Effects of perioperative hypothermia and warming in surgical practice

Abstract

Perioperative hypothermia is common and adversely affects clinical outcomes due to its effect on a range of homeostatic functions. Many of these adverse consequences are preventable by the use of warming techniques. A literature search was conducted to identify relevant published articles on perioperative hypothermia and warming. The databases searched include MEDLINE (1966 to February 2005), EMBASE (1974 to February 2005), CINAHL, the Cochrane library and the health technology assessment database. Reference lists of key articles were also searched. The primary beneficial effects of warming are mediated through increased blood flow and oxygen tension at tissue level. Reduction in wound infection, blood loss and perioperative pain with warming is promising. However, more evidence from good‐quality prospective randomised controlled trials is needed to evaluate the role of warming in improving overall morbidity, mortality and hospital stay as well as to clarify its role as an adjunct to resuscitation and during the pre‐hospital transport phase of critically ill patients. Awareness of the risks of perioperative hypothermia is the key to prevention. Achieving normothermia throughout the patient's journey is a worthwhile goal in surgical patients.

Introduction

Heat as a therapeutic modality has been used in medical practice across cultures for millennia. The use of heat to cauterise wounds and its application as topical compresses in various forms for pain relief are well established. Warm herbal poultices, oil massages, steam baths and moxibustion, a form of local warming used to stimulate acupuncture points are other examples of the use of therapeutic heat in various civilisations (1, 2). In the more recent past, heat, as a prime modality, has been used to treat many infectious diseases. In the early 20th century, induced episodes of fever called fever therapy were used with some success in treating syphilis and gonorrhoea (3). Body and environmental temperature fluctuations have been observed to consistently influence the clinical course of certain infectious, metabolic and immunologic diseases (3).

Perioperative hypothermia is common and is estimated to occur in about half of all surgical patients 4, 5, 6). Over the past few decades, research into perioperative temperature homeostasis has unravelled the wide range of pathophysiological adverse effects that hypothermia could have on surgical patients – some of which may extend well beyond the immediate postoperative period. Maintenance of normothermia has also been shown to be an effective way of avoiding or treating many of these complications and improving outcomes in a variety of clinical contexts.

This review addresses aspects of thermal homeostasis relevant to surgical patients, considers the adverse effects of hypothermia, discusses methods of delivering heat, and appraises laboratory and clinical evidence of the effects of use of therapeutic heat in surgical patients. The use of heat to treat musculoskeletal disorders, sports injuries, chronic wounds and chronic pain is beyond the remit of this review.

Perioperative Thermal Homeostasis

The hypothalamus, which is the controlling and co‐ordinating centre for the autonomic nervous system, acts as a thermostat to maintain body temperature within a narrow physiological range. Vasodilatation and perspiration are the dominant physiological mechanisms of heat dissipation while vasoconstriction, shivering and piloerection are the main heat‐conserving mechanisms.

A combination of abolition of behavioural adaptive regulation while under anaesthesia, impairment of thermoregulatory vasoconstriction, direct peripheral vasodilatory effect of most anaesthetics, exposure of tissues to the cold ambient temperature and cold infusions result in a core temperature reduction of 1–3°C in most surgical patients. Both general and regional anaesthesia can cause hypothermia.

Core temperature changes occur in three recognisable phases during anaesthesia. An initial precipitous drop of 1–1·5°C occurs in the first hour following redistribution of heat from the core to the periphery. A slower linear reduction over the next 2–3 hours occurs due to a higher heat loss relative to metabolic heat production. The final phase, which commences after 3–4 hours, is characterised by a core temperature plateau during which the core temperature is maintained at a relatively steady state.

Effects of Hypothermia

Many biological systems in warm‐blooded animals are temperature dependent. Within physiological limits, increasing the temperature increases the activity within a system.

In certain circumstances, hypothermia may have a protective effect by reducing basal metabolic rate and oxygen consumption, and with it the risk of tissue hypoxia and ischaemia. This effect has been exploited clinically in cardiac surgery, neurosurgery and organ transplantation, wherein maintenance of hypothermia is the standard of care in selected situations. Hypothermia also reduces the risk of developing malignant hyperthermia whilst under anaesthesia.

