Barriers to addressing mental health issues in childbearing women in Ghana

Abstract Aim The aim of the study is to investigate factors hampering the provision of mental health services by nurses and midwives to childbearing women to assist in the prioritization and distribution of limited mental health resources. Design This is a cross‐sectional self‐report study. Methods Data collected from 309 nurses and midwives were analysed using descriptive and inferential statistic, namely chi‐square and analysis of variance (ANOVA). Results The barriers reported by the participants include unavailability of mental health services (77%), lack of knowledge of mental health in women from different tribes (75.7%), lack of a clear mental healthcare pathway (75.1%), heavy workload (74.1%) and lack of knowledge of mental health issue (74.1%). These barriers are less likely to be reported by participants who are males, old and who have practiced for a long time. Conclusions Systematic effort to restructure the healthcare delivery system, including equipping healthcare professionals with requisite knowledge, skills and competencies in maternal mental health, is highly recommended.


| INTRODUC TI ON
Childbearing women, defined as women in the prenatal, perinatal and postnatal periods, are at heightened risk of mental health issues such as depression and anxiety. Globally, it has been estimated that one in five childbearing women is afflicted with mental health issues (Bayrampour, Hapsari, & Pavlovic, 2018). In low-and middle-income countries (LMICs), however, the prevalence rate is reportedly high, with estimates ranging from 15%-57% (Lund et al., 2014). Relating specifically to Ghana, a sub-Saharan African (SSA) country, emerging studies have estimated the prevalence of depression in childbearing women to range from 10%-38% (Bindt et al., 2013;Weobong et al., 2014), whereas anxiety was estimated at 33% (Bindt et al., 2013). Studies have linked poor mental health issues in childbearing women to several adverse outcomes, including prolong labour and peripartum and postpartum complications (Bindt et al., 2012;Sumankuuro, Crockett, & Wang, 2017;Weobong et al., 2014). Mental health issues in childbearing women are among the factors predisposing to maternal and child morbidity and mortality rates that are reportedly high in Ghana and other SSA countries (Bindt et al., 2012;Sumankuuro et al., 2017;Weobong et al., 2014).
In recognition of the gravity of mental health issues in childbearing women, professional organizations (National Institute for Health and Care Excellence, 2014) and researchers (Siu et al., 2016) have recommended routine screening to identify women at risk of mental health problems for early intervention. However, empirical evidence suggests that most childbearing women at risk of mental health problems are undetected, undiagnosed and untreated (Anding, Röhrle, Grieshop, Schücking, & Christiansen, 2015;Bayrampour et al., 2017;Fonseca, Gorayeb, & Canavarro, 2015). This observation provides the impetus for studies to explicate factors that can be improved to promote mental and emotional well-being in this vulnerable population. Consequently, the current study is aimed at investigating potential barriers to addressing mental health issues in childbearing women in Ghana.

| BACKG ROU N D
A recent systematic review has identified several studies, mostly from high-income countries, that have uncovered barriers to addressing mental health issues in childbearing women (Bayrampour et al., 2018). Notable among them are lack of knowledge in mental health, time constraint and lack of local guidelines/policies, unclear pathways of care and referral options (Bayrampour et al., 2018;Higgins et al., 2018). In SSA, however, little information exists with respect to the factors militating against the provision of mental health services to childbearing women (Bayrampour et al., 2018).
Although studies conducted in high-income countries have proffered some useful insights regarding mental health issues in childbearing women, their utilities in SSA may be limited owing largely to the sociocultural differences in healthcare systems with respect to the training of health personnel, as well as the organization and delivery of healthcare services. For instance, relative to high-income countries, SSA countries have few qualified (mental) health professionals (Dussault & Franceschini, 2006;Labonté, Packer, & Klassen, 2006) and insufficient budgetary allocation to meet the health needs of the populace (Raja, Wood, de Menil, & Mannarath, 2010). Additionally, because most of existing studies from high-income countries are qualitative in nature (Bayrampour et al., 2018), little is known about the importance practitioners attach to the various potential barriers (Higgins et al., 2018).
Quantifying potential barriers has the propensity to unearth the enormity of the problem, expand our understanding of the problem and more importantly help in the design of context-specific interventions (McEvoy et al., 2014). Indeed, although previous studies have advocated for continuous professional development programmes for health professionals to acquire the requisite mental health competencies (Bayrampour et al., 2018;Higgins et al., 2018), such programmes can be tailored appropriately by understanding health professionals' perceptions of potential barriers in relation to their local context (Hauck et al., 2015). Consequently, the overarching goal of the present study is to investigate the barriers influencing the delivery of mental health services to childbearing women in Ghana. The study addressed the following research question: What are the barriers and factors influencing the delivery of mental health services to childbearing women in Ghana?

