The efficacy of prophylactic antibiotics in the management of children with kerosene-associated pneumonitis: a double-blind randomised controlled trial

Context. Hydrocarbons, especially kerosene (paraffin), are the most common agents causing childhood poisoning in low and middle income countries (LMICs). Aspiration of kerosene causes an inflammatory sterile chemical pneumonitis, which may increase susceptibility to secondary lower respiratory tract bacterial infection. This study aimed to assess the efficacy of prophylactic antibiotics in the management of kerosene-associated pneumonitis in children and to identify risk factors associated with severity or outcome. Methods. A double-blind placebo-controlled trial of prophylactic antibiotics in the management of kerosene-associated pneumonitis of children presenting to a referral hospital was performed from July 2010 to September 2011. Sequential children with a history of kerosene ingestion and mild respiratory illness were randomised to receive placebo or amoxicillin. Each child was followed-up at Day 3 and Day 5 post-ingestion. The primary outcome measure was the number of treatment failures in each group, defined as any child who deteriorated within this time, necessitating a change in treatment regimen. Secondary outcome measures were length of hospital stay and symptoms and signs at follow-up. Results. Seventy-four patients were enrolled. Thirty-five (47%) received placebo and 39 (53%) active treatment. There was no significant difference in treatment failures between placebo (3/35, 9%; 95% CI, 3-22) and active (2/39, 5%; 95% CI, 1-17) groups (relative risk, 0.60; 95% CI, 0.11-3.37). The median length of hospital stay was identical (placebo 0.5 days; IQR, 0-1.0 and active 0.5 days; IQR, 0.5-1.0). Symptoms and signs at Days 3 and 5 post-ingestion were similar. The only significant risk factor for treatment failure was residence in formal housing. Clinical severity at presentation was similar for treatment successes and failures. Conclusion. Prophylactic antibiotics do not improve the outcome in children with mild respiratory illness after kerosene ingestion.