Apart from these proven advantages of hypothermia, a growing body of evidence supports the view that core hypothermia in surgical patients has deleterious clinical consequences. Studies have identified core hypothermia as an important predictor of a poor outcome both after trauma and in the perioperative setting (6, 7).

Hypothermia has been shown to affect molecular interactions and cellular functions in a number of systems including coagulation, the immune and endocrine systems. These effects, either alone or in combination have also been shown to translate to serious clinical consequences.

Haematological Effects

Viscosity and haematocrit

Hypothermia increases the viscosity of blood, which may contribute to an impairment of perfusion (8). The haematocrit rises by 2% for every 1°C rise in temperature. These falsely high haematocrit values may be misleading in a hypothermic patient with blood loss.

Effects on the coagulation cascade

Hypothermia slows the enzymatic reactions involved in both the intrinsic and extrinsic pathways.

• 1Hypothermia has the effect of prolonging both the prothrombin time and partial thromboplastin times. Rohrer (9) estimated the prothrombin and partial thromboplastin times on pooled samples of plasma from healthy subjects, at different controlled temperatures in vitro (28°C, 31°C, 34°C, 37°C, 39°C and 41°C). There was a statistically significant prolongation of the prothrombin time and partial thromboplastin time at each of the hypothermic temperature ranges when compared to the results at 37°C.
• 2Bunker (10) found a prolongation of clotting time in a study of ten patients undergoing non cardiac operations under hypothermia, when the tests were done at the respective hypothermic in vivo temperatures rather than at 37°C.
• 3Hypothermia may also increase fibrinolysis, but this seems to be a feature only with severe degrees of hypothermia (11, 12).

Effects on platelets

Hypothermia causes thrombocytopenia due to sequestration mainly in the liver, but also in the spleen (11, 13). Hypothermia also impairs platelet function due to defective thromboxane B₂ synthesis. This has been shown to result in reversible prolongation of bleeding times (14, 15).

Effects on the Immune System

Hypothermia is immunosuppressive as evidenced by the reduced resistance to dermal infections (16, 17) and augmented intraperitoneal tumour growth (18) demonstrated in animals. This is mediated by hypothermia impacting on many domains of the immunological system.

Migration and mitogenic response

In vitro incubation of leucocytes at low temperatures suppresses leucocyte migration and mitogenic response 19, 20, 21). Expression of neutrophil‐adhesion molecules CD11b, CD11c and CD18 is delayed by hypothermia (22, 23).

Phagocytosis

In vitro studies have shown a reduction in phagocytic capacity of neutrophils with hypothermia (19). In vitro neutrophil phagocytosis increases slightly with increase in temperature from 32°C to 37°C, but rises markedly when temperature is increased from 37°C to 40°C (24).

Free radical generation

The stress of surgery and anaesthesia has been shown to reduce the generation of oxygen‐free radicals measured in an intraoperative sample of blood to 56% of preoperative levels (24). In this randomised controlled study involving ten patients undergoing curative surgery for colorectal cancer, Wenisch and colleagues (24) showed that intraoperative core temperatures correlated linearly with the capacity of the leucocytes to generate reactive oxygen intermediates. Production decreased fourfold over a 4°C temperature range. In vitro studies have shown that over a temperature range of 0°C to 40°C, the generation of reactive oxygen intermediates by neutrophils harvested from normothermic healthy subjects, increased approximately by 20‐fold. In the more clinically relevant temperature range of 32°C to 40°C, the increase was about sixfold.

Cytokine production

Hypothermia suppresses interleukin‐1 and interleukin‐2 production (21). Generation of interleukin‐6 and tumour necrosis factor‐α, which are pro‐inflammatory cytokines, is also suppressed by hypothermia. 25, 26, 27). Production of interleukin‐8, a very potent and specific neutrophil chemoattractant, is reduced by hypothermia (28). Hypothermia increases the production of interleukin‐10 which is an anti‐inflammatory cytokine with immunosuppressive properties (29).

