Perioperative temperature management

INTRODUCTION — Core body temperature is tightly regulated and is normally maintained within a few tenths of a degree Celsius. Thermoregulation is impaired during either general or neuraxial anesthesia. Consequently, unwarmed anesthetized patients become hypothermic, typically by 1 to 2°C, with core temperatures near 34.5°C. Trials indicate that temperatures <35.5°C are associated with various complications in surgical patients. In contrast, the incidence of substantive complications changes little in the range between 35.5 and 37°C. Thus, core temperature should thus be kept at least 35.5°C. Although intraoperative hyperthermia also occurs, this is less common than hypothermia.

This topic will review thermoregulatory physiology, intraoperative heat balance, consequences of hypothermia, and prevention and management of thermal disturbances.

During cardiac surgery with cardiopulmonary bypass, mild (32 to 35°C), moderate (28 to 32°C), or deep (<28°C) hypothermia may be used in selected patients as a protective strategy for the brain and vital organs. Management of cooling and hypothermia during cardiopulmonary bypass and subsequent rewarming to reestablish normothermia are discussed separately. (See "Management of cardiopulmonary bypass", section on 'Temperature'.)

TEMPERATURE MONITORING — Surgical patients should be kept at a core temperature of at 35.5°C [1]. Temperature derangements (hypothermia or hyperthermia) can only be detected by temperature monitoring. Intraoperative electronic alerts may draw attention to temperature derangements thereby enhancing efforts to maintain normothermia [2].

Whom to monitor — We monitor temperature in patients having general anesthesia lasting more than 30 minutes, major surgery with neuraxial anesthesia, or when clinically significant changes in temperature are intended, anticipated, or suspected, similar to standards of the American Society of Anesthesiologists (ASA) and other professional societies [3-8] (table 1). Temperature monitoring is not typically necessary during mild or moderate sedation, or if a peripheral nerve block is used without general anesthesia because thermoregulatory control is well maintained in these circumstances. (See "Basic patient monitoring during anesthesia", section on 'Temperature monitoring'.)

Site selection for monitoring — When feasible, we monitor core temperature at highly perfused sites where temperature is homogeneous and high compared with the rest of the body. Site examples include the nasopharynx (with the temperature probe inserted 10 to 20 cm into the nares), distal esophagus, tympanic membrane with a thermocouple, sublingual fossa, or pulmonary artery [9,10]. (See "Basic patient monitoring during anesthesia", section on 'Temperature monitoring sites'.)

In some circumstances, it is necessary to use an alternative temperature monitoring site (eg, axillary, bladder, forehead) during general anesthesia with a facemask or supraglottic airway, during neuraxial anesthesia, or during certain surgical procedures such as oral or esophageal surgery [10]. Uncompensated skin temperature is a poor alternative since it is generally lower than core temperature (although forehead temperature is less variable than other cutaneous sites) [10-13]. Specifically, adding a constant to skin temperature does not convert it to core temperature because the skin-to-core temperature difference varies. Temperatures in the external aural canal and rectum also equilibrate poorly with core temperature.

NORMAL THERMOREGULATION — Normally, body temperature is tightly regulated in the core compartment (primarily in the trunk and head), which constitutes approximately one-half the body mass and dominates central thermoregulatory responses [14]. Peripheral tissues (primarily the arms and legs) act as a thermal buffer and vary over a fairly wide range. For example, peripheral tissues are normally 2 to 4°C cooler than the core in patients exposed to hospital ambient temperatures. A consequence is that peripheral tissues can absorb or release heat when necessary, thus obviating the need for shivering or sweating in response to every small change in ambient temperature.

●Response to cold – The primary human thermoregulatory defenses in response to cold are shivering and arteriovenous shunt vasoconstriction in specialized thermoregulatory structures in the fingers and toes that regulate arm and leg blood flow and, thus, heat loss to the environment [15,16]. In adults >80 years old, cold-defense thresholds are reduced approximately 1°C (figure 1) [17]. In infants, but not in adults, non-shivering thermogenesis is an additional thermoregulatory defense [18].

●Response to heat – The primary thermoregulatory defenses to heat are active precapillary vasodilation and sweating [15,19].

