language of the country where the virus originally appeared. Moreover, we recognize the heterogeneity of outcomes, although this is a methodological weakness directly attributable to the content of the original studies published in the literature. However, the approach taken will enable us to assess a large number of studies, offering a broad overview of the phenomenon.

Despite the above-mentioned limitations, we conclude that our findings may be useful for both healthcare professionals and policy makers, as they can assist healthcare professionals in guiding pregnant and breastfeeding women through the decision-making process associated with receiving the COVID-19 vaccine [42–45]. Similarly, our findings could inform public health policies with respect to future vaccine communication strategies.

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

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

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

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

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

#### **References**


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

**Putunywa Zandrina Nxiweni 1, Kelechi Elizabeth Oladimeji 1,2, Mirabel Nanjoh 3, Lucas Banda 1, Felix Emeka Anyiam 4, Francis Leonard Mpotte Hyera 1, Teke R. Apalata 2, Jabu A. Mbokazi <sup>5</sup> and Olanrewaju Oladimeji 1,4,\***


**Abstract:** Access to quality care before, during, and after childbirth remains an effective means of reducing maternal and neonatal mortality. Therefore, the study identified factors influencing the utilization of prenatal care services among women of childbearing age in South Africa. This is a retrospective study based on secondary data from the South African Demographic Health Survey (DHS) conducted from 1998 to 2016. In South Africa, 21.0% of mothers had used ANC services. Higher odds of seeking prenatal care were found in women aged 35 years and older (cOR = 1.26, 95% CI; 1.08–1.47, *p* = 0.003), married or cohabiting (cOR = 1.13, 95% CI; 1.004–1.27) observed, *p* = 0.043), higher level of education (tertiary education: cOR = 0.55, *p* = 0.001), female residents in urban areas (cOR = 1.35, 95% CI; 1.20–1.52, *p* = 0.001), higher wealth index (cOR = 1.32, 95% CI; 1.15–1.51, *p* = 0.001), employed (cOR = 1.48, 95% CI; 1.29–1.70, *p* = 0.001) and media exposure (cOR = 1.27, 95% CI; 1.12–1.44), *p* = 0.001). The findings of this study provide insight into the need to make maternal health services more accessible, more widely used, and of a higher quality. This requires effective strategic policies that promote patronage to reduce maternal mortality and improve newborn outcomes in South Africa.

**Keywords:** antenatal care utilization; antenatal care services; South Africa; maternal health

#### **1. Introduction**

Antenatal care (ANC) is an important factor in reducing maternal morbidity and mortality in pregnant women and in achieving a positive pregnancy experience [1–3]. The essence of this care pathway is to make sure that the health of both the unborn child and the pregnant mother is safe by monitoring the progress of the pregnancy vis-a-vis expected indicators for a normal pregnancy. Access to ANC gives a pregnant woman the opportunity to benefit from care services including health promotion, screening and diagnosis, and disease prevention, required to maintain normalcy and for timely identification of abnormalities that can pose a risk to the life of her unborn child and herself. Unfortunately, many women in developing countries do not have access to such services [1,4].

According to the South African Demographic Health Survey [5], there are approximately 536 prenatal deaths per 100,000 in South Africa. It shows that for every 1000 live births, five (5) women died during pregnancy. A higher proportion of women in South Africa receive prenatal care, also known as antenatal care (ANC) from healthcare professionals; doctors (18%), nurses or midwives (70%). Only a small fraction (2%) are cared for by traditional birth attendants, while 10% receive no prenatal care [5]. The benefits of ANC cannot be overstated, particularly when it comes to reducing maternal and prenatal

**Citation:** Nxiweni, P.Z.; Oladimeji, K.E.; Nanjoh, M.; Banda, L.; Anyiam, F.E.; Hyera, F.L.M.; Apalata, T.R.; Mbokazi, J.A.; Oladimeji, O. Factors Influencing the Utilization of Antenatal Services among Women of Childbearing Age in South Africa. *Women* **2022**, *2*, 285–303. https:// doi.org/10.3390/women2030027

Academic Editors: Mary V. Seeman, Claudio Costantino and Maiorana Antonio

Received: 25 May 2022 Accepted: 26 August 2022 Published: 10 September 2022

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

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

morbidity and mortality. Maternal morbidity refers to any health condition attributed to, or aggravated by, pregnancy and childbirth that negatively affects the woman's wellbeing [6,7].

WHO recommendations prior to 2016 call for at least four ANC visits [8] where a pregnant woman receives focused ANC, if eligible. Currently, a pregnant woman needs at least eight visits [9] to receive any significant evidence-based interventions. The South African Department of Health has classified the appropriate ANC based on the WHO criteria above. If a pregnant woman made at least four and eight visits between April 2006 and April 2017, she was considered booked or received an appropriate ANC. A 2.4% increase in the percentage of South African women who participated in at least four ANC visits from 1998 to 2016 was documented by Global Health data [10,11].

During this period, South Africa recorded 150 maternal deaths per 100,000 live births in 1998 [12] and 119 deaths in 2017 [13]. Despite the observed improvement, the country is far below the required 70 deaths per 100,000 live births to meet the Sustainable Development Goals (SDG) 3.1 [14]. Moreover, the rate of skilled delivery use, a predictor of Maternal Mortality Rate (MMR) in the country increased from 84% in 1998 to 97% in 2016 [15]; although, Bobo et al. [16] reported a higher rate of 96.7 percent. When it comes to pregnant women's health, adequate ANC services are essential.

It has been observed that increasing access to skilled attendants, which has a close link to ANC, emergency obstetric care, and family planning services can significantly reduce maternal mortality in low-income settings such as South Africa [17–19]. Despite the obvious importance of maternity care, including ANC, poor access to and utilization of such services remains an important determinant of maternal mortality and morbidity worldwide [11,17].

Previous research has shown a link between ANC utilization and accessibility, sociodemographic factors, knowledge, and the quality of care provided [20–22], but the extent to which these factors influence ANC utilization has not been adequately documented in the region of South Africa. Consequently, this study investigated the critical factors that influence the utilization of ANC and other maternal health services between the years 1998 and 2016 among women of reproductive age in South Africa. The insights provided by this study will further help to shape the strategic policy that South Africa will use to reduce the number of maternal deaths and improve neonatal outcomes.

#### **2. Results**

#### *2.1. Characteristics of Maternal Household Factors of Women within Reproductive Age in South Africa*

As shown in Table 1, of the 67,645 women included in the analysis, 77.5% were para 1–2, 12% were equally nulliparous, and para ≥3. Timing of ANC (in months) was more among those that have attended between 3–6 months (72.2%), followed by <3 months (17.4%). Almost three-quarters (72.2%) had their first ANC visit between 3–6 months of pregnancy, and slightly above one-sixth (17.4%) attended before three months. More participants resided in the urban area (56.6%), compared to rural (43.4%). The provinces with the most participants were Gauteng (23.5%), Kwazulu-Natal (19.7%), Limpopo (12.6%), and Eastern Cape (12.5%).

**Table 1.** Characteristics of women within reproductive age in South Africa and factors influencing the use of antenatal care among them (*n =* 67,645).



#### **Table 1.** *Cont.*


#### **Table 1.** *Cont.*


**Table 1.** *Cont.*

<sup>γ</sup> indicated variables with missing data.

*2.2. Characteristics of Women within Reproductive Age in South Africa and Factors Influencing the Use of Antenatal Care among Them*

The study analyzed the DHS data of 67,645 women, across South Africa. The majority of the participants (46.6%) were within the age 25–34; 30.9% were within 15–24 years; and 22.4% were 35 years and above. Almost half of the participants (49.8%) were married, and 50.2% were single. Those who have completed secondary education were more with 67.4%, followed by those who have completed primary education (18.5%). Black/African descent was the majority with 86.3%, followed by those of Colored descent (8.4%), White (3.3%), and Indian/Asian (1.9%). Among the women included in the analysis, 77.5% had 1 or 2 parities, 11.6% were nulliparous, and those with 3 or more parities were 11.5%. Almost three-quarters (72.2%) had their first ANC visit between 3 and 6 months of pregnancy while slightly above one-sixth (17.4%) attended before three months. More participants resided in the urban area (56.6%), compared to rural (43.4%) and the provinces with the most participants were Gauteng (23.5%), Kwazulu-Natal (19.7%), Limpopo (12.6%), and Eastern Cape (12.5%) as shown in Table 1.

As shown in Figure 1, the overall prevalence of Utilization of Antenatal care services among women of reproductive age in SA was 79%.

**Figure 1.** Utilization of Antenatal care services among women of reproductive age in South Africa.

As shown in Figure 2, the prevalence of Utilization of Antenatal care services among women of reproductive age in SA was statistically significantly highest in the province of Western Cape (88.6%), followed by Kwazulu-Natal (82.8%), Northern Cape (81.6%) and Northwest (80.5%) (χ<sup>2</sup> = 81.47, *p* = 0.001).

**Figure 2.** Utilization of Antenatal care by women of reproductive age in SA stratified by province.

Women in the poverty band of the wealth index were the majority (41.3%), those in the middle wealth index were 22.8% and barely one-third belonged to the rich band (35.9%). The majority were not employed (69.4%), and 77.1% neither owned a car, a motorcycle/scooter (98.4%), nor a bicycle (85.8%). A little over half owned a refrigerator (58.93%), and most had electricity (71.6%). Barely two-thirds owned a television (68.89%), a radio (65.29%), watched TV every day/week (63.6%), and listened to the radio every day/week (61.4%) and 36.9% read newspapers regularly. The overwhelming majority of the women had no problem getting permission (86.9%) and money (72.2%) to visit the health facility for treatment; had no problem with the distance to the health facility (77.6%); and had no problem going to the health facility alone (88.0%), as seen in Table 1.

As shown in Figure 1, the overall prevalence of the Utilization of Antenatal care services among women of reproductive age in SA was 79%.

As shown in Figure 2, the prevalence of the Utilization of Antenatal care services among women of reproductive age in SA was statistically significantly highest in the province of Western Cape (88.6%), followed by Kwazulu-Natal (82.8%), Northern Cape (81.6%) and North West (80.5%) (χ<sup>2</sup> = 81.47, *p* = 0.001).

