Trends in and Factors Associated With Maternal Autonomy and Its Role in Health Services Utilization and Neonatal/infant Deaths in Bangladesh

Abstract

Introduction: The neonatal/infant mortality rate is high in most low- and middle-income countries, including Bangladesh. Maternal autonomy or decision-making ability improves health care utilization, and adequate health care utilization reduces neonatal/infant deaths. Previous studies reported determinants of neonatal/infant deaths in Bangladesh, but the role of sociocultural factors like maternal autonomy is unclear. Examining the trends (i.e., changes in prevalence) in and factors affecting maternal autonomy, and its role in neonatal/infant deaths could inform future research, programs, or policies to reduce neonatal/infant deaths in Bangladesh. The aims of the study are three-fold: (1) investigate trends in the prevalence of and factors affecting maternal autonomy; (2) identify the association of maternal autonomy with neonatal/infant deaths; and (3) investigate whether health care utilization mediates the association between maternal autonomy and neonatal/infant deaths.

Methods: To investigate Specific Aim 1, data from Bangladesh Demographic and Health Survey (BDHS) 1999-00, 2004, 2007, 2011, 2014, and 2017-18 were analyzed. To examine Specific Aims 2 and 3, BDHS 2017-18 data were analyzed. Maternal autonomy (i.e., exposure variable) was defined as decision-making ability about own health care, large household purchases, and freedom of mobility. To describe prevalence, these three decisions were categorized as no, low, and high with decision-making ability about no, 1-2, and 3 decisions, respectively. Maternal autonomy was also dichotomized as the presence of at least 1 autonomy to investigate the association. Neonatal and infant deaths were defined as deaths within the first one month and one year, respectively. Three health care utilization variables were tested as mediators: at least four antenatal care (ANC) visits, hospital birth, and neonatal post-natal care (PNC). For all aims, the distributions of study samples were compared per exposure, mediators, and outcomes. The prevalence of maternal autonomy was reported, its trends were tested with Cochrane-Armitage test, and the factors associated with maternal autonomy were identified using multilevel logistic regression. Then, the as-sociations of maternal autonomy with neonatal and infant deaths were tested using multi-level logistic regression. Finally, to test mediation, ‘maternal autonomy’s relationships with health care utilization’ and ‘health care utilization’s relationship with neonatal and infant deaths were examined using multilevel logistic regression. Prevalence odds ratios (PORs) were reported with 95% confidence intervals (CIs).

Results: The prevalence of maternal autonomy for all 3 decisions increased from 36.4% (95% CI: 34.8 to 38.1) in 1999-00 to 53.7% (95% CI: 52.1 to 55.2) in 2017-18. The prevalence of no autonomy declined from 27.9% (95% CI: 26.4 to 29.5) to 16.2% (95% CI: 15.1 to 17.3) during that period. These changes were observed regardless of decisions and most background characteristics of the participants. Overall, in the adjusted analysis, mothers with older age, higher education, and employment were more likely to have autonomy than their counterparts (POR > 1, p-value <0.05). Among mothers with no and any autonomy, the neonatal mortality rate was 30.4 (95% CI:19.0 to 48.1) and 15.0 (95% CI: 11.6 to 19.4) per 1000 live births; and infant mortality rate was 32.8 (95% CI: 20.8 to 51.3) and 19.3 (95% CI: 15.4 to 24.2) per 1000 live births, respectively. Compared to mothers with no autonomy, those with any autonomy (adjusted POR: 0.50, 95% CI: 0.28 to 0.88, p = 0.016) had lower odds of neonatal deaths. Similarly, the odds of infant mortality was lower among mothers with any autonomy (adjusted POR: 0.59, 95% CI: 0.35 to 0.99, p = 0.049). Maternal autonomy and neonatal/infant deaths did not have significant association with adequate health care utilization. Therefore, the mediational impact of healthcare utilization was not tested (p>0.05).

Conclusion: During the past two decades, although the level of maternal autonomy has increased in Bangladesh, a significant proportion of mothers do not have autonomy. Expanding educational and earning opportunities may increase maternal autonomy in this country. Considering the association of maternal autonomy with neonatal/infant deaths, more research should be carried out to understand the association of other factors (e.g., socioeconomic factors) that may impact the association and to understand the perspectives of husbands, in-laws, and other family members on the role of autonomy.