Secondary Logo

Journal Logo

Systematic Review Protocol

External Cooling Methods for Treatment of Fever in Adults: a Systematic Review

Chan, E Y RN, MSc (Health Research Methodology), MN (Critical Care), CCRN1; Chen, W T RN, BSc (Honours), Adv Dip (Medical Surgical)2

Author Information
JBI Library of Systematic Reviews: Volume 8 - Issue 8 - p 1-11
doi: 10.11124/jbisrir-2010-848
  • Free


Fever is common among hospitalized patients. Approximately one-third of all medical patients developed fever during their hospitalization.1 The Society of Critical Care Medicine and the Infectious Disease Society of America suggested that a body temperature of 38.3°C be considered a fever although the definition is arbitrary.2

The physiological role of fever is not completely understood. It is unclear if fever is a beneficial or harmful effect of infection and inflammation.3 The value of treating fever is also debatable. Apart from the importance of fever control in patients with neurological injury,4 little evidence exists that treating fevers in adults improves outcome.5,6 Despite this controversy, it is a common practice to cool febrile patients using mechanical/external or pharmacological antipyretic interventions.

Gozzoli et al7 estimated that it costs between $10,000 and $29,000 per year to treat 18 febrile intensive care patients per month with paracetamol, cooling blankets or ice packs. Pharmacological interventions to reduce fever can lead to renal hepatic dysfunction in patients who have underlying kidney or liver diseases.8 Antipyretics such as aspirin and non-steroidal anti-inflammatory drugs (NSAIDS) may also cause unwanted side effects on platelets and gastrointestinal tract.8 Considering these, external cooling methods may be preferred.

There are a variety of external cooling methods. The most commonly used include tepid bathing, ice packs, cooling blankets and fanning.9,10 These methods help the body to lose heat through conduction, convection or evaporation. Interestingly, literature appears to suggest alternative methods of cooling across different cultures. For example, the Japanese used ice pillows11 while the Chinese utilised warm water immersion.12

Many narrative reviews have been published on the importance of external cooling and the different methods adopted.13-16 Unfortunately, these reviews failed to provide a definitive recommendation on the use of external cooling with differing expert opinions on the actual benefits and harms of these methods. To date we are not aware of any systematic review conducted on the topic in adult populations, although they have been done on paediatric populations.17,18 Since there is a lack of evidence to guide practice, the clinical decision-making on the usage of cooling methods is currently left to the individual practitioner's expertise, views and preferences. This is evident from a recent survey conducted in 2007 of neuroscience nurses in the United States. Those surveyed reported that their fever management practices are dependent on their personal judgments.19

As external cooling methods are widely practiced to reduce body temperature, we seek to conduct a systematic review to examine if these methods are effective in managing adults with fever.


To evaluate the effectiveness and complications of external cooling methods in febrile adults in acute care settings.

The following specific research questions were addressed in this review:

  1. What is the effectiveness of external cooling method, compared to no cooling in reducing temperature for adults with fever?
  2. What is the effectiveness of external cooling method, compared to no cooling in sustaining the reduction in temperature?
  3. What are the adverse events or complications with the use of external cooling method compared to no cooling?
  4. What are the patients' accounts of the external cooling methods compared to no cooling?

Criteria for considering studies for this review

Types of studies

This review will consider randomised controlled trials (RCTs), quasi-randomized trials and controlled trials with concurrent control groups, available in full or in abstract form. Expert opinion/comment papers will not be considered for inclusion.

Types of participants

Participants in this study will include adults who have been admitted to an acute care setting who developed elevated body temperature that requires cooling to normothermia. The authors' definitions of adult and fever will be utilised.

Types of interventions

This review will consider any external cooling method compared to no intervention. We will exclude trials reporting invasive cooling methods such as the use of endovascular cooling. Trials directly comparing two external cooling methods without a non-intervention group will also be excluded. Furthermore trials involving induced hypothermia for neurotrauma, management of heat stroke, malignant hyperthermia, or severe hypoxemia will not be considered. The studies will be stratified into those where all participants received antipyretic drugs and those where no participants received antipyretic drugs.

