Secondary Logo

Journal Logo

Original Article

Closing the audit loop - prevention of perioperative hypothermia: audit and reaudit of perioperative hypothermia

Gallagher, G. A.; McLintock, T.; Booth, M. G.

Author Information
European Journal of Anaesthesiology: September 2003 - Volume 20 - Issue 9 - p 750-752
  • Free


Many studies report the need to avoid even moderate perioperative hypothermia. The maintenance of normothermia perioperatively has been shown to reduce the incidence of morbid cardiac events in patients with cardiac risk factors undergoing non-cardiac surgery [1], while hypothermia has been associated with an increased incidence of surgical wound infection and impaired wound healing [2]. Decreased activity of the coagulation cascade associated with increased bleeding and blood transfusion requirements has also been described in moderately hypothermic patients [3].

For all these reasons, it would seem appropriate to prevent perioperative hypothermia. To this end, an audit was undertaken in 1992 to ascertain the extent of hypothermia in postoperative patients at Glasgow Royal Infirmary and to propose changes in practice should they be deemed necessary. Glasgow Royal Infirmary is a large teaching hospital with 13 theatres undertaking some 28 000-30 000 operative procedures per annum. The results of the original audit indicated that postoperative hypothermia, defined as a core temperature of ≤35.5°C, occurred in nearly 50% of the patients audited. A number of changes of practice were introduced to reduce the incidence of hypothermia, including the introduction of forced-air warming devices to be used in all cases where the anaesthetist felt hypothermia was a risk.

Following these changes in practice after the initial audit to reduce the incidence of hypothermia, it was necessary to reaudit our practice for two reasons. First, to see whether these changes in practice had been implemented and, second, to see if they had reduced the incidence of hypothermia. A second audit was therefore undertaken in 1998, thus closing the audit loop.


The 1992 audit recorded measurements from 177 patients undergoing elective surgery. Major procedures were targeted, but there was no formal minimal length of surgery below which patients were not included. Variables recorded were: ASA classification, the duration of operation, the use and description of preventative measures for hypothermia, the amount of blood loss, the use of intravenous (i.v.) fluids, and the core and peripheral temperatures on arrival and discharge from the recovery room, as well as the length of recovery stay. Core temperatures were measured using a Genius® infrared temperature probe (Mallinckrodt, St Louis, MO, USA).

In the 1998 audit, 158 patients were included. Patients undergoing major general, orthopaedic or vascular surgical procedures had core temperatures measured again using an infrared probe. These measurements were recorded preoperatively, immediately upon arrival in the recovery room and just before discharge to the ward. Patients undergoing an operation expected to last >90 min were included in the audit. Other variables recorded were: use of the forced-air warmer, foil hat or fluid warmer, the amount of blood loss, the i.v. fluids infused and the anaesthetic technique performed. The length of stay in the recovery room and the ASA classification of each patient were also recorded.


The raw data indicated that 81 of 177 patients in the first audit had core temperatures ≤35.5°C, whilst in the second audit, only nine of 158 patients were hypothermic by this definition.

The audit results were analysed using a t-test, where P < 0.05 was regarded as significant.

There were several changes in routine practice regarding the prevention of hypothermia between the two audits. The methods used in the first audit consisted of heat and moisture exchange devices, circle anaesthetic systems, water-circulating warming mattresses and foil hats. By the second audit, forced-air warming devices had been introduced, and water-circulating warming mattresses were withdrawn from routine use. The percentages of cases using each method are shown in Table 1.

Table 1
Table 1:
Measures (%) to prevent hypothermia.

The duration of surgery in the two groups was significantly different owing to the exclusion in the 1998 group of any procedure lasting <90 min. This resulted in an average (range) duration of 133.5 (25-330) min in the initial study compared with 154.7 (90-480) min subsequently. This difference reached significance (P < 0.003). Although the duration of operation was longer in the second group of audited patients, they were still warmer on arrival in the recovery room (Table 2).

Table 2
Table 2:
Core temperature.


There were a number of changes in practice that could have contributed to our results. The two groups were not strictly comparable in terms of operative procedure (Table 3). Heat and moisture exchange devices are now used in all anaesthetics for reasons of infection control as well as for hypothermia prevention, and the advent of more expensive inhalational anaesthetic agents means that the use of the circle anaesthetic breathing system is also often routine. Water-circulating warming mattresses have been withdrawn from use as no benefit has been shown [4].

Table 3
Table 3:
Surgical procedures audited.

This audit demonstrates that changes in practice to prevent unintentional hypothermia in patients identified preoperatively as being at risk have now been successfully implemented. Patients arriving in the recovery room have a mean core temperature of 36.6°C, which is 1°C higher than in the previous audit. This is important clinically, as 35.5°C is the core temperature below which adverse effects relating to hypothermia are more likely to occur [1].

We recognize that this audit was limited to an extent, as only those patients deemed likely from the outset to have procedures lasting >90 min were included, thus missing those cases with unforeseen complications prolonging surgery. It may be considered that these are the cases most at risk, as it is less likely that measures of prevention of heat loss will have been taken. The other group of patients not audited was those undergoing emergency surgery. In addition, the procedures audited were different as previously outlined. There were more laparotomies in the first group and more large joint replacements in the second. This may have had some influence on our results. Even given these limitations, this audit demonstrates that simple changes in practice, if widely implemented, can go a long way towards eradicating the problem of inadvertent postoperative hypothermia. This audit has led to an increased awareness amongst anaesthetists and other theatre staff of the possibility of hypothermia and, consequently, increased activity to avoid it. Further forced-air warmers have been purchased for those areas previously lacking such resources.


The authors thank all the nursing staff in the recovery room and the Department of Anaesthesia at Glasgow Royal Infirmary for their help with their audit.


1. Frank SM, Fleisher LA, Breslow MJ, et al. Perioperative maintenance of normothermia reduces the incidence of morbid cardiac events. J Am Med Assoc 1997; 277: 1127-1134.
2. Kurz A, Sessler DI, Lenhardt R. Perioperative normothermia to reduce the incidence of surgical wound infection and shorten hospitalisation. N Engl J Med 1996; 334: 1209-1215.
3. Schmied H, Kurz A, Sessler DI, Kozek S, Reiter A. Mild hypothermia increases blood loss and transfusion requirements during total hip arthroplasty. Lancet 1996; 347: 289-292.
4. Morris RH, Kumar A. The effect of warming mattresses on the maintenance of body temperature of the anesthetized, paralyzed adult patient. Anesthesiology 1972; 36: 408-411.


© 2003 European Academy of Anaesthesiology