Mortality and Morbidity trends in under-five children with severe acute malnutrition at the University Teaching Hospital(UTH), Lusaka, Zambia

Abstract

Introduction: Severe acute malnutrition has continued to cause high disease burden in Zambia. Mortality rates at the University Teaching Hospital (UTH) were as high as up to 50% in 2008. There is paucity of published data on mortality and morbidity trends in under-five children with SAM at UTH. This study aimed to determine mortality and morbidity trends in under-five children with severe acute malnutrition at UTH. Methods: A retrospective cross-sectional study, of all children admitted to ward A07 from January 2009 to December 2013. The study was done to assess trends in mortality, and how they are affected by morbidity, co-morbidity, and length of stay. A total of 9540 under-five children were analysed. The median and inter quartile ranges (IQR) were used to summarize and describe the data, as the data was not normally distributed. The Chi-square test was used to test the difference between the samples of the categorical variables sex, morbidity, co-morbidity and residence. In addition Kruskal Wallis was used to compare if means of the numerical variables length of stay on the ward and age for the different mortality, morbidity, and co-morbidity groups were different. Logistic regression was used to predict the odds of death in the independent variables and to control for confounding since the dependent variable mortality was categorical. Cuzick a non parametric test for trends was used to test mortality and morbidity trends. Furthermore survival analysis was done using Wilcoxon and Cox proportion hazard regression to test of difference of survival for the different groups and illustrated using Kaplan-Meier curves. Results: Overall, 45.9% (4,386) were females and 53.9% males, median age was 17 months (IQR 12 to 22 months), and median length of stay was 8 days, (IQR of 3 to 14). The overall mortality rate was 46.7% with overall prevalence of HIV at 32.2% (2,589). About 62% (5,609) had Kwashiorkor 21.6% had Marasmus and 16.4% had marasmic-Kwashiorkor. Children with Marasmus had the highest prevalence of HIV infection at 40.1% (703). Kwashiorkor was associated with high prevalence of Anaemia 13.2% (121) while, Marasmic-Kwashiorkor had the highest prevalence of Diarrhoea (33.6%). Children with TB had the longest length of stay on the ward (14 days) and those with Septicaemia the shortest stay (4 days). Children who had co-morbidity were 50% more likely to die than those without co-morbidity. Of the children that xiii died, those who were HIV infected were 4 times more likely to die and children with Septicaemia were 3.8 times more likely to die than those with Anaemia respectively. Kaplan Meier survival curves showed also that children with Marasmus, Diarrhoea, Septicaemia, tuberculosis, and those who were HIV infected had reduced survival rates. Mortality and morbidity trends decreased with admissions (from 2009 to 2013) and both trends were significant at p=0.000. Conclusion: Declining mortality and morbidity in children with acute malnutrition at UTH may suggest improved management practices. Nonetheless, limitations to totally prevent malnutrition may be an indicator of complex structural challenges that may be existent in this population thereby needing matching and complex intervention.