The core temperature triggering each response defines the threshold for that defense. Normally, the sweating and vasoconstriction thresholds differ from each other by only a few tenths of a degree Celsius, centered around normal temperature which averages about 37°C. The shivering threshold is 1°C below the vasoconstriction threshold (figure 2). However, intraoperative models of the human thermoregulatory system are not as simple as an "on/off," thermostat because of sequential activation of defenses as well as the effects of anesthesia on thermoregulatory control, as noted below [9]. (See 'Effects of general anesthesia' below and 'Effects of neuraxial or regional anesthesia' below.)

ANESTHETIC EFFECTS ON THERMOREGULATION — Both general and neuraxial anesthesia profoundly impair thermoregulatory control, with the consequence that unwarmed surgical patients become hypothermic [10].

Effects of general anesthesia — Inhalation anesthetics, including volatile agents (eg, sevoflurane, desflurane, isoflurane [17,20,21], as well as nitrous oxide [22]), and intravenous (IV) anesthetics including sedative-hypnotic agents (eg, propofol [23], dexmedetomidine [24-26]) and opioids [27] impair thermoregulatory control in a dose-dependent fashion. Some sedatives such as midazolam do not appreciably impair thermoregulatory control [28]. Although the precise mechanisms by which anesthetic agents impair thermoregulatory control are unknown, the following effects are involved:

●Vasoconstriction and shivering – Overall, general anesthesia decreases the vasoconstriction threshold to approximately 34.5°C, and decreases the shivering threshold yet 1°C below the vasoconstriction threshold. Typical combinations and doses of agents used to maintain general anesthesia synchronously reduce the thresholds for vasoconstriction and shivering in a concentration-dependent manner that is linear for IV anesthetic agents, but disproportionately greater at higher concentrations of volatile anesthetics. In addition to decreasing the vasoconstriction threshold, volatile anesthetic agents reduce gain of the response (ie, the extent to which vasoconstriction increases with further reduction in core temperature) [29]. Once thermoregulatory vasoconstriction is triggered, it effectively constrains metabolic heat to the core thermal compartment even during general anesthesia.

Most general anesthetic agents also obscure the normal pattern of shivering and reduce maximum shivering intensity [30]. However, gain and maximum intensity of shivering are maintained with opioids [31]. Notably, shivering is not possible when patients are paralyzed by neuromuscular blocking agents (NMBAs).

●Sweating – Sweating usually remains nearly intact during general anesthesia [32].

The overall result of these anesthetic effects is that the normally regulated range of core temperatures that spans only a few tenths of a degree Celsius increases 10- to 20-fold during general anesthesia. Anesthetized patients will only activate thermoregulatory defenses when core temperature decreases to below the vasoconstriction threshold (or exceeds the sweating threshold). The figure shows how the interthreshold range (between sweating and vasoconstriction) increases markedly as a concentration-dependent function of various anesthetic agents (figure 3).

Effects of neuraxial or regional anesthesia — Although the local anesthetics used in neuraxial anesthetic techniques (eg, epidural or spinal anesthesia) do not reach the brain, patients receiving neuraxial anesthesia become hypothermic to a degree that is similar to those having general anesthesia. Impairment of thermoregulatory control by neuraxial anesthesia occurs via three mechanisms:

●Hypothermia provokes less thermal discomfort than would be expected in the presence of a neuraxial block [33]. Consequently, patients typically do not complain of feeling cold during epidural or spinal anesthesia even if they are hypothermic. The mechanisms for reduced thermal discomfort are unknown, but are probably due to tonic cold signals from the lower body being blocked, with their absence being interpreted by the hypothalamus as relative warmth.

●Neuraxial anesthesia impairs central thermoregulatory control, reducing the vasoconstriction and shivering thresholds [34]. Changes in the shivering threshold are proportional to the block height (figure 4). The precise mechanisms by which administration of local anesthesia in a location distant from the brain impairs central thermoregulatory control are unknown, but are likely peripherally mediated, and probably due to tonic cold signals from the lower body being blocked, as with the impairment of thermal discomfort by neuraxial anesthesia [35,36]. Central impairment is evident even when epidural anesthesia is induced with a short-acting local anesthetic with a plasma half-life of only seconds such as 2-chloroprocaine, and is thus clearly a peripheral effect [34]. However, the magnitude of central thermoregulatory impairment with neuraxial anesthesia is considerably less than from general anesthesia.

