Systematic Review: Forced air warming versus circulating warm water for the prevention of hypothermia in surgical patients

Joanna Ford and Pete Phillips, Surgical Material Testing Laboratory

21 April 2015



Hypothermia (HT) is defined as subnormal body temperature. Patient core temperature often decreases before and during surgery and this can be detrimental to the patient. Devices designed to warm the patient during surgery are available. Forced air warming (FAW) has been available for many years and is a commonly used method of patient warming during surgery. More recently, a range of circulating water (CW) devices have come onto the market which are designed to keep patients warm during surgery. This review compares the ability of FAW and circulating water devices to prevent hypothermia in surgical patients.


To assess the effectiveness of forced air warming (FAW) versus circulating water (CW) warming in preventing hypothermia (measured by core temperature) in adult patients.

Search method

Searched Pubmed, EMBASE and CENTRAL (Cochrane library) from 2000 to Jan 2015.  Randomised control studies only. Search terms: warming/forced air/FAW/DM Warm/WarmTouch/BairHugger/ or Bair Hugger/ Thermacare/WarmAir/WarmTouch/circulating-water/water mattress/water garment/VitalHEAT (in title, all languages included).

Searches included general web searches, contacting manufacturers and accessing clinical trial  web sites to identify relevant studies (both published and unpublished).

Selection criteria

Once the search for publications had been carried out, abstracts of each paper were searched to identify papers which compared FAW with CW warming devices in patients undergoing elective surgery. Ten papers (all published) were selected for full analysis

Data collection and analysis

The author of the review used a data extraction form to record specific details about each study included in the review. This also included assessing the risk of bias.

Main results

10 studies were included in the analysis in which FAW was compared to a range of circulating water devices (CW garments, mattresses and sleeves (with vacuum)). As well as different types of CW devices, there was variation between studies in types of operations performed, duration of surgery and the different parts of the body warmed by the devices. Studies were considered to possess a moderate to high risk of bias.

Results showed that when FAW was compared with a CW mattress or CW sleeve with vacuum, FAW was better at preventing hypothermia. However, when FAW was compared with a CW garment (or on one occasion, CW mattress with additional CW leg wraps) the CW products demonstrated superior performance.

Authors conclusions

As a method of maintaining patient core temperature it is unclear whether FAW or CW devices perform better. It appears that the important factor in determining a device's performance is the degree to which the body surface area is in contact with the heat (i.e. the more surface area contact, the better the device performance).


Hypothermia (HT) is defined as subnormal body temperature. In practice, when any patient develops a temperature of less than 35oC, it is of major clinical concern. Guidance on maintaining patient normothermia recommends that patients are actively warmed if their core temperature falls below 36oC (NICE, 2008).

The effects of HT on patients are numerous. Surgery increases blood viscosity and drugs can accumulate in the system (Connor and Wren 2000). The colder a patient gets, the less they respond to drugs (their liver metabolism starts to decrease). Other effects include haemoglobin releasing oxygen less readily, sometimes leading to tissue hypoxia (McNeil 1997),  and the disruption of the blood clotting cascade which can cause increased bleeding (Cavallini et al 2005). The heart can also be affected as myocardial ischaemia and arrhythmia can occur (McNeil 1997, Kirkbride and Buggy 2003). Reduced cerebral perfusion can lead to confusion following recovery, as well as an impaired immune system (Connor and Wren 2000), poor wound healing (Kirkbride and Buggy 2003) and increased infection rates (Kurz et al., 1996). NICE carried out a systematic literature review of the topic and concluded that body warming significantly reduces the number of cases of surgical site infection (NICE 2008).

One retrospective analysis of a coronary surgery database demonstrated that patients who were HT upon entering ICU were significantly more likely to die following surgery, had mechanical ventilation for longer periods of time, more transfusions and a longer stay in hospital (Insler et al., 2000).

One German study compared a group of patients who underwent standard peri-operative warming with those who had the addition of upper body warming for half an hour prior to and during surgery. They found that patients in the second group took less time to recover and needed less blood transfusions which they calculated presented a 24% reduction in costs (Bock et al., 1998).



To assess the effectiveness of forced air warming (FAW) versus circulating water (CW) warming in preventing hypothermia (measured by core temperature) in adult patients.

Types of studies

Random controlled trials (RCTs) that assess the effectiveness of intra-operative warming with FAW versus CW warming in adult patients.

Type of participants

Adult patients undergoing elective surgery where hypothermia is not actively induced.

