Background: Access to a complete continuum of maternal and child health care has been recommended globally for better pregnancy outcomes. Hence this study determined the level (pooled prevalence) and predictors of successfully completing continuum of care (CoC) in Rwanda. Methods: We analyzed weighted secondary data from the 2019–2020 Rwanda Demographic and Health Survey (RDHS) that included 6,302 women aged 15 to 49 years who were selected using multistage stratified sampling. We analyzed complete continuum of care as a composite variable of three maternal care services: at least four ANC contacts, SBA, maternal and neonatal post-natal care. We used the SPSS version 25 complex samples package to conduct multivariable logistic regression. Results: Of the 6,302 women, 2,131 (33.8%) (95% CI: 32.8–35.1) had complete continuum of care. The odds of having complete continuum of care were higher among women who had exposure to newspapers (adjusted odds ratio (AOR): 1.30, 95% CI: 1.11–1.52), those belonging to the eastern region (AOR): 1.24, 95% CI: 1.01–1.52), southern region (AOR): 1.26, 95% CI: 1.04–1.53), those with health insurance (AOR): 1.55, 95% CI: 1.30–1.85), those who had been visited by a field health worker (AOR: 1.31, 95% CI: 1.15–1.49), those with no big problems with distance to health facility (AOR): 1.25, 95% CI: 1.07–1.46), those who were married (AOR): 1.35, 95% CI: 1.11–1.64), those with tertiary level of education (AOR): 1.61, 95% CI: 1.05–2.49), those belonging to richer households (AOR): 1.33, 95% CI: 1.07–1.65) and those whose parity was less than 2 (AOR): 1.52, 95% CI: 1.18–1.95). Conclusion: We have identified modifiable factors (exposure to mass media, having been visited by a field health worker, having health insurance, having no big problems with distance to the nearest health facility, belonging to richer households, being married and educated), that can be targeted to improve utilization of the entire continuum of care. Promoting maternity services through mass media, strengthening the community health programmes, increasing access to health insurance and promoting girl child education to tertiary level may improve the level of utilization of maternity services.
Rwanda is a central-eastern African nation of about 12 million people [19, 25] whose health system consists of eight national referral hospitals, four hospitals at provincial level, 35 district level hospitals, 495 health centers, 406 health posts and over 45,000 community health workers (CHWs) [20, 26, 27]. Each village has a male-female CHW pair and one female Agent de Sante Maternelle (ASM) and these CHWs are responsible in delivering the first line of health services including maternal and newborn health services [20, 28]. The country has a universal, community-based health insurance program that has a household subscription and co-payments at the time of care and all citizens are eligible to enroll into it [20, 29]. The 2019-20 Rwanda Demographic Survey (RDHS) was used for this analysis employed a two-stage sample design with the first stage involving sample points (clusters) selection consisting of enumeration areas (EAs) [19]. Between November 2019 and July 2020, a total of 13,005 households were selected from the first stage selected EAs [19]. The household and the woman’s questionnaires provided the data used in this secondary analysis. Eligibility for women to participate in the RDHS was being aged between 15 and 49 years and being either permanent residents of the selected households or visitors who stayed in the household the night before the survey [19]. Out of the total 13,005 households originally sampled, 12,951 had occupants and 12,949 were interviewed [19]. We included only women who had given birth within five years preceding the survey in this analysis. Out of the 14,675 women found eligible for the RDHS, the team was able to interview 14,634 women of which only 6,302 women had given birth within the last five years preceding the survey [19]. Complete continuum of maternal and newborn healthcare was coded 1 and incomplete coded as 0 [6]. Complete continuum was considered as having utilized all the three maternal healthcare services; at least four ANC contacts, SBA, at least one maternal and neonatal PNC checkup within six weeks after childbirth [6, 30, 31]. Nineteen independent variables were categorized into women and household characteristics, and were chosen basing on previous studies [32–34] and availability in the RDHS database as shown in Table 1. Categorization of independent variables Less than 6 and 6 and above. (this was based on the average household size of 5) No education, primary, secondary, and tertiary (highest level of education attended) No big problems and big problems (RDHS had three original categories (no problem, no big problem and big problem) however, after data collection, no woman reported no problem) No big problems and big problems (RDHS had three original categories (no problem, no big problem and big problem) however, after data collection, no woman reported no problem) 2014–2015, 2016, 2017, 2018, 2019 and 2020 (we combined 2014 with 2015 because 2014 only had 9 births) We used the complex sample package of SPSS (version 25.0) statistical software which accounted for the multi-stage cluster study design [35]. Individual sample weight, sample strata In order to ensure representativeness of the survey results at the national and regional level and to minimize the effects of unequal probability sampling in different strata, data were weighted [12]. Initially, we did descriptive statistics. Frequencies and proportions/percentages for categorical variables have been presented. We then used bivariable logistic regression to assess the association between each independent variable and the outcome whose crude odds ratio (COR), 95% confidence interval (CI) and p-values have been presented. Independent variables found significant at p-value less than 0.25 at bivariable level with other independent variables were considered for multivariable logistic regression to assess the independent effect of each variable on the CoC utilisation [6]. Before multivariable analysis, multicollinearity was assessed using variance inflation factor (VIF) and no VIF above 2.5 was observed. Model fitness was assessed with the F-test with a p-value of < 0.001. Adjusted odds ratios (aOR), 95% confidence intervals (CI) and p-values were calculated with statistical significance level set at p-value < 0.05.