As shown in Table 2, the utilization of antenatal care varied across socio-demographic variables. Statistically significant higher prevalence of utilization of antenatal care was observed among those between 25 and 34 years in age (*p* = 0.038); married or cohabiting (*p* = 0.001); had a tertiary level of education (*p* = 0.001), and of the Indian/Asian race (*p* = 0.001). For obstetric and household factors, a significantly higher prevalence of utilization of antenatal care was observed among para 1–2 (*p* = 0.001), attended antenatal <3 months (*p* = 0.001), reside in the urban (*p* = 0.004), and the Western Cape Province (*p* = 0.001). Considering economic status, utilization of antenatal care has statistical significance for those in the richest wealth index (*p* = 0.003), are employed (*p* = 0.001), own a car (*p* = 0.004), own a motorcycle/scooter (*p* = 0.001), own a bicycle (*p* = 0.032), own a refrigerator (*p* = 0.001) and have electricity (*p* = 0.001). Moreover, there was statistically significant higher prevalence of utilization of antenatal care among those who own a television (*p* = 0.001), own a radio (*p* = 0.001), watches television every day/week (*p* = 0.001), listens to the radio every day/week (*p* = 0.038) and reads newspaper regularly (*p* = 0.039) under the media exposure factor. However, no statistically significant association was observed between health institutional factors and the utilization of antenatal care (*p* > 0.05).


**Table 2.** Associated factors for utilization of antenatal care services.


#### **Table 2.** *Cont.*

#### **Table 2.** *Cont.*


Statistically significant (*p* < 0.05); α = A drop in sample population as institutional factor variables were not observed in the 1998 DHS data, rather only in the 2016 DHS data.

#### *2.3. Multilevel Multivariate Logistic Regression Results*

As shown in Table 3**,** statistically significant explanatory variables in the Chi-Square test of association were included for the multilevel multivariate logistic regression.

**Table 3.** Multilevel Multivariable Logistic Regression Results for Factors Associated with the Utilization of Antenatal Care Services among Women of Reproductive Age in South Africa (*n =* 67,645).


#### **Table 3.** *Cont.*


cOR = Crude OR, aOR = Adjusted OR, <sup>R</sup> = reference value, \* significant at *<sup>p</sup>* ≤ 0.05, \*\* significant at *<sup>p</sup>* < 0.01, \*\*\* significant at *p* < 0.001.

Model I: Non-adjusted (crude) aggregate model comprising all explanatory variable categories associated with the utilization of Antenatal care services

#### 2.3.1. Socio-Demographic Factors

The study shows higher odds for the utilization of antenatal care among women aged 35 years and older than those 15–24 years (cOR = 1.26, 95% CI; 1.08–1.47, *p* = 0.003). Moreover, being married or cohabiting had higher odds for utilizing antenatal care than singles (cOR = 1.13, 95% CI; 1.004–1.27, *p* = 0.043). The odds for the utilization of antenatal care among women improved from primary to tertiary compared to those with no education (primary: cOR = 0.38, secondary: cOR = 0.45 and tertiary: cOR = 0.55, *p* = 0.001). White or Indian/Asian descent showed lower odds for the utilization of antenatal care compared to black/African (White: cOR = 0.25, Indian/Asian: cOR = 0.35, *p* = 0.001).

#### 2.3.2. Obstetric and Household Factors

Women with Para 1–2 and Para ≥3 showed increased odds for the utilization of antenatal care compared to those that are nulliparous, with an increased odds in Para ≥3 compared to Para 1–2 (Para 1–2: cOR = 1.33, Para ≥3: OR = 1.63, *p* < 0.05). Women with the timing of ANC at 3 or more months of pregnancy showed increased odds for the utilization of antenatal care compared to those with the timing of ANC less than 3 months (3–6 months: cOR = 41.91, 6+ months: OR = 9.73, *p* = 0.001). Women residing in the urban area showed increased odds for the utilization of antenatal care compared to those in the rural area (cOR = 1.35, 95% CI; 1.20–1.52, *p* = 0.001). Only those residing in Eastern Cape showed increased odds for the utilization of antenatal care compared to Western Cape (cOR = 1.54, 95% CI; 1.09–2.18, *p* = 0.014). The other provinces, Northern Cape (cOR = 0.56, *p* = 0.001), Mpumalanga (cOR = 0.63, *p* = 0.001), Limpopo (cOR = 0.70, 0.001) and Gauteng (cOR = 0.93, *p* = 0.001) showed lower odds for the utilization of antenatal care compared to Western Cape.

#### 2.3.3. Economic Status Factors

The study shows higher odds for the utilization of antenatal care among women in the middle wealth index compared to those in the poorest/poorer wealth index (cOR = 1.32, 95% CI; 1.15–1.51, *p* = 0.001). Higher odds for the utilization of antenatal care were observed among women who are employed (cOR = 1.48, 95% CI; 1.29–1.70, *p* = 0.001), own a car (cOR = 1.44, 95% CI; 1.23–1.69, *p* = 0.001), own a Motorcycle/Scooter (cOR = 2.97, 95% CI; 1.37–6.44, *p* = 0.006), own a refrigerator (cOR = 1.58, 95% CI; 1.40–1.79, *p* = 0.001), and have electricity (cOR = 1.62, 95% CI; 1.43–1.84, *p* = 0.001).

#### 2.3.4. Media Exposure Factors

Higher odds for the utilization of antenatal care was observed among women who own a television (cOR = 1.27, 95% CI; 1.12–1.44, *p* = 0.001), own a radio (cOR = 1.50, 95% CI; 1.33–1.69, *p* = 0.001), watch television everyday/week (cOR = 1.68, 95% CI; 1.49–1.89, *p* = 0.001), listen to radio every day or week (cOR = 1.21, 95% CI; 1.07–1.36, *p* = 0.002), and read newspaper regularly (cOR = 1.47, 95% CI; 1.29–1.68, *p* = 0.001).

#### *2.4. Model II: Household Factors, Economic Factors, and Media Exposure Factors Associated with Utilization of Antenatal Care Services, While Controlling for Their Socio-Emographic/Individual Factors*

#### 2.4.1. Obstetric and Household Factor

Women with Para ≥3 showed increased odds for the utilization of antenatal care compared to those who are nulliparous after adjusting for socio-demographic characteristics (aOR = 1.29, 95% CI; 1.03–1.68, *p* = 0.029). Women with the timing of ANC 3 or more months showed reduced odds for the utilization of antenatal care compared to those with the timing of ANC less than 3 months (3–6 months: aOR = 0.029, 6+ months: aOR = 0.29, *p* = 0.001). Women residing in the urban still showed increased odds for the utilization of antenatal care compared to those in rural areas after adjusting for confounding variables (aOR = 1.24, 95% CI; 1.04–1.49, *p* = 0.016).

All provinces showed lower odds for the utilization of antenatal care compared to Western Cape (*p* < 0.05).

#### 2.4.2. Economic Status Factors

After controlling for socio-demographic characteristics, the study showed no statistically significant association between wealth index and the utilization of antenatal care (*p* > 0.05). Higher odds for the utilization of antenatal care were now observed among women for only those who were employed (aOR = 1.27, 95% CI; 1.10–1.49, *p* = 0.001), own a refrigerator (aOR = 1.20, 95% CI; 1.01–1.41, *p* = 0.036) and have electricity (aOR = 1.27, 95% CI; 1.07–1.50, *p* = 0.006). The variables own a car or own a Motorcycle/Scooter were no longer statistically significant.

#### 2.4.3. Media Exposure Factors

After controlling for socio-demographic characteristics, higher odds for the utilization of antenatal care were still observed among women who own a television (aOR = 1.16, 95% CI; 1.0–1.35, *p* = 0.049) and watch TV every day/week (aOR = 1.44, 95% CI; 1.21–1.72, *p* = 0.001). The variables own a radio, listen to the radio every day or week, and read the newspaper regularly were no longer statistically significant (*p* > 0.05).

*2.5. Model III: Economic Factors and Media Exposure Factors Associated with Utilization of Antenatal Care Services, While Controlling for Obstetric and Household Factors*

#### 2.5.1. Economic Status Factors

After controlling for obstetric and household factors, the study showed a statistically significant association between the wealth index and the utilization of antenatal care. Higher odds for the utilization of antenatal care among women were observed in those in the middle wealth index (aOR = 1.21, 95% CI; 1.0–1.47, *p* = 0.047) and richer/richest wealth index (aOR = 1.23, 95% CI; 1.007–1.51, *p* = 0.043). Higher odds for the utilization of antenatal care were observed among employed women (aOR = 1.20, 95% CI; 1.004–1.44, *p* = 0.046), own a refrigerator (aOR = 1.20, 95% CI; 1.01–1.41, *p* = 0.036) and have electricity (aOR = 1.27, 95% CI; 1.07–1.50, *p* = 0.006). The variables own a car, own a Motorcycle/Scooter, own a refrigerator, and have electricity were no longer statistically significant (*p* > 0.05).

#### 2.5.2. Media Exposure Factors

for maternal household factors.

Only the variable, watches TV every day/week, was statistically significantly associated with the utilization of antenatal care. Those watching TV every day/week showed increased odds for the utilization of antenatal care services (aOR = 1.39, 95% CI; 1.12–1.73, *p* = 0.02).

#### *2.6. Model IV: Media Exposure Factors Associated with Utilization of Antenatal Care Services, While Controlling for Economic Status Factors* Media Exposure Factors

### After controlling for economic status factors, higher odds for the utilization of antenatal were was still observed among women who own a television (aOR = 1.18, 95% CI; 1.02–1.38, *p* = 0.030) and watch TV every day/week (aOR = 1.37, 95% CI; 1.15–1.62, *p* = 0.001). The variables, own a radio, listen to radio every day or week, and read the newspaper regularly, were no longer statistically significant (*p* > 0.05), similar to the findings when controlling

#### **3. Discussion**

Using nationally representative 1998 to 2016 SADHS data, the goal of this study was to assess factors associated with the utilization of ANC services in South Africa. The cluster sampling methodology used ensured sample representativeness and the reliability of the study results. The study included 67,645 mothers of child-bearing age in nine provinces of South Africa whose complete information was available in the survey. In South Africa, 21.0% of mothers had utilized ANC services. There were variations in all the provinces. The highest provinces with the most prevalence were Western Cape and KwaZulu-Natal. The lowest are Eastern Cape and Gauteng, which could be as a result of the demographic and socioeconomic factors associated with both provinces. According to the findings of this study, women in South Africa's rural areas were less likely than women in the country's urban areas to use ANC services. This could be due to the disparities in the availability and accessibility of healthcare facilities, and women's awareness of ANC services in urban and rural areas. This finding was consistent with the findings of other studies conducted in Pakistan and Vietnam where ANC uptake was lower in rural areas [23,24]. This implies that more attention to health awareness, education, and promotion activities in rural areas is needed to improve ANC uptake. According to Rustagi et al. [25], the higher ANC coverage observed in the urban setting may likely be due to ANC accessibility at primary care facilities in these areas, highlighting the need for policy efforts to strengthen primary healthcare. ANC coverage has been found to be linked to primary healthcare availability in similar studies [26,27].