Type of intervention

  • Forced air versus circulating water mattress
  • Forced air versus circulating water garment
  • Forced air versus circulating water sleeve with vacuum

Primary outcome

  • Core body temperature during and /or at the end of surgery

Inclusion criteria

  • Any studies which compare the effectiveness of FAW with a CW warming device (including CW wraps, garments, mattresses or sleeves with vacuum).
  • Studies where any part of the body is subject to warming by the above methods.
  • Studies where primary outcome is core body temperature during and/or at the end of surgery.

Exclusion criteria

  • Studies using healthy volunteers.
  • Studies where hypothermia is induced.
  • Studies with exclusive pre-warming or post-warming.

Search methodology

  • Searched Pubmed, EMBASE and CENTRAL (Cochrane library)
  • Publication type -Randomised control trials (RCT)
  • Publication date 2000-present (focus on recent technology)
  • Search terms: warming/forced air/FAW/DM Warm/WarmTouch/BairHugger/ or Bair Hugger/Thermacare/ WarmAir/WarmTouch/circulating-water/water mattress/water garment/ VitalHEAT (in title)
  • Searched abstracts of each paper to extract papers which compared FAW with CW warming devices in patients undergoing elective surgery.
  • Searches included general web searches, contacting manufacturers for possible study details (published and unpublished) and clinical trial sites to identify relevant studies.

Exclusion details are described below:

  • Number of possible identified via range of search methods described  n = 41
  • Reasons for exclusions:  
    • FAW and CW devices not directly compared n = 23
    • Fluid warming not body warming n = 1
    • CW not a commercial device n = 1
    • Healthy volunteers not patients n = 5
    • Pre-warming only n = 1
  • Number of studies included n = 10


Data extraction

The 10 papers selected for full analysis were scrutinised and the relevant data extracted. Details included methodology, study design and results. Each paper was also assessed for risk of bias. Details for each paper were entered into a data extraction form (an example can be found in the appendix).


Study quality and risk of bias 

With device differences being so obvious, blinding during the operation seems impossible. One study claimed to be 'double blinded' but no details of how this was achieved were included (Ihl 2008). Randomisation was concealed in most cases. One study separately randomised 2 groups of patients (with or without a pulmonary catheters) so that groups were 'equally complex' (Janicki 2001) but in doing so introduced bias. In some cases pre-meds were varied within studies which could affect results (Janicki 2001 and Ruetzler 2011). In addition, a number of studies started warming with one device at a different time than the other (e.g. one before induction of anaesthesia and one afterwards). In one study, it was calculated that there was a 48 minutes difference in start time between the two groups studied (Janicki 2002). Ihl et a., 2008 had a female only population and most studies included a variety of operations (such as a range of open abdominal surgery).  Most papers included in the analyses stated that IV fluids had been warmed in all patients. The one exception was Trentman (2009) where fluids were warmed for some patients but not for others. These sorts of factors introduce limitations in studies and can weaken conclusions made when studies are systematically reviewed together.

A series of questions were asked of the study methodology, as shown in the table below. A study with a 'low' risk of bias would have the answer 'yes' or 'not applicable' (NA) to all the questions. In general, studies appeared to be at 'moderate' to 'high' risk of bias. See table below for summary:



Appropriate random allocation  

 Allocation appropriately concealed*    Losses and drop outs explained     Comparable co-interventions   Outcomes blinded  
Hasegawa et al . '12 yes yes NA NA ?
Negishi et al '03 yes yes NA NA No
Janicki et al. '01 ? ? yes No yes partly
Janicki et al. '02 yes yes NA No ?
Matsuzaki et al.'03 yes yes ? yes No
Ihn et al. '08 ? ? NA NA yes, but no detail
Trentman '09 yes yes yes No ?
Ruetzler et al. '11 yes yes yes ? No
Hofer et al., '05 yes ? yes NA No
Kim et al., 2014 ? ? NA yes No

 * Randomisation results kept in opaque envelope until surgery

Summary of findings

In the ten studies that were investigated, FAWs were compared with a variety of CW warming devices. These were CW garments, mattresses, leg wraps and sleeves (with vacuum). The table below gives a brief summary of results from the 10 papers.

Study Type of surgery Number of patients in study Type of circulating water device being compared with FAW Conclusion of study

Negishi et al., 2003

Open major abdominal surgery



CW group significantly lower core temperature after 150 minutes of surgery compared with FAW

Matsuzaki et al., 2003

laparoscopic cholecystectomy



CW group significantly lower core temperature after 150 minutes of surgery compared with FAW

Ihn et al., 2003

Total abdominal hysterectomy

90 mattress

CW group significantly lower core temperature from 30-120 minutes of surgery compared with FAW

Kim et al., 2014

Total knee arthroplasty

46 mattress

No difference in core temperature between 2 groups. Significantly greater post-op shivering in CW group.

