Review scope

Studies were selected which compared the use of antipyretics with no antipyretics in children with acute febrile infectious illness. Outcomes included author-defined time to fever resolution, and infection-specific indicators of disease resolution.

Review methods

Medline and Embase (to November 2012) were searched for randomised controlled trials (RCTs) and quasi RCTs. Language restrictions were not specified. Quality of studies was assessed using the Cochrane risk of bias tool. Reference lists and Google Scholar were hand searched. Screening and searching was performed by two authors independently. Six studies were included, four recruiting inpatients or patients requiring admission, and two in outpatients.

Main results

The six trials included 657 children, but only five trials provided time to fever resolution (589 children). Antipyretics included oral and rectal paracetamol, rectal naproxen and oral ibuprofen. Studies recruited children with uncomplicated Plasmodium falciparum malaria (three studies), acute febrile respiratory tract infection and febrile illness with no bacterial focus (one study each). All trials were randomised. Risks of bias were described as lack of blinding and use of per protocol analysis in two studies each, prolonged recruitment and missing data in one study each.

In the meta-analysis, use of antipyretics was associated with a mean reduction of 4 h 10 min in time to fever resolution (95% CI 1h 58min to 6h 21min). Despite the clinical heterogeneity in the studies, there was little statistical heterogeneity (I²=17%). The study excluded from the meta-analysis recruited children with varicella zoster and found no significant difference in time to appearance of the last vesicle, time to total scabbing and time to total healing. Malaria studies …