Temperature Management During General Anesthesia

Adverse Effects of Intraoperative Hypothermia

Increased Infection Risk

2-fold risk with 2C drop [Kurz A et al. NEJM 334: 1209, 1996]

Increase in Blood Loss/Transfusions

EBL increased 500 cc with 1.6C drop [Schmied H et al. Lancet 347: 289, 1996, n = 60], 186 cc with 0.5C [Widman J et al. Anesth Analg 95: 1757, 2002, n = 46], and 140 cc with 0.4C drop [Winkler M et al. Anesth Analg 91: 978, 2000, n = 150]

Increased Hospital/PACU Time

2.6 days total hospitalization with 2C drop [Kurz A et al. NEJM 334: 1209, 1996]. 41 minutes PACU with 1.9C drop [Lenhardt R et al. Anesthesiology 87: 1318, 1997, n = 150]. Extubation 10 mins faster with forced air warming [Mahoney]

Increased Cardiac Morbidity

NOT due to shivering, possibly arrhythmias, hypertensive response to stress of cold [Frank SM et al. JAMA 277: 1127, 1997]. Normothermia associated with a 55% reduction in cardiac risk.

Increased Anesthetic Potency

MAC drops 15% for every 1.0C. Vecuronium, atracurium prolonged by up to 60%

Treatment of Intraoperative Hypothermia


Increases core temperature 1.8C [Hynson et al]

Forced Air

Imparts 50W thermal energy, most cost-effective means possible

Hyperthermia Syndromes

Malignant Hyperthermia

Incidence of 1:15,000 episodes of general anesthesia. Inherited disorder. Signs include rigidity (ex. elevated ETCO2), temperature, depressed consciousness, and autonomic instability. Stop all potential medications and give dantrolene 1-2 mg/kg IV bolus, continue for 3 days. Dantrolene causes muscle weakness (particularly grip strength). The risk for liver injury is high in patients with pre-existing liver pathology.

Neuroleptic Malignant Syndrome

Signs include rigidity, temperature, depressed consciousness, and autonomic instability. Drugs include Haldol and other antipsychotics (0.2 – 1.9% incidence with neuroleptics [Hosp Phys 36: 51, 2000]), Reglan and other motility agonists, or cessation of a dopamine agonist (Amantadine, bromocriptine, levodopa). There is no correlation between dose or intensity and risk. Most cases start within 24-72 hours and all will start by 2 weeks. Stop all potential medications (or continue stopped medications) and give dantrolene 2-3 mg/kg per day IV up to 10 mg/kg/day. In liver patients, bromocriptine 2.5-10 mg PO tid is an alternative but hypotension is a potential side effect. Both treatments should be continued for 10 days, and DVT prophylaxis should be added. It is not clear if any of these treatments actually affect mortality [Neurol Clin North Am 22: 389, 2004]

Serotonin Syndrome

Mental status changes, autonomic hyperactivity, and neuromuscular abnormalities. Drugs include SSRIs, MAOIs, TCAs, L-tryptophan, ritonavir.. lithium, sumitriptans, meperidine, and fentanyl. Onset is abrupt, usually within 6 hours of taking the drug. Symptoms are similar to MH and NMS and can include tremor, hyperreflexia, ocular clonus, agitation, and diaphoresis. Hyperkinesis, hyperreflexia, and clonus help distinguish this from MN and NMS. Treat by removing all suspected drugs, control hyperthermia, give benzodiazepines, consider the serotonin antagonist cyproheptadine [J Emer Med 16: 615, 1998], and neuromuscular paralysis in severe cases. Dantrolene is not effective in serotonin syndrome.


Defined as T < 95F (or 35C), can lead to coagulopathies (which are silent because lab assays are run at normal body temperature), electrolyte abnormalities, acid base abnormalities (respiratory or metabolic acidosis), cold diuresis (and increased creatinine), rhabdomyloysis, acute renal failure, and various arrhythmias. J waves (aka Osborn waves) are seen in 80% of EKGs.

Patient hypothermia is frequently observed in the perioperative setting. It is observed in patients undergoing procedures employing both general and regional anesthetic techniques. The multiple deleterious effects of hypothermia have been delineated elsewhere [Sessler DI. Anesthesiology 2000; 92: 578, Sessler Anesthesiology 2001; 95: 531, Sessler NEJM 1997; 336: 1730] and include increased duration of hospitalization, transfusion requirements, and likelihood of cardiac morbidity [Sessler Anesthesiology 2001; 95: 531].

