Maternal mental health as a neglected dimension of primary health care systems responsiveness in low- and middle-income countries: a mixed methods cross-sectional study in Ghana Authors: Elizabeth Awini, Linda Lucy Yevoo, Leveana Gyimah, Anthony Danso-Appiah, Joseph Paul Hicks, Ana Manzano, Richard Afedi Nagai, Sumit Kane, Tolib Mirzoev, Irene Akua Agyepong Published: 2026 ABSTRACT BackgroundPregnancy and postpartum-related mental health problems present serious public health threat to the society; however, this is not prioritized in sub-Saharan Africa (SSA). We aimed to determine the prevalence and analyze perceptions and experiences of maternal mental health (MMH) distress and health systems responsiveness to mental well-being of pregnant and postpartum clients and their relatives, and frontline health workers in two districts of Southern Ghana.MethodsThe study used a cross-sectional mixed methods design and purposively selected study sites to provide insights from rural and urban settings. A structured survey involving antenatal (n = 827), postnatal (n = 618), and delivery (n = 58) maternal health clients was conducted in primary healthcare facilities using a modified World Health Organization health systems responsiveness questionnaire and the 21-item screening tool for depression, anxiety, and stress scale (DASS21). Eighteen focus group discussions (FGDs) were conducted, fourteen involving antenatal and postnatal clients with/without mental health conditions (n = 104), and four with frontline healthcare providers (n = 24). We also conducted six in-depth interviews, four (n = 4) involved postnatal clients with mental health conditions and two (n = 2) involved clients’ relatives. Quantitative data were analyzed using Stata 15 and employing descriptive analysis and generalized linear models. Thematic analysis was done on the qualitative data, with some further analysis of themes using the text sort function in Microsoft Excel.ResultsMild, moderate, and severe MMH distress were experienced in pregnancy and post-partum, but were not addressed as part of routine formal care. The DASS21 showed 19%, 31.6%, and 16.3% of symptoms of mild to severe depression, anxiety, and stress, respectively. Mental health and maternal health services are part of primary healthcare in Ghana, but work in silos. There is an insufficient number of community mental health officers and community psychiatric nurses in the districts, and their work is not synergized with maternal healthcare.ConclusionsMMH is an important dimension of primary care, and interventions to integrate mental health into maternal health are required. Screening using simple tools may enable prompt detection and management of MMH issues among at-risk groups, as part of making the Ghanaian health system more responsive.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12913-026-14704-x. Background Mental health in pregnant and postpartum women Pregnancy and the postpartum period are associated with both psychological and physical challenges that can affect maternal health and well-being. Maternal mental health (MMH) distress can also affect the unborn child [1] and is associated with negative birth outcomes and cognitive development of the infant [2, 3] including low birth weight [4], stunting [5] and childhood infections [6]. However, MMH remains a particularly neglected priority [7–10]. Evidence suggests that many women who experience mental health conditions during and after pregnancy do not get appropriate or needed care [11, 12]. Depression and anxiety are often experienced by pregnant and postpartum women [13–16], and in severe cases, depressed mothers may have symptoms of a psychotic disorder [17]. The problem cuts across countries of all income levels. Globally, MMH problems are estimated to affect about 10% of pregnant women and 13% of postpartum mothers [18–20]. A systematic review estimated the global prevalence of depression among postpartum women to be as high as 17.2% [21]. The prevalence of postpartum depression in high-income countries was estimated to be between 5% and 20% [22]. A review of evidence from low- and middle-income countries (LMICs) on pre- and postnatal depression reported a prevalence of 4.9% to 50% [23], whereas a systematic review involving pre- and postnatal women living in Africa reported the prevalence of depression during pregnancy at 11.3% and 18.3% after delivery [24]. Another review from sub-Saharan Africa reported the prevalence of depression among post-partum women at 18.6% [25], with a range from 7% to 50.3%. In Ghana, where the current study was conducted, recent research shows a similar picture. Although the physical health of pregnant and postpartum mothers is emphasized by Maternal and Child Health (MCH) policies and programs, their mental health is usually not considered [13]. The prevalence of perinatal (mostly third-trimester) depression and anxiety has been found to be between 26.5 and 37.7% and 11.4–14.2%, respectively [4, 26–29]. About 2–12 weeks postpartum, the prevalence of depression has been found to be between 3.5% and 27.5% [28, 30, 31]. And the prevalence of postpartum stress at 3 months and 12 months postpartum has been found to be between 33.1% and 24.4%, respectively. MMH is not routinely assessed by providers or requested to be assessed by women, and screening for MMH conditions is not part of routine maternal healthcare [19]. Primary healthcare in Ghana and maternal mental health Primary Health Care (PHC) delivery system in Ghana is provided by both public and private facilities, and the public healthcare sector provides about 67% of all healthcare services in the country [32]. These facilities serve a large part of the population of both insured and uninsured individuals, due to relative geographical and financial accessibility. PHC services cover health promotion, disease prevention, first-line clinical care services, and primary referrals (up to the district hospital level) for communicable and non-communicable diseases, which include mental health and MCH, with some variations across facility types. The MCH activities have been the priority of donors and the government over the last few decades, with a focus on reducing mortality among newborns, under-fives, and mothers. However, MMH has been less prioritized [33, 34]. Barriers to addressing maternal mental health needs in Ghana include: limited availability of mental health services resulting from inequitable distribution of mental healthcare professionals [35], insufficient awareness among healthcare providers of mental health issues in women from different ethnic and cultural groups, unclear mental healthcare pathways within PHC settings, heavy workloads, and inadequate knowledge of mental health issues among primary care nurses and midwives [36]. Health systems responsiveness is “…when institutions… are cognisant and respond appropriately to the universally legitimate expectations of individuals. safeguarding of rights of patients to adequate… care.” [37]. It is socially constructed, reflecting experiences of service users against the non-medical or social aspects of healthcare [38], and along seven core elements organized into two categories: respect for persons (dignity, autonomy, confidentiality) and client orientation (prompt attention, quality of basic amenities, access to social support networks, choice of provider) [39]. The literature on health system responsiveness to MMH needs in Ghana is severely deficient. While access to social support networks and prompt attention receive moderate documentation, data specific to MMH services are not routinely collected on critical elements, including confidentiality, autonomy, quality of basic amenities, and patient experience and satisfaction data specific to maternal mental health services, which are not routinely collected, limiting accountability for responsiveness [36, 40], and they remain under-researched. The RESPONSE project was a collaborative research entitled “Improving health systems responsiveness to neglected health needs of vulnerable groups in Ghana and Vietnam,” which sought to contribute to improving health systems responsiveness in low- and middle-income countries through case studies of addressing the health needs of vulnerable groups in Ghana and Vietnam [41]. The study we report here was nested in the wider RESPONSE project and aimed to understand the extent of MMH problems among pregnant and postpartum clients using PHC facilities in two mixed rural-urban districts in Southern Ghana. Additionally, it explored maternal health clients’, relatives’, and frontline providers’ perceptions and experiences of responsiveness of primary care services to their needs and expectations. The broader aim was to bridge existing knowledge gaps and provide insights to guide the prioritization of MMH programs and interventions. Methods Study design The study used a cross-sectional design and adopted a convergent parallel mixed methods approach [42]. We used a cross-sectional quantitative survey to collect data on the views of women using ANC, PNC, and delivery services about their perceptions of the responsiveness of the PHC facility in relation to maternal mental healthcare during the provision of these services, and on these women’s mental health status. We used in-depth interviews and FGDs with antenatal care (ANC) and postnatal care (PNC) maternal health clients and frontline healthcare providers. A parallel mixed-methods design was used rather than a sequential design because we had limited time, but wanted to simultaneously analyze and get comprehensive views on these issues. Study setting The study was conducted in Shai-Osudoku (SO) and Ningo-Prampram (NP) districts, two of the 29 districts in the Greater Accra region of Ghana, Greater Accra region, where the larger RESPONSE study was being conducted. These are two adjacent districts, formerly the Dangme West district, which are peri-urban