Predictors of recovery time for anaemic children under five years

Ssuuna James (
Statistics and Applied Economics, Makerere University
June, 2015
I am a statistician and a professional teacher with strong interest in applied statistics/statistical modeling.


The main objective of the study was to investigate the time to recovery of anaemic children under five years. The study used data from Kitovu hospital and Masaka regional referral hospital. Children that were admitted in the two hospitals between June 2011 and June 2013 and had been on treatment for at least 1 day and at most 60 days were followed to determine the possible predictors of their recovery time.

To estimate the recovery function for every sub group of each variable, the Kaplan-Meier Product-Limit estimates were used and the Log-Rank Chi-square to compare the recovery functions across several groups. Finally, the Accelerated Failure Time (AFT) model was considered as an alternative to the Proportional Hazards Model when examining the joint effect of the predictors on the time of recovery, assuming a Weibull distribution for the time variable.

The results of the study revealed that recovery time highly increased among children with HIV (56%) as compared to those without HIV and reduced among children admitted to Kitovu hospital (7%) as compared to those admitted to Masaka regional referral hospital. Having STHIs and pneumonia increased the recovery time of children by 13% and 16%, respectively. Children with either moderate or severe acute malnutrition and those with moderate or severe anaemia had their recovery time increased by 43 or 40% and 11 or 19%, respectively.

Therefore, emphasis should be on development of an integrated multisectorial approach which mainly involves effective management of the major etiologies of anaemia and co-morbidities among anaemic children in order to combat the public health problem of anaemia.

: Predictors, Recovery time,
and Anaemic.