The present study observed a statistically significant relationship between a woman's age and adequate antenatal care utilization. The older the woman (35 years and older), the more likely she will use antenatal care appropriately. This suggests that young women have less experience with pregnancy care than older women. This is similar to findings to research by Adedokun and Yaya [21], who analyzed information obtained from the Demographic and Health Surveys (DHS) carried out in 31 different countries and involving 235,207 women aged 15–49 years old who had given birth to children within 5 years of the surveys. Similar findings were obtained by Joshi et al. [28] in Nepal, Dairo and Owoyokun [29] in Nigeria, Denny et al. [30] in Indonesia, and Ebonwu et al. [31] in South Africa. This may be due to older women placing more value on ANC, as a lack of knowledge about the benefits of ANC or the pregnancy being unwanted, which are common among adolescents, leads to seeking ANC care less frequently among younger women (including teenagers). Another study in Nigeria discovered that being 35 or older consistently increased the odds of using ANC by more than 200 percent [32]. Therefore, it is imperative for the South African Government to formulate policies that will protect adolescent pregnant women and provide for a tailored ANC to ensure utilization and a favorable pregnancy outcome for them. However, studies investigating the association between a woman's age and the use of ANC have not always reached consistent conclusions; as one study observed, the younger age utilization of ANC was found to be adequate because working women tend to postpone their first pregnancy and are more mature in terms of age during pregnancy than unemployed women [33].

The odds for the utilization of antenatal care among women improved from secondary to tertiary compared to those with no education. The findings indicated that women with higher levels of education have a greater likelihood of making appropriate use of antenatal care than women with lower levels of education. This suggests that a woman's likelihood of utilizing antenatal care increases in proportion to the level of education she possesses, which is similar to findings from previous studies [21,34,35]. A plausible explanation is that education fosters better enlightenment on issues, particularly health-related issues. This finding corroborated a study that alluded to increased utilization of maternal healthcare and women's empowerment through education, wealth, and decision making [36]. The girl child education policy needs to be strengthened, ensuring that no girl child is missed, thus improving their educational status and ANC utilization.

In addition, married or cohabiting had higher odds for utilizing antenatal care than singles. Rurangirwa et al. [33] in their study conducted in Rwanda, observed that the risk of poor utilization of ANC services was higher among single women. This may be due to the support that married and cohabitating women receive from their husbands or partners as a result of the ANC attendance sensitization campaign, which equally targets men and encourages them to follow their wife or partner to the clinic [37]. This is also consistent with the data from similar studies [38,39].

This study observed that women with a better economic status (wealth index) and who are employed had more antenatal care utilization than those with lower wealth indexes. Higher odds for the utilization of antenatal care among women were observed in those in the middle wealth index and richer/richest wealth index. When it comes to prenatal care, women from low-income families may not have the financial means to register at clinics or pay for their services. Studies conducted in Ethiopia and Gabon, and evidence from the Demographic Health Surveys data of 31 countries across sub-Saharan Africa corroborated this finding [21,40,41].

Women living in houses equipped with electricity were found to be utilizers of ANC services. It is possible that the presence of electricity in a household may be an indirect measure of accessibility to media services and may be a sign of a better or higher social class [23].

This study found that women exposed to mass media (own a television and watch TV every day/week, or listen to the radio) had a higher chance of ANC utilization than women who were not, as seen in some similar studies, with the propensity to enjoy essential obstetric care from skilled birth attendants [22,42,43]. This may be due to the fact that mass media can reach a large number of people at once, thereby increasing awareness of the benefits of maternal health services and influencing family behavior.

#### *Limitations and Strengths of the Study*

A limitation of this study is that the use of secondary data. One of the strengths of the study is that the DHIS survey is national data with geographical representation; hence, the study results are a true representation of the national data.

#### **4. Materials and Methods**

#### *4.1. Research Design*

This is a retrospective study based on secondary data obtained from the South African Demographic Health Survey (DHS), which was carried out between the years 1998 and 2016.

#### *4.2. Population*

Administratively, South Africa is divided into nine provinces. In 2020, the middleyear population estimated by Statistics South Africa was 59.62 million, of which approximately 51.1% are females. The infant mortality rate for 2020 was estimated at 23.6 per 1000 live births.

#### *4.3. Sample Size and Sampling Frame*

A curated and concatenated dataset on ANC utilization was obtained from demographic and health surveys conducted in South Africa from 1998 to 2016 The targeted study population was women of reproductive age (15–49 years).

The survey involved a two-stage cluster stratified sampling method. In the first stage, the country was divided into clusters, using the enumeration areas (EA); clusters for the study were selected using simple random sampling and the households within each cluster were line listed. Women between 15 and 49 years of age who were citizens or permanent residents were randomly selected from the listed households and enrolled in the study in the second stage [44].

#### *4.4. Instruments*

Data for the DHS were collected through interviewer-administered semi-structured validated questionnaires. Information obtained with this questionnaire includes socioeconomic characteristics, reproductive history, antenatal, delivery, post-natal care, and breastfeeding.

#### *4.5. Validity and Reliability of the Data Collection Instrument*

DHS questionnaire is a validated tool that has been used for many decades. The DHS survey data collection tool's reliability has been tested and established through repeated use by DHS and other experienced research investigators [44].

#### *4.6. Variables of Interest*

The independent variables: These include, sociodemographic characteristics such as age, marital status, education, and race; household factors such as parity (zero, one and two, three or more), the timing of ANC, place of residence, and region; economic status factors such as wealth index, employment, own a car/truck, own a motorcycle/scooter, own a bicycle, own a refrigerator, and electricity; media exposure factors such as own a television, own a radio, watch television regularly, listen to the radio regularly and read newspapers regularly; institutional factors such as access to a health facility and distance to a health facility.

The dependent (outcome) variable: ANC utilization during the women's pregnancy period was the outcome variable. This was categorized as 'not utilized—women who did not attend ANC', and 'utilized'—women who utilized. ANC not utilized was defined by <4 clinic visits and ANC utilized by ≥4 clinic visits across the study years.

#### *4.7. Data Analysis*

The data were analyzed using the SPSS package for data analysis. Descriptive analyses such as count, frequencies, and percentages are presented using a frequency table and bar/pie charts where appropriate. Pearson chi-square test was used to establish relationships between the independent and outcome variables, using a statistical significance of *p*-value less than or equal to 0.05 (*p* ≤ 0.05).

Bivariate and multivariate logistic regression analyses were used to measure the associations between the independent and outcome variables.

The study further used a regression model expression to simulate a nested approach in which a non-adjusted aggregate model comprising all explanatory variable categories and utilization of Antenatal care services would be iterated to generate Model 1. Model 2 was simulated using obstetric and household factors; economic factors; and media exposure factors while controlling for their socio-demographic/individual factors. Simulation using economic factors and media exposure factors while controlling for household factors was for Model 3 and lastly, Model 4 was simulated using only media exposure factors and controlling for economic status factors. The primary benefit of the model selected is avoiding confounding effects by analyzing the association between all variables simultaneously. Confounding effects were tested in the four models among different factors. After defining the technique, the fundamental interpretation of the results was emphasized. A *p*-value set at 0.05 was considered statistically significant.

#### **5. Conclusions**

The study uncovered factors that influence women's use of antenatal care in South Africa. Age, marital status, having a tertiary education, living in an urban area, and socioeconomic factors, such as being in the richest wealth index and employed, having electricity, and media exposure, all influenced antenatal care utilization. Antenatal care enables the early detection and treatment of diseases that may affect both the mother and the child. It also allows a pregnant woman to be cared for during prenatal, antenatal, childbirth, and post-natal periods, reducing the chances of complications leading to maternal and neonatal death. Introducing targeted health promotion and education programs in communities would empower young and illiterate rural women to use available ANC services more often during pregnancy. Strengthening antenatal care visits becomes critical to the government in promoting and improving the health of the mother and child. This will lead to improved maternal and neonatal outcomes and minimize rural–urban reproductive health indices in South Africa. Maternal health services need to be accessible, used more frequently, and of higher quality. In addition, strengthening girl child education is paramount, not only to improve women's empowerment, but also to improve ANC utilization among those who are pregnant. Further, health promotion in the primary and secondary levels of education needs to be intensified to change the narrative of poor ANC utilization among these categories of people.

**Author Contributions:** Conceptualization, O.O. and P.Z.N.; methodology, O.O.; validation, O.O., K.E.O. and L.B.; formal analysis, O.O. and F.E.A.; resources, O.O.; writing—original draft preparation, O.O.; writing—review and editing, O.O. and K.E.O.; visualization, M.N.; F.L.M.H., T.R.A. and J.A.M.; Supervision, O.O.; funding acquisition, O.O. All authors have read and agreed to the published version of the manuscript.

**Funding:** Financial support for OO from the Fogarty International Center and National Institute of Mental Health, National Institutes of Health Award (D43 TW010543), National Research Foundation Grant (132385) to Incentive Funding for Rated Researchers (IPRR), and The APC was funded by the Walter Sisulu Seed Funding. The content is solely the responsibility of the authors and does not necessarily reflect the opinion(s) of the sponsors and affiliated institutions.

**Institutional Review Board Statement:** The DHS team ensured that ethical approvals were obtained from the national department's health ethics committee before the surveys were conducted. Ethical approval was obtained from the WSU IRB (033/2021) before the secondary data were obtained from the DHS for retrospective analysis. Respondents in the DHS data were informed that participation in this study is voluntary and they were asked to sign a voluntary consent form prior to enrollment in the study. Confidentiality and privacy were respected.

**Data Availability Statement:** The datasets analyzed for this study are available at https://dhsprogram. com (accessed on 15 June 2021).

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

#### **References and Note**


## *Article* **Knowledge and Perception of Risk in Pregnancy and Childbirth among Women in Low-Income Communities in Accra**

**Patricia Anafi 1,\* and Wisdom Kwadwo Mprah <sup>2</sup>**


**Abstract:** Perception and knowledge of risk factors for pregnancy influence health behaviors during pregnancy and childbirth. We used a descriptive qualitative study to examine the perception and knowledge of risk factors in pregnancy and childbirth in low-income urban women in Ghana. Over the course of three-months, 12 focus group discussions and six individual interviews were conducted with 90 participants selected from six communities in the study area. Data were analyzed using inductive-thematic content analysis. Findings revealed that participants had knowledge of some risk factors, although some had superstitious beliefs. Participants viewed pregnancy as an exciting and unique experience, but also challenging, with a host of medical and psychological risks. Pre-existing medical conditions (e.g., diabetes), lack of physical activity, poverty, poor nutrition, and lack of social support were identified as conditions that could lead to negative pregnancy outcomes. Superstitious beliefs such as exposure to "evil eye" during pregnancy, as well as curses and spells, were also identified as risk factors for pregnancy complications. This research has implications for policies and programs to improve pregnancy outcomes for low-income women in Ghana. Thus, we recommend social and economic support programs as well as health education to change misperceptions about pregnancy risk and to support other efforts being made to improve maternal health outcomes.