●Since active vasodilation, sweating, vasoconstriction, and shivering are autonomic thermoregulatory defenses that are primarily neurally mediated, each requires intact nerve conduction. Neuraxial anesthesia prevents most efferent and afferent neural activity to and from the lower body to the brain including afferent pain signals and the efferent signals controlling vasoconstriction and shivering. With impaired neurally mediated signaling, the gain and maximum intensity of thermoregulatory defenses are substantially reduced [37].

In contrast with neuraxial anesthesia, peripheral nerve blocks do not have clinically important thermoregulatory effects. Peripheral blocks only prevent local neurally-mediated thermoregulatory responses, which can be compensated for in other regions.

Combined general and neuraxial techniques — Patients having combined general and neuraxial anesthetic techniques are at highest risk for intraoperative hypothermia due to the additive thermoregulatory impairment from each technique [34]. With combined techniques, the vasoconstriction threshold temperature, the gain of vasoconstriction, and the maximum intensity of vasoconstriction are all reduced by the sum of the independent effects of each individual technique. Thus, patients become colder before thermoregulatory defenses are activated and once activated, defenses are less effective in preventing further decreases in core hypothermia compared with patients having either technique alone [34].

INTRAOPERATIVE HYPOTHERMIA

Causes — Intraoperative hypothermia develops in nearly all unwarmed surgical patients and results from the combination of anesthesia-induced thermoregulatory impairment, cool operating room ambient temperature, and exposure of open body cavities during certain surgical procedures. The most important of these factors is thermoregulatory impairment since unanesthetized adults would otherwise resist surgical heat loss via vasoconstriction and shivering. Intraoperative hypothermia develops with a characteristic three-phase pattern (figure 5):

●During the initial hour after induction of general anesthesia or activation of a neuraxial block, core temperature decreases rapidly due to redistribution from the core to peripheral tissues (figure 6 and figure 7) [38,39]. This initial rapid reduction in core temperature is primarily due to anesthetic-induced vasodilation resulting from impairment of central thermoregulatory control rather than direct peripheral effects of anesthetics (see 'Effects of general anesthesia' above). Although this redistribution of heat does not alter mean body temperature, it substantially reduces core temperature [40].

●Redistribution hypothermia is typically followed by a slower, linear reduction in core temperature that results from heat loss exceeding metabolic heat production. The main routes of heat loss to the environment are usually radiation and convection (figure 8). Normally, only small amounts of heat are lost via conduction and evaporation. Unknown amounts of heat are lost through surgical incisions, but the amount may be substantial when incisions are large, and especially during open abdominal surgery. The rate at which temperature decreases is a function of the difference between metabolic heat production and heat loss, which in turn depends on ambient temperature, extent of surgical exposure of skin or body cavities, and whether the patient is insulated or actively warmed.

●Once patients become sufficiently hypothermic to activate thermoregulatory vasoconstriction (typically at approximately 34.5°C during general anesthesia), core temperature reaches a plateau and does not decrease further regardless of the incision size or duration of surgery. A plateau in core temperature may also result from a passive process if heat production is balanced against heat loss. Once activated, arteriovenous shunt vasoconstriction effectively retains metabolic heat in the core tissues, thus preventing further decreases in core hypothermia. However, heat loss from peripheral tissues continues. Consequently, body heat content continues to decrease although core temperature remains constant [41].

Consequences — Most cellular functions are temperature-dependent, and hypothermia also provokes systemic responses. Thus, even mild hypothermia (eg, ≥1 to 2°C reductions in core temperature) may have adverse consequences. Randomized trials evaluating complications of mild perioperative hypothermia are listed in the table (table 2).

Coagulopathy — Hypothermia causes coagulopathy, primarily due to reversible impairment of platelet aggregation via reduced release of thromboxane A3, which impairs formation of an initial platelet plug [42,43]. Hypothermia also reduces the activity of enzymes in the coagulation cascade, which in turn reduces clot formation [44]. The combination of platelet and enzyme impairment typically increases perioperative blood loss and the need for transfusion [45]. In a meta-analysis, even mild hypothermia increased transfusion requirement by approximately 35 percent (figure 9) [46]. However, a recent trial of 5056 patients that was not included in the meta-analysis found that transfusion requirements were nearly identical in patients randomized to 35.5 or 37°C [1].