Antibody production

T‐cell‐mediated antibody production has been shown to be impaired by hypothermia in animal experiments (30).

Complement activation and C‐reactive protein production

Activation of complement and the levels of C‐reactive protein are attenuated by hypothermia (22, 31).

Effects on the Cardiovascular System and Solid Organs

Reduced splanchnic blood flow and suppression of liver function caused by hypothermia causes prolongation of certain anaesthetic drugs, such as pancuronium, d‐tubocurare and morphine, which depend on the metabolic or excretory functions of the liver (32). Reduction in renal blood flow and glomerular filtration rate, caused by hypothermia, similarly prolong the effects of pancuronium and d‐tubocurare.

Hypothermia results in a decrease in cardiac output. At 30°C, the cardiac output is reduced by 30%. Thermoregulatory vasoconstriction has been shown to reduce leg blood flow, cause venous stasis and result in greater desaturation of venous blood in a study of five volunteers (33). Hypothermia triggers vasoconstriction. During rewarming, the capacitance of the circulatory system increases. This may unmask a relative hypovolaemia, which had been compensated by vasoconstriction, and a distributive shock may result (34).

Patients with intraoperative hypothermia have higher serum norepinephrine concentrations, more pronounced peripheral vasoconstriction and higher arterial pressures in the early postoperative period (35). In one study, when compared to normothermic patients, patients with intraoperative hypothermia had a threefold higher incidence of myocardial ischemia on ECG and a 12‐fold higher incidence of angina postoperatively (36).

Measuring and Delivering Heat

Core temperature is measured reliably at the tympanic membrane, nasopharynx, distal oesophagus or pulmonary artery. Though axillary, oral, rectal, bladder and forehead skin temperature measurements can be performed clinically, they may not reflect core temperatures very accurately. Rectal temperature tends to have a time lag with core temperature changes, while oral and axillary measurements underestimate it. Skin‐surface temperatures tend to be on an average, 2–4°C lower than core temperature and vary widely with the site being sampled and the environmental temperature.

Though a core temperature below 36°C is generally accepted as the standard definition of hypothermia, some authors have used a cut‐off of 36·4°C. Hypothermia is further arbitrarily classified into mild (36–32°C), moderate (31·9–28°C) and severe (<28°C). Kirkpatrick (37) has suggested a system that classifies hypothermia into four classes: Class I (36–35°C), Class II (34·9–32°C), Class III (31·9–28°C) and Class IV (<28°C).

Heat may be delivered to the body by conduction, convection and radiation. Passive warming supports heat retention primarily by providing insulation and prevention of further heat loss. Active warming increases the total heat content of the body by increasing its production or by net transfer of heat from an external source.

Heat delivery systems may further be classified into systemic or local. Depending on the mechanism of heat generation and transfer, they may be classified into mechanical, chemical or electrical systems, which may be applied intermittently or continuously.

Table 1 summarises the various methods of heat delivery which have been used clinically. Invasive methods, such as cavity irrigations and extracorporeal techniques, are used in special circumstances such as cardiac surgery and in the treatment of severe accidental hypothermia. In general surgical patients, passive warming with blankets, active external warming with forced air, circulating water blankets or mattresses, mattresses or blankets with electrical heating elements and warming of infusions are commonly used and are appropriate.

Table 1.

