Covariates PS balance before and after weighting.

Background
The prevalence of Short Birth Interval (SBI) is higher in Low- and Middle-Income countries (LMICs), including Bangladesh. Previous studies in LMICs have estimated the effects of SBI on child mortality by comparing two unequal groups of mothers based on their socio-economic status. This approach may lead to overestimation or underestimation of the true effect of birth interval on child mortality, particularly when sample sizes are relatively small.
Objective
We determined the effects of SBI on several forms of child mortality in Bangladesh by comparing two equal groups created by applying the propensity score matching technique.
Methods
This study analyzed data from 5,941 mothers and 1,594 health facilities extracted from the 2017/18 Bangladesh Demographic and Health Survey and the 2017 Bangladesh Health Facility Survey. The exposure variable was SBI (defined as the interval between two subsequent births <33 months: yes, no), while the outcome variables were neonatal mortality (defined as mortality within 28 days of birth: yes, no), infant mortality (defined as mortality within 1 year of birth: yes, no), and under-five mortality (defined as mortality within 5 years of birth: yes, no). Multilevel Poisson regression based on inverse probability treatment weights was used to determine the association between exposure and outcome variables.
Results
The prevalence rates of neonatal, infant, and under-five mortality were 48.8, 30.8, and 23.1 per 1000 live births, respectively. Newborns of SBI mothers were found to have a 63% higher likelihood of neonatal mortality (aPR, 1.63; 95% CI, 1.08–2.46) compared to newborns of non-SBI mothers. Furthermore, the prevalence of infant mortality and under-five mortality was 1.45 times higher (aPR, 1.45; 95% CI, 1.01–2.08) and 2.82 times higher (aPR, 2.82; 95% CI, 2.16–3.70), respectively, among babies born in a short interval of their immediately preceding sibling as compared to babies born in a normal interval of their immediately preceding sibling.
Conclusions
Findings of this study indicate that SBI is an important predictor of child mortality. Consequently, around 1 million children born in a short interval every year in Bangladesh are at risk of dying before reaching their fifth birthday. This indicates a challenge for Bangladesh to achieve the SDG 3 target to reduce neonatal and under-five mortality to 12 and 25 deaths per 1000 live births, respectively. Hence, awareness-building programs about the adverse effects of SBI and strengthening existing healthcare facilities are important.