Types of outcome measures

The primary outcome of interest is resolution of fever using the authors' definitions. The secondary outcomes of interest are:

  1. Mean drop in temperature after starting treatment
  2. Mean time to fever resolution
  3. Continued drop in temperature after discontinued treatment
  4. Persistent febrile state after starting treatment
  5. Complications such as shivering, chills, vomiting, irritability, headache and muscle pain
  6. Adverse events defined as those leading to discontinuation of the intervention or defined as ‘adverse events’ using the authors' definitions
  7. Patient's account of the treatment regimen

Search strategy for identification of studies

The search strategy for this review will aim to find all relevant published or unpublished studies up to Aug 2009 regardless of language.

A three-step search strategy will be utilised for the database searches. Firstly, an initial limited search of MEDLINE and CINAHL will identify key words or terms contained in the titles and abstracts. Secondly, using all identified key words or terms a more extensive search of databases will be undertaken. Thirdly, the reference lists and bibliographies of all relevant articles are to be searched for additional studies.

The OVID version of all databases will be searched and incorporate a highly sensitive search strategy (HSSS) to retrieve studies from MEDLINE and modify it for other databases. The search will not apply any language restriction

Initial search terms using the MeSH (Medical Subject Heading) and free text are:

  • Fever
  • Pyrexia
  • Temperature
  • Heat/Cold application
  • Hyperthermia
  • Hypothermia
  • Cryotherapy
  • Cool
  • Sponges
  • Fan
  • Baths
  • Blanket
  • Ice
  • Antipyretic
  • Antipyresis
  • External cooling
  • Physical cooling

The following databases will be searched:

  • MEDLINE 1950 to Aug week 1 2009
  • CINAHL 1982 to Aug week 1 2009
  • CENTRAL 2nd quarter 2009
  • EMBASE 1998 to Aug 2009
  • Expanded Academic Index 1998 to Aug 2009
  • TRIP
  • EBM Reviews
  • Biomed Central
  • WorldCat
  • Chinese Databases

The search for unpublished studies shall include:

  • Conference proceedings
  • Dissertation Abstracts International
  • Grey Literature Report
  • NLM Gateway
  • Index to Theses
  • Networked Digital Library of Theses and Dissertations (NDLTD)
  • DIVA Academic Archieve Online
  • SIGLE (System for Information on Grey Literature in Europe)
  • Health Technology Assessment

We will also conduct hand searches of relevant journals and contact experts in the field in order to find additional unpublished trials. Methods of the review

Assessment of methodological quality

Two reviewers will independently screen titles and abstracts, select studies for potential inclusion (using the criteria listed in Appendix 1), and appraise the methodological quality of included studies. Disagreements in all phases shall be resolved by discussion or consultation with a third reviewer.

The reviewers will independently appraise the methodological quality of included trials using the Meta Analysis of Statistics Assessment and Review Instrument (MAStARI) Critical Appraisal Checklist for Randomised and Pseudo-randomised Studies developed by the Joanna Briggs Institute (JBI) (Appendix 1). The criteria assess methodological quality with respect to allocation concealment, baseline similarity of treatment and control groups, blinding of outcomes assessors, analysis of primary outcome following the intention-to-treat principle, and completeness of follow-up (Appendix II).

Data extraction

The data extraction will involve a a modified version of the JBI Data Extraction Form for Randomised and Pseudo-randomised Controlled Trials which will be piloted by the two reviewers prior to utilisation (Appendix III).

The data to be extracted will include:

  1. Study design
  2. Method of sample selection
  3. Patient characteristics
  4. Inclusion and exclusion criteria
  5. Description of the main intervention (external cooling methods used)
  6. Description of the confounding factors (antipyretics used)
  7. Duration of follow-up and attrition
  8. Effects on outcomes

The authors will be contacted for clarification of data and additional data.

Data synthesis

For each external cooling method, the results of studies where appropriate will be quantitatively pooled. It is anticipated there will be variability in the reporting of complications, adverse events, and each patient's account of treatment regimen, and thus these will be described narratively.

Quantitative comparisons, treatment and control groups will be compared using relative risk (for categorical outcomes) and weighted mean difference (for continuous outcomes), and their 95% confidence intervals, using the random effects model20 in Review Manager 4.2 and all results will be double entered.

