The role of perioperative warming in surgery: a systematic review

ABSTRACT OBJECTIVE: The objective of this review was to systematically analyze the trials on the effectiveness of perioperative warming in surgical patients. METHODS: A systematic review of the literature was undertaken. Clinical trials on perioperative warming were selected according to specific criteria and analyzed to generate summative data expressed as standardized mean difference (SMD). RESULTS: Twenty-five studies encompassing 3,599 patients in various surgical disciplines were retrieved from the electronic databases. Nineteen randomized trials on 1785 patients qualified for this review. The no-warming group developed statistically significant hypothermia. In the fixed effect model, the warming group had significantly less pain and lower incidence of wound infection, compared with the no-warming group. In the random effect model, the warming group was also associated with lower risk of post-anesthetic shivering. Both in the random and the fixed effect models, the warming group was associated with significantly less blood loss. However, there was significant heterogeneity among the trials. CONCLUSION: Perioperative warming of surgical patients is effective in reducing postoperative wound pain, wound infection and shivering. Systemic warming of the surgical patient is also associated with less perioperative blood loss through preventing hypothermia-induced coagulopathy. Perioperative warming may be given routinely to all patients of various surgical disciplines in order to counteract the consequences of hypothermia.


Introduction
Hypothermia, defined as core temperature below 36 °C1-3 is common in operating theaters and has often been disregarded as an inevitable consequence of general anesthesia and surgery. 2,4,5The body's core temperature is determined by the balance between heat loss and heat gain.Exposure to a cold operating theater environment and anestheticinduced impairment of thermoregulatory control are two of the commonest contributing factors that tip the balance in favor of heat loss, thereby leading to hypothermia in surgical patients. 1,6ypothermia confers distinct benefits as well as severe complications in surgical patients.The potential benefits include protection against the deleterious effects of cerebral ischemia and malignant hyperthermia. 7However, hypothermia may increase susceptibility to perioperative wound infection by causing vasoconstriction and impaired immunity.Vasoconstriction decreases the partial pressure of oxygen in tissue, which lowers the resistance to infection. 8The other commonly known adverse effects of hypothermia include shivering, 9 prolonged duration of drug action, 10 coagulopathy, 11 myocardial ischemia and decreased resistance to surgical infections. 12Perioperative warming has been shown to reduce perioperative complications. 13,14Several prophylactic and therapeutic measures have been tried with the aim of reducing or abolishing the development of perioperative hypothermia.Various perioperative warming techniques like simple cotton blankets, carbon-fiber sheets, circulating hot water mattresses, forced air warming, warm fluid infusion and esophageal heat exchange systems 9,15,16 are in use in all surgical disciplines.These perioperative warming systems are being used during the preoperative, intraoperative and postoperative phases with variable efficacy.The duration of perioperative warming is also under review and prolonged exposure of surgical patients to warming systems has proven to be quite effective in major elective abdominal surgery. 17he aim of this systematic review was to compare the efficacy of perioperative warming of surgical patients aimed at reducing the con-sequences of wound infection, coagulopathy, blood loss, postoperative pain and postoperative shivering, in relation to no warming.