**Keywords:** perception; knowledge; risk in pregnancy; women; low-income communities

#### **1. Introduction**

Risk is defined as a factor that presents eminent danger or increases the probability of experiencing adverse outcomes [1,2]. Perception, on the other hand, refers to a mental image or subjective ideas about a potential occurrence of a phenomenon [3]. The concept of perception is important because it is a driver of health-seeking behavior and the management of health outcomes [4]. Risk perception, therefore, refers to risk interpretations or understanding, as well as subjective judgements about risk [2]. In the field of health, risk perception denotes subjective judgements about the likelihood of negative or adverse outcomes of conditions such as illnesses, injuries, diseases, or death. Perceptions and knowledge about risk are important determinants of health behaviors and risk-related decision-making, such as whether to adopt or not to adopt healthy behaviors and reject or accept a certain level of health risks [2,5].

Risk perception has two dimensions: the cognitive aspect, which relates to how much individuals know and understand risks; and the emotional aspect, which relates to how people feel about themselves and a potential risk [6]. Research on risk perception often begins with the presumption that how people feel about danger is determined by the level of knowledge and certainty they have regarding that risk. This idea is founded on the rational choice model of decision-making, which presents individuals as rational beings who evaluate the possibilities of health outcomes after first estimating the prospective costs and benefits. On the other hand, laypersons usually assess risk by using heuristics (for

**Citation:** Anafi, P.; Mprah, W.K. Knowledge and Perception of Risk in Pregnancy and Childbirth among Women in Low-Income Communities in Accra. *Women* **2022**, *2*, 385–396. https://doi.org/10.3390/ women2040035

Academic Editors: Claudio Costantino and Maiorana Antonio

Received: 17 October 2022 Accepted: 14 November 2022 Published: 22 November 2022

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

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

example, previous experiences) and other informal ways of thinking [2,7]. This means that, when people are aware of certain risks or potential risks, they tend to believe that those risks could happen more frequently than the risks actually occur, and vice versa. These risk misinterpretations, rooted in heuristics, can cause people to overestimate or underestimate the occurrence and severity of potential health threats [2].

According to Lennon [8], risk perception in pregnancy entails both the objective medical evaluation of risk as well as a subjective, socially constructed risk, guided by a complex web of personal, psychological, and cultural factors. It includes the assessment of the possibility of harm or negative health outcomes to either the mother or the newborn or both. However, sometimes this harm or risk is not related to a particular medical condition; instead, women may see pregnancy and childbirth as inherent risks rather than conditions that cause the risk. In recent years, risk perception and the way risks are construed in pregnancy have received a lot of scholarly attention [8]. This can be explained in part by the growing prevalence of a medicalized perspective on pregnancy and the social pressure placed on women to behave in a way that lessens the perceived risks [8]. Increasingly, medical interventions are seen as both necessary and desirable for successful pregnancies, as the state of pregnancy has become more and more medicalized. Thus, seeking preventive services such as antenatal screening, genetic screening, and testing, as well as health behavior modifications, are considered necessary to reduce any potential risk of pregnancy complications [8,9].

Furthermore, considerable evidence shows that perceptions and understanding (or knowledge) of risk are shaped by many factors, including socio-cultural background [10], levels of literacy [11], and religious or traditional beliefs [11,12]. An individual's knowledge and interpretation of risk is also dependent on their personal life philosophy and previous experience [9]. It has been documented that sometimes women perceive risk in relation to pregnancy from the social perspective, where risk is seen as being influenced by the social, cultural, and political milieu in which they reside [8,9]. Wheeler and colleagues [13], for example, reported women's employment experience during pregnancy as an important factor in determining their perception of risk in pregnancy. Furthermore, other scholars have reported that maternal age [14–16], personal and family history [17], knowledge about pregnancy and childbirth [18], as well as level of risk or complication in pregnancy [19] are important determinants of perception and understanding of risk in pregnancy. These factors can influence opinions, interpretations of and values put on the risks and even benefits associated with pregnancy [12]. In a qualitative study on pregnancy risk perception and preterm birth, Silva and colleagues [17] found that personal negative experiences in previous pregnancies, such as stillbirths, miscarriages, having a preterm birth or neonatal death, informed risk perceptions of the current pregnancy. Women's perception of risk in pregnancy has also been attributed to their knowledge-base of risk [20].

Perception and understanding of risk in pregnancy have the effect of influencing various critical pregnancy and childbirth decisions. They affect decisions such as when and where to seek antenatal care, where to give birth (or choice of maternity site), who supervises the birth and even the mode of delivery [8,21]. Silva and colleagues [17] also argued that a patient's risk perception of health guides his or her decisions on treatment and can further influence health-seeking behavior during serious conditions such as preterm births, one of the most critical potential risks faced during pregnancy. Janson [22], on childbirth decisions and traditional structures in Ghana, explained that women who viewed childbirth as a natural process, and without potential risks or ill health, may consider home delivery as the best option. Traditionally, in Ghanaian society, pregnancy and childbirth are viewed as a vulnerability and potentially dangerous experiences that do not only require biomedical care, but also spiritual intervention. Therefore, many women seek a combination of care: biomedical care, traditional care (which involves the use of herbal medicine) and spiritual support (faith healing and prayer) in the pregnancy period with the hope of averting any perceived or actual health risk to the pregnancy and during childbirth [23]. The focus of this paper, therefore, is to present findings from a study

that investigated women's perception and knowledge of risk in pregnancy and childbirth, and how these affect their maternal healthcare seeking behavior in selected low-income communities in Accra. The implications of the findings for maternal and newborn health policies and programs are discussed.

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

#### *2.1. Study Design*

This was a descriptive qualitative study that targeted women in selected low-income communities in Accra, Ghana. The data were gathered over a three-month period through focus group discussions (FGDs) and interviews to examine women's perception and knowledge about risk in pregnancy and childbirth.

#### *2.2. Study Area*

The study was conducted in the Ashiedu-Keteke sub-metropolitan district in the city of Accra, Ghana's capital. Ashiedu-Keteke is one of the 3 sub-metropolitan districts of Accra Metropolitan City and has some of the poorest communities in the city. Using the Greater Accra region population growth of 3.1%, the sub-metropolis has an estimated population of about 143,990 in 2018 [24]. The sub-metropolis has the Central Business District (CBD), and it is the center of major commercial activities within the city of Accra, with an influx of approximately 2 million people from all parts of the country daily [24]. As the center of commerce, it houses the Markola and Agbogbloshie markets, which are two major markets within the Greater Accra region. The population of Ashiedu-Keteke is made up of the indigenous Ga people, who live in the coastal communities along the Gulf of Guinea (the Atlantic Ocean), and other migrant populations, who reside a bit further away from the coast. The main occupation of the indigenous Ga people is fishing; the men do the fishing, and the women smoke the fish for the market. The migrant population engages in small-scale commercial activities, mostly trading.

In terms of health care, the Ussher Polyclinic, the Prince Marie Louise Children's Hospital, and the James Town Maternity are among the major health facilities that provide care to residents in the sub-metropolitan district. Maternal health continues to be a health challenge as Ashiedu-Keteke has one of the poorest maternal health outcomes in Accra and has a persistently high teenage pregnancy rate in the city [25,26]. There is high prevalence of the use of informal maternal health care (traditional birth attendance, faith healing, and prayer) by pregnant women in the area, especially among the teenage and indigenous mothers. From available data, about 40% of pregnant women who reside in the sub metropolitan district do not seek antenatal care until the second and third trimesters, a situation quite a bit higher than the national average of 36% [27].

#### *2.3. Sampling of Participants and Data Collection*

Purposive sampling was utilized to select six communities within the study area for the FGDs and individual interviews. Altogether, 90 women between the ages of 17 and 45 years were recruited for 12 focus group discussions, and six individuals were involved in the interviews. For the FGDs, a snowball sampling approach was employed to recruit 84 women who had at least one child. We conducted two FGDs in each of the communities. We divided the women in the FGDs into two age cohorts: those aged 17 to 29 and those aged 30 to 45. Each FGD had an average of seven participants and lasted between 45–60 min. We conducted six FGDs with each age cohort. Additionally, six women with at least one child from each of the six communities (one mother from each community) were interviewed for about 30 min to elicit in-depth perspectives on the subject matter. Two community leaders assisted with the recruitment of the study participants.

The FGDs and individual interviews both used comparable questions, which were framed around our research objectives. We asked participants about their knowledge and risk perceptions with regard to pregnancy and childbirth and how these factors influenced their healthcare seeking decisions and behaviors during pregnancy and childbirth. Both FGDs and interviews were carried out with the assistance of our two fieldworkers and in two local languages spoken in the study area. The two field assistants were native speakers of the two local languages (Ga and Twi). The FGDs and interviews were audio recorded with the participants' permission.

#### *2.4. Ethical Consideration*

The Human Subject Review Committee of the University of Massachusetts Amherst School of Public Health granted the ethical approval of the research. Additionally, we received authorization from the local health administration in the study area and verbal consent from all study participants prior to their enrollment in the study. Since most of our study participants had only a basic formal education, we only sought verbal consent from them prior to recruitment into the study and assured them that information obtained from the FGDs and interviews would be kept confidential, and only the lead researcher would have access to the raw data. Participation in the study was voluntary, and they had the option of withdrawing at any time.

#### *2.5. Data Analysis*

The data analysis began during the data collection phase and continued following FGDs and interview sessions. With the assistance of our two fieldworkers, audio recordings of both FGDs and interviews were transcribed from the local languages into English. Following transcription, we cross-checked the transcripts to confirm that the responses had been accurately transcribed and translated from the local languages to English. We reviewed the final transcripts in order to determine which words, phrases, and statements were pertinent to our primary research questions and objectives. We classified the data into nine themes using an inductive content analysis approach proposed by Corbin and Strauss [28] and Miles and colleagues [29].

We first identified and cataloged the major concepts and recurring ideas in each interview transcript. Second, we compared and classified the significant concepts and emerging themes from the interview transcripts. We did the same thing with the transcript data from the FGDs; we compared significant concepts and developing themes within and across FGDs. The raw text section containing the essential concepts and ideas were coded manually and classified according to the emergent themes. To ensure the findings' internal validity, we compared and contrasted data segments from the FGDs with the individual interviews [28,29]. In this article, we present data on nine themes that emerged from our analysis.

