Aetiology and outcome of severe community-acquired pneumonia in children admitted to a paediatric intensive care unit.

OBJECTIVE: To determine the aetiological agents and outcome of severe community-acquired pneumonia (SCAP) in children admitted to the paediatric intensive care unit (PICU) at Kalafong Hospital, Pretoria. PATIENTS AND METHODS: An audit was done after a protocol was implemented to identify the aetiological agents in children with life-threatening SCAP admitted to the PICU from the emergency room. The following investigations were done as per protocol: blood culture, culture of the tracheal aspirate, immunofluorescence and culture of the nasopharyngeal aspirate, microscopy and culture of the gastric juice for Mycobacterium tuberculosis, and determination of HIV status. The following data, documented prospectively, were obtained from patient records: date of admission, age, gender, weight, duration of ventilation, duration of stay in the PICU, survival or death, and severity of illness as determined by means of the score for acute neonatal physiology (SNAP) or paediatric risk of mortality (PRISM) score depending on the child's age. RESULTS: Twenty-three children were admitted over a 1-year period (1 November 1994-30 October 1995). Their median age was 10 weeks (range 2 weeks-5 years) and the sex distribution was equal. Two children were HIV-infected. Twenty children received mechanical ventilation for a median period of 6.5 days (range 2-16 days). Aetiological agents were identified in 15/23 children (65%). Respiratory syncytial virus (RSV) was the most common pathogen, identified in 7/23 children, Klebsiella pneumoniae was the most common bacterial pathogen, identified in 5 children (2 blood cultures and 3 tracheal aspirates). Tuberculosis was not diagnosed. The mean PRISM score was similar in survivors and children who died. The case fatality rate was 30%. The 7 children who died had a median arterial oxygen tension/fraction of inspired oxygen (PaO2/FiO2) ratio of 94 (range 32-111) and the 16 survivors had a median ratio of 146 (range 51-252) (P = 0.01) on admission. Both HIV-infected children died and postmortem examination showed a pneumonia due to Pneumocystis carinii and cytomegalovirus. CONCLUSIONS: SCAP occurs in very young children. One or more pathogens were isolated in 65% of cases. Viral pathogens predominated, with RSV being the most common. The yield of positive blood cultures was low at 17%. Streptococcus pneumoniae and Haemophilus influenzae were not found. The case fatality rate was 30% and death was more likely with a low PaO2/FiO2 ratio on admission.