Warming strategies

Passive warming	Active external warming	Active core warming
General	Systemic	Endogenous
		Amino acid infusion*
1. Simple blankets	1. Warming by contact
Single	Warming mattress/blankets/pads	Exogenous
Multiple (These may be warmed or unwarmed)	• Circulating water mattress/pad	Warm infusion
	• Incorporated heating elements	Warm inhalation
	Electric blankets	Warm lavage*
2. Reflective blankets	Resistive carbon‐fibre mattresses	• Thoracic
Have an additional layer of insulation	(Inditherm Medical Products, Rotherham, UK)	• Peritoneal • Gastric
	Forced air warming devices	• Colonic
	• Bair‐Hugger (Arizant Health care, Eden Prairie, MN, USA)	• Bladder
	• Thermacare convective warming system (Gaymar Industries, New York, NY, USA)	Haemodialysis* Venovenous bypass* Arteriovenous bypass*
	2. Radiant warming	Extracorporeal
	Radiant room heaters Infrared lamps*	Cardiopulmonary bypass*
Local	Local
Gloves	Radiant heat dressing
Socks	• Warm‐up therapy (Augustine Medical)
Stockings	Exothermic wound dressings
	• Warm‐up acute wound care† (Arizant Medical, Eden Prairie, MN, USA)
	Incorporated heating elements
	• Conductive carbon polymer pads (Inditherm Medical Product)†
	Gloves with warming elements‡
	Wax bath‡
	Water bath‡
	Hot fomentation (soaked towels, water bottles, microwaveable gels, etc)‡
	Infrared source‡

^* Items used in special circumstances.

^† Emerging technologies or products.

^‡ Techniques in use, but generally out with the perioperative setting.

While passive warming used alone is not an efficient way of preventing or treating hypothermia in surgical patients, active warming has been shown to be effective both in prevention and treatment in the perioperative setting.

Pre‐emptive skin‐surface warming for 1–2 hours preoperatively is effective in reducing the initial redistribution hypothermia during anaesthesia (38). Combined preoperative and intraoperative surface warming is better than intraoperative warming alone in preventing hypothermia in the first 2 hours of anaesthesia (39).

Forced air warming has been shown to be superior to passive warming and warming with water flow systems or radiant heat. Resistive heating with carbon‐fibre‐incorporated blankets is as effective as forced air warming (40).

Effects of Warming on Cells and Tissues

Warming can have both systemic and local effects, which may serve to prevent or reverse the consequences of hypothermia. Locally, warming is thought to exert its effects primarily by improving local blood flow and consequently oxygen availability. However, it also has direct effects at a cellular level (by which it influences metabolism and proliferation) and molecular level (by which it may influence many enzyme systems, e.g. the coagulation cascade).

Intermittent radiant warming has been shown to increase fibroblast (41) and endothelial cell (42) proliferation and metabolic activity in vitro when compared to cells not exposed to warming. Application of radiant warming has been claimed to reduce the inhibitory effect of chronic wound fluid on the growth of dermal fibroblasts (43).

Rabkin and Hunt (44) in a landmark study in 1987, which evaluated the effect of local hyperthermia on subcutaneous oxygen tension and blood flow, found a linear correlation between subcutaneous tissue temperature and oxygen tension. The mean increase in subcutaneous temperature was 4°C. The mean increase in subcutaneous oxygen tension (PsqO₂) on warming was 80% of baseline values. The increase in PsqO₂ per degree rise in subcutaneous temperature a veraged 10·7 mmHg. Estimated blood flow in the heated tissues was three times that in unwarmed tissues under the study conditions. Ikeda (45) in a similar study in healthy human volunteers demonstrated a 50% increase in subcutaneous oxygen tension with an intermittent radiant heat source set at 38°C. Interestingly, the increase in the PsqO₂ was sustained for up to 3 hours after the cessation of heating.

The increase in subcutaneous oxygen tension produced by systemic warming is not completely reversed even after 1 hour of local cooling as shown by Sheffield and colleagues (46) in a study of five volunteers. While systemic cooling significantly reduced tissue oxygen tension, local warming of an extremity resulted in the return of oxygen tension to baseline values (46).

Clinical Applications of Perioperative Warming

Effect of Warming on wound infection

Body and ambient temperature have been observed to be important determinants in not only the occurrence of certain infections but also the time patterns and spatial distribution of the manifestations of the infections (3). A range of pathogens including bacterial, viral, fungal, protozoal, rickettsial and spirochaetal organisms have been shown to be susceptible to increases in body temperature (3, 47, 48). In addition to the positive effect of warming on the cells of the immune system, a mechanism by which warming could interfere with bacterial iron metabolism has been observed (49, 50). Some clinically important pathogens that cause wound infections such as E. coli and Pseudomonas, are dependent on iron. Elevated temperature, through a number of mechanisms, causes an increase in iron demand by bacteria as well as a decrease in availability of free iron resulting in an inhibitory effect on the organisms.