#### **3. Results**

The nine themes were identified as the following factors: first trimester experience; medical conditions; lack of physical activity; antenatal care; lack of social support; poverty and poor nutrition; sleeping posture and hot showers; exposure to certain conditions; and spells and curses.

#### *3.1. First Trimester Experience*

According to most of the participants, pregnancy during the first trimester period is associated with both joy and risk. The participants reported that the usual thought of being pregnant brings "joy" and "hope", but the experience during this period could be challenging for many women. Participants explained that during the first trimester, the pregnant woman is unable to eat, as she experiences general body weakness, morning sickness, fevers, severe headaches, dizziness, heart palpitations, and sleeplessness. Although these are normal, they could pose serious health risks to the woman during this period of the pregnancy. The women described psychological problems such as fears, anxiety, and stress that a woman could experience during early months of the pregnancy. Two participants explained it this way:

"When a woman is pregnant, she does not feel well as she used to be, the whole experience can make you feel sick and you know, this feeling of sickness can last during the whole pregnancy period for many women, the first three months can be hard on you, you become anxious and afraid . . . " [FGD participant 1]

"In my case, I always felt like I was going to fall down in my early months. But this feeling happens to many pregnant women, so its normal to feel that way because your body is adjusting, it begins to go away after the third months for most pregnant women." [Interview participant 1]

#### *3.2. Medical Conditions*

Both FGD and interview participants were of the view that complications could occur in pregnancy when a pregnant woman already has "a pre-existing disease". They explained that pre-existing diseases or infections such as "malaria", "diabetes", candidiasis ("odeepu") or "HIV" can increase the vulnerability of a pregnant woman. This is because "it is not easy to carry a pregnancy when you are already sick or have a disease." [Interview participant 2] Having a chronic condition " ... like AIDS or sugar disease [diabetes] and becoming pregnant could be challenging as you don't sit at home. You will need to go to the clinic for regular and proper care." [FGD participant 2] These medical conditions could expose the woman to the risk of experiencing a difficult pregnancy, labor, and delivery, and could affect the baby's health as well.

According to the participants in the FGDs, other medical conditions indicating that the pregnant woman is at risk of complications are high blood pressure, swelling in the hands and feet, anemia, and delay in the delivery of the afterbirth or the placenta. Participants also mentioned bleeding during pregnancy and childbirth as a major health risk for the mother and the baby. These views were also expressed by the women in the individual interviews, as exemplified by a personal experience of an interview participant who said:

"I don't have an easy pregnancy. My second pregnancy was the worst of all. I had swollen feet and hands, and they [nurses] said my blood pressure was going up at some point, and I thought those were not good signs for my pregnancy, especially when they told me my pressure was going high and I had to come in for regular review." [Interview participant 3]

#### *3.3. Lack of Physical Activity*

Another risk factor identified was being physically inactive during pregnancy. According to some of the participants, a pregnant woman who is physically inactive or does not exercise could likewise be at risk for a difficult and prolonged labor, as explained in the following statement by an FGD participant.

"You see, there are pregnant women who don't do any work, they don't walk, they just sit at one place, and they stop coming to the market. But there are others who work with their pregnancy. They go to the farm and do everything. When you are not active, the baby will not be active. The baby cannot turn or move in the womb, and you will have problem during delivery ... " [ FGD participant 3]

#### *3.4. Antenatal Care*

Most of the participants said women who do not seek medical care or attend antenatal clinics could be at risk of pregnancy complications. During the FGDs, participants recounted that a pregnant woman should seek regular health care to prevent potential complications that could affect the pregnancy and to ensure a successful birth. A participant in the FGDs noted, "a situation where the mother failed to seek antenatal screening for early risk detection and treatment, it leads to conditions such as fetal malpresentation or malposition, a critical condition for prolonged labor." [FGD participant 4]

Similar views were also shared by participants in the interviews, who identified prolonged labor as a risk a pregnant woman could be exposed to during childbirth if she fails to attend an antenatal clinic.

#### *3.5. Lack of Social Support*

According to the participants, pregnant women who do not have adequate spousal or family support could experience difficult pregnancy and childbirth because pregnancy could be stressful and emotionally demanding. As a result, if the pregnant woman is not adequately supported, it could negatively affect her health and that of the unborn baby. Two FGD participants explained the effect of a lack of social support on pregnancy outcomes in the quotation below.

"I know a young woman that the man who made her pregnant refused to accept the pregnancy. Her parents are not living here [referring to her community] and she does not have any other family member in Accra here. The baby she delivered was very small. You know ... she didn't have any support; she was always by herself and only got a little help from neighbors and that affected the baby." [FGD participant 5]

"Sometimes, these young mothers work long hours. The kayaye [teen girls head porters] who are pregnant ... they carry heavy loads in the market for people who come to shop or do groceries. They don't have their families or anybody here to help them ... It is stressful to do this kind of job when you are pregnant . . . " [FGD participants 6]

These findings were corroborated by the interview participants who reported a lack of social support for young migrant pregnant women in the study area as a major risk factor for negative pregnancy outcomes. The following remark was made by a participant in support of this assertion, "when you go to the market right now, you see them [young female migrant porters] carrying big bowls full of load for people with their pregnancy. Some of them don't even know the fathers of their babies, and their parents are not here to help them, and this is not good for the pregnancy." [Interview participant 4]

#### *3.6. Poverty and Poor Nutrition*

Poverty emerged as an important risk factor for pregnancy and childbirth complications. Both FGDs and interview participants agreed that poverty could lead to poor nutrition among pregnant women, leading to poor pregnancy outcomes because "a pregnant woman should ensure she eats well, on time, and the right portions", which is a major problem for some women. According to the participants, some pregnant women "don't have the money to buy enough food" and that could affect their health and the pregnancy as well. A pregnant participant in one of the FGDs confirmed this and said, "the poor mothers who don't have families here, their general health begins to become worse after delivery. This is because when they go home, there is nothing there to eat and their babies cannot grow because they don't feed them well since they themselves don't eat well". [FGD Participant 7] According to some participants, the young pregnant migrant mothers who lived in the study area relied on neighbors for meals when they became pregnant. One FGD participant also put it this way:

" ... if they don't get this assistance, they don't eat. Some pregnant mothers eat small portions of meal in the morning, and they don't eat until evening because they don't have families here [Accra] and the men who made them pregnant didn't accept the pregnancy." [FGD participant 8]

These findings were supported by some of the interview participants who said that the young pregnant migrant mothers usually worked for long hours, carrying heavy loads to make a living, and to save towards delivery, which exposes them to pregnancy-related risks. The quotation below highlights this issue.

"Oh, when you go there (markets) right now you will see them working in the market, they carry loads in the market. Some sell under the hot sun [and], this is not too good for the pregnancy, but they need the money, so they and their babies can have food to eat the weeks following delivery." [Interview participant 5]

Furthermore, most of the participants thought that poverty made it difficult for pregnant mothers to attend antenatal clinics and delayed care during labor, exposing the pregnant woman and the baby to life-threatening complications.

#### *3.7. Sleeping Posture and Hot Showers*

Both the FGDs and interview participants agreed that when a pregnant mother sleeps on her back, it can affect the health of the mother and the unborn baby. They explained that when a pregnant mother sleeps on her back, it obstructs the flow of oxygen from the mother to the unborn baby, and this could be fatal for the unborn baby as well as the mother.

"This [sleeping on your back during pregnancy] is not good for your pregnancy ... when you are pregnant, you don't sleep on your back. You sleep by your side. People (pregnant women) who sleep on their back can hurt their baby, the baby cannot breath because air will not flow from you the mother to your baby ... so they tell us not to sleep on our back. They are not good practice for the baby, so you can kill your baby because the baby cannot get air from your when you sleep on your back . . . " [FGD participants 9]

Some participants also reported that taking hot water showers and baths could negatively affect the pregnancy. Two participants explained:

" ... they say bathing with hot water is not good, warm water is okay ... hot water is not good for the baby too. It can make the baby temperature go up" [Interview participant 4]

"Bathing with hot water can make you the mother feel hot ... and you can pass heat on to the baby in your womb, so you don't take hot showers when you are pregnant." [FGD participants 10]

#### *3.8. Exposure to Certain Conditions*

One major perceived risk in pregnancy mentioned by the participants is the exposure of the unborn baby to conditions that can cause the child to acquire certain health defects such as cerebral palsy (called "asram" in the local language). This often happens if the pregnant woman does not take "very good care" of herself and her pregnancy. For example, if the pregnant woman does not dress well, she could expose herself and her pregnancy (unborn baby) to individuals with such conditions. This is based on the belief that cerebral palsy is transferrable spiritually from a person who has the condition to the mother, and then to the baby to be born. "Asram" is a spiritual disease that can be passed on through eye contact when a pregnant mother comes into contact with an individual who spreads the disease. One interview participant elaborated on this perception in the following quotation.

"For a disease like asram, when you are pregnant you don't have to eat everywhere. When some people see the pregnancy, they can transfer the disease to the unborn baby. You have to dress decently, so that you don't expose yourself and your pregnancy. Some people dress exposing their body when they are pregnant ... ... So when you dress like that and you come across someone with the disease, that individual can transfer the disease to the baby." [Interview participant 4]

Some participants also reported that cerebral palsy can kill, and that it can only be treated with traditional or herbal medicine. For example, during the FGDs, one mother recounted her experience with "asram", "it [asram] affected me and my baby, and I was taken to a certain woman for herbal treatment for almost a year before my baby was ok." [FGD participant 11].

#### *3.9. Spells and Curses*

Participants also narrated that a woman can be at risk of pregnancy complications and/or even death if she is cursed during pregnancy. According to most FGD participants, when a pregnant woman is disrespectful or often picks quarrels with neighbors, a curse could be cast on her, and this may lead to stillbirth and death during childbirth. This view was confirmed by a participant in the interviews in the following remarks:

" ... do you know that evil people can harm your pregnancy? They can cast a curse spell on you and your unborn baby, so the young pregnant women here who like quarreling and fighting, some ended up having a difficulty childbirth." [Interview participant 4]

This finding was supported by other participants who spoke of some young pregnant mothers visiting spiritual churches to seek protection for fear the effects of curses by people they might have wronged.