Clinical homogeneity will be assessed by considering the settings, populations, interventions and outcomes. For each pooled comparison, heterogeneity will be 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.21 Values of I2 greater than 50% are considered to represent substantial between-study heterogeneity.21,22 When pooled analyses contain at least five studies, funnel plots will then be constructed to assess for the potential of publication bias.

The following additional analyses will be conducted for the outcomes of resolution of fever and reduction in body temperature at regular time intervals:

  1. Subgroup analyses of treatment effect across different adult populations; and
  2. Sensitivity analysis in trials with adequate blinding of outcomes assessors compared to all trials.

Potential conflict of interest



1. Cunha BA, Shea KW. Fever in the intensive care unit. Infect Dis Clin North Am. 1996; 10: 185-209.
2. O' Grady NP, Barie PS, Bartlett JG, Bleck T, Carroll K, Kalil AC, et al. Guidelines for evaluation of new fever in critically ill adult patients: 2008 update from the American College of Critical Care Medicine and the Infectious Diseases Society of America. Crit Care Med. 2008; 36(4): 1330-1349.
3. Kluger MJ, Kozak W, Conn CA, Leon LR, Soszynski D. Role of fever in disease [review]. Ann N Y Acad Sci. 1998; 856:224-233.
4. Polderman K. Induced hypothermia and fever control for prevention and treatment of neurological injuries. Lancet. 2008; 371: 1955-1969.
5. Kiekkas P, Brokalaki H, Theodorakopoulou G, Baltopoulos GI. Physical antipyresis in critically ill adults. AJN 2008; 108:40-49.
6. MacLaren G, Spelman D. Fever in the intensive care unit. UpToDate. Version 16.2:May 2008.
7. Gozzoli V, Schottker P, Suter PM, Ricou B. Is it worth treating fever in intensive care unit patients? Preliminary results from a randomized trial of the effect of external cooling. Arch Intern Med. 2001; 161(1):121-123.
8. Plaisance KI. Toxicities of drugs used in the management of fever. Clin Infect Dis. 2000; 31(Suppl 5): S219-223.
9. Thompson HJ, Kirkness CJ, Mitchell, PH, Webb DJ. Fever management practices of neuroscience nurses: National and regional perspectives. J Neurosci Nurs. 2007; 39(3): 151-162.
10. Axelrod P. External cooling in the management of fever. Clin Infect Dis. 2000; 31 Suppl 5:224-229.
11. Ikematsu Y. Characteristics of and interventions for fever in Japan. Int Nurs Rev. 2004; 51: 229-239.
12. Zhang Y, Wang X, Xu J. Observations on warm water feet immersion to lower down the body temperature. Shanxi Nursing Journal. 2000; 14(1): 35-36.
13. Axelrod YK, Diringer MN. Temperature management in acute neurologic disorders. Neurol Clin. 2008; 26(2): 585-603.
14. Bernard SA, Buist M. Induced hypothermia in critical care medicine. Crit Care Med. 2003; 31(7): 2041-2051.
15. Henker R, Carlson K. Fever. Applying research to bedside practice. AACN Advanced Critical Care. 2007; 18(1): 76-87.
16. Price T, McGloin S. A review of cooling patients with severe cerebral insult in ICU (part1). Nurs Crit Care. 2003; 8(1): 30-36.
17. Meremikwu M, Oyo-Ita A. Physical methods for treating fever in children (Review). 2007; The Cochrane Library. Issue 4.
18. Watts R, Robertson J, Thomas G. Nursing management of fever in children: A systematic review. Int J Nurs Pract. 2003; 9: S1-S8.
19. Thompson HJ, Kirkness CJ, Mitchell PH. Fever management practices of neuroscience nurses: Nurse, Patient and Barriers. J Neurosci Nurs. 2007; 39(4): 196-201.
20. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986; 7:177-188.
21. Higgins JP, Thompson SG. Quantifying heterogeneity in a metaanalysis. Stat Med. 2002; 21:1539-1558.
22. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003; 327:557-560.

Appendix I: Selection Criteria for Study Inclusion

No Caption available.

Appendix II MAStARI Critical Appraisal Tool for Randomized Controlled / Pseudo-randomized Trials

No Caption available.

Appendix III

Data Extraction Form for Randomised and Pseudo-Randomised Controlled Trials

No Caption available.
No Caption available.
© 2010 by Lippincott Williams & Wilkins, Inc.