Hypothermia-induced coagulopathy is often missed with routine laboratory testing because these tests are performed at 37°C, and are thus reported as normal although they would not be if temperature-adjusted.

Infection — There are at least three mechanisms by which perioperative hypothermia impairs host defenses against surgical wound contamination:

●Reduced tissue perfusion to wounded tissue with reduced access for key immune cells (due to the vasoconstriction that occurs to constrain metabolic heat to the core). (See 'Causes' above.)

●Decreased motility of key immune cells (eg, macrophages).

●Reduced scar formation, which is necessary to prevent wound dehiscence and recontamination.

Available evidence suggests that risk of infection is increased if temperature is allowed to drop to ≤34.5°C. For example, a 1996 trial in 200 patients having colorectal surgery noted that incisional infections were less likely in patients maintained at normothermic temperature (36.5°C) compared with those who developed hypothermia ≤34.5°C (6 versus 19 percent) [47]. However, in a 2022 randomized trial that included 5056 patients noted that maintaining target core temperature at 35.5°C was not associated with increased serious or superficial infections compared with maintaining target temperature at 37°C [1].

Prolongation of drug effects — Even mild hypothermia prolongs the action of drugs used during anesthesia, particularly neuromuscular blocking agents (NMBAs) [48]. For example, hypothermia doubles the duration of action of vecuronium (2°C) [49,50], and prolongs the duration of atracurium by 60 percent (3°C) [51]. Similarly, plasma propofol concentrations are increased by 28 percent with hypothermia (3°C), primarily due to reduced hepatic blood flow [51]. Also, hypothermia decreases the minimum alveolar concentration of volatile inhalation anesthetic agents necessary to prevent movement in response to a surgical stimulus. Consequences of delayed drug disposition include delayed reversal or neuromuscular blockade and delayed emergence in hypothermic patients [52]. (See "Delayed emergence and emergence delirium in adults", section on 'Consider prolonged drug effects' and "Overview of the management of postoperative pulmonary complications".)

Shivering — The shivering threshold is a full degree Celsius (°C) below the vasoconstriction threshold. Vasoconstriction is effective even during anesthesia. Consequently, core temperature rarely decreases to the 1°C lower temperature necessary to reach the shivering threshold. Intraoperative shivering is rare for this reason, and because anesthesia impairs activation of thermoregulatory defenses including shivering. By contrast, postoperative shivering is common in hypothermic patients when anesthesia has worn off and no longer prevents activation of thermoregulatory defenses including shivering. Shivering should be avoided since it acutely augments metabolic rate [53], and because patients find it uncomfortable. (See 'Shivering' below.)

Myocardial ischemia — Hypothermia results in sympathetic stimulation with augmentation of plasma norepinephrine concentration (eg, approximately 700 percent at 1.3°C of hypothermia [54]) with consequent increased metabolic rate and hypertension [55,56]. These effects increase myocardial oxygen consumption, particularly if shivering occurs [16,57,58]. In patients with ischemic heart disease, these effects may lead to myocardial ischemia and arrhythmias although there is scant evidence that mild hypothermia causes myocardial injury [59-61]. In a trial in 5056 surgical patients randomized to aggressive rewarming to maintain perioperative target core temperature at 37 °C (actual temperature 37.1 ± 0.3 °C) versus conventional rewarming to maintain 35.5 °C (actual temperature 35.6 ± 0.3 °C), the primary outcome of myocardial injury and a composite outcome of myocardial injury, nonfatal cardiac arrest, or all-cause mortality at 30 postoperative days were similar between groups [1]. Secondary adverse outcomes (serious wound infections, transfusions) and duration of hospital stays were also similar. Keeping core temperature at least 35.5 °C throughout the perioperative period may be sufficient to prevent temperature-related complications.

