O2070: Derivation and Validation of a Novel Risk Assessment Tool to Identify Young Children at Risk for Post-discharge Mortality in Dar Es Salaam, Tanzania and Monrovia, Liberia

Background: Rates of post-discharge mortality among children in sub-Saharan Africa are as high as 3-18% within months of hospital discharge versus < 1% among children in the US. Accurately identifying infants and young children at risk for post-discharge mortality is a crucial step towards reducing mortality in this period of vulnerability following a hospitalization. Our objective was to derive and internally validate a risk assessment tool to identify infants and children at risk for all-cause mortality within 60 days of discharge from the pediatric wards of two referral hospitals in Dar es Salaam, Tanzania and Monrovia, Liberia.

Methods: We performed a prospective observational cohort study of infants and children aged 1-59 months who were discharged from pediatric wards at Muhimbili National Hospital in Dar es Salaam, Tanzania and John F. Kennedy Medical Center in Monrovia, Liberia from October 2019 to January 2022. Caregivers of enrolled participants received phone calls at 7, 14, 30, 45, and 60 days after discharge to ascertain participants’ vital status. We collected socioeconomic, demographic, clinical, and anthropometric data for each enrolled participant during their hospital admission. We included candidate variables with a P value < 0.20 in bivariate analyses into a multivariable logistic regression model with best subset selection to identify risk factors for all-cause, 60-day post-discharge mortality. We selected the model with the lowest Akaike Information Criterion and Bayesian Information Criterion as the final model. We calculated adjusted log coefficients for each candidate variable and assigned weighted points to derive our risk assessment tool. We used bootstrap validation with 500 repetitions to internally validate our tool.

Results: There were 2,047 infants and children discharged. 1,933 (94.4%) enrolled and 1,895 (98.0%) had 60-day outcomes available. The median (interquartile range) age at discharge was 11 (4, 23) months and 1,135 (59.8%) were male. In total, 67 (3.5%) died within 60 days of discharge. While cancer, pedal edema at discharge, leaving against medical advice, and cyanosis were most predictive of post-discharge mortality, 10 variables contributed to our tool for a total possible score of 82 (Table). A score of ≥11 was the optimal cut point to identify infants and children at risk for post-discharge mortality. Our risk assessment tool demonstrated good discriminatory value (optimism corrected area under the receiver operating characteristic curve 0.77; Figure) and excellent calibration.

Conclusion: Our risk assessment tool to predict all-cause, 60-day post-discharge mortality among infants and children in Tanzania and Liberia had good discriminatory value. Risk assessment tools, such as this, must be externally validated before implementation. After validation, this tool may be used in sub-Saharan Africa to identify young children at risk for post-discharge mortality and direct focused resources and follow up for high-risk children.