Holgersson J, Ceric A, Sethi N, et al. Fever therapy in febrile adults: systematic review with meta-analyses and trial sequential analyses. BMJ. 2022 Jul 12;378:e069620. doi: 10.1136/bmj-2021-069620. (Systematic review)

OBJECTIVE: To investigate the effects of fever therapy compared with no fever therapy in a wide population of febrile adults.

DESIGN: Systematic review with meta-analyses and trial sequential analyses of randomised clinical trials.

DATA SOURCES: CENTRAL, BIOSIS, CINAHL, MEDLINE, Embase, LILACS, Scopus, and Web of Science Core Collection, searched from their inception to 2 July 2021.

ELIGIBILITY CRITERIA: Randomised clinical trials in adults diagnosed as having fever of any origin. Included experimental interventions were any fever therapy, and the control intervention had to be no fever therapy (with or without placebo/sham).

DATA EXTRACTION AND SYNTHESIS: Two authors independently selected studies, extracted data, and assessed the risk of bias. Primary outcomes were all cause mortality and serious adverse events. Secondary outcomes were quality of life and non-serious adverse events. Aggregate data were synthesised with meta-analyses, subgroup analyses, and trial sequential analyses, and the evidence was assessed using the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach.

RESULTS: Forty two trials assessing 5140 participants were included. Twenty three trials assessed 11 different antipyretic drugs, 11 trials assessed physical cooling, and eight trials assessed a combination of antipyretic drugs and physical cooling. Of the participants, 3007 were critically ill, 1892 were non-critically ill, 3277 had infectious fever, and 1139 had non-infectious fever. All trials were assessed as being at high risk of bias. Meta-analysis and trial sequential analysis showed that the hypothesis that fever therapy reduces the risk of death (risk ratio 1.04, 95% confidence interval 0.90 to 1.19; I2=0%; P=0.62; 16 trials; high certainty evidence) and the risk of serious adverse events (risk ratio 1.02, 0.89 to 1.17; I2=0%; P=0.78; 16 trials; high certainty evidence) could be rejected. One trial assessing quality of life was included, showing no difference between fever therapy and control. Meta-analysis and trial sequential analysis showed that the hypothesis that fever therapy reduces the risk of non-serious adverse events could be neither confirmed nor rejected (risk ratio 0.92, 0.67 to 1.25; I2=66.5%; P=0.58; four trials; very low certainty evidence).

CONCLUSIONS: Fever therapy does not seem to affect the risk of death and serious adverse events.


Discipline Area Score
Hospital Doctor/Hospitalists 7 / 7
Internal Medicine 7 / 7
Family Medicine (FM)/General Practice (GP) 5 / 7
General Internal Medicine-Primary Care(US) 5 / 7
Emergency Medicine 4 / 7
Comments from MORE raters

Family Medicine (FM)/General Practice (GP) rater

Very interesting and well performed systematic review concludes that fever therapy is not related to death and serious adverse events. More relevant and better designed studies are needed for additional confirmation. It appears that fever itself is not a relevant outcome and we have to treat the symptoms.

Family Medicine (FM)/General Practice (GP) rater

Important questions: Does high fever worsen outcomes? Does reducing the temperature improve outcomes? Traditionally, patients and doctors believe that fevers need to be treated. This review supports the lack of benefit in reducing fevers, so takes the urgency and concern out of treating them. However, the study was not designed to answer the question of whether a fever helps the body fight off infections sooner. So, is it best not to treat a fever except for patient comfort, etc? This may be a more difficult question to answer.

General Internal Medicine-Primary Care(US) rater

The hackneyed saying 'garbage in, garbage out' is applicable here: All trials included in the meta-analysis had high risk of bias!! While anti-pyretic agents can treat a bothersome symptom and reduce distress, the only reason to NOT use them is if they worsened distal outcomes (time to recovery from infection, etc). This article does not answer this more relevant question.

General Internal Medicine-Primary Care(US) rater

Nice study that pretty much supports what most of us suspect about treating fever: fever is a physiological rather than pathological process as a consequence of inflammation.

General Internal Medicine-Primary Care(US) rater

I stopped treating elevated temperatures decades ago. I noticed that nurses on the floors feel like a failure unless their patients have normal temperatures. When the temperature is rising, the patient feels a chill or, at worst, a rigor (you feel miserable). After an antipyretic, you feel miserable because you're drenched with sweat. So, instead of one or, at most, 2 chills and sweats during the day even with an infection, we cause our patients to sustain multiple chills and multiple sweats. I don't want to make my patients feel worse and cause further cardiovascular instability. Not to mention ventilatory overdrive. So, my orders say symptoms of fever, not just fever or some number. I've suspected for a long time that we have no effect on mortality except in patients with severe cardiovascular disease where elevated fever and it's attendant metabolic expense can result in cardiovascular collapse or ventilatory insufficiency.

Internal Medicine rater

Important analysis. In our department, nurses and doctors use a lot of paracetamol to lower temperature in fever patients. This article suggests this is an unnecessary effort.
Comments from EvidenceAlerts subscribers

No subscriber has commented on this article yet.