Post-discharge morbidity and mortality in children admitted with severe anaemia and other health conditions in malaria-endemic settings in Africa: a systematic review and meta-analysis

Abstract

Background: Severe anaemia is associated with high in-hospital mortality among young children. In malaria-endemic areas, surviving children also remain at increased risk of mortality for several months after hospital discharge. We aimed to compare the risks of morbidity and mortality among children discharged from hospital after recovery from severe anaemia versus other health conditions in malaria-endemic settings in Africa.

Methods: Following PRISMA guidelines, we searched PubMed, Scopus, Web of Science, and Cochrane Central from inception to Nov 30, 2021, without language restrictions, for prospective or retrospective cohort studies and randomised controlled trials that followed up children younger than 15 years for defined periods after hospital discharge in malaria-endemic countries in Africa. We excluded the intervention groups in trials and studies or subgroups involving children with sickle cell anaemia, malignancies, or surgery or trauma, or those reporting follow-up data that were combined with the in-hospital period. Two independent reviewers extracted the data and assessed the quality and risk of bias using the Newcastle Ottawa Scale or the Cochrane Collaboration's tool. The coprimary outcomes were all-cause death and all-cause readmissions 6 months after discharge. This study is registered with PROSPERO, CRD42017079282.

Findings: Of 2930 articles identified in our search, 27 studies were included. For children who were recently discharged following hospital admission with severe anaemia, all-cause mortality by 6 months was higher than during the in-hospital period (n=5 studies; Mantel-Haenszel odds ratio 1·72, 95% CI 1·22-2·44; p=0·0020; I²=51·5%) and more than two times higher than children previously admitted without severe anaemia (n=4 studies; relative risk [RR] 2·69, 95% CI 1·59-4·53; p<0·0001; I²=69·2%). Readmissions within 6 months of discharge were also more common in children admitted with severe anaemia than in children admitted with other conditions (n=1 study; RR 3·05, 1·12-8·35; p<0·0001). Children admitted with severe acute malnutrition (regardless of severe anaemia) also had a higher 6-month mortality after discharge than those admitted for other reasons (n=2 studies; RR=3·12, 2·02-4·68; p<0·0001; I²=54·7%). Other predictors of mortality after discharge included discharge against medical advice, HIV, bacteraemia, and hypoxia.

Interpretation: In malaria-endemic settings in Africa, children admitted to hospital with severe anaemia and severe acute malnutrition are at increased risk of mortality in the first 6 months after discharge compared with children admitted with other health conditions. Improved strategies are needed for the management of these high-risk groups during the period after discharge.

Funding: Research Council of Norway and US Centers for Disease Control and Prevention.

Figure 1:. PRISMA flow diagram

Reasons for exclusion exceed the number of articles excluded because some articles had more than one reason for exclusion. Children in IPD represents admitted children. IPD=in-patient department. OPD=outpatient department. *Including a prospective follow-up study of children who were initially enrolled in a prospective case-control study.

Figure 2:. Crude mortality risk for any follow-up period in the first 6 months after discharge, all studies

Data from our unpublished study, Zucker 1996, and Opoka, 2016 are included in more than one subgroup, each representing a mutually exclusive group. The malaria studies did not include any post-discharge chemoprophylaxis groups. Summary statistics are not shown because of considerable heterogeneity (overall I² is 95·5%) between and within admission health-condition groups.

Figure 3:. Relative risk of mortality 6 months after discharge among studies reporting results by several health conditions

Weights are from the random-effects analysis. The figure only includes studies that reported enough detail to allow direct comparisons of mortality after discharge by health condition among children from the same cohort study. Diamond shapes depict pooled effect size. The crude RR is calculated by comparing the exposure condition versus other conditions. For example, in the first section under severe anaemia, the random-effects summary crude RR of 2·69 (95% CI 1·59–4·53) represents the relative risk of mortality after discharge when comparing children who were recently admitted with severe anaemia versus all other children that were recently admitted for any other conditions that excluded severe anaemia (other conditions), such as severe acute malnutrition, severe malaria, severe pneumonia, or other unspecified conditions. Similarly, the second section under malnutrition (summary random-effects RR 3·12) represents the RR of mortality after discharge when comparing children who were recently admitted with severe acute malnutrition versus children that were admitted for any other conditions that excluded severe acute malnutrition. In the case of children that were admitted for any other conditions, other conditions includes children with severe anaemia, severe malaria, severe pneumonia, or other unspecified conditions. D + L=DerSimonian and Laird random effects. M-H=Mantel-Haenszel fixed effect. RR=relative risk.

Figure 4:. In-patient mortality versus mortality after discharge within 6 months by health condition on admission

For each study, the MHOR was obtained by comparing the number of deaths after discharge versus in-patient deaths during the initial hospital admission. D + L=DerSimonian and Laird random effects. I-V=inverse-variance fixed effect. MHOR=Mantel-Haenszel odds ratio.