in nature. Together, they cover a land area of 1528.9km2. The total population of the two districts was estimated at 310,283, of which NP had 204,673 and SO had 105,610. The female-to-male ratio was 50% each according to the 2021 Population and Housing Census [43]. The two districts have a total of 81 health facilities, comprising 56 public, 24 private, and one non-profit facility. Only 42 of these health facilities, 23 in Ningo-Prampram and 19 in Shai-Osudoku, provide ANC, PNC or routine delivery services. For this study, antenatal clients refer to pregnant women who attend the health facility for ANC services, postpartum mothers/postnatal clients refer to women who had given birth and are coming back to the facility for PNC services and delivery clients refer to women who have just given birth at the facility and have been discharged or are waiting in the lying-in for discharge on the day of the survey. This group was separated from the postnatal clients to get their perceptions of responsiveness regarding delivery services. Inclusion and exclusion criteria Inclusion criteria Pregnant women who attended the health facilities for ANC services.Postpartum mothers who attended the facilities for PNC services.Women who just delivered and were discharged or waiting in the lying in to be discharged. Pregnant women who attended the health facilities for ANC services. Pregnant women who attended the health facilities for ANC services. Postpartum mothers who attended the facilities for PNC services. Postpartum mothers who attended the facilities for PNC services. Women who just delivered and were discharged or waiting in the lying in to be discharged. Women who just delivered and were discharged or waiting in the lying in to be discharged. Pregnant and postpartum women who have lived in the two study districts for a minimum of 3 months (qualitative interviews). Exclusion criteria Women who were sick and needed to see a physician or had just gone through Cesarean section (CS) were excluded. Maternal clients clinically assessed as unable to consent or participate in the qualitative study. Women who were sick and needed to see a physician or had just gone through Cesarean section (CS) were excluded. Women who were sick and needed to see a physician or had just gone through Cesarean section (CS) were excluded. Maternal clients clinically assessed as unable to consent or participate in the qualitative study. Maternal clients clinically assessed as unable to consent or participate in the qualitative study. Sample size and sampling procedure The sample size for the quantitative survey was determined using Cochran’s formula [44]. To maximize the generalizability of the sample size calculation, we assumed an outcome proportion of 50% and a desired 95% confidence interval with a margin of error (α) of 5% in the calculation. The sample sizes were determined separately for ANC and PNC clients, as they were considered distinct populations. However, because the delivery numbers were small, we decided not to estimate a sample size and instead interviewed any client who had delivered and was being discharged at the time of data collection in the facility. The sample size for ANC was determined using the two-week service load as the population, while the PNC sample size was determined using the PNC service load during the same period. The estimated sample size was 706 for ANC clients and 362 for PNC clients, after inflating the sample size by a design effect of 1.5 to account for clustering within facilities [45] and adding a 10% inflation factor to adjust for non-response. The design effect was selected based on the national level Ghana Demographic and Health Survey (GDHS) and other studies [46–49]. The sample size for each facility was then determined based on their size. When this was done, some facilities had fewer than 30 clients. However, a minimum of 30 clients will be required to allow the analysis to be conducted for each facility. This was based on the fact that most statistical texts consider a sample size of 30 as the borderline between small and large samples [50], and the aim is not to do any comparisons at the facility level but to describe or analyze simple proportions. Therefore, the sample size for health facilities with fewer than 30 clients was increased to at least 30 for both ANC and PNC. In the end, the sample sizes were 827 for ANC and 618 for PNC. Purposive sampling was used to select 16 health facilities, eight from each district, out of the 42 health facilities that provide MCH services for the survey. The selection was based on a service utilization rate of 10 or more maternal health clients per week. This approach was taken because some facilities were severely under-utilized, and the cost of data collection outweighed the benefits of collecting data from those facilities. It would have required many more resources to get the required sample size from each facility. Data collection Interviewers administered a modification of the WHO Health systems responsiveness questionnaire purposively for this study (see file 1) to elicit perceptions of responsiveness on the seven WHO dimensions of responsiveness (dignity, autonomy, confidentiality, prompt attention, access to networks, quality of amenities, and choice of providers, with trust as additional dimension [38]. This was followed by the administration of the 21-item depression, anxiety, stress scale (DASS21) screening tool to the same women. The responsiveness questionnaire used a Likert scale ranging from “Strongly disagree” to “Strongly agree” or “Not confident” to “Very confident”, which were used to prime the individuals to think about various aspects of each domain so they could rate it more consistently and on informed basis. Each domain was then given an overall rating from 0 to 10. We used the DASS21 [51], which has been used in several studies in LMICs [52–54] and in other settings in Ghana [55–57] to screen for symptoms of depression, anxiety, and stress in all the women who participated in the exit surveys. The DASS21 questionnaire was validated in Ghana in a previous study among health workers and a community sample [58]. The DASS21 has 21 questions equally divided among three scales: depression, anxiety, and stress. Each item allows for a Likert scale response ranging from “Did not apply to me at all” to “Applied to me very much and scored as 0–3”, thus resulting in a total score of 0–21. The decision to use DASS21 is because it can be used for examining the distribution of depression, anxiety, and stress symptoms. The facility ANC record books of the participants were also reviewed to confirm information on ANC attendance and the number of pregnancies reported by these participants. We also collected data on background characteristics of each health facility (type of facility, ownership of facility) and clients (maternal age, gravida, employment, education, language spoken, and religion). The Open Data Kit (ODK) application was used to capture all exit survey data on computer tablets. Study quantitative variables Outcome variables: The outcome variables were probable depression, anxiety, and stress which were measured using the DASS21. Subscales of depression, anxiety, and stress had seven questions each. In addition, responsiveness, which is measured using the modification of the WHO health systems responsiveness questionnaire with dimensions of dignity, autonomy, confidentiality, prompt attention, access to networks, quality of amenities, choice of providers, and trust. Covariates: The covariates are age, sex, occupation, education, facility type, education gravida. For the association between responsiveness and mental health distress, the covariate is mental health distress measured using DASS21. Data quality and assurance To maximize data quality, we trained all data collectors via a six-day training and pretest, where they were taken through an overview of the study and given refresher training on data collection techniques and conducting fieldwork. The tools were extensively discussed and translated into the local languages (Ga-Dangme, Twi, Ewe). The meaning of each question was explained and asked in the local dialect. The data collectors then took turns practicing administering the questions to their peers. Each of the role plays were done repeatedly until we were comfortable that the tools were well understood, and the questions asked correctly. A full day was assigned for pretesting and discussion of feedback. The pretesting was conducted in two health facilities which were not part of the study facilities. This was done on the 5th day of the training. The data collectors administered the questionnaires to MCH clients when they had received service and were leaving the facility. We built automated checks into the ODK data collection application to minimize errors such as inconsistencies and invalid entries. The supervisors checked and edited the completed questionnaires to ensure that errors were rectified before the data were uploaded. The data manager also checked the data everyday as they uploaded after the day’s work for errors. The data was rechecked during data analysis. Full access to the data was restricted to only the data manager and the statistician who coordinated the survey. Mixed-methods analysis A parallel convergent model for mixed-methods analysis was used, with quantitative and qualitative data being collected simultaneously and analyzed separately before being triangulated during the interpretation phase [42]. After the separate analyses, findings were compared to establish relationships to each other, identifying areas where the qualitative and quantitative results converged, diverged, or complemented each other. Statistical analysis The DASS21 is a brief version of the DASS, which has 42 items. Total scores for each of the sub-scales were thus multiplied by 2 to get the scores for the 42 items [51]. To be able to