| Design and data collection procedure
The study design and data collection procedure has been described elsewhere (Adjorlolo, Aziato, & Akorli, 2019). Briefly, a cross-sectional self-report survey was used to gather data from the participants across the various facilities after institutional permission was obtained from management. The participants were recruited from antenatal, delivery and postnatal units of their respective health institutions using both convenience sampling approach. The research team approached nurses and midwives on the various duty shifts (i.e. morning, afternoon and night) for their consent to participate in the study. They were informed that their participations in the study will help in understanding strategies to promote mental health issues in childbearing women. Because heavy workload at the facilities could interfere with the data collection process, effort was made to administer the questionnaires when the facilities were less busy, in this case mostly in the afternoon. They were given time to ask questions or raise issues they have with the research. Those who expressed willingness and consented to be part of the study were handed a pack of questionnaires containing several measures, including those measuring barrier and demographic variables. The questionnaires were handed over to the research team on completion. Of the 320 questionnaires administered, 309 were returned, representing a response rate of 97%.

| Setting and participants
Data for this study form part of a large-scale study investigating strategies to promote mental health in childbearing women in Ghana (Adjorlolo et al., 2019). Data were collected from 309 primary care nurses and midwives recruited from randomly selected health facilities that are highly patronized by childbearing women in Greater Accra, the capital of Ghana. These include Pentecost Hospital, Abokobi Health Center and University of Ghana hospital.
To be included in the study, the participants must be fully registered with the appropriate licensing authority (i.e. Nursing and Midwifery Council) and should have worked with childbearing women for at least 12 months, a period within which the participants might have encountered childbearing women with or in need of mental health services. A large proportion of the participants were females (N = 296, 96%). The participants designated as young (i.e. 21-30 years) were 140 (45.3%) whereas those designated as old (i.e. 31 years and above) were 169 (54.7%). A large proportion were females (N = 296, 96%), public health nurses (N = 138, 44.7%), general nurses (88, 28.5%) and midwives (83, 26.9%). The percentage of participants who reportedly practiced for 1-5 years and 6 years and above were 174 (56.3%) and 135 (43.7%), respectively. In terms of qualification, most participants had diploma (i.e. 3 years of nursing education; N = 171, 55.3%), followed by certificate (i.e. <3 years of nursing education; N = 116, 37.5%) and degree or advanced qualification (i.e. more 3 years of nursing education; N = 22, 7.1%).

| Measures and instrument
The participants' responses to measures of demographic variables namely gender, age, years of practice and professional category and barrier items were used in the current study. The barrier items, culled predominantly from the existing literature (Bayrampour et al., 2018;Higgins et al., 2018), were included in the study on the recommendation from health practitioners (e.g. midwives) and scholars in Ghana.
A total of 25 barrier items (shown in Table 1) were used in this study.
In keeping with the response format of existing studies (Higgins et al., 2018), the items were scored on a 4-point Likert-type scale, ranging from 1 (i.e. strongly disagree)-4 (strongly agree). The total score on the "barrier scale" ranges from 28-100, with high score indicating high perceived barriers. The internal consistency of the "barrier scale" was 0.88.

| Data analytic strategy
All data were analysed with SPSS Version 23 (IBM.corp). Cases with missing data on the barrier items represent less than 1% of the total cases, and because data were missing completely at random (Little's MCAR chi-square > 0.05), the expectation maximization algorithm was used to impute the missing data points. A two-tailed statistical significance was set at 0.05. To ascertain the usefulness of each of the 25 barrier items, the corrected item-total correlations, squared multiple correlations (SMCs) and Cronbach's alpha when items were deleted were examined. The results indicate that the items are measuring barrier to the provision of mental health services to childbearing women. Responses on the barrier items were subsequently dichotomized to reflect No (i.e. Strongly Disagree and Disagree) and Yes (i.e. Strongly Agree and Agree). This was subjected to descriptive statistics, namely frequency and percentages, to determine the proportion of "No" and "Yes" responses. Chi-square was used to investigate the bivariate relationship between the barrier items and demographic variables (i.e. age, gender, years of practice and education), with effect size estimated with Phi and Cramer's V coefficients.
Additionally, the 25 barrier items, rated on the four-point Likert response format, were subjected to principal component analysis (PCA) to determine their underlying latent component(s). Analysis of variance (ANOVA) was subsequently conducted to determine the effects of demographic variables, using the extracted component(s) as the dependent variable.

| Barriers to addressing mental health issues in childbearing women
The results of the analyses regarding the barriers to addressing mental health issues in childbearing women are summarized in Table 1.  Table 1 further shows that the rating of the barrier items appeared not to differ in any systematic way based on the category of profession (e.g. midwives versus PHNs) and qualification (e.g. diploma versus degree).

| Demographic correlates of barriers
For simplicity, the demographic variables that correlate significantly with the barriers to addressing mental health issues in childbearing women are presented here (the full result is available on request).  The result suggests that participants who have practiced for 6 years and above are less likely to rate the above as barriers.