Role in prophylaxis of wound infections

In addition to inherent patient factors, the intraoperative and immediate postoperative periods are critical in determining the occurrence of wound infection. While antibiotics have been shown to be useful in prophylaxis in clean‐contaminated and contaminated surgical wounds, their efficacy in clean non implant wounds has not been consistent. Warming, by virtue of its effects on improving blood flow and oxygen tension in the wounds, has the potential to be useful as an adjunct or as a substitute to antibiotics in prophylaxis. Both systemic and local warming used in the perioperative environment have been shown to be effective in prophylaxis of infections.

Systemic warming

Kurz and colleagues (51) evaluated the effect of systemic intraoperative warming on wound‐infection rates after colorectal surgery in 200 patients by a randomised controlled trial with a blinded outcome assessment. The hypothermia group had a final mean core temperature of 34·7°C compared to 36·6°C in the warmed group. Follow up was for 3 weeks and culture‐positive pus defined a wound infection. The wound‐infection rate was three times higher in the hypothermia group (19%) than the warmed group (6%).

There are studies that have failed to find any significant differences in wound outcomes between hypothermic and normothermic patients 52, 53, 54, 55). However, these were either of a retrospective design, used inappropriate models or were of a poor quality, making interpretation of data difficult and less reliable.

Local warming

Melling et al. (56) recruited 421 patients undergoing clean surgical procedures (hernia repairs, breast operations and surgery for varicose veins) in a randomised controlled trial into three groups: local warming (30 minute preoperative warming with non contact radiant heat), systemic warming (30 minute preoperative warming with a forced warm air blanket) and control (unwarmed). Follow up was for 6 weeks by a blinded observer. There was a statistically significant decrease in wound‐infection rates in the two warmed groups compared to the unwarmed group (4 and 6% versus 15%).

Warming and tissue viability

Tissue ischaemia is an important factor in pressure ulcer development. It is claimed that the seeds for pressure sore development are often sown in the operating theatre, as many of the risk factors converge in this environment.

Intraoperative systemic warming, by improving cutaneous blood flow and oxygen tension, may confer some protection. This hypothesis has been tested in a randomised controlled trial involving patients undergoing a range of orthopaedic, general surgical and urological procedures (57). The warmed group (n = 161), who received perioperative forced air warming, using the Bair‐Hugger device (Augustine Medical, Eden Prairie, MN, USA), and the control group (n = 163) were comparable in terms of age, comorbidity and length of surgery. The incidence of pressure ulcers was 10·4% in the control and 5·6% in the warmed group corresponding to an absolute risk reduction of 4·8% and a relative risk reduction of 46%. This, however, did not reach statistical significance. The numbers needed to treat was 21.

It might be worthwhile to investigate if extending the period of warming into the perioperative period would result in a further reduction in the incidence of pressure ulcers.

Effect of warming on morbidity, mortality and hospital stay

In a retrospective review of outcomes in 262 patients who had elective abdominal aortic aneurysm repair, patients who had intraoperative hypothermia, were found to have a significantly higher fluid, transfusion, vasopressor and inotrope requirements with eventually higher incidence of organ dysfunction and death. Length of stay in the intensive care unit and total hospital stay were also significantly higher in the hypothermic patients (58).

Kurz and colleagues (51) reported that in patients undergoing colorectal procedures, the duration of hospitalisation was prolonged by 2·6 days in their patients who were hypothermic when compared to the normothermic group. The time to oral feeding and suture removal was also delayed by a day in the hypothermic group.

Another randomised controlled trial has compared standard intraoperative forced air warming (control) with an extended period of perioperative warming (warmed group). The patients in the warmed group were placed on a resistive carbon fibre warming mattress set at 40°C for 2 hours preoperative in the ward. They were kept on the mattress during surgery and for 2 hours postoperatively. The outcome measures were postoperative morbidity (wound related, regional and systemic morbidity) and 30 day mortality. This was compared to the POSSUM‐predicted morbidity and mortality. Interim results are reported to favour the warmed group with a significant reduction in overall morbidity (59).