Hasegawa et al., 2012

Open major abdominal surgery

36 mattress plus leg wrap/garment

CW group significantly higher core temperatures at the end of surgery compared with FAW.

Janicki et al., 2001

Open major abdominal surgery

53 garment

CW group significantly higher core temperatures at the end of surgery compared with FAW.

Hofer et al., 2005

Off-pump coronary artery bypass

80 garment

CW group significantly higher core temperatures at the end of surgery compared with FAW

Janicki et al., 2002

liver transplant patients

24 garment

CW group significantly higher core temperatures at the end of surgery compared with FAW.

Trentman et al., 2009

total knee replacement


sleeve plus vacuum

CW group had lower core temperature at 60 minutes peri-operative and more hypothermia in this group.

Ruezler et al., 2011

Open major abdominal surgery


sleeve plus vacuum

CW group had slightly lower core temperature up to 4 h peri-operative. However, statistically non-inferior to FAW.

Effect of interventions

Forced air versus circulating water mattress (n=4)

Four studies compared FAW (1 lower body, 3 upper body) with circulating water mattresses (under patient). In one of the studies, patients were undergoing a range of major abdominal surgery (Negishi et al., 2003), in one group all patients were having key hole gall bladder surgery (Matsuzaki et al., 2003) in another, total hysterectomy (the only all-female patient group) (Ihn et al., 2008) and in the most recent, total knee replacement (Kim et al., 2014). One of the studies involved lower body FAW warming (Negishi et al., 2003) and three used upper body FAW warming (Matsuzaki et al., 2003, Ihn et al., 2003, Kim et al., 2014).  Three studies used a tympanic membrane probe and the fourth used oesophageal and rectal probes to measure core body temperature (Ihn et al., 2003). Three of the RCT studies found that FAW groups had significantly higher core temperatures than CW mattress groups during and at the end of surgery (Negishi et al., 2003, Matsuzaki et al., 2003, Ihn et al., 2003). Kims study found no difference in core temperature between FAW and CW mattress groups although patients exhibited significantly more shivering in the CW group than FAW group (Kim et al., 2014).

Forced air versus circulating water sleeve with vacuum (n=2).

In two of the ten studies, the circulating water device was a sleeve over one arm that exerted a vacuum which is thought to increase vasodilation and increase contact between the skin and the warming device. Both used a vitalHEAT product, possibly different versions. These were compared with upper body FAW. One study was of patients undergoing total knee replacement whereas the other was of patients undergoing a range of open abdominal surgery. The FAW performed better than the CW sleeve in both studies.

Forced air versus circulating water garment (n=4)

Two of the studies which used total body garments were by same authors. One compared upper and lower FAW with the CW garment (liver transplantation patients) (Janicki et al., 2002) and the other compared upper FAW with CW garment in patients undergoing open abdominal surgery (Janicki et al., 2001). Both studies concluded that patients in the CW group were able to 'maintain normothermia' better than the FAW group. A third study by a different investigator compared a CW garment (covering the patient's back and most of their extremities) to lower body FAW in patients undergoing off-pump coronary artery bypass and found that core temperatures were significantly higher in the CWG group than the FAW group. The final study combined a CW mattress with CW leg wraps/garment in one group and compared them to patients in a lower body FAW group, all undergoing major abdominal surgery. The CW group maintained their core body temperature better than the FAW group.


The results of this systematic review suggest that CW is a useful method of perioperative patient warming and in most cases, is as good, if not better than FAW in preventing hypothermia. However, the results suggest that it is not so much the mechanism of warming that is crucial but the surface area that is in contact with the source of warmth. This would explain why a water mattress alone or a negative pressure sleeve does not warm the patient as effectively as a full CW garment.  As only 4 papers were found that looked at CW garments and two of them were by the same author, this area certainly warrants more investigation but these preliminary results are promising. Another RCT study comparing different warming methods on a group of cardiac patients  (Zangrillo et al., 2008) was not included in the review as the control group combined FAW with a CW mattress rather than FAW alone. However, the results of this study showed that normothermia was maintained and core temperature was even slightly raised in the CW garment patient group compared to the control group and differences reached significance after 2 hours of surgery.

