Consequences of post-discharge hospitalisation on the growth of young Bangladeshi children hospitalised with diarrhoea: a secondary case-control analysis of Antibiotics for Children with Diarrhea (ABCD) trial.

Background: Due to the scarcity of published data on growth among children with severe diarrhoea requiring readmission during post-discharge follow-up, we aimed to investigate the potential impact of post-discharge readmission at day-90 follow-up on growth in diarrheal children aged 2-23 months.

Methods: We performed a secondary analysis using Bangladesh site data from the Antibiotic for Children with Diarrhea (ABCD) trial, a multi-country, randomised, double-blind, placebo-controlled study conducted from July 2017 to July 2019. Children aged 2-23 months who had severe diarrhoea defined as having acute diarrhoea with some/severe dehydration, or severe stunting, or moderate wasting, were admitted to the facility. In this analysis, we classified children who were re-hospitalised within a 90-day post-discharge follow-up period as cases and randomly selected controls who did not require re-hospitalisation, matching them by similar ages and sexes in a 1:3 ratio. We gathered anthropometric data on enrolment and day 90 follow-up. The outcome variables were changes in nutritional indicators height-for-age (ΔHAZ), weight-for-age (ΔWAZ), weight-for-height (ΔWHZ), and mid-upper arm circumference (ΔMUAC). We assessed for growth changes at day 90 post-discharge follow-up using multivariate linear regression.

Results: Among 1431 diarrhoeal children enrolled, we identified 145 cases and 435 controls. In terms of the baseline admission characteristics, the cases were less likely to be immunised (81% vs. 72%; P = 0.031), vomit (11% vs. 22%; P = 0.001), and have dehydrating diarrhoea (26% vs. 36%; P = 0.026) than the controls. After adjusting for potential covariates, the cases had a significant reduction in growth than the controls at 90 days of post-discharge follow-up, according to anthropometric indices: ΔHAZ (β = -0.11; 95% confidence interval (CI) = -0.21, -0.01; P = 0.029), ΔWAZ (β = -0.24; 95% CI = -0.35, -0.14; P < 0.001), ΔWHZ (β = -0.25; 95% CI = -0.39, -0.12; P < 0.001), and ΔMUAC (for children 6-23 months, β = -0.17; 95% CI = -0.29, -0.04; P = 0.011).

Conclusions: Diarrhoeal children aged 2-23 months requiring readmission during the 90-day post-discharge follow-up period had a significant deterioration of ponderal and linear growth, compared with those who did not require readmission. This finding underscores the importance of early identification of children with risks of post-discharge readmission and designing a package of post-discharge trials, including social and nutritional interventions that may help to reduce post-discharge readmissions as well as subsequent growth faltering.