Warming and blood loss

Hypothermia, by virtue of its effects on coagulation and platelets, is associated with an increased operative blood loss. Several studies have documented the effectiveness of rewarming in reversing the coagulopathy induced by hypothermia in animals (11, 13, 14).

The value of warming in reducing blood loss has been evaluated in a randomised controlled trial of 60 patients undergoing total hip arthroplasty (60). Patients in the warming group had a statistically significant lower blood loss than the hypothermic group (Mean of 1·7 ± 0·3 l versus 2·2 ± 0·5 l, respectively). Significantly higher transfusion requirements were noted in the hypothermia group.

Widman et al. (61) used an amino acid infusion to induce thermogenesis 1 hour before and during spinal anaesthesia in 22 patients undergoing hip arthroplasty and compared core temperature changes and perioperative blood loss with 24 controls. The warmed group suffered a 0·4°C drop in end‐operative core temperature compared to a 0·9°C reduction in the control group (P < 0·05). Blood loss during surgery was significantly larger in the control patients (702 ± 344 ml) than in the patients receiving amino acids (516 ± 272 ml; P < 0·05). While at least one additional study has corroborated these results (62), another study (63) did not find any difference in blood loss between the warmed and control groups.

There are concerns that perioperative blood transfusion may have immunosuppressive properties. Whilst this is perceived as an advantage in selected contexts like transplantation surgery, in most clinical situations, it is detrimental. For example, perioperative blood transfusion has been claimed to increase the chance of tumour recurrence and reduce survival rates in colorectal cancer (64, 65). Blood transfusion may also amplify the immunosuppressive effects of perioperative hypothermia. This may potentially increase the postoperative morbidity, particularly infectious complications, and affect survival rates in patients with malignancy who are exposed to both these insults in the perioperative period. While further studies are required to clarify the magnitude of the effects of combined hypothermia and blood transfusion on outcomes, it is encouraging that achieving normothermia may have the potential to prevent the synergistic effects of both the adversities.

Warming and perioperative pain

Poorly controlled pain causes sympathetically mediated vasoconstriction, which has been shown to reduce the oxygen tension in wounds (66). Reduction of pain may potentially be another mechanism by which warming might serve to improve wound outcomes.

In chronic wounds, a few studies have reported significant reductions in pain when patients were placed on warming therapy (67, 68). Continuous warming has also been demonstrated to be effective in treating back pain (69), wrist pain (70) and dysmenorrhea (71). At least four randomised controlled trials involving patients with symptomatic cholelithiasis (72), renal colic (73), acute back pain (74) and minor trauma (75) have shown a significant reduction in admission pain scores when patients were warmed during emergency transport in the pre‐hospital phase.

With reference to postoperative pain in general surgical patients, interim results from one randomised controlled trial have shown reduced pain scores in patients after inguinal hernia surgery, when local warming was applied in the immediate postoperative period and intermittently thereafter for 3 days (76). However, studies have not found any significant differences in overall analgesic requirement between warmed and hypothermic patients (51, 77).

While it is unlikely that warming would be used as the as a prime modality for pain control in the perioperative period, the additional positive effects on wound perfusion, oxygen dynamics and infection prophylaxis seem persuasive enough to consider using warming as an adjunct in pain relief.

Warming and thermal comfort

Hypothermia causes an uncomfortable perception of cold and is often remembered by patients as one of the worst aspects of their perioperative experience. It may, in addition, result in shivering which is also unpleasant in the postoperative period. Shivering increases myocardial work load and oxygen consumption, raises blood and intraocular pressure, which may all contribute to morbidity. While pharmacological interventions remain the cornerstone of management of shivering, maintaining normothermia in the intraoperative and postoperative period has been shown to improve thermal comfort and reduce the incidence and severity of shivering 78, 79, 80, 81, 82).

Active warming during transfer to hospital has also been shown to result in better thermal comfort, reduction in anxiety and improvement in overall patient satisfaction in a randomised controlled trial, when compared to passive warming (75).

