Abstract
Background Stillbirths during pregnancy do not only trigger maternal psychological trauma but also affect the economic growth in developing countries. Since the predictors of stillbirths and their dynamics are diverse in different settings, it requires adequate knowledge to mitigate these predictors in reducing perinatal deaths in Ghana. This study aimed to determine the prevalence and determinants of stillbirths in a referral hospital in Northern Region of Ghana.
Methods This study is a portion of a retrospective cohort study conducted from April to July 2020 in Savelugu municipality of Northern Ghana. Sociodemographic and obstetric data were collected using a structured questionnaire from antenatal and delivery records. We used binary logistic regression to identify independent predictors of stillbirths.
Results Four percent (4.0%) of the 422 deliveries were stillborns. Anaemia in the first trimester of pregnancy (aOR: 3.15; 95%CI: 0.94 – 10.6; p=0.034), self-employed mothers (aOR: 7.75; 95%CI: 2.33 – 25.8; p=0.018), and non-use of insecticide-treated bed nets during pregnancy (aOR: 12.9; 95%CI: 1.55 – 107; p=0.001) were independent predictors of stillbirths.
Conclusion The relatively high stillbirth prevalence at the referral facility is affected by gestational anaemia and lifestyle factors including non-use of insecticide-treated bed nets and self-employed jobs. Management of Savelugu Municipal Hospital (Ghana Health Service), program planners, and other stakeholders who are engaged in improving perinatal and maternal health should be made aware of these predictors in order to reduce the stillbirth rate and its psychological effect on families in Savelugu.
Introduction
Stillbirths (also called stillborns) during pregnancy do not only trigger maternal psychological trauma but also contribute to maternal mortalities which affect the economic growth and development in low and middle-income countries (1,2,3). Stillbirths denoted as the death of foetuses occurring after 28 weeks of gestation before or during childbirth (3), account for more than half of all perinatal mortalities worldwide (4).
In 2015, an estimated 2·6 million neonates were stillborn globally, which indicates an approximately 19.0% reduced rate since 2000 with a sluggish reduction among Sahelian Africans. About 98.0% of all stillbirths happen in developing countries with 77.0% occurring among African and Asian pregnancies (5). In China (6), South Africa (7), and Nigeria (8), the prevalence of stillbirths were found to be 0.9%, 3.5%, and 4.0% respectively. Among the Ghanaian population, higher stillbirth rates of 5.9% and 10.6% were respectively recorded in 2013 and 2018 at two popular tertiary hospitals (9,10) which makes these stillbirth occurrences very worrisome in the country.
The aetiology and antecedents of stillbirths are multifaceted and are affected by sociological, physiological, and physical factors (1). It is reported that stillbirth rate and its determinants are dependent on the geographic and sociodemographic characteristics of the population (3,5). Therefore, the study findings from one setting might not be applicable to another. Universally, earlier studies acknowledged that the factors associated with stillbirths include congenital foetal deformities, foetal distress (11), preterm births, breech delivery (10), abruptio placentae, and male foetus (12). The use of insecticide-treated bed net (13), maternal age less than 20 years, unmarried mothers (6), self-employment status, multiparity, and vaginal delivery (8) were also regarded as the determinants of stillbirths in other studies.
To achieve the World Health Assembly’s national target of stillbirth prevalence of 10 or less per 1000 deliveries by 2035 in every country through the Every Newborn Action Plan (3), it is essential that the determinants or predictors of stillbirths and their dynamics in different settings are well understood and appreciated. Although stillbirths are higher in developing countries including Ghana (9,10), relevant information on predictors of stillbirths are inadequately documented and/or readily available in the district, regional, or national databank. Hence, adequate awareness and proper documentation of the sociodemographic and obstetric factors of stillbirths in Northern Ghana could lead to preventable perinatal mortalities including stillbirths, and subsequently contribute to improving maternal and child health. This study sought to determine the sociodemographic and obstetric predictors associated with stillbirths in a peri-urban referral hospital in the Northern Region of Ghana.
Materials and methods
Study design
This report mirrors findings from a segment of retrospective study conducted in Savelugu municipality of Northern Region of Ghana to investigate abnormal haemoglobin levels and birth outcomes. The current study is also a cross-sectional survey and analysis of retrospective data of mothers who delivered at Savelugu Municipal Hospital from June to July 2020 in Northern Ghana.
