EVIDENCE-BASED ANSWER
IT DEPENDS ON WHICH OUTCOMES YOU LOOK AT. Treating fever significantly increases comfort, activity, feeding, and fluid intake and decreases the patient’s temperature compared with placebo (strength of recommendation (SOR): A, multiple randomized controlled trials [RCTs]). It doesn’t shorten or prolong the overall duration of illness or reduce the recurrence of febrile seizures (SOR: A, multiple RCTs). In patients with varicella, reducing fever prolongs the time it takes for lesions to crust, but doesn’t appear to cause group A streptococcal necrotizing fasciitis (SOR: B, multiple prospective cohorts).
Ibuprofen and acetaminophen reduce fever effectively and safely; their side effect levels are similar to placebo (SOR: A, multiple RCTs). Physical methods of cooling also are effective for lowering fever (SOR: A, high-quality meta-analysis).
Evidence summary
The possible risks and benefits of lowering fever with antipyretics have prompted much discussion in the medical literature. Most evidence supporting the useful effects of fever comes from animal models. A few human studies in special populations (patients with brain trauma) suggest that antipyretics may worsen certain aspects of disease processes.1 However, many studies show that antipyretics do clinically benefit, and don’t harm, the typical child with an upper respiratory illness.2-4
Acetaminophen improves some symptoms, but doesn’t shorten fever
A double-blind, randomized, placebo-controlled trial of 210 children with upper respiratory illness showed that giving acetaminophen significantly increased short-term temperature reduction compared with placebo (0.36°F vs 0.09°F/hr; P<.001). Acetaminophen also improved subjective symptoms at 6 hours over placebo: activity (60% vs 16%; P<.001), alertness (58% vs 21%; P<.001), comfort (37% vs 7%; P<.001), mood (36% vs 12%; P<.001), appetite (20% vs 1%; P<0.001), and fluid intake (22% vs 2%; P<.001).2 Total time until complete resolution of fever (defined as no fever with or without medication) didn’t differ between treatment and placebo groups (32 hr for acetaminophen vs 36 hr for placebo; P=.23).2
A similar study failed to show a statistically significant difference between acetaminophen and placebo in fever clearance time (34.7 vs 36.1 hr), mood, comfort, or appetite. However, the study did find significant improvement in activity (38% vs 11%; P=.005) and alertness (33% vs 12%; P=.036) based on parents’ reports.3
Ibuprofen is a better antipyretic than acetaminophen
A meta-analysis (N=84,192) comparing acetaminophen with ibuprofen found both drugs to be equally efficacious analgesics; ibuprofen was a more effective antipyretic. Both drugs had the same level of adverse outcomes when compared with each other and placebo.5
Lowering fever doesn’t reduce recurrence of febrile seizures
Antipyretics don’t appear to affect the recurrence of febrile seizures. In a well-designed, randomized, double-blind, placebo-controlled trial, 230 patients 1 to 4 years of age with 1 or more risk factors for febrile seizures were randomized to fever reduction with ibuprofen or placebo. The estimated 2-year probability for recurrent febrile seizure was 32% in the ibuprofen group and 39% for the placebo group. The difference between the groups was not significant (relative risk=0.9; 95% confidence interval [CI], 0.6-1.5).6
Varicella lesions crust more slowly with acetaminophen
An RCT involving 68 children with varicella found that acetaminophen increased the time for lesions to crust compared with placebo (6.7 days for acetaminophen vs 5.6 days for placebo) and that acetaminophen didn’t alleviate itching or improve appetite. However, patients taking acetaminophen were more active by day 2.7
NSAIDs don’t cause necrotizing fasciitis
A case-control study of 48 children with varicella suggested an association between ibuprofen and increased incidence of necrotizing soft tissue infections (odds ratio [OR]=11.5; 95% CI, 1.4-96.9),8 but larger reviews haven’t borne out this finding. A review of 5 prospective studies of nonsteroidal anti-inflammatory drugs and invasive soft-tissue infections (N=912; 114 cases of necrotizing fasciitis) failed to show any connection. The largest study (474 cases) suggested a trend that was not statistically significant (OR=3.5; 95% CI, 0.8-16).9