Perioperative Hypothermia—Friend or Foe?

Abstract & Commentary

Synopsis: Prevention of hypothermia during major operative procedures resulted in 55% fewer post-operative cardiac events in this large prospective study of vascular surgery patients.

Source: Frank SM, et al. JAMA 1997;277:1127-1134.

All anesthetics produce changes in temperature preservation mechanisms, which, when combined with cold operating rooms and open body cavities, result in some degree of hypothermia in most patients. Adverse effects of operating room-induced hypothermia have been reported and include: increased wound infection rate, coagulopathy, higher incidence of shivering, more pain and discomfort, and a longer hospital stay. This prospective, randomized study compared the incidence and severity of cardiac events in a group of patients with known or suspected cardiac disease who were treated routinely or aggressively with surface warm-air circulation to maintain body temperature during major vascular procedures. In the treatment group, the application of surface warm-air circulation was continued for two hours into the post-operative period.

A total of 300 patients were entered into the randomized trial. A definite or highly suggestive history of coronary artery disease was required to be entered into randomization. Due to the nature of their surgery, all patients were monitored with indwelling arterial lines and admitted to the ICU directly from the OR. A mixture of regional and general anesthetic techniques was employed as indicated by the surgical procedure and anesthesiologist choice. In addition to continuous blood pressure monitoring, ECG with ST-segment analysis, pulse oximetry, intermittent pain scores, and urine output, body temperature was measured at several sites for 24 hours following surgery. Adverse cardiac events studied included ischemia or ventricular arrhythmias, cardiac arrest, or myocardial infarction. The amount of time that blood pressure was outside the normal range was also calculated.

The patient groups were similar in demographics, severity of cardiac risk factors, and magnitude of operations. The routine treatment group entered the ICU with a mean core temperature 1.3°C less than that in the treatment group (35.4°C vs 36.7°C, P < 0.001). Intraoperatively, the groups had similar incidences of ischemia and ECG events. Postoperatively, the routine (hypothermic) group exhibited more hypertension, electrocardiographic abnormalities, and myocardial events. There was no difference in mortality or hospital length of stay between the groups. Only one myocardial infarction was detected in any patient (hypothermia group) and two cardiac arrests occurred (both in the hypothermia group). The largest difference between the groups was noted in non-sustained ventricular tachycardias, which were detected in 8% of the hypothermic patients and only 2% of the normothermic patients.


Hypothermia following surgery is common. Severe hypothermia is known to induce coagulopathy and increase bleeding in the perioperative period. Mild hypothermia is associated with increased wound infection rates. Shivering increases oxygen consumption several fold and this may stress a compromised delivery system resulting in myocardial ischemia or infarction. In this study shivering was more frequent in the hypothermic group, but adverse cardiac events were not associated with shivering. This study implies a better cardiac outcome with maintenance of normothermia in patients undergoing vascular surgery who have underlying coronary artery disease. However, the only significant finding to support this contention was that the incidence of non-sustained ventricular tachycardia was higher in the hypothermic group. This problem was transient and required no treatment.

An additional interesting association with adverse cardiac events was the use of a beta-blocker preoperatively. Half of the patients with cardiac events were on beta blockers compared to only 27% of those without events. This may be a confounding issue affecting the results of this study. Beta blockers may increase the incidence of non-sustained ventricular tachyarrhythmias due to their effects on the sinus node and conduction system. The use of beta blockers may also be a marker for more severe or a different kind of cardiac disease. Other known cardiac risk factors were similar in both groups. Beta blockers may be associated with a lower incidence of perioperative myocardial ischemia and infarction but may increase the incidence of arrhythmias.

Mild hypothermia may have beneficial effects. Following head injury, recovery may be improved with mild hypothermia. The two patients sustaining cardiac arrests in this study may have benefitted from the presence of hypothermia at the time of the arrest.

The costs associated with preventing hypothermia are not discussed in this paper. The disposable component of the warming device employed by the investigators costs at least $10.00 per patient. The capital costs are many thousand dollars per ICU bed and OR suite. The potential cost savings of maintaining normothermia in this study were negligible (a shortened ICU stay of 5 hours) due to identical hospital lengths of stay.

The answers about intraoperative hypothermia are not yet complete. This study is an example of how therapy begun prior to ICU admission can have significant positive or negative impact in ICU outcomes. Consideration of this possibility should be kept in mind when evaluating ICU therapies. Cost implications must be considered as well as patient risks in all outcome studies.