Prevention and management — Hypothermia develops in nearly all unwarmed surgical patients. Even patients who are actively warmed initially develop transient core hypothermia due to redistribution [40]. Typically, core temperature decreases during the initial hour of anesthesia, and then decreases further or increases depending on ambient temperature, size of the operation, patient morphometric characteristics, and efficacy of insulation or active warming.

Various perioperative warming strategies are available. These can broadly be categorized as prewarming strategies (before induction of anesthesia), passive insulation, or devices that actively warm the skin surface, heat fluids, warm inspired gases, or are endovascular systems that directly exchange heat with the circulating blood.

Prewarming before induction of anesthesia — Redistribution hypothermia can be ameliorated by prewarming patients. Although warming patients before induction of anesthesia or placement of a neuraxial block does not substantially increase core temperature (which remains tightly regulated), the absorbed heat increases peripheral tissue temperature, thereby reducing the normal core-to-peripheral tissue temperature gradient [14,62,63]. Typically, core temperature in prewarmed patients remains approximately 0.4°C warmer than those who are not (figure 10) [64].

Passive insulation — A single layer of passive insulation reduces cutaneous heat loss 30 percent at typical operating room temperatures [65]. The type of insulation is relatively unimportant since the layer of still air trapped below the insulator is actually the major heat-loss barrier. Fully covered skin approximately compensates for the 30 percent reduction in metabolic heat production that accompanies general anesthesia (see 'Effects of general anesthesia' above). Adding additional layers of insulation provides only slight additional benefit. For example, three layers of passive insulation only reduces heat loss by approximately 50 percent (figure 11) [66]. Most surgical patients become and remain hypothermic with only passive insulation.

Active warming devices — In a 2020 meta-analysis of randomized trials in noncardiac surgical patients, active body surface warming to maintain normothermia was associated with reduced risk for shivering (odds ratio [OR] 0.2 95% CI 0.1-0.4), wound infections (OR 0.3, 95% CI 0.2-0.7), and blood transfusion (OR 0.6, 95% CI 0.4-1.0), compared with nonactive warming controls (3974 patients; 54 studies) [67].

Skin surface warming — Most surgical patients are actively warmed via the skin surface because skin is easily accessible, can be safely warmed, and most heat loss is from the skin. Forced-air [68], resistive heating [69], and circulating-water garment devices [69] are commercially available and have comparable efficacy.

Forced-air warming is the most common approach in conventional surgery and procedures that employ laminar flow [47,70]. These systems are effective because modest heat intensity is distributed over a large body surface area. Since the air does not reach dependent regions of the skin, the dangerous combination of heat and pressure on skin is avoided. In the supine position, forced-air warming is more effective on the lower body because a larger body surface area is covered compared with upper body warming [71,72]. Conversely, in the lateral decubitus position, warming is more effective on the upper body [73].

Theoretical concerns have been raised regarding the potential for forced-air warming devices to disturb laminar flow and promote infection risk. However, laminar flow itself paradoxically increases infection risk, apparently by detaching bacterial particles from bedside personnel and driving them directly into the incision. There is no clinical evidence that forced-air warming increases infection risk [74,75], and all major trials showing warming and infection risk were conducted with forced-air warming [1,47,70]. Furthermore, there is no effect of forced-air warming on the number of bacterial colonies that grow when bacterial culture plates are placed near patients who are having surgery with laminar flow [76].

Fluid warming — Fluid warming does not significantly warm patients because intravenous (IV) fluids can only slightly exceed core temperature. Thus, fluid warming does not compensate for initial hypothermia due to redistribution from core to peripheral tissues, nor for subsequent losses from skin surface and surgical incisions. Similarly, warming of irrigation fluids used in the peritoneal cavity does not transfer meaningful amounts of heat to the patient [77].

Fluids of any type should be warmed prior to administration in large volumes (ie, more than 1 L/hour). Patients can be cooled considerably by infusion of unwarmed IV fluid or exposure to unwarmed peritoneal irrigation fluid. For example, each liter of fluid infused at ambient temperature reduces mean body temperature by 0.25°C in a 70 kg patient. A unit of refrigerated blood also reduces mean body temperature by 0.25°C since the unit is approximately one-half a liter but is twice as cold as room temperature crystalloid. Warming blood also renders it less viscous, and thus easier to infuse.