Study settings and population
This study was conducted in Savelugu Municipal Hospital, one of the five major public health facilities in Savelugu Municipality located in Northern Ghana. Savelugu Municipal Hospital is one of the 16 district/municipal hospitals in the Northern Region that serves as a referral facility in the Savelugu municipal and also a referring facility to some health facilities in five neighbouring districts including Gushegu, Karaga, Kumbungu, Nanton, and Tolon. The hospital provides comprehensive maternal and newborn services including free delivery services (14) to a population of approximately 130,000. Due to the high average monthly delivery count of about 250 which frequently comes out with adverse birth outcomes (15), this only referral hospital in the municipal was selected to investigate predictors of stillbirths among mothers who delivered across two months in 2020.
Sample size, inclusion, and exclusion criteria
A total of 498 mothers visited the Savelugu Municipal Hospital for childbirth or delivery purposes in June and July 2020. All mothers who owned antenatal record books and delivered at the hospital were eligible for the study. However, 76 mothers who were further referred to seek secondary or tertiary health services and/or with cardiovascular or kidney problems were ineligible. The study therefore employed 422 mothers who delivered at the Savelugu Municipal Hospital in the months of June and July 2020.
Data collection
Data were collected using a pretested questionnaire by well-trained data collectors (midwives) through structured postpartum interviews with all recruited mothers from June 1 to July 31, 2020. This interview was done within the first 12 hours after delivery while the data was anonymously collected. Based on existing literature, previous studies, sociological and biological plausibility (1,5,15,16), sociodemographic and obstetric information were collected from antenatal and delivery records not excluding age, marital status, education level, job status, first-trimester body mass index (since foetal weight at first trimester is negligible (17)), antenatal visits, parity, gravidity, pre-eclampsia diagnosis, postpartum haemorrhage, delivery mode, birth/delivery prognosis, gestational age at delivery, haemoglobin levels across pregnancy trimesters, use of pre-pregnancy family planning, and insecticide-treated bed nets (15).
Statistical analysis
The outcome variable was birth prognosis categorized into binary outcomes for stillbirths (coded as 1) or livebirths (coded as 0) while exposure variables included all selected sociodemographic and obstetric variables which were categorical. According to World Health Organization, stillbirth is a foetal death that occurs after 28 weeks of pregnancy usually before or during childbirth while livebirth is regarded as newborns with signs of life (3). Haemoglobin levels were trichotomized into anaemia (< 11g/dl), normal (11 – 13.1g/dl), and polycythaemia (≥ 13.2g/dl) while gestational age at delivery was grouped into < 37 weeks (preterm), 37 – 41 weeks (normal term), and ≥ 42 weeks (post-date). Parity was classified as primiparity (one delivery), and multiparity (two or more deliveries) while gravidity was grouped into primigravida (one pregnancy) and multigravida (two or more pregnancies). Standardized categories of pre-pregnancy body mass index of underweight (<18.5 kg/m2), normal (18.5–24.9 kg/m2), and overweight (≥25.0 kg/m2) were used. Antenatal visits were also dichotomized into ≥ 8 visits and < 8 visits (15,16).
Chi-square/ Fischer’s analysis test of proportions was employed to determine the statistical association between birth prognosis and all exposure variables. Multicollinearity among the significant exposure variables was then assessed with variance inflation factor where variables with variance inflation factor less than 5 were selected into logistic regression analyses (15).
Applying simultaneous variable entry, the significant exposure variables with no multicollinearity were forwarded into the binary logistic regression model and run to determine significant predictors of stillbirths. All analyses were computed using STATA 17.0 (Stata Corporation, Texas, USA) and a p-value of < 0.05 was considered significant during the analysis.
Ethics approval and consent to participate
This study was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving research study participants were approved by Navrongo Health Research Centre Institutional Review Board with approval number NHRCIRB373. Permission was also given by Ghana Health Service. Written informed consent/assent was obtained from all study participants and/or legal representatives and also from parents/guardians of participants under 18 years of age.
Results
Prevalence of stillbirths among study participants
About 498 mothers visited the facility for childbirth in June and July 2020. However, 76 of the mothers were excluded due to our exclusion criteria. Hence, 422 deliveries and/or mothers were used for analysis. Of the 422 mothers, the prevalence of stillbirths was 4.0% (Figure 1).