There are a number of papers relating to this topic that do not fit the criteria of this systematic review but are interesting nevertheless. Sury et al., 2006 (cross matched, non RCT study) found that CW garments were significantly better at maintaining normothermia in infants undergoing abdominal surgery compared with standard warming methods (warm air mattresses or duvets). Another study anaesthetised 9 healthy adult volunteers and compared heat transfer between FAW and CW garments and found that patients were warmed faster and heat transfer was greater in the CW group than the FAW group (Tauchi et al., 2004). In contrast, a study using healthy volunteers to compare negative pressure rewarming with the Vital Heat sleeve did not warm patients any faster than those covered with a cotton blanket (Tauchi et al., 2001). 

So in conclusion, this systematic review indicates that circulating water wraps/garments (but not negative pressure sleeves or CW mattresses alone) are more efficient than FAW in maintaining normothermia in surgical patients. This is thought to be related to the surface area available for heat transfer rather than the mechanism of warming itself. 


Conflict of Interest

No conflict of interest is reported.


The author used a systematic review 'check list' produced by Specialist Unit for Review Evidence (SURE) 2013. Questions to assist with the critical appraisal of a systematic review.

Available at

The author also used the 'Cochrane Handbook for Systematic Reviews of Interventions'. Eds. Higgins J and Green S (2011). The Cochrane Collaboration version 5.1.0



Bock M; Muller J; Bach A; Bohrer H; Martin E; Motsch J.  Effects of pre-induction and intraoperative warming during major laparotomy. British Journal of Anaesthesia. 1998; 80:159-163

Cavallini M; Baruffaldi Preis F W; Casati A. Effects of mild hypothermia on blood coagulation in patients undergoing elective plastic surgery. Plast Reconstr Surg 2005; 116(1):316-21

Connor E; Wren K. Detrimental effects of Hypothermia: A systems analysis. Journal of Peri Anesthesia Nursing. 2000; 15(3):151-5

Hasegawa K; Negishi C; Nakagawa F; Ozaki M. Core temperatures during major abdominal surgery in patients warmed with new circulating-water garment, forced-air warming, or carbon-fiber resistive-heating system. J Anesth. 2012;26:168-173

Hofer C, Worn M, Tavakoli R, Sander L, Maloigne M, Klaghofer and Zollinger A. Influence of body core temperature on blood loss and transfusion requirements during off-pump coronary artery bypass grafting: A comparison of 3 warming systems.The Journal of Thoracic and Cardiovascular surgery. 2005; 129(4): 838-843

McNeil B. Inadvertent hypothermia in the operating theatre. Prof Nurse. 1997; 12(6):418-21

NICE Inadvertent perioperative hypothermia : the management of inadvertent perioperative hypothermia in adults. NICE Guidance 65. 2008

Ihn C; Joo J; Chung H; Choi J; Kim D; Jeon Y; Kim Y; Choi W
Title: Comparison of Three Warming Devices for the Prevention of Core Hypothermia and Post-Anaesthesia Shivering. Journal of International Medical Research.2008;36:923-931

Insler S R; O'Connor M S; Leventhal M J; Nelson D R; Starr N J. Association between postoperative hypothermia and adverse outcome after coronary artery bypass surgery. Ann Thorac Surg.2000;70(1):175-81

Janicki P; Higgins M; Janssen J; Johnson R; Beattie, C.Title: Comparison of two different temperature maintenance strategies during open abdominal surgery: upper body forced-air warming versus whole body water garment. Anesthesiology. 2001;95(4):868-74

Janicki P; Stoica C; Chapman W; Wright J; Walker G; Pai R; Walia A; Pretorius M; Pinson C. Water warming garment versus forced air warming system in prevention of intraoperative hypothermia during liver transplantation: a randomized controlled trial [ISRCTN32154832]. BMC Anesthesiol  2002 ;2(1): 7-11

Kim H, Lee K, Lee M, Kim M-N, Kim J-S, Lee W, Lee J. Comparison of the efficacy of a forced-air warming system and circulating water mattress on core temperature amd post-amesthesia shivering in elderly patients undergoing total knee arthroplasty under spinal anesthesia.

Kirkbride D, Buggy D. Thermoregulation and mild peri-operative hypothermia. British Journal of Anaesthesia. 2003;3(1):24-27

Kurz A, Sessler D,  Lenhardt R. Perioperative normothermia to reduce the incidence of surgical-wound infection and shorten hospitalization. The New England Journal of Medicine. 1996; 334(19):1209-1215

Matsuzaki Y; Matsukawa T; Ohki K; Yamamoto Y; Oshibuchi T. Warming by resistive heating maintains perioperative normothermia as well as forced air heating.  British Journal of Anaesthesia. 2003;90(5):689-91

McNeil B. Inadvertent hypothermia in the operating theatre. Prof Nurse. 1997; 12(6):418-21

Negishi C; Hasegawa K; Mukai S; Nakagawa F; Ozaki M; Sessler D. Resistive-heating and forced-air warming are comparably effective. Anesth Analg.