Methods
Relevant prospective randomized controlled trials on perioperative warming among surgical patients published between January 1980 and June 2007 were identified through the Medical Literature Analysis and Retrieval System Online (Medline), Excerpta Medica (Embase), Cumulative Index to Nursing and Allied Health Literature (CINAHL), Cochrane library and Pubmed databases.The search strategy for target articles was not limited by time, age or gender.However, through frequent and thorough searching, it was noticed that there were no published comparative, non-randomized or randomized trials in the literature before 1980.The terms "randomized trials on perioperative warming", "trials on perioperative warming" and "warming in surgical patients" were used in combination with the headings "surgical patients", "forced air warming", "thermoregulation in anesthetized patients" and "warming blankets".Relevant articles referenced in these publications were obtained.The "related article" function was also used to widen the search criteria.All abstracts, comparative studies, randomized trials, non-randomized trials and citations that were firstly scanned through were reviewed comprehensively in accordance with the Quality of Reporting of Meta-analyses (QUORUM) template for the literature search.Each article was critically reviewed to assess its eligibility for inclusion or exclusion in this review.
Statistical analysis was performed by a senior statistician, using the Statistics for Windows software in Microsoft Excel 2007  .The methods used were Hedges G statistic for the calculation of standardized mean difference (SMD), the inverse variance method for the fixed effect model and the DerSimonian/Laired method for the random effect model.The estimate of the difference between the two techniques was pooled depending on the effect weights in the results, which were determined by the variance in each trial estimate.Forest plots were used for graphical displays of results from the meta-analysis: the square around the estimate represents the accuracy of the estimation (sample size) and the line represents the 95% confidence interval.

Abdominal surgery
Temperature, blood loss, shivering, extubation time Vanni et al. 24 30 NA Abdominal surgery Temperature Persson and Lundberg 25 59 Forced warming blanket Gynecological Temperature, pain Melling et al. 14 421 Forced warming blanket General surgery Wound infection El-Rahmany et al. 26 149 Forced warming blanket Cardiothoracic Temperature, cardiovascular vital signs Bock et al. 27 40 Forced warming blanket Abdominal Surgery Temperature, blood loss, stay, cost, transfusion.Wongprasartsuk et al. 28 26 Chart 1. Causes of heterogeneity sidered inadequate and the results from our review may be considered biased.Heterogeneity (clinical and methodological diversity) was seen among all these trials (Chart 1).Limited availability of data on various outcome variables and lack of a major multicenter double blind randomized controlled trial restricted this review with regard to detailed sub-group analysis.However, a subgroup analysis of trials with clearly reported allocation concealment was performed.We felt that performing sensitivity analysis was not relevant due to limited numbers of studies.We attempted to assess publication bias by using funnel plots, but this was difficult to compute due to the small numbers of patients.

Myocardial dysfunction, coagulopathy and stress hormone imbalance
There was insufficient data in the trials available to assess hypothermia-induced myocardial dysfunction, coagulopathy and stress hormone imbalance.

Wound infection
Three trials 14,18,29 contributed towards the combined analysis on the postoperative wound infection rate.In the fixed effect model, the warming group was associated with lower risk of developing postoperative wound

Discussion
Patients in various surgical disciplines are exposed to numerous factors that may alter thermoregulatory mechanisms and result in postoperative hypothermia, including a cold operating theater, cold intravenous fluids, cold blood transfusions, cold antiseptic skin preparations and anesthesia. 1,38,39The latter obliterates behavioral responses and inhibits afferent input, thereby lowering the temperature threshold for thermoregulatory responses to hypothermia and preventing efferent responses. 40Some patients are particularly at higher risk of developing hypothermia: the factors involved include surgery lasting for more than two hours, extremes of age, trauma, abdominal surgery, thoracic surgery, massive transfusions of intravenous fluids or blood and massive blood or fluid loss. 38,39Inadvertent perioperative hypothermia prolongs the recovery time and also increases blood loss, surgical site infection and total hospital stay. 8,39erioperative skin warming has been shown to reduce the initial postinduction hypothermia, intraoperative hypothermia and postoperative shivering, even for procedures lasting for more than three hours. 9urthermore, a single hour of preoperative skin surface warming has been reported to reduce the rate at which core hypothermia developed during the first hour of anesthesia. 33Our analysis shows that the no-warming group is at significant risk of developing perioperative hypothermia, which in turn can give rise to significant perioperative morbidity.
Perioperative systemic warming, in addition to standard forced warm air intraoperative warming, significantly reduces blood loss and complications in patients. 17These findings corroborate those from the independent studies of Schmied et al. 11 and Winkler et al. 34 In the latter study on blood loss following total hip arthroplasty, even a small difference in median core intraoperative temperature of 0.5 °C resulted in significantly less blood loss among the patients who were warmed.This excessive blood loss in hypothermic patients is due to hypothermia-induced coagulopathy 41,42 that results from impaired platelet aggregation and prolonged bleeding time.Bleeding time depends on several variables, including the number and function of platelets, white and red cell counts, vascular factors, hormones and temperature.Although studies have been widely conducted, the bleeding time test does not strictly correlate with surgical bleeding. 41,43Nonetheless, with standardized techniques and knowledge of the merits and limitations of the bleeding time test, it is useful for diagnosing hemostasis disorders, guiding their therapy and warning of unexpected bleeding complications in surgical patients. 44tensrud et al. 45 evaluated the effects of intraoperative hypothermia on blood transfusion during cardiac surgery.They reported that even though no differences in total blood requirements were reported between pa-tients receiving a normothermic cardiopulmonary bypass and those receiving a hypothermic bypass, the hypothermic patients showed an activated partial thromboplastin time that was prolonged by nearly 8%, compared with patients who were actively warmed.No differences were observed in prothrombin time and fibrinogen concentrations.Our study confirms that perioperative warming can significantly reduce bleeding following surgery and that it may be recommended for regular use.
The risk of wound infection in patients undergoing colonic surgery ranges from 9-27% 46 and it may be reduced by two-thirds among patients who receive perioperative warming. 8,46By extending the warming Figure 6.Blood loss: combined analysis of the randomized controlled trials in this review.