Prevalence of stillbirths among study participants
Sociodemographic, and obstetric characteristics of study participants
The average age of the mothers was 27.6 (sd=6.0) years while teenage deliveries were estimated at 9%. Majority of the participants could not read nor write (no formal education) (43%), were informal sector employees (54%), and were married (92%) (Table 1). Anaemia was highest in the second (71%), followed by the first (64%) and third (45%) trimesters of pregnancy. Few of the mothers had delivered once (28%), attended antenatal clinic at least eight times (28%), ingested sulphadoxine-pyrimethamine at least three times (35%), and did not use insecticide-treated bed nets during pregnancy (36%) (Table 2).
Predictors of stillbirths
Among the selected characteristics of study participants, maternal occupation status (p=0.003), sulphadoxine-pyrimethamine intake (p=0.012), insecticide-treated bed nets use (p<0.001), parity (p=0.035), and third-trimester haemoglobin levels (p=0.030) were significantly associated with birth prognosis (stillbirths) during bivariate analysis (Tables 1 and 2).
In the multiple logistic regression analysis, maternal occupation status, insecticide-treated bed nets use, and third-trimester haemoglobin levels remained significant (Table 3).
Mothers who were self-employed had 12.9 (95%CI: 1.55 – 107; p=0.018) increased odds of stillbirths compared with their unemployed counterparts. Mothers who did not use insecticide-treated bed nets throughout pregnancy also had 7.75 (95%CI: 2.33 – 25.8; p=0.001) odds of stillbirths compared with those who used it. The odds of mothers with anaemia in the third trimester of pregnancy having stillbirths was 3.15 times (95%CI: 0.94 – 10.6; p=0.018) compared to mothers with normal haemoglobin levels.
Discussion
The global occurrence of stillbirth as a critical public health issue not only increases the rate of perinatal mortalities but also affects the mental health of mothers and their families (1,2,3). In poor economic communities like the Ghanaian context, most stillbirths are generally higher and usually go unrecorded (18). The prevalence of stillbirths in the present study was found to be 4.0% (95%CI: 2.4% – 6.4%) which is in line with a multicenter study in Nigeria (8). This similarity might be due to the uniformity in the healthcare system and provision of services for pregnant mothers in Ghana and Nigeria. This study’s stillbirth prevalence was lower (10.6%) than the estimation found in a cohort study conducted in a tertiary hospital in Southern Ghana (9), but higher (0.9%) than the findings from a Chinese report (6). The discrepancy might be due to variations in the study area and the inclusion and exclusion criteria (methodologies) used. Moreover, the lesser prevalence of stillbirths in this study as compared to Southern Ghana might be due to the deployment of trained traditional birth attendants as agents of facility deliveries, and trained community health nurses to perform the duties of midwives in the Savelugu municipality. This was an initiative of the Catholic Relief Services (development partner) which was adopted by the Northern Regional Health Directorate of Ghana Health Service to improve maternal and child health outcomes. Nonetheless, the relatively low prevalence of stillbirths in the study area is still worrying as postpartum mothers and their families suffer psychological trauma after experiencing stillbirths throughout this nine-month journey (1,3). It could also indicate that the care for women during pregnancy, labour, and birth might not have received appropriate and standardized obstetric practices and protocols.
In this study, maternal anaemia in the third trimester of gestation was independently associated with stillbirths. Whilst a retrospective study in Columbia (19) was in agreement with the present study, a similar study in Pakistan (20) was not in support. Anaemia is known to be a significant predictor of malnutrition among pregnant mothers as iron deficiency type is apparently the most prevalent cause of anaemia during gestation (21). As anaemia causes inadequate oxygen and nutrient supply to the foetus during the late stage of pregnancy (22,23), it could lead to deficient supply of these nutrients to the foetus which subsequently causes death of the foetus either in-utero or within short period after birth (24).
This study also found that self-employed mothers have increased odds of stillbirth delivery. A study in Nigeria chronicled that self-employed mothers were strong predictors of stillbirths (8) which is in support of this study’s findings. Self-employed mothers in Northern Ghana are always busy with their job activities and give little attention to healthcare services during pregnancy period which predisposes these mothers to poor birth outcomes like stillbirths. Additionally, due to the multiple responsibilities of these self-employed mothers for taking care of their businesses and other household activities throughout the long day period, it may reduce healthcare visits making them susceptible to stillbirth delivery.