NICE Inadvertent perioperative hypothermia : the management of inadvertent perioperative hypothermia in adults. NICE Guidance 65. 2008

Ruetzler K; Kovaci B; Gu loglu E; Kabon B; Fleischmann E; Kurz A; Mascha E; Dietz D; Remzi F; Sessler D. Forced-Air and a Novel Patient-Warming System (vitalHEAT vH 2 ) Comparably Maintain Normothermia During Open Abdominal Surgery. Anesthesia and analgesia. 2011;112:3:608-14

Sury M, Scuplak S. Water-filled garment warming of infants undergoing open abdominal or thoracic surgery. Pediatr Surg Int. 2006;22:182-5

Taguchi A; Arkilic C; Ahluwalia A; Sessler D.; Kurz A. Negative pressure rewarming vs. forced air warming in hypothermic post anesthetic volunteers. Anesth Analg. 2001; 92(1):261-6

Taguchi A; Ratnaraj J; Kabon B; Sharma N; Lenhardt R; Sessler D; Kurz A. Effects of a Circulating-water Garment and Forced-air Warming on Body Heat Content and Core Temperature. Anesthesiology. 2004; 100(5):1058-64

Trentman T; Weinmeister K; Hentz J; Laney M; Simula D. Randomized non-inferiority trial of the vitalHEAT TM Temperature Management System vs the Bair Hugger warmer during total knee arthroplasty. J Can Anest. 2009;56:914-20

Zangrillo A, Pappalardo F, Corno C, Landoni G, Scandroglio A, Rosica C, Crescenzi G. Temperature management during off-pump coronary artery bypass graft surgery:a randomised clinical trial on the efficacy of a circulating water system versus a forced-air system. J Cardiothoracic and Vasc. Anaesthesia. 2008;20(6):788-792


Example of data extraction form


Ruetzler et al., 2011



Devices that were compared

Forced air warming (FAW) and circulating water (CW) sleeve with vacuum

Type of intervention (model name in brackets)

CW sleeve set at 41oC (VitalHEAT)

Type of control (model name)

FAW with upper body blanket (Bair Hugger) set 'high' 43oC

Inclusion criteria

18-75 y patients with ASA I-III and 20-36 Kg/m2, operation duration >2h

Exclusion criteria

Bilateral vascular catheters distal to the

elbow, serious skin lesions on the hands or arms, history of vascular

conditions including Reynaud’s Syndrome, preoperative fever,

contraindication to sevoflurane endotracheal anesthesia, or a preexisting


Patients undergoing major scheduled surgery? Details

Yes. Open abdominal surger-various.

Use of premeds

1 of 2 types of pre-meds used depending on clinician's preference

Time of intervention (pre/peri etc)


CW sleeve started after induction of anaesthesia, FAW started

before induction.

Part of the patient's body in contact with warming device


FAW- front of upper body

CW sleeve -one arm

Were IV fluids warmed? Details.


yes, 40-42oC

List primary outcomes


Average core temperature during surgery

Short summary of results


4h peri-op CWS 36.3oC +/- 0.6oC vs FAW 36.4oC +/- 0.5oC

CW sleeve non-inferior to FAW as core temp. no >0.5oC lower

when measured every hour (up to 4h) peri-op.

 Risk of bias table

How was treatment allocation schedule generated?

Computer generated codes

Was allocation sequence concealed properly?

Opaque envelopes

No. of patients randomised


No. of patients analysed

71 (37 in CW group, 34 in FAW group)

Did paper state that power analysis was carried out to

determine no. patients required for statistical analysis?


Were losses and drop-outs properly described?


Any co-interventions? (e.g. heating fluids, additional blanket)


switched from CW sleeve to FAW after 4 hours op

(although only analysed up to 4 h). Also carried out rescue

warming with FAW if any in CW group <35oC

Are co-interventions comparable between 2 groups?

Rescue warming not comparable. No further details given.

Did they try and assess outcomes in a blinded manner?


Any other type of bias (details).

none detected

Risk of Bias *


* If all the issues in bold are considered adequate, bias is 'low', if 1 issue is not adequate the risk of bias is considered 'moderate' and if more than one is not adequate risk is considered 'moderate' to 'high'. 

Joomla Template: from JoomlaShack