Effect of warming on postoperative recovery

Hypothermia delays the metabolism of many anaesthetic agents (e.g. vecuronium, neostigmine and propofol) and may potentiate the toxicity of others (e.g. bupivacaine). Hypothermia has been shown to delay immediate postoperative recovery, resulting in an extended stay in the recovery room.

Both prospective and retrospective studies have found a correlation between length of stay in the recovery room and admission temperatures (83, 84). A prospective controlled study also found a 33% increase in the time spent in the recovery room, which was directly attributable to hypothermia, for a control group when compared with warmed patients (85). Maintaining normothermia intraoperatively may contribute towards efficient management of resources in the recovery room.

Warming as an adjunct to resuscitation

Patients with surgical emergencies such as peritonitis and haemorrhagic shock are often metabolically and haemodynamically compromised. Conventionally resuscitative measures have focused on cardiorespiratory and renal optimisation consisting of replacing fluids, correcting acid – base and electrolyte abnormalities and supplementing oxygen. Hypothermia, a frequent accompaniment in such situations is an often neglected aspect of resuscitation and may contribute to the poorer outcomes.

The value of systemic warming aimed at preventing hypothermia during resuscitation was evaluated in a randomised controlled trial involving patients presenting with peritonitis (86). Twenty‐seven patients with generalised peritonitis were randomly allocated to either a warmed group, who were placed on a resistive carbon‐fibre mattress set at 40°C on admission or a control group which was not warmed. Both groups were comparable, in terms of their admission Apache II scores and received the same standard resuscitative measures. Warming was continued for 24 hours or up to the time of surgery, whichever was earlier. The Apache II score was used as an outcome measure. There was an improvement in the median Apache II scores in the warmed group at the end point of the study which translated to a 3·5% reduction in predicted mortality. In contrast, the Apache II scores deteriorated in the control group with a corresponding increase in predicted mortality of 1·5%. Though these changes were not statistically significant, the study underscores the potential for warming as an adjunct to standard resuscitation tools.

Warming during patient transfer

Surgical patients who are critically ill may have disordered thermoregulatory mechanisms, making them vulnerable to the onset of hypothermia. This is particularly significant if they are likely to be exposed to low environmental temperatures which can occur during the process of medical care. This may, for example, happen in the pre‐hospital phase or during intrahospital transfer for therapeutic or diagnostic radiological procedures, such as CT or MRI scans which are usually performed in colder ambient temperatures than a ward environment.

The efficacy of active warming in preventing core hypothermia in critically ill patients during their intrahospital transfer has been evaluated in a randomised controlled trial recruiting 30 patients (87). The actively warmed group maintained their core temperatures, whereas the mean core temperature of the passively warmed group sustained a fall of 2°C.

Another study evaluated the value of active warming with a resistive heating blanket compared to passive insulation during pre‐hospital transfer in patients with minor trauma. In the unwarmed group the mean core temperature decreased by 0·4°C, while in the actively warmed group, the mean core temperature increased by 0·8°C (75). These studies highlight the fact that even small windows of exposure in the pre‐hospital phase or within a hospital environment could cause hypothermia and that it can be effectively prevented by warming.

Conclusion

Thermal homeostasis seems to be closely linked to many other homeostatic mechanisms crucial for survival, and this integration is evident across cellular and extracellular levels involving many organ systems. Non physiological deviations of temperature in surgical patients whose homeostatic systems are often challenged in the perioperative period are hence poorly tolerated. Fortunately, the very same thermoregulatory mechanisms which are so vulnerable to environmental and iatrogenic influences are also amenable to simple supportive manipulation, as has been shown by the several studies in diverse clinical contexts. Reduction in wound infection, blood loss and perioperative pain with warming is promising. However, more evidence from good‐quality prospective randomised controlled trials is needed to evaluate the role of warming in improving overall morbidity, mortality and hospital stay as well as to clarify its role as an adjunct to resuscitation and during pre‐hospital transport phase of critically ill patients. With a range of options available for its application in different situations, therapeutic heat is a useful addition to the therapeutic armamentarium not only in the operating theatre but throughout the patient's journey.