Finally, mothers who did not sleep under insecticide-treated bed nets during pregnancy were more likely to deliver stillbirths. In line with the present findings, mothers’ sleep practices including those who did not sleep under insecticide-treated bed nets during pregnancy were exposed to stillbirths in a cross-sectional study in Ghana (13) but was contrasting to a meta-analysis by Flenady and his colleagues (25). Despite the constant supply of free insecticide-treated bed nets to pregnant mothers during their first antenatal visits in Ghana, it is reported that most mothers do not sleep under the nets due to unfavourable weather conditions (hot weather) while others also use the nets for subsistence farming purposes in the municipality (26). This exposes them to malaria infections thereby increasing their risk of stillbirths delivery (27).
Coincidentally, large proportion of these mothers who served as possible predictors of stillbirths (thus self-employed mothers and mothers who did not use insecticide-treated bed nets) made less than eight ANC visits and had anaemia during each trimester of pregnancy. This implies that, lesser ANC visits and low haemoglobin levels in each trimester during pregnancy could precipitate unfavourable intra-uterine and gestational conditions (22,23) leading to an increased maternal risk for stillbirths delivery (27,28) in the study area.
This study had limitations in the generalizability of the predictors of stillbirths due to the employment of a cross-section of data in the year 2020. The prevalence of stillbirths in the hospital may be under and/or overestimated since the span of the data was about two months. We must also admit that the use of previous antenatal and obstetric data (maternal weight, haemoglobin levels, gestational age) from antenatal records could be a limitation of the study as a result of possible measurement and documentation errors.
Conclusion
The prevalence of stillbirths is relatively high among mothers who delivered at the Savelugu municipal hospital in June and July 2020. These stillbirths were independently influenced by gestational anaemia, and lifestyle factors including non-use of insecticide-treated bed nets and self-employment status.
Timely identification of these determinants is vital to prompt intervention strategies to strengthen facility obstetric triage and indirectly mitigate the higher risk of maternal psycho-emotional problems from stillbirths in Savelugu.
Regarding the global incidence of stillbirths, World Health Assembly recommends a national target of stillbirth prevalence of 10 or less per 1000 deliveries by 2035 in developing countries like Ghana (3). In order to achieve this goal, education on nutrition and lifestyle, early detection, and swift management of gestational anaemia are very necessary to prevent stillbirths and consequently reduce perinatal mortalities in Savelugu.
Management of Savelugu Municipal Hospital (Ghana Health Service), program planners, and other stakeholders who are engaged in improving perinatal and maternal health should be made aware of these predictors in order to reduce the stillbirth rate and its psychological effect on parents in Savelugu.
Data Availability
The datasets collected, generated, and/or analyzed during the present study are available from the corresponding author upon reasonable request.
Supporting information
S1 Fig. Prevalence of stillbirths among study participants
Ethics approval and consent to participate
This study was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving research study participants were approved by Navrongo Health Research Centre Institutional Review Board with approval number NHRCIRB373. Permission was also given by Ghana Health Service. Written informed consent/assent was obtained from all study participants and/or legal representatives and also from parents/guardians of participants under 18 years of age.
Consent For Publication
Not applicable
Availability of data and materials
The datasets collected, generated, and/or analyzed during the present study are available from the corresponding author upon reasonable request.
Competing interests
The authors declared that they have no competing interests
Funding
There was no funding support
Authors’ contributions
SAG, AA, and PAA designed, and conceptualized the study. SAG, AA, MG, and WP carried out the data collection while SAG, WP, and PAA analyzed and interpreted the data. All authors contributed to drafting the initial manuscript, reviewing, and approving the final manuscript.
Authors’ acronyms
Silas Adjei-Gyamfi (SAG); Abigail Asirifi (AA); Mamudu Gafaru (MF); Wisdom Peprah (WP); Paul Armah Aryee (PAA)
Acknowledgments
The authors are thankful to all research assistants, study participants, and volunteers who contributed in diverse ways to the successful implementation of the study. We are also grateful to University for Development Studies for supportive supervision during this study.
References
