Setting: In the recent past, there have been reports of rising treatment failure rates for non-severe pneumonia. It is felt that World Health Organization (WHO) criteria for therapy failure are too sensitive and that many children are unnecessarily classified as failures. We studied alternative, less sensitive therapy failure criteria.
Methods: In this nested study we followed the clinical course of non-severe pneumonia in children aged 2-59 months using alternative therapy failure criteria. All children received amoxicillin and were followed up on days 3, 5 and 14 after enrollment. On day 3, children were labelled as therapy failure only if their condition had deteriorated. These failure rates were compared with those using WHO definitions.
Results: During the study period, 876 children with non-severe pneumonia were followed up until day 14. On day 3, using alternative therapy failure criteria, 31 (3.5%) children were labelled as therapy failure compared to 95 (10.8%) using current WHO criteria. The difference was statistically significant (P = 0.001).
Conclusions: The alternative therapy failure criteria work reasonably well, without causing any higher risk to children with non-severe pneumonia. Antibiotics should be changed only in those children who show signs of deterioration on day 3. This would prevent unnecessary changes in antibiotic treatment in many children.