Antibiotic use in infants hospitalised with HIV-related pneumonia in Harare, Zimbabwe.

OBJECTIVE

To describe the clinical features of infants admitted with HIV-related pneumonia and to describe antibiotic use in relation to recommended treatment guidelines.

DESIGN

Case series.

SETTING

Paediatric medical wards of two University Teaching Hospitals, Parirenyatwa and Harare Central Hospitals.

SUBJECTS

100 infants aged one to 12 months admitted with HIV-related pneumonia.

MAIN OUTCOME MEASURES

Mortality and antibiotic use in the two hospitals.

METHODS

Records of 100 infants admitted for 48 hours or more with features of HIV-related pneumonia were analysed for clinical features and antibiotic use.

RESULTS

77% of patients were in the first six months of life with a peak age of two months and a median of four months (Q1 = 2, Q3 = 6). The median age of children admitted to Parirenyatwa hospital was 5.5 months (Q1 = 3, Q3 = 7) and in Harare hospital it was three months (Q1 = 2, Q3 = 6). The difference was statistically significant, p = 0.035. Fifty four percent of cases received penicillin, aminoglycoside and cotrimoxazole and overall only 30% of prescriptions complied with Essential Drug List of Zimbabwe (EDLIZ) recommendations for treatment of severe pneumonia in children with HIV infection. The overall mortality was 27.0%. The mortality in Harare Central Hospital was 40.4% and 15.7% in Parirenyatwa. The difference was statistically significant p = 0.005.

CONCLUSION

The difficulties in establishing the cause of the pneumonia in infants with HIV infection was a contributory factor to lack of adherence to standard treatment guidelines. In countries with a high prevalence of HIV infection and with limited resources, a clinical case definition for Pneumocystis carinii pneumonia (PCP) is required as a measure to provide treatment for infants with HIV related pneumonia which is evidence based. This approach will also promote rational antibiotic prescribing and will contain cost.