It is unclear if the use of external cooling to treat fever contributes to better patient outcomes. Despite this, it is a common practice to treat febrile patients using external cooling methods alone or in combination with pharmacological antipyretics.
The objective of this systematic review was to evaluate the effectiveness and complications of external cooling methods in febrile adults in acute care settings.
Types of patients
We included adults admitted to acute care settings and developed elevated body temperature.
Types of interventions
We considered any external cooling method compared to no cooling.
Types of study designs
We considered randomised control trials (RCTs), quasi-randomised trials and controlled trials with concurrent control groups
We searched relevant published or unpublished studies up to October 2009 regardless of language. We searched major databases, reference lists, bibliographies of all relevant articles, and contacted experts in the field for additional studies. Two reviewers independently screened titles and abstracts, and retrieved all potentially relevant studies.
Two reviewers independently conducted the assessment of methodological quality of included studies.
The results of studies where appropriate was quantitatively summarised. Relative risks or weighted mean difference and their 95% confidence intervals were calculated using the random effects model in Review Manager 5. For each pooled comparison, heterogeneity was assessed using the chi-squared test at the 5% level of statistical significance, with I2 statistic used to assess the impact of statistical heterogeneity on study results. Where statistical summary was not appropriate or possible, the findings were summarised in narrative form.
We found six RCTs that compared the effectiveness and complications of external cooling methods against no external cooling. There was wide variation in the outcome measures between the included trials. We performed meta-analyses on data from two RCTs totalling 356 patients testing external cooling combined with antipyretics versus antipyretics alone, for the resolution of fever. The results did not show a statistically significant reduction in fever (relative risk 1.12, 95% CI 0.95 to 1.31; P=0.35; I2 =0%).
The evidence from four trials suggested that there was no difference in the mean drop in body temperature post treatment initiation, between external cooling and no cooling groups. The results of most other outcomes also did not demonstrate a statistically significant difference. However summarising the results of five trials consisting of 371 patients found that the external cooling group was more likely to shiver when compared to the no cooling group (relative risk 6.37, 95% CI 2.01 to 20.11; P=0.61; I2 =0%).
Overall this review suggested that external cooling methods (whether used alone or in combination with pharmacologic methods) were not effective in treating fever among adults admitted to acute care settings. Yet they were associated with higher incidences of shivering. These results should be interpreted in light of the methodological limitations of available trials.
Implication for practice
Given the current available evidence, the routine use of external cooling methods to treat fever in adults may not be warranted until further evidence is available. They could be considered for patients whose conditions are unable to tolerate even slight increase in temperature or who request for them. Whenever they are used, shivering should be prevented.
Implication for further research
Well-designed, adequately powered, randomised trials comparing external cooling methods against no cooling are needed.