estimate the prevalence levels of probable maternal depression, anxiety, and stress, each sub-scale was categorized into normal, mild, moderate, severe, and extremely severe using the categorizations proposed in the DASS manual [51]. The categories for severe and extremely severe were merged as severe for the analysis, as the numbers in these categories, especially the extremely severe, were exceedingly small. We performed the analysis in two stages. The first stage was descriptive analysis to estimate the prevalence of depression, anxiety, and stress symptoms. These are used as a measure of MMH problems. Background information of the health facilities (type of facility, ownership of facility) and participants (educational status, employment status, maternal age, gravida) was summarized via appropriate descriptive statistics. In the next stage, we examined the descriptive association between clients’ characteristics, including participants’ educational status, employment status, gravida, and MMH distress (depression, anxiety, and stress) scores by estimating differences in prevalence and means within the category of each factor. In the first part of this analysis, we generated a binary variable for each of the sub-scales. For depression, a score of 10 and above indicates the presence of symptoms, a score of 8 and above indicates symptoms of anxiety while a score of 15 and above indicates symptoms of stress. Presence of symptoms was coded as 1 while no symptoms was coded as 0. To estimate the prevalence and difference in prevalence between the covariate categories, generalized linear models (GLM) with logit links and binomial errors were fitted for each of the DASS21 sub-scales (depression, anxiety, stress) with each covariate of interest including in a separate model given the descriptive focus of the analysis [59] The results of this analysis are presented in Tables 4, 5 and 6 in the results section. Overall responsiveness was estimated by calculating the mean of all the ratings of all the responsiveness domains considered. Responsiveness for any of the domains was estimated by calculating the mean rating of that domain. The overall responsiveness was estimated for those with symptoms of MMH distress (depression, anxiety, stress) and those without. It was also estimated separately for the population groups. We then fitted generalized linear models with identity links gaussian errors to estimate the mean scores of the overall DASS21 and the difference in mean scores between the categories of each of the covariates of interest, along with their 95% confidence intervals. Again, as the goal was the description of association, each model only included one covariate of interest at a time. All quantitative analyses were conducted using complete case analysis. Analysis of variables with missing data was limited to records that had complete data for those specific variables. All analyses adjusted for the clustered sampling design by using clustered standard errors. Qualitative methods The qualitative study aimed to triangulate findings from the quantitative survey, as well as describe and analyze perceptions and experiences of mental health and health systems responsiveness to mental health among maternal health clients and their relatives, and frontline health workers in the two purposively sampled districts. A total of 161 individuals participated in the qualitative study, which involved six in-depth interviews (IDIs) and eighteen focus group discussions (FGDs) with maternal health clients and healthcare professionals. Two separate group interview guides (for frontline health workers and pregnant women with or without mental health conditions) were developed for this study and used as guides during the interviews (See files 2 and 3). Six IDIs were held with maternal health clients and relatives who agreed to participate in the study. Four of the IDIs were conducted with postnatal clients with mental health conditions, and two of the IDIs were conducted with relatives of the maternal health clients who had a severe mental health condition. Additionally, ten FGDs were held with pregnant and postnatal women without mental health conditions with varied socio-economic status (n = 102). Two FGDs were initially planned for each category of maternal health clients (pregnant women and postnatal women without mental health issues) from the nine purposively sampled communities across the two districts. However, due to the wide geographical distribution of the maternal health clients and the fact that some of the women lived in hard-to-reach communities (especially communities such as Duffour and Old Ningo communities), the research team was able to conduct FGDs with women who were able to make it to the scheduled discussion venues. One FGD was held with pregnant and postnatal women with mental disorders (n = 10), three FGDs were held with marginalised and vulnerable groups such as adolescent mothers (n = 19), and four FGDs were held with healthcare providers who provide maternal or mental health services at the PHC level (n = 24). The FGDs comprised an average of 6–7 participants and were held at private venues within a district hospital. Most discussions with the clients were in the local languages, specifically Twi, and occasionally Ga-Dangme or Ewe, and lasted 45–60 min. Participants were compensated for travel expenses (up to GHS 50, or US$5 equivalent) and were also served refreshments. Interviews and FGD were recorded following informed consent and transcribed verbatim, while also translated into English. Thematic analysis [60] was used to identify emerging themes and subthemes by reading each transcript line-by-line. Thematic analysis was manually done but with some further analysis using the text sort function in Microsoft Excel. Ethics and informed consent The wider RESPONSE study within which this study is nested received approvals from the Ghana Health Service Ethics Review Committee (ref: GHS-ERC:012/03/20), London School of Hygiene and Tropical Medicine (ref 22981), and the University of Leeds (ref MREC 19–051). Written informed consent was obtained from all participants before any data collection. and anonymity and confidentiality of the participants were maintained throughout. Results Participants characteristics Fig. 1Flowchart for quantitative study participants Flowchart for quantitative study participants Exit survey participants As presented in the flowchart (Fig. 1), a total of 1,503 women participated in the exit surveys. The final numbers by health facility type and ownership are presented in Table 1. The majority of the clients were interviewed at public PHC facilities. The background characteristics of the women at the exit survey are summarized in Table 2. Most of them were pregnant women, while the rest were postpartum mothers, including those who had just delivered and been discharged but had not yet left the facility. Table 1Respondents by health facility characteristicsIndicatorDistrictShai-Osudoku (N = 883)Ningo-Prampram (N = 620)Total Service point N % n % n % Antenatal48054.434756.082755.0 Delivery273.1315.0583.9 Postnatal37642.624239.061841.1 Ownership Public75885.858494.21,34289.3 CHAG8810.000.0885.9 Private374.2365.81734.9 Facility type Hospital52459.3365.856037.3 Polyclinic00.012920.81298.6 Health Centre20122.810817.430920.6 CHPS Compound15817.934756.050533.6 Table 1Respondents by health facility characteristicsIndicatorDistrictShai-Osudoku (N = 883)Ningo-Prampram (N = 620)Total Service point N % n % n % Antenatal48054.434756.082755.0 Delivery273.1315.0583.9 Postnatal37642.624239.061841.1 Ownership Public75885.858494.21,34289.3 CHAG8810.000.0885.9 Private374.2365.81734.9 Facility type Hospital52459.3365.856037.3 Polyclinic00.012920.81298.6 Health Centre20122.810817.430920.6 CHPS Compound15817.934756.050533.6 Respondents by health facility characteristics Table 2Characteristics of exit survey respondentsIndicator N %Indicator N % Total number of respondents 1503100 Service point *Cost of healthcare provided by Antenatal82755.0Husband37889.2Delivery583.9Parents204.7Postnatal61841.1Other relatives266.1 Age group of participants Religion 14–19 years1288.5Christian1,36390.720–24 years34422.9Moslem1379.125–29 years39226.1Traditional30.230–34 years36724.4 Ethnicity 35–49 years27218.1Ga-Dangme75350.1 Highest education level attained Ewe35123.4Never been to school1308.7Akan21114.0Primary school26217.4Other18812.5Junior secondary school70246.7Senior secondary /Technical/Vocational32521.6 Languages spoken Tertiary845.6English91861.1 Occupation Ga-Dangme92761.8Does any work that brings in income1,07971.8Ewe50033.3 Primary occupation ( N = 1079) Twi1,24783.0Farmer615.7Other23115.4Fishmonger100.9 Gravida (1186) Trader57052.8One pregnancy33422.2Artisan (e.g. Hairdresser, Seamstress)35833.2Two pregnancies3 1721.1Civil Servant/Gov’t worker585.4Three pregnancies30720.4Student10.1Four pregnancies22915.3Other (specify)212.0Five or more pregnancies31621.0 *Cost of living provided by ( N = 424) Obstetric history Husband34581.4Pregnancy ended in abortion52334.8Parents6114.4Pregnancy ended in stillbirths624.1Other relatives184.2First ANC visit (N = 827)18922.9*For participant who do not do any work that brings in income Table 2Characteristics of exit survey respondentsIndicator N %Indicator N % Total number of respondents 1503100 Service point *Cost of healthcare provided by Antenatal82755.0Husband37889.2Delivery583.9Parents204.7Postnatal61841.1Other relatives266.1 Age group of participants Religion 14–19 years1288.5Christian1,36390.720–24 years34422.9Moslem1379.125–29 years39226.1Traditional30.230–34 years36724.4 Ethnicity 35–49 years27218.1Ga-Dangme75350.1 Highest education level attained Ewe35123.4Never been to school1308.7Akan21114.0Primary school26217.4Other18812.5Junior secondary school70246.7Senior secondary /Technical/Vocational32521.6 Languages spoken Tertiary845.6English91861.1 Occupation Ga-Dangme92761.8Does