Respiratory gas warming — The heat capacity of air is low. The heat of vaporization (required to humidify dry gases) is higher, but still low compared with the metabolic rate of patients. Consequently, little metabolic heat is lost through the airway, and airway heating and humidification does not transfer clinically important amounts of heat into patients [77].

Endovascular heat-exchange catheters — Invasive endovascular heat-exchange catheters transfer far more heat than surface systems [78]. Their use has been limited to establishing deliberate therapeutic hypothermia because they are invasive and expensive.

INTRAOPERATIVE HYPERTHERMIA — Intraoperative hyperthermia is far less common than intraoperative hypothermia.

Causes — Intraoperative hyperthermia can result from passive causes including (table 3 and table 4):

●Excessive heating with commercial warming systems, particularly during long procedures. Hyperthermia consequent to use of routine warming devices is rare in adults, but occasionally occurs in infants and children [79].

●Inadequate heat loss (eg, high ambient temperature, high humidity, prevention of sweating by impervious body coverings such as hazardous material suits).

●Peritoneal lavage with heated solutions including chemotherapy.

●Excessive heat production (eg, malignant hyperthermia). (See "Malignant hyperthermia: Diagnosis and management of acute crisis".)

●Fever is a type of hyperthermia defined by a regulated increase in the internal thermostat setpoint. Fever is mediated by circulating pyrogenic cytokines including interleukins [80] and interferon [81]. Endogenous pyrogens activate the vagus nerve, which triggers release of prostaglandin E2 in the hypothalamus, thereby increasing the setpoint [82]. Causes of intraoperative fever include:

•Acute transfusion reaction (see "Approach to the patient with a suspected acute transfusion reaction", section on 'Patient with fever/chills')

•Infections, including those that were present before surgery (see "Fever in the surgical patient", section on 'Nosocomial infection' and "Fever in the surgical patient", section on 'Surgical site infection')

•Adverse effects of certain medications (eg, neuroleptic malignant syndrome, serotonergic syndrome) (see "Drug fever")

Notably, fever rarely occurs during general anesthesia because both volatile anesthetics [83] and opioids [84] blunt the febrile response. Fever is more likely to occur in the postoperative period when the thermoregulatory effects of anesthesia dissipate. (See 'Fever or hyperthermia' below.)

Consequences — Hyperthermia increases metabolic rate and may provoke shivering when the febrile shivering threshold exceeds actual body temperature. These effects may lead to adverse consequences including increased myocardial oxygen consumption. It is critically important to determine the underlying causes of fever since some are serious and require urgent treatment (table 3 and table 4).

Prevention and management — External causes of hyperthermia are easy to treat by eliminating causes of excessive heating and/or promoting of heat loss.

In contrast, fever is internally regulated and more difficult to manage. Treatment of the underlying cause of fever (ie, infection) or administration of antipyretic agents such as acetaminophen to block central nervous system mechanisms altering the setpoint is appropriate [85]. However, in some cases, underlying causes of fever are unknown or untreatable, and even theoretically effective drugs may not adequately blunt fever [86,87], possibly because not all fever is prostaglandin-mediated [88].

Passive cooling measures such as promoting heat loss can be employed to treat fever, but usually fail to overcome the centrally controlled new higher temperature setpoint [89]. Active cooling often provokes undesirable autonomic nervous system activation and shivering, as well as thermal discomfort (if the patient is awake) (table 5) [90,91].

Malignant hyperthermia is an emergency necessitating rapidly implemented specific treatments (table 6).

POSTOPERATIVE TEMPERATURE DERANGEMENTS

Hypothermia — Patients who are kept normothermic during surgery usually remain normothermic postoperatively. Thus, prevention of intraoperative hypothermia is the optimal clinical strategy, although ongoing warming in the post-anesthesia care unit (PACU) is often necessary [9,59,92]. All patients are warmed to 36°C, with treatment initiated within 15 minutes of arrival in the PACU [93]. Forced-air warming is the most common postoperative approach to treat hypothermia [9,92,94]. Those with only slight hypothermia (eg, 35.5°C) will rapidly return to normothermia once anesthetic-induced thermoregulatory impairment dissipates.

