Chloramphenicol in paediatrics: current prescribing practice and the need to monitor

Abstract

Two hundred and fifty-five neonates, infants and children, from 45 hospitals, who were receiving chloramphenicol therapy for serious infections were the subject of this study. Samples of serum and cerebrospinal fluid (CSF) were assayed for chloramphenicol and the patient's treatment regimens analysed. Less than 50% of neonates and 25% of infants received the “recommended” dose of chloramphenicol. In older children the recommended dose was used. Only 34% babies under 1 year of age and 50% older children had serum concentrations within the therapeutic range (15–25 mg/l). Thirty-one percent of neonates and infants had potentially toxic serum concentrations. Forty-three percent of neonates receiving chloramphenicol every 6h had subtherapeutic peak serum levels compared to 20% of those receiving the antibiotic every 12h. Concomitant administration of phenobarbitone or phenytoin had no effect on mean serum chloramphenicol levels. Serum concentrations of chloramphenicol were significantly higher in patients also receiving penicillin. CSF levels in 77 samples (39 patients) ranged from 1–60 mg/l. CSF from 44% patients contained less than 4 mg/l. Twelve neonates and infants (5.5%) suffered toxic side effects, four died. A further eight babies received an accidental 2- to 10-fold overdose and in three others an overdose was assumed following assay. No overdoses or toxic effects were reported in children over 1 year of age. Eight patients with impaired renal function had elevated serum levels and three showed toxic effects. In 22% patients dosage regimens were altered following assay. Even when the recommended dosage regimen for chloramphenicol is followed serum from all babies under 1 year of age should be assayed every 48–72 h if safe and effective levels are to be maintained.