Pneumocystis carinii severe pneumonia among human immunodeficiency virus-infected children in Thailand: The effect of a primary prophylaxis strategy

Abstract

Background. A knowledge of the epidemiology of Pneumocystis carinii pneumonia (PCP) is important for the development of a strategy for primary PCP prophylaxis and empiric treatment for severe pneumonia in HIV-infected children. However, little is known about the epidemiology of PCP in developing countries. Objective. To measure the relative rate of PCP among hospitalized HIV-infected children with severe pneumonia in Bangkok and evaluate the effect of a strategy of primary PCP prophylaxis in HIV-exposed infants. Methods. All HIV-infected children hospitalized from January, 1996, to December, 1997, for severe pneumonia were investigated for PCP with the use of specimens obtained from bronchoalveolar lavage, endotracheal aspiration or lung tissue necropsy. Characteristics associated with severe pneumonia were described, and the differences between PCP and non-PCP in these severely ill children were analyzed. In June, 1996, a strategy of primary PCP prophylaxis using trimethoprim-sulfamethoxazole in all HIV- exposed infants from 1 to 6 month of age was initiated in our institution. The effect of this strategy was evaluated. Results. Of 279 hospitalized HIV- infected children 128 (46%) were diagnosed with pneumonia and 26 (20%) of these had severe pneumonia. P. carinii was identified in 9 (35%) children with severe pneumonia. After June, 1996, the rate of severe pneumonia among all hospitalized children decreased from 16% from January through June, 1996, to 7% from July, 1996, through December, 1997 (P = 0.02). Cases of PCP decreased from 9 in 1996 to zero in 1997. The percentage of HIV-infected children receiving PCP prophylaxis at the time of admission increased from 53% before June, 1996, to 72% in late 1997 (P = 0.04). The overall percentage of patients with severe pneumonia receiving PCP prophylaxis at the time of admission was 34%. Breakthrough PCP occurred in 2 children with poor compliance. Patients with PCP were significantly younger than those without PCP (mean age, 10.6 ± 10.6 vs. 29.8 ± 28.3 months, P = 0.02). Conclusion: PCP occurred in one-third of cases of severe pneumonia in HIV-infected children in Bangkok. The data suggest that PCP prophylaxis can prevent both PCP and non-PCP.