"They [young pregnant mothers] go the spiritual churches to pray for successful pregnancy because they are afraid somebody that they have quarreled with or disrespected might have cursed them and the pregnancy." [FGD participants 12]

#### **4. Discussion**

Although the findings from this study do not represent the views of all Ghanaian women, they provide some insight into the general perception and knowledge of risk factors for negative pregnancy and childbirth outcomes among women in Ghana. The findings indicated that participants view pregnancy as an exciting and unique experience; however, they acknowledged the risks, both medical and spiritual, associated with it. This perception of risk could serve as a motivation for women to adopt positive health-seeking behaviors such as attending antenatal clinics, exercising, and eating good food during pregnancy. However, this perception of risk can also induce serious emotional experiences such as fear, anxiety, and stress that will require both informal and professional support. On the contrary, it seems that some women in the study area were not adequately receiving this support during pregnancy. This might be one of the reasons why maternal health outcomes in the study area are reported to be poor [26]. The findings of the study are consistent with previous studies where women were found to hold such perceptions of risks [30–32]. Anxiety and fear due to perceived risks in pregnancy and childbirth, such as prolonged and painful labor, lack of social support, as well as economic uncertainty, leading to stress, were identified among women in studies by Erickson et al. [33], Lyberg and Severinson [34], and Saisto and Halmesmaki [35]. Thus, pregnancy risk perceptions among pregnant women can affect their health and health care decisions and treatment options [17].

The findings that women perceived a lack of social support as a risk factor for pregnancy is relevant. As confirmed by previous studies, people with a high quality or quantity of social networks and economic stability have a decreased risk of mortality in comparison with those who have a low quality or quantity of social network engagement and are economically unstable [36–38]. In pregnancy, in particular, Hotelling and colleagues [39] found that women with continuous support, either emotionally or socially, were less likely to have complications in pregnancy that could lead to Caesarean deliveries than those without any support. Likewise, evidence from Ghana indicated that lack of support from friends and extended family, being abandoned by one's husband, and being compelled to live with unfriendly in-laws are risk factors that could expose pregnant women to psychological problems during pregnancy [12].

Several studies have illustrated that poor women are at higher risk of food insufficiency, insecurity, and poor feeding practices, leading to malnutrition and maternal morbidity [40–42]. Poverty hinders access to sufficient and nutritious food, and at the same time, acts as a barrier to accessing quality and timely maternal health care [43,44]. In Kenya, for example, Izugbara and Ngilangwa [45] found that poverty compelled

pregnant women in slums to engage in tedious work for long hours in order to save enough money for delivery, risking their lives. Women in our study shared similar views regarding young migrant pregnant women who carry heavy loads and work longer hours for a living and in preparation for delivery.

The findings showed that participants had knowledge of some common medical risks associated with negative pregnancy outcomes in Ghana. In Ghana, like many other tropical countries, malaria is endemic, and it is known to be a major contributing factor to stillbirths [46,47]. Medical conditions such as diabetes, candidiasis, high blood pressure, HIV, anemia, and delayed placenta as well as lack of physical activity are also risk factors that were identified as dangers to pregnancy and childbirth. These findings are consistent with a study conducted in Kenya where participants identified similar conditions as threats to positive pregnancy outcomes [48].

Our participants had a strong belief in religious-spiritual factors such as curses as risk factors in pregnancy. These beliefs are rooted in the community and are not only widespread among Ghanaians, but also exist in many cultures in sub-Saharan Africa. This belief has an influence on health care choices and decisions during pregnancy and childbirth and motivates most women to resort to herbal medicine, spiritual care (faith healing and prayer), or a combination of medical and traditional treatments during pregnancy [12,49]. In rural Zimbabwe, for example, it has been discovered that women fear being bewitched because they are thought to be vulnerable to witchcraft in the early months of pregnancy. As such, they preferred to seek protection from faith healers who are believed to possess supernatural powers to protect them and their pregnancy [50]. Similar findings have also been observed in studies in Southern America [51] and indigenous Pilipino in Southeast Asia [52]. Interestingly, this notion, according to the authors, is parallel to the biomedical perspective, which describes the early months of pregnancy as the most critical period. Unlike in many western cultures, women in Ghana fully cover their pregnancy during the entire pregnancy period. This practice is believed to protect the pregnancy from a curse, spell, or witchcraft [53].

#### **5. Implications**

As governments are making efforts to address the challenges of poor maternal health outcomes, the need for research that has practical applications is essential. Our findings revealed many issues that are relevant for maternal and child healthcare policies and programs for low-income women in Ghana. First, the findings suggest that many pregnant women go through psychological issues during pregnancy due to the perceived risks associated with it. Although medical interventions such as antenatal care are important during pregnancy, psychological support is equally important. Pregnancy induces physical and emotional change and increases the risk of mental illness [54]. Psychological support for women during pregnancy is therefore very important for positive outcomes. The Ministry of Health, in consultation with other relevant stakeholders such as the department of Social Welfare, should consider the implementation of social and emotional support programs that can help to alleviate the emotional stress of pregnant women, especially for poor and single women. Counseling units, specifically focusing on pregnant women, could be set up at the maternal and childcare units of health care centers to support women who need help.

Second, the belief that pregnant women could be cursed has implications for safe motherhood policies and programs in Ghana. As the findings indicated, some pregnant women use traditional medical practitioners, including faith healers, due to the belief that they could be cursed. This does not promote safe motherhood practices and should be addressed. Public health education must focus on explaining to women and communities the risk and non-risk factors associated with pregnancy and childbirth, as well as correcting misperceptions that may have a negative impact on pregnancy outcomes.

Third, the impact of poverty was a major risk factor identified by the participants. Current socio-economic conditions in the country are having dire consequences on the poor, and this could seriously affect Ghana's quest to achieve its Sustainable Development Goal 1. Poverty among women needs serious attention through protective social interventions. The Ministry of Health and Ministry of Gender, Children and Social Protection could partner with community organizations, including churches and other religious groups, to augment government efforts to reduce poverty among deprived communities.

#### **6. Conclusions**

The findings of the study revealed that low-income urban women in Ghana have a wide range of knowledge and perceptions of risk factors for negative pregnancy and childbirth outcomes. The women viewed pregnancy as a unique experience, but they acknowledged that it could be affected by a host of medical and non-medical issues, which could lead to adverse pregnancy outcomes, including negatively affecting the health of the mother and the baby. Though the study focused on urban women, these risk perceptions and knowledge about pregnancy and childbirth are held by many Ghanaian women. The findings also suggest that despite efforts being made to reduce maternal morbidity and mortality in Ghana, many pregnant women are still at risk of pregnancyrelated complications. This study therefore brings to the fore the need to employ multiple approaches to safe motherhood programs—social, economic, religious, and psychological to help poor pregnant women. While these activities will reduce barriers to quality maternal healthcare, health education should be ramped up to address superstitious beliefs as well as cultural misperceptions about pregnancy risk factors.

**Author Contributions:** P.A. was the lead investigator; led the research design, data collection and analysis. P.A. and W.K.M. prepared the original manuscript, reviewed, and edited. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by Compton Foundation International Dissertation Fellowship.

**Institutional Review Board Statement:** The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Boards (or Ethics Committees) of University of Massachusetts-Amherst School of Public Health [SPHHS-HSRC file # 10:35].

**Informed Consent Statement:** Verbal informed consent was obtained from all subjects involved in the study. Additionally, authorization was granted from the sub metropolitan district health administration to conduct the research.

**Data Availability Statement:** The study data are available upon request from the corresponding author and with approval from the author's institution's Research Ethics Committee.

**Acknowledgments:** Authors profound gratitude goes to David R. Buchanan in the School of Public Health and Health Sciences at the University of Massachusetts-Amherst who supervised the original research. We also acknowledge our research participants and two field research assistants.

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

#### **References**


**Ritsuko Shirabe 1,\*, Tsuyoshi Okuhara 2, Hiroko Okada 2, Eiko Goto <sup>2</sup> and Takahiro Kiuchi <sup>2</sup>**


**Abstract:** Support needs for pregnancy-related anxiety among low-risk pregnant women remain unclear. This study aimed to clarify the kinds of support for anxiety that women seek during pregnancy in Japan. Data were collected in a semi-structured focus group interview involving five pregnant women who were not in specific risk groups, recruited from three facilities in Tokyo. We generated themes using inductive thematic analysis. This paper adhered to the consolidated criteria for reporting qualitative research. From the data on support needs for anxiety during pregnancy, three themes were derived: (1) seeking tailored professional support; (2) seeking continuous support within informal relationships; and (3) seeking others' success stories in the same situation. These three types of support gave participants a sense of reassurance or raised concern, depending on the situation. We proposed a model comprising the three derived themes using social cognitive theory. We discussed how these three types of support influenced pregnant women's self-efficacy, which is the core concept of the social cognitive theory. Our findings may help to plan theory-based research and effective interventions to provide support for women's anxiety during pregnancy using a population approach. Our results also demonstrated the importance of collaboration with pregnant women in developing further research and interventions.

**Keywords:** pregnancy; pregnancy-related anxiety; needs assessment; social support; self-efficacy; social cognitive theory; qualitative study; health communication

#### **1. Introduction**

Maternal mental health problems have become a major issue worldwide. Japan is no exception, as 10–20% of mothers become depressed after childbirth [1] and many more experience milder symptoms than depression. Support for maternal mental health should start from the antenatal period because more than one in five pregnant women experience anxiety or depression symptoms during pregnancy [2], which can predict the deterioration of their mental health in the following postpartum period [3]. Support during pregnancy may also be beneficial in terms of the feasibility of seamless care, because pregnant women in Japan can receive about fourteen antenatal checkups at public expense [4] and have opportunities to meet with specialists during pregnancy rather than postpartum.

To consider perinatal maternal mental health, numerous studies have explored psychosocial factors, such as parenting stress and social support. There is evidence that social support, which is considered an environmental factor in social cognitive theory [5], may be a major protective factor against perinatal anxiety and depression [6–11]. This theory has previously been used in the context of perinatal mental health problems [12–14]. In this theory, self-efficacy is the core concept for cognitive factors and interacting with environmental and behavioral factors [5], with one study indicating that self-efficacy for nurturing mediated the association between social support and postnatal depression [14]. However,

**Citation:** Shirabe, R.; Okuhara, T.; Okada, H.; Goto, E.; Kiuchi, T. Support Needs for Anxiety among Pregnant Women in Japan: A Qualitative Pilot Study. *Women* **2023**, *3*, 95–106. https://doi.org/10.3390/ women3010008

Academic Editor: Mary V. Seeman

Received: 7 December 2022 Revised: 30 January 2023 Accepted: 31 January 2023 Published: 8 February 2023

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

we cannot ignore the negative aspects of social support, because support that did not meet needs was associated with an increased likelihood of postnatal depression [15]. Therefore, we should improve our understanding of the needs of pregnant women to ensure the provision of effective support in terms of mental health.

