Cohort profile: The WHO Child Mortality Risk Stratification Multi-Country Pooled Cohort (WHO-CMRS) to identify predictors of mortality through early childhood

Abstract

ABSTRACTPurposeTo provide details of a pooled dataset that will be used to estimate absolute and relative mortality risks and other outcomes among children less than 59 months of age and the predictive performance of common risk exposures, both individually and in combination.ParticipantsChildren from birth to five years of age recruited at health facilities or community settings into 33 longitudinal observational or intervention studies in 18 low-and middle-income countries.Findings to dateThe dataset includes 75,287 children with a median age of 3 months (IQR 1, 12) at first measurement. In the pooled sample, 2,805 (3.7%) of the study children died. Data on birthweight was recorded in 18 studies, and gestational age in 13 studies. Among these, 14% of the included children were reported with low birthweight and 14% preterm birth. At first measurement, 33% of the children were stunted, 24% were wasted, and 35% underweight. 13% and 7% of caregivers reported that their child had acute diarrhoea or acute lower respiratory tract infection before the study visit, respectively. The proportion of children being breastfed at any study visit decreased from 99% at age <6 months to 77% in the age group 12-23 months. Child characteristics differed considerably between studies in the community and health care settings. The median study period was 15 months (IQR 7.6 to 18.4 months).Future plansThe WHO Child Mortality Risk Stratification Multi-Country Pooled Cohort (WHO-CMRS) provides a large dataset including child, parental, and household characteristics from a diverse range of geographical, community and health system settings; planned analyses will examine knowledge gaps with the aim of informing global guidelines and their derivatives such as clinical management tools and implementation guidance, and to inform future research agendas.STRENGTHS AND LIMITATIONSStrengths– A large dataset that integrates diverse geographical areas in LMICs, with children recruited from community and health system settings.– Prospective data collection and availability of key time-varying exposures, such as nutritional status, common childhood morbidities, and feeding practices and non-time varying exposures, such as birth characteristics and socio-demographics.– Ability to estimate age-specific absolute mortality risks from different exposure domains and combinations thereof.Limitations– Non-systematic selection of included datasets.– Heterogeneity of inclusion criteria, exposures collected and follow-up schedules across studies.