Mortality among low birth weight neonates after hospital discharge in a resource-limited setting

Published: 31 March 2022| Version 1 | DOI: 10.17632/75rjpsfy96.1
Ronald Nsubuga


This was a prospective cohort study of 220 participants aimed at determining mortality and its predictors among low birth weight neonates after discharge from Mulago National Referral Hospital, in Uganda. Captured data were entered into Epidata version 3.1. During the study period, 4 neonates were lost to follow-up, 6.5% of the 216 that completed follow-up died. Mortality was highest among those who were had discharge weight < 1200g. The predictors of mortality were discharge weight <1200g (AHR 23.47, 95% CI: 4.34-126.8), 5-minute Apgar score <7 (AHR 4.25, 95% CI: 1.31-13.8) and a diagnosis of neonatal sepsis during admission (AHR 7.93, 95% CI: 1.67-37.6). The causes of death were presumed neonatal sepsis (71.4%), suspected aspiration pneumonia (14.3%) and suspected cot death (14.3%). Majority of deaths occurred from home (54.1%). This data was


Steps to reproduce

Study population All neonates discharged from the Special Care Baby Unit at Mulago National Referral Hospital whose birth weight was <2500g. Sample size Sample size was estimated using an OpenEpi formula for frequency in a population . Since there were no similar studies addressing post-discharge mortality among LBW neonates, we used an anticipated frequency of 50%. Sample size (n) for 95% Confidence Interval of 95% was 200. Assuming a 10% loss to follow up, the final sample size (n)= 200 + (200*10%) = 220. Procedure All neonates with birth weight <2500g were identified by a research assistant at discharge. Their parents were then introduced to the study. For neonates who met the inclusion criteria, the research assistant obtained written consent after providing details of the study to them. Recruitment was conducted daily until the desired sample size was achieved. Recruited neonates received a study identification number and their details were captured in a logbook. Each parent submitted 3 phone contacts. Phone contacts of the principal investigator (PI) and research assistants were shared with parents for them to call in case of any concerns. Data were captured at discharge, and at every clinic visit. Study participants were followed up for 28 days. Parents of study participants were contacted by research assistants via phone once a week and significant events (death, illness) were registered. Those who missed their scheduled follow-up visits were contacted by phone and requested to bring the neonate for review on the next clinic day. A neonate who missed two consecutive clinic visits and whose parents were not reachable by phone throughout the study period were considered loss to follow up. Data management and analysis Captured data were checked for accuracy and completeness by the principal investigator. Data sheets were kept in a lockable cabinet. Captured data were entered into Epidata version 3.1, and exported to STATA Version 14 for analysis.


Makerere University College of Health Sciences


Neonates, Mortality, Low-Birth Weight Birth