Even mild untreated postoperative hypothermia may cause complications. A decrease in body temperature as little as 2°C slows metabolism of anesthetics and neuromuscular blocking agents (NMBAs), prolonging their residual effects and rendering the patient more susceptible to adverse respiratory events in the PACU [52,95]. (See 'Prolongation of drug effects' above and "Delayed emergence and emergence delirium in adults", section on 'Consider temperature and metabolic derangements' and "Overview of the management of postoperative pulmonary complications".)

Hypothermia also results in sympathetic stimulation with hypertension and/or tachycardia with increased myocardial oxygen consumption, particularly if shivering occurs [16,57,58] (see 'Shivering' below). Patients with ischemic heart disease are at risk for myocardial ischemia and arrhythmias [59]. As noted above, other adverse consequences of persistent hypothermia include coagulopathy and decreased platelet function, as well as surgical site infection or sepsis. These adverse effects are associated with increased length of hospital stay and possibly increased mortality [96,97]. (See 'Consequences' above.)

Fever or hyperthermia — Postoperative fever (core body temperature ≥38°C) may be present in a patient with pre-existing fever that was blunted by general anesthesia, or may develop due to factors causing hyperthermia in the operating room that persist in the postoperative period (see 'Intraoperative hyperthermia' above). Increased body temperature produces a hypermetabolic state with increased respiratory rate and heart rate, which can cause shivering and may exacerbate myocardial ischemia.

Typically, fever is treated with acetaminophen (paracetamol). For adults ≥50 kg, the oral or rectal dose is 650 mg every four to six hours; the intravenous (IV) dose is 650 mg every four hours or 1000 mg every six hours. Dosing considerations include timing of previously administered acetaminophen on the day of surgery (eg, as part of a multimodal strategy to control pain), as well as weight <50 kg or renal impairment. Use is contraindicated in patients with severe hepatic impairment.

Details regarding evaluation and management of persistent postoperative fever are discussed separately. (See "Fever in the surgical patient".)

Shivering — In contrast with the intraoperative period, postoperative shivering is common in hypothermic patients, and may also occur in febrile patients [16]. Such rhythmic involuntary muscular activity after surgery is largely thermoregulatory, and is aggravated by volatile anesthetics. Occasionally, patients also have low-intensity shivering-like muscular activity that is not thermoregulatory [98] and appears to be aggravated by pain [99].

Shivering from any cause increases metabolic rate and provokes sympathetic stimulation; thus, it increases myocardial oxygen consumption, which may lead to myocardial ischemia [16,57]. Also, shivering and related phenomena cause considerable discomfort in awake postoperative patients.

Several medications to prevent postoperative shivering have been studied; however, significant heterogeneity in dosing regimens and type of surgery has led to uncertainty regarding their efficacy [100,101]. A 2023 meta-analysis of randomized trials noted that prophylactic intraoperative administration of a steroid reduced overall risk of shivering compared with placebo [102]. The risk ratio [RR] for IV dexamethasone was 0.67 (95% CI 0.52-0.87; 18 trials with 2558 patients), while the RR for IV hydrocortisone was 0.51 (95% CI 0.32-0.80; 4 trials with 390 patients). However, steroids were not more effective than active control interventions (eg, meperidine, ketamine, tramadol, ramosetron, dexmedetomidine, oral paracetamol). Prophylactic IV dexamethasone is frequently used to prevent postoperative nausea and vomiting (PONV), and these findings support a potential benefit for the prevention of shivering as well. Thus, it is reasonable to consider this option for patients at risk for myocardial ischemia or other adverse consequences of shivering. (See "Postoperative nausea and vomiting", section on 'Glucocorticoids' and "Postoperative nausea and vomiting", section on 'Prophylaxis based on risk'.)

Postoperative shivering can be treated with a single dose of meperidine 12.5 to 25 mg IV [103]. A reasonable alternative is an alpha₂ agonist such as clonidine 150 mcg or dexmedetomidine 0.5 mcg/kg administered IV over three to five minutes, with monitoring for side effects of bradycardia and/or hypotension [104,105]. One study suggests that perioperative administration of acetaminophen may reduce the incidence of postoperative shivering [106].