Multiple factors can lead to perinatal mental disorders, and it is difficult to identify high-risk individuals; therefore, a universal approach for preventing the deterioration of maternal mental health that extends beyond identified-risk groups is crucial. Previous qualitative studies have explored and clarified needs in medical situations among pregnant women, leading to one review which indicated that routine antenatal services might help only a small proportion of what matters to pregnant women without identified risks [16]. If routine antenatal services are not enough to help pregnancy-related anxiety, it is important to identify what kind of support pregnant women seek and what level of satisfaction and dissatisfaction they have in their daily lives in terms of their anxiety; however, at the present time, such details remain unclear.

We conducted this pilot study to explore pregnancy-related anxiety and support needs for anxieties among women residing in Japan, regardless of their specific risks. We believe these findings may be useful to conduct further quantitative studies and generate instruments or programs for a universal intervention to support pregnant women in Japan in the future.

#### **2. Results**

#### *2.1. Participants' Characteristics*

The participants' mean age was 31.8 years (range 27–36 years), and the mean gestational period was 28 weeks (range 23–33 weeks) at the time of the interview. Four of the five women were expecting their first baby. All participants were married and living with their husbands. All participants had completed university education and were working (Table 1).


**Table 1.** Participants' characteristics.

GA, gestational age; IVF, in vitro fertilization.

Participants expressed various anxieties based on their personal background and by hearing other participants' narratives. Table A1 shows the coded anxieties of this focus group.

#### *2.2. Support Needs*

We found three themes that described women's support needs for anxiety during pregnancy.


Table 2 shows the three derived themes with supporting quotes. The women's positive affects, derived from having received support that met their needs or was more than they expected, were coded as positive examples. Experiences where support for their needs was lacking (including negative affects caused by the support they received) were coded as negative examples.

**Table 2.** Themes and supporting quotes.


<sup>1</sup> Participants' narratives were translated into English for this publication.

#### 2.2.1. Theme 1: Seeking Tailored Professional Support

In the professional sector, participants did not expect experts to only be involved with them in a manualized way. They wanted experts to see them not as "a pregnant woman" but as "myself". This did not have to involve difficult technology, as participants indicated that simple measures such as a friendly attitude toward "me", talking to "me", or listening to "my" minor physical problems and gaining a sense of the professionals' involvement was helpful.

In fact, it's like ... when I gave birth, the people around me were ... well ... only the midwives and the doctors, so I felt very reassured, and the doctors and midwives who were there and kept calling out to me were of course much more powerful than the stories I had heard. That's right. It was very reassuring. (Participant D)

Participants also expressed concerns about the professional sector, such as unclear explanations by medical professionals, uncertainty, and lack of options that allowed them to focus on their values, even if there was insufficient evidence. Beyond the manual or guideline, they needed guidance on how to bridge the gap between the scientific evidence of "correct answers" and their own real-world situations.

I tried to talk to a nurse or a doctor at the hospital about such things [partner's smoking], but all they said was that I should definitely stop him from smoking . . . (Participant E)

2.2.2. Theme 2: Seeking Continuous Support within Informal Relationships

All participants agreed on the value of the existence of continued support within informal relationships. They especially valued support from people from the same generation who had previously experienced pregnancy or childbirth. Because of their high expectation of such support, one participant expressed concern about the lack of it.

I don't have any friends who are pregnant or have children, so I don't really have anyone to talk to about something ... like events during pregnancy, I guess I'm a little nervous about it. (Participant B)

Participants could get new information and felt reassured by just talking within such relationships without any specific purpose. Participants indicated they could talk easily and honestly within such relationships about what they hesitated to talk to experts about (e.g., weight gain during pregnancy and fear of childbirth).

When I'm in the hospital, I feel a bit rushed and I don't feel like I can take my time to talk to the doctor, so I tend to ask [another] mother who is ahead of me such things about the pain of childbirth and daily life after childbirth. (Participant C)

Participants also reported encountering unexpected information or messages within informal relationships that they did not want.

My number of gestational weeks is now 33, so I can't run any more tests of a definitive diagnosis at this week, well, like NIPT (noninvasive prenatal genetic testing). I'll tell you what ... oh, I guess I should have taken it, although I can't take it anymore, when I've heard that my friend who is also pregnant got it, I was a little worried that maybe I should have taken the test, well, I made my own decision, but I think I [will] have anxiety about whether my decision was right all the way until my baby is born. (Participant C)

In particular, they valued support from their friends and relatives rather than cohabiting family, perhaps because they wanted "reasonable" nosiness. Some participants felt worried and anxious about the excessive involvement or indifferent attitudes of partners.

My husband is a smoker, and we've talked about it, but the results haven't been to my satisfaction . . . (Participant E)

2.2.3. Theme 3: Seeking Others' Success Stories in the Same Situation

Most participants sought information about "success stories" of other people in the same situation, especially in their trouble. To seek this support, they often chose sources that did not have limitations in terms of time or access (i.e., the Internet, social networking services, and magazines), because information from people around them was not enough. They wanted to read or hear such stories at their convenience.

Well, I'm not sure if I can really take care of my children properly ... of course my niece and nephew are cute but ... yes, this is my first baby so I guess I'm a little nervous about having to raise my baby day in and day out by myself, so ... although I think the only way to solve it is to have a baby and raise it, I've been looking at pictures of cute babies on the Internet, or been reading blogs, by those things I can think babies are pretty cute, and it's kind of comforting. (Participant E)

Participants with concerns sought answers and solutions, even if they were not medically correct. Information biases, text-based one-way information, and inaccurate information in this type of support could damage participants physically and emotionally.

(By reading information about non-evidence-based treatments on the Internet) Well, that's the emotional part, the bleeding may not be treatable, but if I receive some kind of treatment, and it might be good for my baby ... that would make me feel like I'm doing my best, it's better than doing nothing ... on the contrary, I was doing nothing and just waiting to see what happens ... oh, well, I think I was more worried. (Participant B)

#### *2.3. Proposed Model*

Based on our findings, we proposed a model of the needs of support for anxiety among pregnant women (Figure 1). Pregnant women in the focus group sought three types of support: tailored professional support; continuous support within informal relationships; available success stories.

**Figure 1.** Needs of support for anxiety during pregnancy. White circles: type of support; Black arrows: needs.

In our study, participants sought these three types of support based on their individual preferences and environment. They used different types of support depending on their situation. To paraphrase using the Figure 1, black arrow thickness varies and fluctuates among individuals.

When it comes to normal checkups, I don't really have that much to talk about ... In other words, I thought that it would be easier to ask someone closer to me. (Participant D)

Some participants relied on online information when they could not solve their own problems by talking to experts or to the people around them.

At first, I was very shocked, it is my first pregnancy, and no way, I was told that there was a possibility of a miscarriage. After that, I talked to the doctor about various things, let me see, like how likely it was, I asked a lot of questions about such things, but I was said the doctor didn't know ... After I went home, I did a Google search and then it said that in my case, almost 100% my baby was going to be miscarried, so my shock got worse. (Participant B)

#### **3. Materials and Methods**

We used a narrative research design, whereby researchers study the lives of individuals and ask them to provide stories about their lives [17]. We conducted a semi-structured focus group discussion with pregnant women to explore a specific set of issues [18]. This is considered a useful method to elicit information on patient priorities and needs, with the aim of improving the quality of healthcare by collecting rich and detailed data in an interactional group structure [19,20]. One or two focus groups are said to be sufficient in exploratory studies [20], therefore we planned to conduct one focus group involving 4–12 women to analyze data over time [18]. This study adhered to the consolidated criteria for reporting qualitative research (COREQ) [21].

#### *3.1. Participant Recruitment*

This pilot study used convenience sampling to rapidly recruit participants for the following quantitative survey. To ensure that we recruited women from various backgrounds, this study was conducted across three facilities in Tokyo, Japan: a university hospital, a hospital, and a clinic. The university hospital is a regional perinatal center that manages normal and painless deliveries as well as high-risk pregnancies (e.g., maternal complications and fetal diseases). The hospital is also a regional perinatal center and manages various delivery types, including socially high-risk women, but does not deal with painless deliveries. The clinic only conducts pregnancy checkups and does not manage deliveries. Women who were pregnant, living in Japan, and fluent in Japanese were eligible for this study. In October 2020, the first author or research collaborators (obstetricians/midwives) approached eligible women who visited each facility face-to-face (e.g., "Would you like to talk about your anxieties during pregnancy?") and introduced this study using an information leaflet. The leaflet included a QR code that led to the first author's email address for further explanation, which was conducted face-to-face or in an online meeting. Of the 11 women who were approached and informed about this study (including the rationale) by the first author, eight women agreed to participate. Three women declined to participate; one did not want to show her home online; one believed that the research was useless and that policies (e.g., improving nursery schools) were the only support she needed; and the reason given by the third woman was unclear. After recruitment, three more women dropped out because of emergency hospitalization, schedule inconvenience, and inability to be contacted; therefore, we conducted a focus group with five women.

All participants provided written informed consent to participate in the study and to the publication of this paper. Each woman received a JPY 2000 (USD 20) gift certificate after the interview as a gesture of appreciation for their time. This study was approved by the Institutional Review Board of The University of Tokyo (approval code: 2020154NI).

#### *3.2. Interview Procedure*

Before the interview, we collected the participants' demographic information using an online questionnaire. This information included marital status, cohabiting family, job, education background, economic comfort (four-point Likert-scale; almost none to pretty much comfortable), and participation in any prenatal education class. Other information was collected from their medical records (age, expected delivery date, number of past deliveries and abortions, infertility treatment, planned delivery facility, and identified risk) by research collaborators at each facility.

The semi-structured focus group interview was held in an online meeting room (Zoom meetings: https://explore.zoom.us/ja/products/meetings/ (accessed on 2 February 2023)) to prevent infection during the current COVID-19 pandemic. The first author (female, MD, MPH) had conducted in-person medical interviews with pregnant women for seven years as an obstetrician. She felt challenged that the anxieties of pregnant women were unable to be fully addressed in their clinic visits. Therefore, this author facilitated the interview but concealed her occupation. Only the participants and the facilitator were present during the focus group. Following established focus group interview procedures [20], we created the goals of this pilot study: (1) to identify the needs of pregnant women without specific risks regarding the support that can be provided for their anxieties during pregnancy; (2) to get some idea and information from participants which lead to support and initiatives that pregnant women want to participate in. We designed guidelines and semi-structured questions for the interview based on these goals. The interview began with the general question, "Do you have any anxieties about this pregnancy or childbirth?" followed by us showing them a developing scale about pregnancy-related anxiety and discussing it. After clarifying and sharing each participant's anxieties during pregnancy, the facilitator gave semi-structured questions such as: "How are you dealing with your anxieties?"; "Do you want to share your anxieties with someone else?"; "How much are you satisfied with support during pregnancy?"; and "What are you satisfied or dissatisfied with?" The facilitator took field notes throughout the interview with particular attention to the speaker's tone of voice and to the facial expressions and physical posture of speakers and listeners. The interview was audio- and video-recorded and lasted around one hour. Repeat interviews were not conducted because we intended to hear narratives of pregnant women in real time and all participants were no longer pregnant after data analysis.