any work that brings in income1,07971.8Ewe50033.3 Primary occupation ( N = 1079) Twi1,24783.0Farmer615.7Other23115.4Fishmonger100.9 Gravida (1186) Trader57052.8One pregnancy33422.2Artisan (e.g. Hairdresser, Seamstress)35833.2Two pregnancies3 1721.1Civil Servant/Gov’t worker585.4Three pregnancies30720.4Student10.1Four pregnancies22915.3Other (specify)212.0Five or more pregnancies31621.0 *Cost of living provided by ( N = 424) Obstetric history Husband34581.4Pregnancy ended in abortion52334.8Parents6114.4Pregnancy ended in stillbirths624.1Other relatives184.2First ANC visit (N = 827)18922.9*For participant who do not do any work that brings in income Characteristics of exit survey respondents *For participant who do not do any work that brings in income Background characteristics of qualitative study participants A total of 137 maternal health clients (those with or without mental health issues) participated in the qualitative study across the nine sampled communities in the two districts (Fig. 2). Fig. 2Participants across the sampled communities Participants across the sampled communities The maternal health clients who participated in the qualitative study were women aged 15–49 years, with the largest number in the 25–29 year group (Fig. 3). The majority of them had a basic level of education (95% had primary to senior high school levels of education), and 5% had no formal education (see Fig. 4). Also, most of the maternal health clients were employed in the informal sector (53.7%) as petty traders (15%), and artisans (32.5%) (e.g., seamstress and hairdresser etc.), and about 41% of them were unemployed (See Fig. 5). Fig. 3Age range of participants Age range of participants Fig. 4Participants educational level Participants educational level Fig. 5Respondents’ occupation Respondents’ occupation Most healthcare professionals who participated in the qualitative study were females. Midwives formed the majority of the PHC providers, followed by community health nurses, with personnel from the records department forming the lowest number (see Table 3). In terms of experience, many of the healthcare professionals had practiced for one to ten years. There was, however, one outlier (a midwife), who had practiced for 34 years. Two of them had a first degree, while the rest had either a diploma or certificate levels of education. Table 3Respondent, profession, years of practice and educational levels districtProfessional trainingShai-OsudokuNingo-PrampramMidwife47CHN44Psych nurse22Clinical nurse01 Total 10 14 Years of experience < 1 year121-10years91011 years and above01 Total 10 14 Highest level of education Certificate12Diploma910Degree02 Total 10 14 Table 3Respondent, profession, years of practice and educational levels districtProfessional trainingShai-OsudokuNingo-PrampramMidwife47CHN44Psych nurse22Clinical nurse01 Total 10 14 Years of experience < 1 year121-10years91011 years and above01 Total 10 14 Highest level of education Certificate12Diploma910Degree02 Total 10 14 Respondent, profession, years of practice and educational levels district Reliability test results for DASS21 Reliability test was done to assess internal consistency by estimating Cronbach’s alpha and McDonald’s omega/Raykov’s rho for the DASS21 tool as a whole and the DASS21 sub-scales. An alpha value of 0.7 and above is considered acceptable [61]. Though Cronbach’s alpha has been accepted and used widely for assessing internal consistencies, it has very restrictive assumptions that are seldom met in practice so the omega, which is similar to the alpha, is being preferred [62, 63]. Both alpha and omega values have been estimated for the overall DASS21 and the DASS21 subscales. The overall Cronbach’s alpha and McDonald’s omega/Raykov’s rho for the DASS21 were about 0.87. The DASS21 subscales also had good internal consistencies of between 0.699 and 0.80. The Cronbach’s alphas for the DASS21 subscale were depression subscale = 0.80, anxiety subscale = 0.699, and stress subscale = 0.712. In the following sections, we present our findings in relation to our stated objectives. Firstly, we present the assessment of levels of mild, moderate and severe symptoms of probable depression, anxiety and stress among clients using PHC clinics for ANC, PNC and delivery care and perception of the extent of the problem by pregnant and postpartum women. This is followed by our analysis of background characteristics of the respondents that were associated with depression, anxiety and stress, and finally our analysis of the perceptions of PHC responsiveness between clients with and those without symptoms of depression, anxiety and stress. The estimates of each of the probable outcomes of depression, anxiety, and stress are not exclusive of each other. Levels, perceptions, and experiences of common MMH distress among antenatal and postnatal clients The distribution of MMH distress (depression, anxiety, stress) symptoms among the exit survey participants is shown in Fig. 6. Anxiety was the most prevalent symptom, with 19.5% of respondents scoring moderate to severe anxiety. This was followed by moderate to severe depression (11.2%) and stress (9.5%). That moderate and severe anxiety had the highest scores in the DASS21 assessment correlates with findings from the qualitative interviews in which antenatal and postnatal women described experiencing varying levels of mental health distress symptoms ranging from mild to severe symptoms during pregnancy and childbirth. The women mentioned that ‘thinking a lot’ due to concerns about possible negative birth outcomes caused them excessive anxiety. Another source of ‘thinking a lot’ (excessive anxiety) stemmed from distressing news received from maternal care providers during antenatal visits, such as concerns about the physical well-being of their unborn baby, or the potential need for assisted delivery, particularly a caesarean section. Anxiety in pregnancy is also linked to the perception that pregnancy and childbirth involve significant vulnerabilities and can be as potentially dangerous due to medical and supernatural factors. Concerns include beliefs in being bewitched out of envy, fears of medical complications and mortality, which are thought to require spiritual intervention. A participant said:In our community, the moment you become pregnant witches, evil spirits here or somewhere, even get to know you are pregnant before you the woman get to know. Then they start planning evil things against you and the baby, such as the baby or you the mother should die, due to mere envy and wickedness…. All these things which happens in the spiritual realm make most pregnant women worry, think a lot. That is why most pregnant women go to church for spiritual intervention during this time of vulnerability Postnatal mother without mental disorder, Doryumu, FGD. In our community, the moment you become pregnant witches, evil spirits here or somewhere, even get to know you are pregnant before you the woman get to know. Then they start planning evil things against you and the baby, such as the baby or you the mother should die, due to mere envy and wickedness…. All these things which happens in the spiritual realm make most pregnant women worry, think a lot. That is why most pregnant women go to church for spiritual intervention during this time of vulnerability Postnatal mother without mental disorder, Doryumu, FGD. Depression, anxiety and stress appear to have significant comorbidity, occurring together in the same client (Fig. 7). About 12% of respondents had symptoms of depression and anxiety together while 8.8% had symptoms of depression, anxiety and stress together. This aligned with the qualitative data where some women who described symptoms of anxiety also described symptoms of depression. They mentioned that occurrence of traumatic events such as the death of a loved one, partner neglect, rejection of the pregnancy or unreasonable demands such as wanting a male child, triggered depression and anxiety, sometimes severe to the point of suicidal ideation or attempted suicide.I was in a romantic relationship with a man, and we were going to have a second baby together. … He denied responsibility of it because when I realized I was pregnant he gave me a drug to abort the pregnancy. He witnessed that I had started bleeding after taking the medication. For him the feotus had been aborted. With time, I realized that I was still pregnant, but he said he would not accept responsibility of the pregnancy because he felt it belonged to another man. His stance became an issue between his family and mine. I was always thinking, “what would I do?” because I am not working. How can I take care of the baby and myself? I nearly killed myself because of this. I even drunk poison but luckily nothing happened to me. (Pregnant woman without mental disorder, Ayikuma, FGD). I was in a romantic relationship with a man, and we were going to have a second baby together. … He denied responsibility of it because when I realized I was pregnant he gave me a drug to abort the pregnancy. He witnessed that I had started bleeding after taking the medication. For him the feotus had been aborted. With time, I realized that I was still pregnant, but he said he would not accept responsibility of the pregnancy because he felt it belonged to another man. His stance became an issue between his family and mine. I was always thinking, “what would I do?” because I am not working. How can I take care of the baby and myself? I nearly killed myself because of this. I even drunk poison but luckily nothing happened to me. (Pregnant woman without mental disorder, Ayikuma, FGD). Fig. 6Prevalence of depression, anxiety, stress symptoms among survey participants (pregnant and postpartum women) Prevalence of depression, anxiety, stress symptoms among survey participants (pregnant and postpartum women) Fig. 7Prevalence comorbidity of depression, anxiety and stress symptoms Prevalence