| Principal component analysis and effects of demographic variables on barriers
As noted previously, PCA was conducted to investigate possible clustering of the barrier items. The results of the initial PCA showed that the clustering of the barrier items did not yield any interpret-

| D ISCUSS I ON
In Ghana and other low-and middle-income countries, there are no-  (Bayrampour et al., 2018;Higgins et al., 2018) although the challenges are more pervasive in Ghana owing to funding and logistics constraints, as discussed below.
Relating to the unavailability of mental health services, it is interesting to note that the provision of mental health services in Ghana is largely concentrated at the three public psychiatric hospitals in the country. Unfortunately, recent efforts to decentralize mental health service delivery in the country have been limited largely to regional and teaching hospitals. Importantly, because several health facilities located in urban and rural areas of the country are yet to boast of psychiatric units, it follows that mental health services are not readily available to childbearing women visiting these facilities.
Similarly, the limited coordination among healthcare institutions and units or departments in Ghana (Adjorlolo, 2015) suggests that even in health facilities with psychiatric units, there is the possibility that midwives and primary care nurses may be unaware of their existence. In addition is the lack of a structured protocol detailing the pathway of addressing mental health issues in childbearing women.
This development could put health professionals in a helpless situation whereby they are not able to and adequately use any existing mental health services effectively, including referral to psychiatric hospitals (Jomeen, Glover, Jones, Garg, & Marshall, 2013).
Additionally, heavy workload has been rated among the major Closely related to the above is the endorsement of lack of knowledge of mental illness in women from different tribes. A similar concern, referred to as cultural competency issue, has been noted by previous studies conducted in high-income countries (Almond & Lathlean, 2011;Borglin et al., 2015;Nithianandan et al., 2016). Relating specifically to Ghana, there are over 46 different tribes and cultures in Ghana, suggesting that the country is not necessarily culturally homogenous (Adjorlolo, Abdul-Nasiru, Chan, & Bambi, 2018). Because cultures tend to differ with respect to socialization, worldviews and beliefs, there is the possibility that the acquisition and expression of mental health issues may vary as a function of different cultures (Adjorlolo, Abdul-Nasiru, Chan, & Bentum, 2017;Yamagishi, Hashimoto, & Schug, 2008). More importantly, because the conceptualization of mental illness may differ from culture to culture, health professionals with dissimilar cultural backgrounds with their clients or patients may have difficulty for instance, initiating a conversation around mental and emotional problems (Almond & Lathlean, 2011;Nithianandan et al., 2016). They may also have doubt about whether a particular behaviour pattern should be granted attention as a potential mental health issue. Unfortunately, empirical studies are yet to explicate cultural influences on mental health issues in Ghana, including cultural differences in the prevalence and incidence rates of mental illness, disclosure and conceptualization of mental illness.

| Implications for practice and research
The It is also recommended that the government of Ghana and other stakeholders in the provision of healthcare in Ghana take the necessary steps to reduce the nurse-population ratio by recruiting more health professionals to reduce the workload of the few nurses and midwives. This would make it relatively easier for midwives and primary care nurses to devote some amount of their time and attention to addressing mental and emotional well-being of childbearing women.
The study has suggested several areas of empirical research in the quest to promote the provision of mental health services to childbearing women. This includes investigations of the possible impacts of cultural tendencies on the distribution, manifestation, assessment, diagnosis and treatment of mental health issues in childbearing women from different cultural background. Given the indication of negative attitudes towards mental illness among mental health professionals in Ghana (Adjorlolo et al., 2018), it is also recommended that future studies investigate the attitudes of midwives and primary care nurses towards mental health in childbearing women. Likewise, studies should investigate culturally appropriate strategies to promote positive attitudes towards mental health in child bearing. Lastly, it would be informative to determine the extent to which other factors not investigated in this study, such as geographic location and type of health facility (i.e. government versus mission), could potentially hinder the provision of mental health services to childbearing women.

| Limitations of the study
The relatively small sample size and the use of non-probability sampling limit the applications of the findings beyond the participants of the study. The use of only questionnaire does not provide mechanism for checking the accuracy or biasness of the responses gathered from the participants. The above notwithstanding, the findings reported here do not differ systematically from those reported by previous studies, as noted previously.

| CON CLUS ION
This study has found that there are barriers to midwives and primary care nurses addressing mental health concerns in childbearing women, including heavy workload, lack of clear pathways and lack of knowledge about mental health in childbearing women. Prioritizing limited resources to mitigating the impacts of the barriers reported in this study could help equip midwives and primary care nurses to render basic and low-cost mental health services to childbearing in need of these services.

ACK N OWLED G EM ENTS
We are grateful to those who have supported in the design of the questionnaire, including Dr Mary Ani-Amponsah.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no competing interests.

AUTH O R CO NTR I B UTI O N S
SA collected and analysed data. LA wrote the manuscript.

E TH I C A L A PPROVA L
The study received ethical clearance from the Institutional Review