Impact of home-based management of fever on progression to severe illness in children under five years undergoing treatment for malaria at Kericho District Hospital, Kenya

Abstract

Background: Mortality in children under five years of age is contributed substantially by - malaria and 75% of estimated malaria deaths occur in African children. Home-based management of fever is common as the first step in treatment of malaria. It has been adopted as one among strategies for Roll-Back Malaria (RBM) by the WHO through training of caregivers at community level to enhance prompt and correct use of antimalarial drugs. Objective: To establish the impact of home-based management in progression of fever to severe illness in children under 5 years undergoing treatment for malaria. Method: A case-control study was conducted between so" July and 31st August 2007 at Kericho District Hospital in Rift Valley Province on children under five years undergoing treatment for malaria. Diagnosis was based on National Guidelines for Diagnosis, Treatment and Prevention of Malaria for Health Workers in Kenya (MOH, 2006) and for purpose of this study both confirmed and probable cases of malaria were included in the study. Cases were children seeking inpatient treatment for malaria whereas controls consisted of febrile children treated for malaria as - outpatients at the hospital. Mothers for both cases and controls were interviewed using a standard questionnaire collecting data on home-based management characteristics. Results: 45 children (22 cases and 23 controls), mean age of 22.8 months, diagnosed with malaria were consecutively enrolled. Out of all the children that were studied, 55.6% received home-based management of fever. None of home-based management reported having administered antimalarial drugs to their febrile childrep but 73.1% used analgesics (panadol and paracetamol). The practice of home-based management was detected in 31.8% and 78.3% of cases and controls respectively with an odds ratio of 0.13, P=O.OOand indicating that it was not associated with increased risk of severe illness but on the contrary was protective against severe illness. Majority (61.1%) of the severely ill children were treated at the dispensary or private clinic before being brought to the hospital while none of the children with mild illness reported the same. Treatment in the dispensary/private clinic seemed to have caused delay in coming to the hospital, 113.5 hours. after detection of illness in the children compared to 70.4 hours. P= 0.12. Due to the small numbers of controls that were treated in dispensary or private clinic, it was impossible to obtain odds ratio associated to it; consequently, though a possible confounding factor, was not controlled for in obtaining association between home-based management of fever and severe illness. In addition, duration of illness, place of residence and presence of training in prevention of malaria were not significantly different between the cases and the controls. Though statistically insignificant, place of residence and presence of training in prevention of malaria increased the risk by two fold (OR= 2.8 and 2.5 respectively). Furthermore, duration of illness. place of residence and presence of training in prevention of malaria that were possible independent risk factors did not need to be controlled for because of their lack of association with severe illness. Conclusions: The practice of home-based management of fever in children under 5 years was prevalent and it reduces the likelihood of progression from mild fever to severe illness in malaria. This may appear as a spurious finding given that the type of medications administered during home-based management did not include antimalarial drugs. Treatment in dispensary or private clinic that resulted in delay among the cases is a factor to investigate given that none of the children treated with mild illness reported having utilized it hence resulting to an infinitely high odds ratio in its association with severe illness. Place of residence, though not significantly associated, cannot be ruled out as a contributing factor and requires further investigation as well.