Hypothermia and Surgical Site Infection

In 1996, Kurz et al. published the results of a randomized controlled trial examining the effects of hypothermia on the incidence of SSI. Two hundred patients undergoing colorectal surgery were randomized to either standard intraoperative thermal care (the hypothermia group) or additional warming (the normothermia group). The patients’ anesthetic care was standardized, and they were all given appropriate perioperative antibiotics. Patients in the hypothermia group had a mean intraoperative core temperature of 34.7°C whereas patients in the normothermia group had a mean intraoperative core temperature of 36.6°C. This small ~2°C difference in core temperature resulted in a 3-fold higher incidence of SSI in the hypothermia group (19% vs. 6%, p = 0.009). In addition, sutures were removed one day later in the patients assigned to hypothermia than in those assigned to normothermia and the duration of hospitalization was prolonged by a mean of 2.6 days in the hypothermia group [Kurz NEJM 1996; 334: 1209]. Perioperative hypothermia has also been established as a risk factor for SSI in several retrospective studies [Flores-Maldonado Archives of Medical Research 2001; 32: 227, McAnally Pediatric Infectious Disease Journal 2001; 20: 459, Hedrick Surgical Infections 2007; 8: 425]

The mechanisms by which hypothermia increases the incidence of SSI have been also defined. Hypothermia suppresses phagocytic activity by decreasing migration of PMN’s, reducing superoxide anion production, and reducing oxidative bacterial killing by neutrophils [van Oss Journal of the Reticuloendothelial Society 1980; 27: 561, Clardy Pediatric Infectious Disease 1985; 4: 379]. Bacterial killing by neutrophils is significantly reduced as temperature decreases from 40 to 32 °C [Wenisch Anesthesia & Analgesia 1996; 82: 810].

Even after surgery and anesthesia have concluded, hypothermic patients remain “cold” for several hours [Kurz NEJM 1996; 334: 1209]. In order to rewarm the core, normal compensatory reflexes, including peripheral vasoconstriction, are employed by hypothermic patients. This vasoconstriction results in significant decreases of blood flow to the surgical wound. Indeed, Kurz et al. observed significant vasoconstriction in 74% of the hypothermic patients but only 6% of the normothermic patients [Kurz NEJM 1996; 334: 1209].

Vasoconstriction results in decreased delivery of PMN’s to the wound and, perhaps more importantly, decreases the delivery of oxygen to the wound. Oxidative killing is dependent on the production of bactericidal superoxide radicals from molecular oxygen by NADPH-linked oxygenase. NADPH-linked oxygenase has a Km of approximately 60 mmHg for oxygen and thus oxidative killing is pO2-dependent from a range of 0 to 300 mmHg [77]. Wound pO2 is a powerful predictor of SSI. In 1997, Hopf et al. measured subcutaneous pO2 perioperatively and found that no patient with a subcutaneous oxygen tension greater than 90 mmHg developed a SSI. In stark contrast, the incidence of SSI among patients with a subcutaneous oxygen tension of 40-50 mmHg was 43% [Hopf Archives of Surgery 1997; 132: 997].

Approximately 90% of the body’s heat is lost through the skin with radiation and convection usually contributing far more than evaporation or conduction [Sessler Anesthesiology 2000; 92: 578, Sessler NEJM 1997; 336: 1730], making radiation and convective losses ideal targets. The two most efficacious methods for preventing hypothermia are pre-warming and forced-air systems.

Hypothermia and Neuroprotection

Means of Combating Hypothermia


Pre-warming is based on two principles. A study of six volunteers, who each underwent two episodes of general anesthesia (once with forced-air prewarming, once without), showed that 120 minutes of prewarming attenuated the temperature drop following general anesthesia (from 36.7 to 34.9C after one hour in the control group, from 36.7 to 36.1C in the prewarmed group) [Hynson Anesthesiology 1993; 79: 219].

Forced Air Systems

Forced air systems are capable of delivering 30-50W of thermal energy to an anesthetized patient [Sessler Anesthesiology 1990; 73: 218]. Passive insulators, by comparison, reduce heat losses from approximately 100W to 70W [Sessler Anesthesiology 1991; 74: 875, Sessler DI et al. Anesth Analg 1993; 77: 73]. Therefore, the net benefit of a forced air system, in terms of thermal energy, ranges from 100 – 120W.

Some investigators have expressed concern regarding the potential for forced-air warming systems to increase bacterial contamination of surgical wound [Tumia Journal of Hospital Infection 2002; 52: 171] and have advocated not activating the system until after the patient has been completely draped. There is absolutely no empiric support for this practice. Denying the patient access to active warming—especially during the beginning of the procedure before the application of surgical drapes when heat loss and redistribution are greatest—puts the patient at an increased risk for hypothermia, which will increase the risk of SSI. Further, all forced-air warming units include filters that essentially eliminate bacteria in the heated air: the number of colony-forming units recovered from operating rooms is not increased by forced-air blowers [Zink RS et al. Anesth Analg 1993; 76: 50]. Finally, the principle means by which Kurz et al. prevented hypothermia and reduced the incidence of SSI was forced-air warming [Kurz NEJM 1996; 334: 1209].

Other Means of Increasing Temperature

Other, less effective means of maintaining thermal equilibrium include modifications of the surface environment through the use of passive insulation, resistive heaters, circulating water, radiators, and pharmacologic vasodilation, as well as heating IV fluids and airway gases.