#### *3.3. Data Analysis*

The unit of analysis was the individual who was pregnant. The first author transcribed the audio recording verbatim immediately after the interview. Participants' facial expressions and physical actions (e.g., nods, laughter, and raising their hands) were captured by the video recording and added to the transcript in words. The transcript was returned to participants for comment, with no corrections necessary. We applied inductive thematic analysis as proposed by Boyatzis [22]. The first author read the transcript carefully and repeatedly, and then coded it manually irrespective of the goals of the interview. The unit of coding was one sentence of the transcript. All derived codes were labeled with definitions and recorded in Microsoft® Excel® for Microsoft 365 MSO (version 2212 build 16.0.15928.20196) 64 bits as a code book. After coding, all similar codes were identified and grouped. Next, the first author and the second author (who had experience of qualitative studies) discussed the data in order to generate themes focused on the types of support participants wanted or perceived as helpful. Thematic codes were generated with consideration for the conditions proposed by Boyatzis, such as clear definitions, conditions for inclusion/exclusion, and specific positive and negative examples [22]. We classified codes as "positive" or "negative" focusing on the speaker's affects. Finally, the thematic codes were adapted to all derived codes. Codes that were judged to be irrelevant for the goals of the interview (e.g., anxiety about some specific risks) were then removed from the analysis. The first author wrote a report of the overall analysis and discussed the generated codes and themes among authors to reach a consensus. This report was returned to all participants, and no objections or changes were made. Two weeks after the final coding, the first author repeated the coding using non-marked transcripts. Minor inconsistencies were modified through discussion. All coding procedures were recorded in the code book.

#### **4. Discussion**

We extracted three types of support needs for anxiety during pregnancy based on the perceptions of the focus group participants: tailored professional support, mainly from the medical field; existence of informal relationships, especially with those in the same situation with adequate influence; and successful examples that could be easily accessed. With these three types of support, participants could have both positive and negative feelings depending on their situation. Even if there was an apparent supply of support, the support was not always appropriate for their needs.

To achieve positive pregnancy experience, the previous review questioned the tendency of routine antenatal care which focused on biomedical tests and treatment [16]. Pregnant women mostly sought healthcare support, such as access to healthcare services and experiences within medical/healthcare settings (i.e., positive interpersonal relationships with providers, skills and competencies of providers, and getting physiological, biomedical, and behavioral information) [16]. While our study also showed that the involvement of professionals can be both a positive and negative emotional experience for pregnant women, it also suggested that what they are looking for may differ depending on the place and person of support supplied. In other words, while they naturally sought biomedical tests and treatments from professionals, the degree to which they sought psychological support from professionals varied from person to person; some women were likely to turn to professional support, while others were more likely to turn to other sources, possibly due to the nature of their concern or because of a previous experience of seeking but not getting support. Our study covered both support in medical situations and their

experiences in daily life to highlight their preferences and conflicts, details which have been difficult to ascertain in previous reviews and quantitative studies.

#### *4.1. Theoretical Implications*

Bandura outlined four factors that influenced self-efficacy: (i) previous experiences; (ii) vicarious experiences; (iii) verbal persuasion; and (iv) emotional arousal [23]. From this perspective, self-efficacy is developed by one's own experiences and by seeing successes and failures of other people (vicarious experiences). Verbal persuasion, which encompasses direct encouragement from a trustworthy person, is also effective in building self-efficacy. The three themes extracted in this study corresponded to two of Bandura's four factors (vicarious experiences and verbal persuasion). In other words, pregnant women in this study may have acted or sought support for building their self-efficacy. Therefore, the selfefficacy of pregnant women may be improved by creating a desirable support environment.

Our proposed model (Figure 1) suggests pregnant women can build their own selfefficacy if they can obtain the three types of support that meet their needs. However, the impact of support on self-efficacy may vary by the type of support, because each of Bandura's four factors has a different strength of impact on self-efficacy in various fields [23,24]. In addition, people appear to live in their own unique psychological world [23]. In other words, each person perceives, understands, and remembers events through their own unique "lens" [25]. The types of support to offer and focus on therefore depend on the unique lens of that individual. As seen in our study, pregnant women may choose (and change) the type of support they rely on based on their preferences and experiences. Therefore, to support every pregnant woman, interventions covering all three directions may be needed.

#### *4.2. Practice Implications*

The benefits and best methods of education or support programs during pregnancy remain unclear for a few reasons: it is difficult to conduct high quality research (e.g., randomized, controlled trials) in this population because of ethical considerations; and previous studies were conducted for various purposes using different methods [26]. Our findings also indicated that intervention with a single type of support may not be effective, which may explain the inconsistent results previously reported [26].

As in a previous review [16], the present study found examples in which pregnant women's needs were not met in the professional sector. Our results indicated that professional support has different functions from other types of support (e.g., friendship) because pregnant women may hesitate to talk about all of their concerns with experts. Consistent with the review that found that the provision of relevant, appropriate and timely information was a key factor in positive pregnancy experience [16], our study suggested that accurate descriptions, specific measures to resolve troubles, and reliable sources of information should be provided in the professional sector. Pregnant women may also need to share their values with professionals during treatment decisions and lifestyle transformations.

All participants in this study perceived the value of support from other experienced mothers, even if there were some disadvantages in encountering unexpected support. Although the lack of close relatives and friends with whom to talk and share similar problems was reported to predict the deterioration of mental health among pregnant women more than a decade ago [27], the present study revealed that not all pregnant women had such support. Previous interventions of lay-person-offered support only investigated the effect against adverse mental health outcomes (e.g., depression) among high-risk women [28]. However, our study suggested that such support may be theoretically effective for improving self-efficacy among pregnant women, regardless of specific risks. From the perspective of either vicarious experiences or verbal persuasion, further studies and interventions should be planned to connect anxious pregnant women with women who have had similar experiences to share real-life success or failure stories. In terms of informal relationships, it is worth noting that the present study found only negative statements

about support from partners. Participants wanted support from their husbands. However, if their needs were not met, they may have remembered this as a negative experience that conflicted with the perception that they deserved support from their husbands. A systematic review found that marriage or cohabiting with the baby's father had no effect against maternal depressive symptoms after controlling for potential confounders [29], and support from partners that did not meet needs could raise the risk [15]. Education or support programs during pregnancy should therefore include pregnant women's partners.

As most women of reproductive age have smart phones and can easily access the Internet anytime and anywhere in Japan, online information has become a big source of support among pregnant women. Our study supported the previous finding that pregnant women often gained reassurance from other people's experiences online [30]. However, many pregnant women also felt scared by the information they read online [31]. We encountered similar cases in this study where participants were psychologically damaged by online information. Our study also indicated that pregnant women may be confused by information that may not be accurate. To offer effective support to any pregnant women in Japan, it may be valuable for professionals to create or recommend reliable websites from which women can find correct and unbiased information or read about other pregnant women's experiences.

Several limitations of this study should be acknowledged. First, although this study was conducted as a preliminary study to identify meaningful variables in the following large observational study, our findings were based on one focus group and additional focus groups may find other themes or key concepts. Although one or two focus groups are said to be sufficient in exploratory studies [20], we should use these results as a starting point to know what support needs may be needed during pregnancy in today's local Japanese context and should evaluate them in the next quantitative phase in a larger sample. Comparing this study with focus groups in different samples would also bring meaningful insights in the future. Second, by using convenience sampling, we might have failed to capture important perspectives from hard-to-reach women [32]. Although we tried to recruit a heterogeneous group, we found that some of the items we considered were homogeneous (such as parity, educational background, and marital status). A possible explanation for this was that this study was conducted in the urban area. Even though some items appeared to be homogeneous, all participants had different backgrounds that could not be measured. The results of this study would be helpful for considering better support in the urban areas in Japan or for populations with similar backgrounds, such as newly pregnant women. Further studies are needed to explore the other targeted groups that have different characteristics from our study. Third, the information leaflet used in recruiting the participants might have encouraged more anxious women to participate and therefore this study was not representative of the needs among pregnant women in Japan. However, gathering parties who have some opinions can stimulate discussion in focus groups and yield rich data. Despite these limitations, to our knowledge, this was the first report in the last decade of a focus group exploring support needs for anxiety during pregnancy among women without specific risks in an economically prosperous country.

#### **5. Conclusions**

Using a focus group among pregnant women without specific risks, we found three types of support needs for anxiety during pregnancy: (1) seeking tailored professional support; (2) seeking continuous support within informal relationships; and (3) seeking others' success stories in the same situation. We proposed a model of needs of support for pregnancy-related anxiety backed up by social cognitive theory. Because an individual's needs for each type of support may be influenced by their preferences and circumstances, professionals should be aware of all three types of support when considering effective universal support for pregnant women's mental health. Further research and interventions should also consider the narratives and collaborate with pregnant women, because a unilateral supply of support may cause negative feelings in pregnant women.

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

**Funding:** This research was funded by The Health Care Science Institute Research Grant.

**Institutional Review Board Statement:** This study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Institutional Review Board of The University of Tokyo (approval code: 2020154NI, date of approval: 22 July 2020).

**Informed Consent Statement:** Informed consent was obtained from all participants involved in this study. Written informed consent was obtained from participants to publish this paper.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author. The data are not publicly available, for ethical reasons.

**Acknowledgments:** We thank all participants in the focus group as well as the women who could not participate even though they agreed. In addition, we thank Mie Yamada, Yuka Yamamoto, and Yo Takemoto as research collaborators, and the staff in each facility for their help in recruiting some of the participants in this study. We thank Audrey Holmes, MA, from Edanz (https://jp.edanz.com/ac (accessed on 2 February 2023)), for editing a draft of this manuscript.

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

#### **Appendix A**


**Table A1.** Coded anxieties and supporting quotes.


**Table A1.** *Cont.*

IVF, in vitro fertilization; COVID-19, coronavirus disease 2019. <sup>1</sup> Participants' narratives were translated into English for this publication.

#### **References**


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