The etiology and outcome of pneumonia in human immunodeficiency virus-infected children admitted to intensive care in a developing country.

In developing countries, many human immunodeficiency virus (HIV)-infected children require intensive care unit (ICU) resources for pneumonia, but there is little information on the etiology of pneumonia or the impact of ICU intervention.

OBJECTIVE

To compare the etiology and outcome of pneumonia in HIV-positive and seronegative children admitted to ICU.

DESIGN

Prospective study.

SETTING

Two pediatric ICUs linked to the University of Cape Town, South Africa. PATIENTS: Consecutive children admitted for pneumonia during 1998. MEASUREMENTS AND MAIN

RESULTS

Clinical, demographic, ventilatory, and laboratory data were collected. Blood for testing was obtained. Induced sputum or nondirected bronchoalveolar lavage was performed for culture and Pneumocystis carinii identification; gastric lavage (GL) provided specimens for mycobacterial culture. Seventy-six children (21 [27.6%, 95% confidence interval {CI} = 18-39.1] HIV-positive) were enrolled. At admission, HIV infection was diagnosed in 15 of the 21 (71.4% [47.8-88.7]) HIV-positive patients. P. carinii pneumonia occurred in eight HIV-positive children (38% of HIV-infected patients) and one HIV-negative child. It was the acquired immunodeficiency syndrome (AIDS)-defining illness in seven children (47%). The incidence of bacteremia (15.3%) was similar in HIV-positive (15.8%) and HIV-negative children (15.1%), p =.94; Streptococcus pneumoniae and Staphylococcus aureus were the predominant isolates. Bacterial and viral isolates from sputum or bronchoalveolar lavage, including Mycobacterium tuberculosis in six (8%) children, did not differ by HIV status. Intermittent positive pressure ventilation was used in 8 of 21 (38%) HIV-positive children and 28 of 55 (51%) HIV-negative children, p =.32. Median days of intermittent positive pressure ventilation (3 [2-6]), ICU (5 [3-9.5]), and hospital (11 [7.5-19]) did not vary by HIV status. The in-hospital mortality rate for HIV-positive children (6 of 21 [28.6%]) was double that for seronegative patients (8 of 55[14.5%], relative risk [RR] 1.96 [0.77-4.99], p =.16).

CONCLUSION

More than a quarter of children admitted to ICU for pneumonia in this geographic area are HIV-positive; most are diagnosed with HIV at admission. P. carinii pneumonia is a common AIDS indicator disease. HIV-infected children admitted with pneumonia had a worse outcome than seronegative children, a difference that is rendered statistically insignificant by the small sample size.