comorbidity of depression, anxiety and stress symptoms Clients’ characteristics associated with depression, anxiety and stress symptoms The relationships between clients’ characteristics and depression, anxiety and stress among the participants are presented in Tables 4, 5 and 6. Table 4Prevalence and difference in prevalence of depression symptoms level and client characteristicsIndicatorDepression symptoms, n; % (95% CI)Difference in prevalence percentage points (95% CI) Works to bring in any kind of income Yes162; 16.0% (8.2, 23.9)RefNo105; 26.4% (16.2, 36.6%)10.3PP (5.2, 15.5) Gravida: Number of pregnancies One pregnancy22.3% (11.6, 32.9)RefTwo pregnancies18.3% (7.8, 28.8)-3.9PP (-9.5, 1.7)Three pregnancies17.9% (10.4, 25.5)-4.3PP (-11, 2.4)Four pregnancies17.5% (6.9, 28.2)-4.7PP (-11.4, 2)Five or more pregnancies18.2% (9.7, 26.6)-4.1PP (-9.4, 1.3) Age at last birthday 14-19yrs42; 33.9% (22.9, 44.8)Ref20-24yrs70; 22.2% (10.8, 33.5)-11.7PP (-24.2, 0.8)25-29yrs66; 18.1% (10.6, 25.6)-15.8PP (-27.1, -4.5)30-34yrs52; 15% (5.8, 24.3)-18.8PP (-29, -8.7)35-49yrs37; 14.4% (5.3, 23.5)-19.5PP (-27.7, -11.3) Highest education level attained Never been to school32; 25.2% (10.9, 39.5)RefPrimary school65; 26.3% (15.5, 37.1)1.1PP (-7.3, 9.5)Junior secondary school116; 17.7% (10.7, 24.6)-7.5PP (-16.6, 1.5)Senior secondary /Technical/Vocational45; 15% (4.6, 25.4)-10.2PP (-19.4, -1)Tertiary9; 11.7% (4.4, 19)-13.5PP (-25.6, -1.4) Has any pregnancy ended in abortion? Yes88; 18.1% (9.6, 26.6)RefNo179; 19.4% (10.3, 28.6)1.4PP (-1.5, 4.3) Has any pregnancy ended in a still birth? Yes11; 19% (5.9, 32)RefNo256; 19% (10.1, 27.9)0PP (-10.7, 10.7) Service point used by client ANC163; 21.4% (11.9, 30.9)RefDelivery14; 25.5% (8.4, 42.5)4PP (-8.3, 16.3)PNC90; 15.2% (7, 23.4)-6.3PP (-11, -1.6) District Shai-Osudoku115; 13.7% (4.8, 22.6)RefNingo-Prampram152; 26.8% (18.9, 34.7)13.1PP (1.2, 25)PP = percentage points. Data are n; % (95% CI) or PP (95% CI) Table 4Prevalence and difference in prevalence of depression symptoms level and client characteristicsIndicatorDepression symptoms, n; % (95% CI)Difference in prevalence percentage points (95% CI) Works to bring in any kind of income Yes162; 16.0% (8.2, 23.9)RefNo105; 26.4% (16.2, 36.6%)10.3PP (5.2, 15.5) Gravida: Number of pregnancies One pregnancy22.3% (11.6, 32.9)RefTwo pregnancies18.3% (7.8, 28.8)-3.9PP (-9.5, 1.7)Three pregnancies17.9% (10.4, 25.5)-4.3PP (-11, 2.4)Four pregnancies17.5% (6.9, 28.2)-4.7PP (-11.4, 2)Five or more pregnancies18.2% (9.7, 26.6)-4.1PP (-9.4, 1.3) Age at last birthday 14-19yrs42; 33.9% (22.9, 44.8)Ref20-24yrs70; 22.2% (10.8, 33.5)-11.7PP (-24.2, 0.8)25-29yrs66; 18.1% (10.6, 25.6)-15.8PP (-27.1, -4.5)30-34yrs52; 15% (5.8, 24.3)-18.8PP (-29, -8.7)35-49yrs37; 14.4% (5.3, 23.5)-19.5PP (-27.7, -11.3) Highest education level attained Never been to school32; 25.2% (10.9, 39.5)RefPrimary school65; 26.3% (15.5, 37.1)1.1PP (-7.3, 9.5)Junior secondary school116; 17.7% (10.7, 24.6)-7.5PP (-16.6, 1.5)Senior secondary /Technical/Vocational45; 15% (4.6, 25.4)-10.2PP (-19.4, -1)Tertiary9; 11.7% (4.4, 19)-13.5PP (-25.6, -1.4) Has any pregnancy ended in abortion? Yes88; 18.1% (9.6, 26.6)RefNo179; 19.4% (10.3, 28.6)1.4PP (-1.5, 4.3) Has any pregnancy ended in a still birth? Yes11; 19% (5.9, 32)RefNo256; 19% (10.1, 27.9)0PP (-10.7, 10.7) Service point used by client ANC163; 21.4% (11.9, 30.9)RefDelivery14; 25.5% (8.4, 42.5)4PP (-8.3, 16.3)PNC90; 15.2% (7, 23.4)-6.3PP (-11, -1.6) District Shai-Osudoku115; 13.7% (4.8, 22.6)RefNingo-Prampram152; 26.8% (18.9, 34.7)13.1PP (1.2, 25)PP = percentage points. Data are n; % (95% CI) or PP (95% CI) Prevalence and difference in prevalence of depression symptoms level and client characteristics PP = percentage points. Data are n; % (95% CI) or PP (95% CI) Table 5Prevalence and difference in prevalence of anxiety symptoms level and client characteristicsIndicatorAnxiety symptoms n; % (95% CI)Difference in prevalence percentage points (95% CI) Works to bring in any kind of income Yes298; 28.7% (24.3, 33.1)RefNo159; 39% (29.6, 48.4)10.3PP (3.3, 17.3) Gravida: Number of pregnancies One pregnancy106; 33% (23.1, 42.9)RefTwo pregnancies87; 29.3% (24.7, 33.9)-3.7PP (-12.5, 5.1)Three pregnancies93; 31.3% (25.4, 37.2)-1.7PP (-10.7, 7.3)Four pregnancies73; 32.9% (25.6, 40.2)-0.1PP (-8.5, 8.3)Five or more pregnancies98; 31.6% (23.7, 39.5)-1.4PP (-9.3, 6.5) Age at last birthday 14-19yrs51; 41.5% (30.4, 52.5)Ref20-24yrs117; 35.9% (26.7, 45.1)-5.6PP (-15.1, 3.9)25-29yrs109; 28.9% (24.4, 33.4)-12.6PP (-23.5, -1.6)30-34yrs100; 28.2% (23.7, 32.7)-13.3PP (-23, -3.6)35-49yrs80; 30.1% (22, 38.2)-11.4PP (-18.7, -4.1) Highest education level attained Never been to school48; 37.2% (26.6, 47.8)RefPrimary school97; 38.5% (25.7, 51.3)1.3PP (-9.7, 12.3)Junior secondary school205; 30.3% (25, 35.6)-6.9PP (-16.6, 2.8)Senior secondary /Technical/Vocational89; 29% (24.4, 33.6)-8.2PP (-18.1, 1.7)Tertiary18; 22% (11.6, 32.3)-15.3PP (-28, -2.6) Has any pregnancy ended in abortion? Yes174; 34.1% (28, 40.2)RefNo283; 30.2% (23.3, 37.1)-3.9PP (-10.6, 2.7) Has any pregnancy ended in a still birth? Yes21; 34.4% (16.4, 52.5)RefNo436; 31.5% (25.9, 37)-3PP (-18.9, 13) Service point used by client ANC274; 34.7% (28.5, 40.9)RefDelivery21; 36.8% (27.9, 45.8)2.1PP (-7.3, 11.5)PNC162; 27% (20.1, 33.8)-7.8PP (-14.2, -1.4) District Shai-Osudoku219; 25.7% (22.2, 29.2)RefNingo-Prampram238; 40% (32.4, 47.6)14.3PP (5.9, 22.7)PP = percentage points. Data are n; % (95% CI) or PP (95% CI) Table 5Prevalence and difference in prevalence of anxiety symptoms level and client characteristicsIndicatorAnxiety symptoms n; % (95% CI)Difference in prevalence percentage points (95% CI) Works to bring in any kind of income Yes298; 28.7% (24.3, 33.1)RefNo159; 39% (29.6, 48.4)10.3PP (3.3, 17.3) Gravida: Number of pregnancies One pregnancy106; 33% (23.1, 42.9)RefTwo pregnancies87; 29.3% (24.7, 33.9)-3.7PP (-12.5, 5.1)Three pregnancies93; 31.3% (25.4, 37.2)-1.7PP (-10.7, 7.3)Four pregnancies73; 32.9% (25.6, 40.2)-0.1PP (-8.5, 8.3)Five or more pregnancies98; 31.6% (23.7, 39.5)-1.4PP (-9.3, 6.5) Age at last birthday 14-19yrs51; 41.5% (30.4, 52.5)Ref20-24yrs117; 35.9% (26.7, 45.1)-5.6PP (-15.1, 3.9)25-29yrs109; 28.9% (24.4, 33.4)-12.6PP (-23.5, -1.6)30-34yrs100; 28.2% (23.7, 32.7)-13.3PP (-23, -3.6)35-49yrs80; 30.1% (22, 38.2)-11.4PP (-18.7, -4.1) Highest education level attained Never been to school48; 37.2% (26.6, 47.8)RefPrimary school97; 38.5% (25.7, 51.3)1.3PP (-9.7, 12.3)Junior secondary school205; 30.3% (25, 35.6)-6.9PP (-16.6, 2.8)Senior secondary /Technical/Vocational89; 29% (24.4, 33.6)-8.2PP (-18.1, 1.7)Tertiary18; 22% (11.6, 32.3)-15.3PP (-28, -2.6) Has any pregnancy ended in abortion? Yes174; 34.1% (28, 40.2)RefNo283; 30.2% (23.3, 37.1)-3.9PP (-10.6, 2.7) Has any pregnancy ended in a still birth? Yes21; 34.4% (16.4, 52.5)RefNo436; 31.5% (25.9, 37)-3PP (-18.9, 13) Service point used by client ANC274; 34.7% (28.5, 40.9)RefDelivery21; 36.8% (27.9, 45.8)2.1PP (-7.3, 11.5)PNC162; 27% (20.1, 33.8)-7.8PP (-14.2, -1.4) District Shai-Osudoku219; 25.7% (22.2, 29.2)RefNingo-Prampram238; 40% (32.4, 47.6)14.3PP (5.9, 22.7)PP = percentage points. Data are n; % (95% CI) or PP (95% CI) Prevalence and difference in prevalence of anxiety symptoms level and client characteristics PP = percentage points. Data are n; % (95% CI) or PP (95% CI) Table 6Prevalence and difference in prevalence of stress symptoms level and Client characteristicsIndicatorStress n; % (95% CI)Difference in prevalence percentage points (95% CI) Works to bring in any kind of income Yes153; 15.2% (10, 20.4)RefNo75; 19.4% (10.3, 28.4)4.2PP (-1.7, 10.1) Gravida: Number of pregnancies One pregnancy45; 14.7% (6.4, 22.9)RefTwo pregnancies41; 13.7% (7.7, 19.6)-1PP (-5.7, 3.7)Three pregnancies48; 16.7% (11.2, 22.3)2.1PP (-4.7, 8.9)Four pregnancies30; 14.8% (7.8, 21.7)0.1PP (-8.2, 8.4)Five or more pregnancies64; 21.5% (13.2, 29.8)6.8PP (2.7, 11) Age at last birthday 14-19yrs29; 23.8% (16, 31.5)Ref20-24yrs54; 17.3% (9.2, 25.3)-6.5PP (-12.3, -0.7)25-29yrs51; 13.9% (7.7, 20)-9.9PP (-17.6, -2.2)30-34yrs56; 16.7% (9.5, 23.8)-7.1PP (-14, -0.2)35-49yrs38; 14.8% (8.7, 21)-8.9PP (-16.1, -1.8) Highest education level attained Never been to school19; 15.4% (8.1, 22.8)RefPrimary school50; 20.4% (10.3, 30.5)5PP (-0.7, 10.6)Junior secondary school119; 18.1% (12, 24.3)2.7PP (-4, 9.4)Senior secondary /Technical/Vocational36; 12.1% (5.9, 18.3)-3.3PP (-10.4, 3.7)Tertiary4; 5.4% (2.2, 8.6)-10PP (-18.5, -1.6) Has any pregnancy ended in abortion? Yes103; 21.2% (13.1, 29.3)RefNo125; 13.8% (8.2, 19.3)-7.4PP (-11.6, -3.3) Has any pregnancy ended in a still birth? Yes7; 12.3% (1.2, 23.4)RefNo221; 16.5% (10.2, 22.9)4.2PP (-7.9, 16.4) Service point used by client ANC149; 19.6% (12.8, 26.3)RefDelivery9; 17% (1.8, 32.2)-2.6PP (-13, 7.8)PNC70; 12.1% (6.4, 17.7)-7.5PP (-11.7, -3.2) District Shai-Osudoku127; 15% (7.3, 22.7)RefNingo-Prampram101; 18.4% (10.3, 26.5)3.4PP (-7.8, 14.6)PP = percentage points. Data are n; % (95% CI) or PP (95% CI) Table 6Prevalence and difference in prevalence of stress symptoms level and Client characteristicsIndicatorStress n; % (95% CI)Difference in prevalence percentage points (95% CI) Works to bring in any kind of income Yes153; 15.2% (10, 20.4)RefNo75; 19.4% (10.3, 28.4)4.2PP (-1.7, 10.1) Gravida: Number of pregnancies One pregnancy45; 14.7% (6.4, 22.9)RefTwo pregnancies41; 13.7% (7.7, 19.6)-1PP (-5.7, 3.7)Three pregnancies48; 16.7% (11.2, 22.3)2.1PP (-4.7, 8.9)Four pregnancies30; 14.8% (7.8, 21.7)0.1PP (-8.2, 8.4)Five or more pregnancies64; 21.5% (13.2, 29.8)6.8PP (2.7, 11) Age at last birthday 14-19yrs29; 23.8% (16, 31.5)Ref20-24yrs54; 17.3% (9.2, 25.3)-6.5PP (-12.3, -0.7)25-29yrs51; 13.9% (7.7, 20)-9.9PP (-17.6, -2.2)30-34yrs56; 16.7% (9.5, 23.8)-7.1PP (-14, -0.2)35-49yrs38; 14.8% (8.7, 21)-8.9PP (-16.1, -1.8) Highest education level attained Never been to school19; 15.4% (8.1, 22.8)RefPrimary school50; 20.4% (10.3, 30.5)5PP (-0.7, 10.6)Junior secondary school119; 