Case management of childhood fevers in the community : : Exploring malaria and pneumonia care in Uganda

Abstract: Background: Acute respiratory infections (ARI), especially pneumonia, are leading causes of death in children under-five. Symptoms often overlap with those of malaria. In Uganda, the Home Based Management of fever (HBM) strategy recommends treating all febrile children with antimalarials provided by local community health workers (CHWs) - in Uganda called drug distributors (DDs). However, HBM overlooks the pneumonia symptom overlap, with potentially adverse effects for the affected children. Main aim: To explore aspects of home and community care for childhood fevers in Uganda and devise recommendations for integrated community based management of malaria and pneumonia. Methodology: Five sub-studies (I-V) were performed using a triangulation of qualitative (II& V) and quantitative (I, III, IV and V) methods in households (III& IV), communities (II, IV & V), health centres (I& IV) and a hospital (IV & V). Study I was cross-sectional in 14 health centres where 3,671 child consultation records were analysed for symptom overlap. Study II used 10 focus group discussions (FGDs) with mothers, fathers and grandparents. Study III was a cross-sectional household survey where mothers of 3,249 children were interviewed using 2 week recall. Study IV used case-series in the community, interviewing caretakers of 117 referred children and tracking the child in the outpatient records of nearby health facilities. Study V used performance assessment of 96 DDs in a hospital, 4 FGDs with mothers in the community and unstructured interviews with 2 key informants. Results: Thirty percent of children seen in health facilities (I) and 19% of sick children in the community (III) had symptoms compatible with both malaria and pneumonia. Some febrile conditions were perceived to require urgent allopathic treatment, and others were first treated with traditional remedies (II). Of children with cough and difficult/rapid breathing in the community, 35% were treated with antibiotics but when fever was present, antibiotic use dropped (p=0.12) and antimalarial use increased (p<0.001) (III). Among caretakers of children referred by DDs, 82% stated having completed referral but 52% had delayed >=2 days (IV). DD assessment of rapid breathing was adequate with 75% sensitivity and 83% specificity (V). Many biomedically relevant terms for ARI existed in the local language but most were related to fever and perceived to need antimalarial treatment (V). Discussion: Addressing only malaria in community management strategies may increase treatment delays for potential pneumonia. More comprehensive community management covering also pneumonia could potentially increase child survival. While management of pneumonia in the community involves dispensing of antibiotics by non-medically trained distributors, antibiotic drugs are already widely used in the community. With adequate training, supervision and support for DDs there is potential to rationalise antibiotic use while concurrently increasing access to treatment. To achieve high community uptake and minimise drug misuse, the local illness concepts and treatment actions need to be addressed. While DDs can operate to identify and treat children early in the disease, the high rate of referral completion demonstrates that also links to the formal health system can be maintained. With these findings, the feasibility and impact of full-scale implementation of integrated management of malaria and pneumonia in the community should be tested.