Assessing the geographical variation of antenatal care and maternal mortality rates among pregnant women in Sierra Leone based on DHIS2 data from 2020-2022: a retrospective, descriptive, ecological study.

backgroundSierra Leone reported a high maternal mortality rate (MMR) of 717 per 100,000 live births in 2019 due to disparities in use, quality, and access to skilled birth attendants and antenatal care (ANC) services. Our study aimed to assess the geographical variation in MMR and ANC among pregnant women in Sierra Leone, based on District Health Information System-2 (DHIS2) data from 2020 to 2022.

methodsThis retrospective ecological study analysed aggregated DHIS2 data from health facilities across all five administrative regions of Sierra Leone from 2020-2022. The data analysis included all pregnant women and maternal deaths captured in DHIS2. The MMR was evaluated using descriptive statistics, while the geographical disparity between the MMR and ANC was assessed through factorial analysis. Multivariate regression and bivariate association analysis were performed to determine the relationship between ANC and MMR across the five administrative regions of Sierra Leone.

resultsThe MMR reported in 2020 was 760 per 100,000 live births, which significantly decreased to 648 in 2021, and the lowest rate (288 per 100,000 live births) was reported in 2022. However, regional disparities persisted, particularly in the northern, western, and northwest regions. This study found that infection deaths, haemorrhage, and malaria were the significant causes of maternal deaths from 2020-2022. The correction coefficient for the causes ranged from 0.627 to 0.960. This study found that at the 1st ANC visiting time point, all regions, excluding the southern region, showed a non-significantly decreased relative number of screened pregnant women in 2021 and 2022 compared to 2020. However, at the 4th and 8th ANC visits, all regions, excluding the western region, showed a significantly increased relative number of screened pregnant women in 2021 and 2022 than in 2020. This pattern is consistent with growing awareness and adaptive strategies, though temporal trends alone cannot establish causality among pregnant women and healthcare systems in Sierra Leone. This study found that regional differences were less pronounced each year despite significant increases over time in haemoglobin testing, syphilis screening, and IPT doses. The multivariate regression results indicated that the IPT-1st dose showed the strongest negative association (standard regression coefficient = -1.307, P < 0.001) with MMR. This study found a significant negative association between ANC-1st visit, iron folate, and malaria in the 2nd or 3rd trimester in pregnant women treated with artemisinin-based combination therapies.

conclusionsThis ecological analysis revealed significant regional disparities in MMR and ANC utilisation in Sierra Leone. Regions with lower ANC utilisation rates were associated with higher MMRs, highlighting the need for targeted interventions. However, these associations do not imply causation, and individual-level studies are needed to confirm these relationships. Improving access to and quality of ANC services in the northern, northwest, and western regions may be critical for efforts to reduce maternal mortality and for progress toward Sustainable Development Goal 3.1. However, the ecological associations reported here require confirmation through individual-level intervention studies.