18.1% (12, 24.3)2.7PP (-4, 9.4)Senior secondary /Technical/Vocational36; 12.1% (5.9, 18.3)-3.3PP (-10.4, 3.7)Tertiary4; 5.4% (2.2, 8.6)-10PP (-18.5, -1.6) Has any pregnancy ended in abortion? Yes103; 21.2% (13.1, 29.3)RefNo125; 13.8% (8.2, 19.3)-7.4PP (-11.6, -3.3) Has any pregnancy ended in a still birth? Yes7; 12.3% (1.2, 23.4)RefNo221; 16.5% (10.2, 22.9)4.2PP (-7.9, 16.4) Service point used by client ANC149; 19.6% (12.8, 26.3)RefDelivery9; 17% (1.8, 32.2)-2.6PP (-13, 7.8)PNC70; 12.1% (6.4, 17.7)-7.5PP (-11.7, -3.2) District Shai-Osudoku127; 15% (7.3, 22.7)RefNingo-Prampram101; 18.4% (10.3, 26.5)3.4PP (-7.8, 14.6)PP = percentage points. Data are n; % (95% CI) or PP (95% CI) Prevalence and difference in prevalence of stress symptoms level and Client characteristics PP = percentage points. Data are n; % (95% CI) or PP (95% CI) Employment was negatively associated with depression (16% among those who work to bring income and 26.4% among those who do not work) and anxiety (29% and 39%, respectively) but not with stress. The difference in prevalences for depression and anxiety were statistically significant. This concurs with findings from the qualitative interviews where the participants mentioned that poor emotional, social and economic support from partners and limited other alternatives for social support make the ‘thinking a lot’ worse. They described feeling sad and worrying a lot about how to meet their basic needs, such as feeding themselves and their children. Pregnant adolescents (ages 14–19 years) were more likely to experience mental health distress, symptoms of depression, anxiety and stress with higher proportions of 33.9%, 41.5% and 23.8% respectively compared to the older age groups (Tables 3, 4 and 5). Apart from the age category 20–24 years, the difference in prevalences between all the age groups and that of 14–19 were statistically significant for all the three symptoms. The age group that had the lowest proportion experiencing symptoms of depression was those aged 35–49 years while that of anxiety was in those aged 30–34 years and stress those aged 25–29 years. Higher proportions of those with lower education experienced depression, anxiety and stress symptoms than those with higher education. The differences in prevalences between those above secondary school level and those who have never been to school were statistically significant. For anxiety and stress, the difference in prevalence was statistically significant at the tertiary level. Also, a higher proportion of those who just delivered experienced depression and anxiety symptoms while the lowest was among those attending PNC. For stress symptoms, the highest proportion was among those attending ANC. History of miscarriage or abortion was however, associated with stress symptoms level of the participants but depression and anxiety were not (Tables 4, 5 and 6). The difference in prevalence of stress symptoms level among those who reported history of miscarriage or abortion was significantly higher than those who never reported history of miscarriage or abortion. Client characteristics with maternal mental health distress as per the DASS scale Table 7Mean scores and mean score differences for DASS Scale for Client characteristicsVariable/category leveln; mean (95% CI)Difference in mean score (95% CI) Works to bring in any kind of income Yes945; 16.6 (14.3, 19.0)No368; 21.2 (16.5,25.9)4.6 (1.1, 8.1) Gravida: Number of pregnancies One pregnancy288; 18.4 (13.3, 23.6)Two pregnancies277; 17.1 (14.3, 19.9)-1.3 (-4.5, 1.8)Three pregnancies270; 18.0 (15.3, 20.6)-0.4 (-4.3, 3.5)Four pregnancies194; 17.7 (14.0, 21.4)-0.7 (-4.8, 3.4)Five or more pregnancies284; 18.4 (15.6, 21.1)-0.1 (-3.5, 3.4) Age at last birthday 14-19yrs117; 23.7 (18.2, 29.1)20-24yrs287; 19.7 (15.3, 24.2)-4.0 (-8.1, 0.1)25-29yrs342; 17.4 (15.1, 19.7)-6.3 (-11.1, -1.5)30-34yrs325; 16.6, (13.6, 19.6)-7.1 (-11.0, -3.2)35-49yrs242; 15.6 (12.8, 18.3)-8.1 (-12.4, -3.8) Has any pregnancy ended up in abortion? Yes464; 19.0 (16.2, 21.7)No849; 17.4 (13.8, 21.0)-1.6 (-4.2, 1.1) Has any pregnancy ended up in still birth? Yes53; 18.8 (13.3, 24.3)No1,260; 17.9 (14.8, 21.0)-0.9 (-6.0, 4.3) Service point used by client ANC710; 19.3 (16.2, 22.5)Delivery51; 21.3, (12.6, 30.0)2.0 (-5.2, 9.2)PNC552; 15.8, (12.6, 19.0)-3.6 (-6.2, -0.9) District Shai-Osudoku795; 16.3 (13.3, 19.3)Ningo-Prampram518; 20.4 (16.8, 24.0)4.1 (-0.6, 8.8) Highest education level attained Never been to school121; 20.9 (16.4, 25.4)Primary school234; 21.1 (15.1, 27.1)0.2 (-3.5, 4.0)Junior secondary school609; 17.9 (15.6, 20.2)-3.0 (-6.3, 0.3)Senior secondary /Technical/Vocational278; 16.0 (13.4, 18.6)-4.9 (-8.7, -1.0)Tertiary71; 10.1 (8.0, 12.3)-10.7 (-15.9, -5.5) Table 7Mean scores and mean score differences for DASS Scale for Client characteristicsVariable/category leveln; mean (95% CI)Difference in mean score (95% CI) Works to bring in any kind of income Yes945; 16.6 (14.3, 19.0)No368; 21.2 (16.5,25.9)4.6 (1.1, 8.1) Gravida: Number of pregnancies One pregnancy288; 18.4 (13.3, 23.6)Two pregnancies277; 17.1 (14.3, 19.9)-1.3 (-4.5, 1.8)Three pregnancies270; 18.0 (15.3, 20.6)-0.4 (-4.3, 3.5)Four pregnancies194; 17.7 (14.0, 21.4)-0.7 (-4.8, 3.4)Five or more pregnancies284; 18.4 (15.6, 21.1)-0.1 (-3.5, 3.4) Age at last birthday 14-19yrs117; 23.7 (18.2, 29.1)20-24yrs287; 19.7 (15.3, 24.2)-4.0 (-8.1, 0.1)25-29yrs342; 17.4 (15.1, 19.7)-6.3 (-11.1, -1.5)30-34yrs325; 16.6, (13.6, 19.6)-7.1 (-11.0, -3.2)35-49yrs242; 15.6 (12.8, 18.3)-8.1 (-12.4, -3.8) Has any pregnancy ended up in abortion? Yes464; 19.0 (16.2, 21.7)No849; 17.4 (13.8, 21.0)-1.6 (-4.2, 1.1) Has any pregnancy ended up in still birth? Yes53; 18.8 (13.3, 24.3)No1,260; 17.9 (14.8, 21.0)-0.9 (-6.0, 4.3) Service point used by client ANC710; 19.3 (16.2, 22.5)Delivery51; 21.3, (12.6, 30.0)2.0 (-5.2, 9.2)PNC552; 15.8, (12.6, 19.0)-3.6 (-6.2, -0.9) District Shai-Osudoku795; 16.3 (13.3, 19.3)Ningo-Prampram518; 20.4 (16.8, 24.0)4.1 (-0.6, 8.8) Highest education level attained Never been to school121; 20.9 (16.4, 25.4)Primary school234; 21.1 (15.1, 27.1)0.2 (-3.5, 4.0)Junior secondary school609; 17.9 (15.6, 20.2)-3.0 (-6.3, 0.3)Senior secondary /Technical/Vocational278; 16.0 (13.4, 18.6)-4.9 (-8.7, -1.0)Tertiary71; 10.1 (8.0, 12.3)-10.7 (-15.9, -5.5) Mean scores and mean score differences for DASS Scale for Client characteristics Table 7 presents mean scores and mean scores difference of the DASS21 scale for background information of the participants. Employment status, maternal age, educational level, and service point (whether participant was still pregnant, at delivery or postnatal) were all associated with variation in the mean DASS21 scores of the participants. Participants who said they work to bring home income had a lower mean DASS21 score (16.6) compared to those who do not work (21.2) and the difference in mean score was statistically significant. Mean scores of the DASS scale decreased as maternal age increased. The 14–19 age group had the highest mean score (23.7) and the lowest mean score was for those aged 35–49 years. The teenage scores were confirmed with the qualitative data where some healthcare providers reported that through observation, adolescent pregnancy was often associated with anxiety and depression due to parental and partner neglect. One of the healthcare providers stated.Most of the pregnant women sometimes they go into depression because of anxiety, and more especially among adolescent mothers. The reason is most adolescent girls are neglected by their partners and parents immediately they find out that they are pregnant. These girls have no jobs to fend for themselves and their babies, which makes them start thinking about how to survive with their babies. Sometimes, parents even neglect their teenage girls who have delivered. I know this because when I have an encounter with them, I ask if they have any help at home… Most of them have not returned to school due to this neglect. … it’s only few that get support from parents and have ensured that their children returned to school after the baby (Community psychiatric Nurse, New Ningo CHPs, FGD). Most of the pregnant women sometimes they go into depression because of anxiety, and more especially among adolescent mothers. The reason is most adolescent girls are neglected by their partners and parents immediately they find out that they are pregnant. These girls have no jobs to fend for themselves and their babies, which makes them start thinking about how to survive with their babies. Sometimes, parents even neglect their teenage girls who have delivered. I know this because when I have an encounter with them, I ask if they have any help at home… Most of them have not returned to school due to this neglect. … it’s only few that get support from parents and have ensured that their children returned to school after the baby (Community psychiatric Nurse, New Ningo CHPs, FGD). Apart from the 20–24 years age group, the differences in mean scores between the 14–19 years group and any of the year groups were statistically significant. The participants who reported that they have never been to school or ended their education at the primary school level had the highest mean score for DASS scale while those who reported they schooled up to tertiary level had the lowest mean score. Apart from those with primary education, the differences in mean scores between those who have never been to school and the other levels of education was significant. Though those who reported ever having miscarriage or abortion and those with history of still birth had higher mean scores, the differences in mean scores were not statistically significant. Mental health distress and primary healthcare responsiveness The responsiveness domains (dignity, autonomy, confidentiality, prompt attention, quality of basic amenities, choice of health provider, communication, and trust) were rated high by the clients in the exit surveys. With a total score