Thermal discomfort — Postoperative hypothermia causes thermal discomfort in awake patients. Although not life-threatening, thermal discomfort is typically intense and may have a prolonged duration. In one study, untreated patients who were 2°C hypothermic at the conclusion of surgery required two hours to return to normothermia and thermal comfort [107]. Anecdotal experience suggests that, unlike pain or nausea, memories of postoperative thermal discomfort persist for years after surgery. Active cutaneous warming markedly improves thermal comfort and simultaneously speeds rewarming. It is distinctly preferable to prevent thermal discomfort and shivering by warming adequately during surgery than to treat hypothermia postoperatively. (See 'Prevention and management' above.)

SUMMARY AND RECOMMENDATIONS

●Temperature monitoring – Patients should be kept at a core temperature of at least 35.5°C during anesthesia and surgery. We monitor body temperature in patients having general anesthesia lasting more than 30 minutes, major surgery with neuraxial anesthesia, or when clinically significant changes in temperature are intended, anticipated, or suspected. (See 'Temperature monitoring' above.)

●Normal thermoregulation – Normal core body temperature is tightly regulated to within a few tenths of a degree Celsius. The primary thermoregulatory defenses in humans are arteriovenous shunt vasoconstriction and shivering in response to cold, and active precapillary vasodilation and sweating in response to heat (figure 2). The thresholds for vasoconstriction and sweating differ by only a few tenths of a degree celsius, while the shivering threshold is 1°C lower. (See 'Normal thermoregulation' above.)

●Effects of general or neuraxial anesthesia on thermoregulation – Volatile and intravenous (IV) anesthetics markedly impair thermoregulatory control, leading to hypothermia in unwarmed surgical patients (figure 3). Patients having neuraxial anesthesia become nearly as hypothermic as those having general anesthesia. The combination of general and neuraxial anesthetic techniques results in the highest risk for intraoperative hypothermia due to the additive effects of each technique. (See 'Effects of general anesthesia' above and 'Effects of neuraxial or regional anesthesia' above and 'Combined general and neuraxial techniques' above.)

●Intraoperative hypothermia

•Causes – Intraoperative hypothermia develops in nearly all unwarmed surgical patients with a three-phase pattern (figure 5), characterized by (see 'Causes' above):

-Initially, hypothermia results from core-to-peripheral redistribution of body heat after induction of general anesthesia (figure 7), or activation of a neuraxial block.

-Subsequent heat loss to the environment occurs due to radiation and convection (figure 8).

-A core temperature plateau occurs when patients become sufficiently hypothermic (typically approximately 34.5°C) and trigger thermoregulatory vasoconstriction to effectively constrain metabolic heat to the core thermal compartment.

•Consequences – Intraoperative core temperatures of at least 35.5°C do not provoke serious complications. Lower temperature may cause coagulopathy, infection, prolonged drug action, thermal discomfort, and shivering (table 2). (See 'Consequences' above.)

•Prevention and management – Prewarming before induction of anesthesia reduces intraoperative hypothermia, but by <0.5°C. Active intraoperative warming is needed in most patients. (See 'Prevention and management' above.)

●Intraoperative hyperthermia – Intraoperative hyperthermia is much less common than hypothermia.

•Causes – Causes include excessive heating, inadequate heat loss, excessive heat production, and fever (table 3 and table 4). Fever differs from other thermal perturbation as it is a regulated increase in the internal thermostatic setpoint consequent to infection, acute transfusion reaction, or adverse effects of certain medications. (See 'Causes' above.)

•Consequences – Hyperthermia increases metabolic rate and may provoke shivering and increased myocardial oxygen consumption. (See 'Consequences' above.)

•Management – External causes of hyperthermia are treated by eliminating the source of excessive heating and/or promoting heat loss. Management of internally regulated fever includes treatment of the underlying cause and administration of antipyretic agents such as acetaminophen to block central nervous system mechanisms altering the setpoint. (See 'Prevention and management' above.)

●Postoperative hypothermia or hyperthermia – Patients who are adequately warmed during surgery will remain normothermic postoperatively. Rarely, fever which was blunted by general anesthesia may appear postoperatively. (See 'Postoperative temperature derangements' above.)