Improvements in nutritional management as a determinant of reduced mortality from community-acquired lower respiratory tract infection in hospitalized children from rural central Africa

Background: In-hospital mortality from lower respiratory tract infections (LTRI) is unacceptably high in developing countries where LTRI are still a leading cause of death. Objective: To identify new approaches to reduce in-hospital mortality of LRTI through the improvement of its management. Methods: The prospectively collected database of children admitted during an 11-year period with LRTI in a pediatric rural hospital in Central Africa was reviewed to determine the predictors of death and to evaluate the impact on mortality of 4 different protocols for the management of malnutrition. Results: During the study period, 859 children were admitted with a nonmeasles severe LRTI. In the 3-year period during which blood cultures were obtained, 29.0% of the children with LRTI were bacteremic, and multiresistant Enterobacteriaceae were recovered in 81.4% of positive blood cultures. Independent predictors of death in children without edema were age <24 months, dehydration and hepatomegaly with adjusted odds ratios (numbers in parentheses, 95% confidence interval) of 3.47 (1.70-7.08), 4.24 (2.11-8.50) and 2.90 (1.43-5.85), respectively. In those with edema, a significantly increased risk of death was noted for girls [4.31 (1.71-10.90)], in children with z-score of weight to height less than or equal to -3 [5.45 (1.67-17.79)] and when the serum albumin was <16 g/l [2.58 (1.01-6.58)]. The improvement in the management of malnutrition was followed by a reduction of LRTI-related mortality in children with edema from 32.4 to 8.9% (P < 0.001). In children without edema, the LRTI-related mortality decreased from similar to 12% to 3.5% when their diet was supplemented with micronutrients. Conclusion: This study indicates that the improvement of the management of underlying nutritional deficiencies is crucial for the reduction of the high in-hospital case fatality rate associated with severe nonmeasles LRTI. The empiric antibiotic regimen should be modified to cover for multiresistant Enterobacteriaceae.