of 10 for each domain, the mean scores ranged from 6.2 (choice of service provider) to 8.85 (dignity), with the overall mean score for responsiveness being 64.85 out of a total score of 80. It was found that those with some symptoms of mental health distress per our screening rated the health system very slightly less responsive (Figs. 8 and 9), reporting a mean score of 63.83 compared to a mean score of 65.4 for those reporting no symptoms of mental health distress. The association of mental health distress and responsiveness (the mean score difference) was statistically significant (p = 0.009). The relationship was similar when the analysis was done disaggregated by population groups. The differences in mean score responsiveness were statistically significant for antenatal and postnatal clients. The difference in mean score for those who have just delivered, though, was more than for the others, was not statistically significant. This could be due to the few numbers in this group (Fig. 10). The perception that the health system is less responsive among those experiencing mental distress aligns with findings from qualitative data, where participants expressed feelings that their mental health concerns were neglected by healthcare providers. Many women who had not been clinically diagnosed with a mental health condition but reported experiencing symptoms that were suggestive of being stressed, worried, anxious, and depressed stated they did not mention these symptoms to providers or seek help from the formal health system because they were unaware that mental health services were available at the PHC level. Mental health concerns are not typically addressed as part of routine maternal healthcare in Ghana.So, you (researchers) were asking questions about our mental health issues before some of the information you have had about our mental health situation started coming out. When you go to the ANC or PNC, they (maternal care providers) won’t ask you any questions to voice out your worries or confide in them. (Pregnant woman without mental health condition Ayikuma, FGD) So, you (researchers) were asking questions about our mental health issues before some of the information you have had about our mental health situation started coming out. When you go to the ANC or PNC, they (maternal care providers) won’t ask you any questions to voice out your worries or confide in them. (Pregnant woman without mental health condition Ayikuma, FGD) Although the confidentiality aspect of responsiveness was generally rated highly by respondents in the exit survey, most participants in the qualitative data expressed concerns about confidentiality. They reported based on personal experiences or those of other women, that healthcare providers do not consistently maintain client confidentiality. One maternal health client explained that they do not discuss their mental health issues with the healthcare providers, as reflected in the following quote:You (researchers) gave us assurance that the information we will tell you will be held in confidence, but for the nurses, they can broadcast your personal information like it is no one’s business. Why will you confide in them if they cannot keep their mouth shut?. (Pregnant woman without mental health condition, Natriku, FGD) You (researchers) gave us assurance that the information we will tell you will be held in confidence, but for the nurses, they can broadcast your personal information like it is no one’s business. Why will you confide in them if they cannot keep their mouth shut?. (Pregnant woman without mental health condition, Natriku, FGD) Healthcare providers in the FGDs reported that confidentiality issues in healthcare facilities sometimes stem from organizational and systemic challenges, including the low prioritization and minimal investments in mental health infrastructure at the primary care level. As a result, mental health units in PHC facilities often compete for operational space with other healthcare departments, leading to privacy lapses for maternal mental healthcare services, which, in turn, impact on stigma. A community mental healthcare provider expressed this concern during an FGD, stating:In our facility, the psychiatry unit is also a changing room, disease control and nutrition office and an injection room. Whilst you are doing an assessment for pregnant women, and another client will enter the place to receive an injection. Because of stigma, people with mental health condition do not want others to know about their condition (Community Psychiatric Nurse, FGD). In our facility, the psychiatry unit is also a changing room, disease control and nutrition office and an injection room. Whilst you are doing an assessment for pregnant women, and another client will enter the place to receive an injection. Because of stigma, people with mental health condition do not want others to know about their condition (Community Psychiatric Nurse, FGD). Fig. 8Perception of health system responsiveness by pregnant and postnatal mothers Perception of health system responsiveness by pregnant and postnatal mothers Fig. 9Perception of health system responsiveness by status of mental health distress Perception of health system responsiveness by status of mental health distress Fig. 10Perception of health system responsiveness by status of mental health distress by population group Perception of health system responsiveness by status of mental health distress by population group Discussion with frontline healthcare providers of maternal healthcare also revealed that MMH issues have not received the needed attention. They indicated that mental health nurses are few and most PHC facilities do not have them. The focus of nurses who provide ANC, delivery, and PNC services is primarily MCH service, which are typically limited to physical aspects of health. Mental health is not included in their duties, and they received no specialized training beyond the basics covered in their initial education. Although they recognized that MMH distress is a significant issue that needs attention, they did not have the capacity to do so. They had long forgotten the foundational mental health knowledge from their training. As one midwife emphasized:Concerning mental health, we all learnt a little about mental health at basic training but currently we have forgotten all that we learnt because we do not practice it, and when there is an in-service training for mental health providers, we are not included because it is not core to our work (Midwife, Old Ningo CHPs, FGD). Concerning mental health, we all learnt a little about mental health at basic training but currently we have forgotten all that we learnt because we do not practice it, and when there is an in-service training for mental health providers, we are not included because it is not core to our work (Midwife, Old Ningo CHPs, FGD). Healthcare providers reported that they do not actively screen for mental health issues among maternal health clients due to the absence of screening tools and the lack of detailed MMH information in the healthcare records. Consequently, they typically only intervene when mental illness becomes visibly severe, after progressing from mild symptoms. In such cases, they conduct basic assessments and refer patients to mental health nurses or community psychiatric officers. The few community mental health nurses available in the districts are primarily involved in case finding and home visits for maternal clients with severe mental health issues. Frontline providers of maternal care also highlighted that one barrier preventing pregnant and postpartum mothers from seeking and accessing formal care for mental distress is the influence of socio-cultural beliefs surrounding the causes of mental health problems, which often stigmatize mental distress and mental illness. The healthcare providers mentioned that often, community members have the belief that witches and malevolent spirits cause mental health problems. As a result, (MMH) clients prefer seeking care from shrines, traditional healers or prayer camps although these alternative healthcare practitioners are unable to manage their condition. As illustrated in one of the providers’ responses below.In this district people believe that mental health problems are the result of bewitching the person. So, instead of them to seek proper medical care for the mental health issue they run to the prayer camps or shrines, thinking that they will be completely cured in prayer camps or shrines (FGD, Community mental healthcare provider). In this district people believe that mental health problems are the result of bewitching the person. So, instead of them to seek proper medical care for the mental health issue they run to the prayer camps or shrines, thinking that they will be completely cured in prayer camps or shrines (FGD, Community mental healthcare provider). These complex socio-cultural and systemic factors seem to contribute to the reliance on traditional remedies and alternative services when women seek and accept care for their mental health conditions and can hamper effective diagnoses, treatment and management of mental health conditions for this population. Discussion Our mixed-methods study in Shai-Osudoku (SOD) and Ningo-Prampram (NP) districts in Ghana examined the prevalence, perceptions, and experiences of mental health distress (depression, anxiety, and stress) among maternal health clients and the responsiveness of the health system towards them. It also explored associations between client characteristics and MMH views of frontline providers at PHC on MMH and the health system response. Results from screening a sample of antenatal, delivery and postnatal follow-up visit clients with the DASS21 screening tool for mental health distress (probable depression, anxiety and stress) revealed levels of MMH distress ranging from mild to severe, which aligned with findings from the