Objective: Acute severe pneumonia with respiratory failure in human immunodeficiency virus-infected and -exposed infants carries a high mortality. Pneumocystis jiroveci is one cause, but other organisms have been suggested to play a role. Our objective is to describe the coinfections and treatment strategies in a cohort of human immunodeficiency virus-infected and -exposed infants with respiratory failure and acute respiratory distress syndrome, in an attempt to improve survival.
Design: Prospective intervention study.
Setting: Steve Biko Academic Hospital, Pretoria, South Africa.
Patients: Human immunodeficiency virus–exposed infants with respiratory failure and acute respiratory distress syndrome were recruited into the study.
Interventions: All infants were treated with routine therapy for Pneumocystis jiroveci and bacterial coinfection. However, in addition, all infants received ganciclovir from admission until the cytomegalovirus viral load result was demonstrated to be <log 4.
Measurements: Routine investigations included human immunodeficiency virus polymerase chain reaction, cytomegalovirus viral load, blood culture, C-reactive protein, and white cell count. Tracheal aspirates for Pneumocystis jiroveci detection, bacterial culture, tuberculosis culture, and viral identification were performed.
Main Results: Sixty-three patients met the recruitment criteria. The mortality rate was 30%. Pneumocystis jiroveci was positive in 33% of infants, while 38% had cytomegalovirus viral load ≥log 4. Only 7.9% of infants had a positive tuberculosis culture. Nineteen deaths occurred, 13 of which had a cytomegalovirus viral load ≥log 4. Bacterial coinfection and CD4 count were not predictors of mortality.
Conclusions: A case fatality rate of 30% is achievable if severe pneumonia with respiratory failure and acute respiratory distress syndrome is managed with a combination of antibiotics and ventilation strategies. Cytomegalovirus infection appears to be associated with an increased risk of death in this syndrome. This may, however, be a marker of as yet undefined pathology.