Favours warming
Favours no-warming Zhao et al 21 Xu et al. 23 Persson and Lundberg 25 Bock et al. 27 Schmied et al. 11 Total (fixed effect) Total (random effects) -3.50 -3.00 -2.50 -2.00 -1.50 -1.00 -0.50 0.00 0.50 Standardized mean difference period, to two hours before and after surgery, the incidence of wound infection can be further reduced from 27% to 13% and overall complications can be reduced from 54% to 32%. 17 Our review concludes that perioperative warming can significantly reduce the incidence of wound infection.
There was significant heterogeneity among the trials (Chart 1).There may be many reasons for heterogeneity, including combined analysis on trials from various surgical disciplines, combined analysis on trials in which different types of anesthesia (general, spinal or combined epidural and spinal) are used in variable doses and inclusion of trials in which warming was given to different parts of the body.The results from the studies included in this review were also inconsistent.No major multicenter, randomized, controlled trial was reported in the literature.Thus, it was difficult to find high quality, unbiased data for analysis.Nonetheless, this is the only reported systematic review on the role of perioperative warming among surgical patients.

Conclusion
In conclusion, perioperative warming of surgical patients is effective for reducing postoperative wound pain, wound infection and shivering.Systemic warming of surgical patients is also associated with less perioperative blood loss, by preventing hypothermia-induced coagulopathy.Perioperative warming may be given routinely to all patients in various surgical disciplines in order to counteract the consequences of hypothermia.

Favours warming
Favours no-warming Zhao et al 21 Xu et al. 23 Vanni et al. 24 Camus et al. 32

Figure 1 .
Figure 1.Quality of Reporting of Meta-Analyses (QUORUM) diagram template used in this review and results from the retrieval of randomized controlled trials (RCT).

Figure 2 .
Figure 2. Hypothermia: combined analysis of the randomized controlled trials in the review.

Figure 3 .
Figure 3. Postoperative pain: combined analysis of the randomized controlled trials in this review.

Figure 4 .
Figure 4. Wound infection: combined analysis of the randomized controlled trials in this review.

Figure 5 .
Figure 5. Shivering: combined analysis of the randomized controlled trials in this review.

Table 1 .
Characteristics of included trials

Table 5 .
Wound infection: combined analysis

Table 6 .
Trials on postoperative shivering: combined analysis * limb a of trial; † limb b of trial.
limb a of trial; † limb b of trial.