qualitative data where women indicated experiencing varying levels of mental health symptoms ranging from mild to severe during pregnancy, delivery and postpartum. Expressions used to describe such experiences were: “thinking a lot”, “feeling sad”, “sleeplessness”, and “suicidal thoughts” and in extreme cases, attempts at suicide. These are consistent with findings of other studies that reported experiences such as feeling anxious or stressed [64], suicidal ideation [65], and a feeling of sadness [66] during pregnancy and the postpartum period. These findings indicate that health (MMH) is an important issue and confirm observations from other studies [67]. MMH is not well integrated into MCH programs in the study setting, a situation that mirrors that of many other low- and middle-income countries (LMICs) [68]. The levels of probable anxiety, depression, and stress we found in this study are a little higher than the global average from some studies [11, 19]; but lower than those reported in the systematic review when Wang and colleagues estimated the global prevalence of postpartum depression to be 17.2% [17]. Though antenatal depression symptoms were estimated to be higher than postnatal depression symptoms in this study, the level recorded is lower than what was found in other studies in Ghana [4, 26–31]. Though in most of the studies, postnatal depression is higher than antenatal depression, this study found higher levels of depressive symptoms in antenatal women. This could be due to the fact that antenatal depression is a predictor of postnatal depression [69–71]. Antenatal depression can develop if women are living in stressful conditions or have experienced significant adverse life events without adequate support [72, 73]. However, some of these stressors may resolve or reduce after childbirth, thereby reducing the prevalence of postnatal depression compared to antenatal depression. The probable anxiety level recorded in this study using the DASS21 screening (32%) is higher than what was found in other studies conducted in Ghana that reported values between 11.4% and 14.2% [4, 27, 28, 30]. Also, in contrast to these studies, the probable anxiety level in this study is higher than that of probable depression. Our findings are, however, consistent with other studies that show that anxiety levels are higher than depression [74–76]. The high anxiety may be due to fear of the unknown, worries about the pregnancy, or economic issues. The study also found that probable depression, anxiety, and stress symptoms were significantly associated with years of completed schooling, maternal age, and service point (whether participants were still pregnant, at delivery or postnatal). However, employment status was only significantly associated with depression and anxiety, while history of miscarriage or abortion was only significantly associated with stress. Those not employed or not engaging in any income-generating activity were more likely to be anxious or depressed. This was corroborated by the narratives from the qualitative data, where participants attributed their experiences of depression and anxiety to their low socio-economic status. They mentioned that not knowing how to cater for themselves and their children was one of the reasons for “thinking a lot”. The findings of this study concur with other studies where low socio-economic status has been found to be a risk factor for antenatal and postnatal mental disorders, such as depression and anxiety [77–79]. Maternal age and educational level were also found to be associated with MMH distress, similar to findings in other studies [80, 81]. Loss of a wanted pregnancy is a devastating adverse event that brings about stress due to its traumatic nature. A history of abortion or miscarriage was a risk factor for stress symptoms in this study, as has also been found in other studies [82, 83]. Other drivers of mental health stress mentioned by the women included social circumstances such as having abusive or unsupportive partners, receiving distressing news about their pregnancy or unborn babies, and facing poor socioeconomic conditions. These findings are consistent with other studies where women reported having perinatal distress because of intimate partner violence [84–86], feeling that they have failed [87], access to resources [88], having insufficient social support [89], and being stigmatized [90]. Although generally, maternal health clients rated the health systems responsiveness domains highly, those experiencing mental distress on the DASS21 screening rated it less responsive compared to those without such distress. This highlights a significant gap in maternal healthcare at the primary level, where mental health needs are often unmet. Pregnant or postpartum women with mental health distress may go unnoticed until their symptoms become severe, resulting in inadequate attention and support from the healthcare system [91]. Given the lack of a formal health system response to the mental health needs of these clients, they may resort to traditional healers, prayer camps, and others to fill this gap. It will be important for the formal health services not only to improve the screening and early detection of MMH conditions but also improve their capacity to respond. Additionally, to explore collaboration with well-established and trusted alternative providers, such as traditional and faith-based practitioners, in the provision of MMH. Interviews with frontline providers (midwives) revealed that while they appreciate the mental health needs of the maternal health clients, they are unable to address them due to lack of capacity and resources. Their focus remains primarily on physical maternal healthcare, often overlooking early signs of mental health issues unless they become severe enough to require a referral to the mental health unit. The few mental health nurses or community psychiatric officers (a total 27 for the two districts, comprising a Clinical Psychologist, Clinical Psychiatric Officer (Physician Assistant in Psychiatry), Registered Mental Health Nurses, and Community Mental Health Officers) do attend to the severe cases, do case finding, and home visits. They are, however, very few and need the assistance of their colleagues in the maternal care units who are willing to support but lack the capacity. This finding concurs with a study on healthcare providers’ knowledge, attitude, and perceptions of screening for MMH during and after pregnancy [19]. Other studies have revealed similar findings where integration of maternal and mental health has been suggested [79, 92–94] and its feasibility demonstrated [92] if maternal health providers are trained and supported to identify and help provide services to those in need [1, 95]. These findings resonate with the call for mental health to be integrated into the maternal healthcare system to meet the needs of pregnant and postpartum women with mental health distress [79]. Policy and practice implications Our interactions with clients and frontline healthcare providers revealed the need for the health system in Ghana to integrate mental health in maternal healthcare. This integration could be supported by providing training in mental health conditions and symptoms to health (MCH) providers. Screening and early identification of common mental health problems such as depression, anxiety, and stress could help minimize, if not prevent severe mental health disorders among pregnant and postpartum mothers. Using interventions such as simple screening tools can facilitate prompt detection and management of maternal mental health conditions among at-risk groups identified across all levels of care, making the Ghanaian health system more responsive in ensuring a comprehensive service delivery. Integration of maternal health and mental health services could be supported through the collaboration of the Mental Health Authority and GHS Family Health and Institutional Care Divisions. Study strengths and limitations This study has several strengths. We used a mixed-method design, which allowed triangulation of the findings from both qualitative and quantitative data. Qualitative interviews and FGDs provided deeper insights into the experiences captured in the quantitative data, offering a more comprehensive understanding of women’s experiences. Additionally, the study benefited from rigorous pre-fieldwork training and intensive supervision during the data collection, coupled with robust quality assurance processes. The study also had some limitations. First, the cross-sectional design limits our ability to establish causal relationships between MMH distress and background characteristics. Second, conducting exit survey at health facilities may have influenced some responses, especially those concerning perceptions of responsiveness. Third, most of the pregnant and post-partum women were illiterate and so relied on data collectors for interpretation of the questionnaires. To mitigate these limitations, data collectors received extensive training, and the tool was translated and back-translated during role plays. Participants were assured of confidentiality during the consent process, and interviews were conducted away from health workers. Conclusions Maternal mental health is an important but neglected public health problem. Pregnant and post-partum mothers experience varying levels of anxiety, stress, and depression, as well as other mental health conditions, which are not screened for, detected, and managed as part of routine maternal healthcare at primary health care. Interventions are required to integrate mental health into maternal health in primary care, including screening, early detection, and prompt management to improve health outcomes of pregnant and postpartum women. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1 Supplementary Material 2 Supplementary Material 2 Supplementary Material 3 Supplementary Material 3