Ramos G.E., Resta M., Patiño O., Bolgiani A.N., Prezzavento G., Grillo R.,Chacón Pazmiño G., Benaim F., Rutan R.L.

Benaim Burn Foundation Burn Centre, Aleman Hospital, Buenos Aires, Argentina

SUMMARY. Burn patients are especially susceptible to hypothermia. The purpose of this study was to determine the incidence, risk factors, and sequelae of intra-operative hypothermia in our burn population. This prospective observational study was conducted between May 1998 and February 1999. During that time, fourteen burn patients (body surface area [BSA] burned, 10-64%) underwent 66 operative debridements of 1-15% BSA. The patients experienced hypothermia (core temperature, ± 36 °C) in 13 of the 66 procedures (19.7%). The hypothermia was mild in all cases (35.2 ± 0.3 °C; mean ± SD). BSA, age, and surgical room temperature were identified as risk factors for hypothermia, but not the duration of surgery. Bleeding time was longer in hypothermic patients, but without clinical consequences.

Introduction

In the general surgical population, approximately one-half of patients in routine peri-operative thermal care develop a core body temperature of less than 36 °C during the peri-operative period, and a further one-third exhibit core temperatures of less than 35 °C. It is known that this form of hypothermia may cause disturbances in blood clotting, increase the potential for haemodynamic instability and the risk of peri-operative cardiac events, prolong anaesthetic recovery time, and delay wound healing. Monitoring should therefore be used and hypothermia must be prevented.

Intra- and post-operative hypothermia has been described in trauma and burns and in cardiac, orthopaedic, and other surgical patients. However, in all these groups, burn patients are by far the most susceptible to hypothermia, since the damaged skin is no longer able to prevent the loss of body heat. In addition, the injury itself instigates a true hypothalamic reset of core temperature control, challenging burn patients with significant injuries to strive to maintain a temperature at least one degree higher than normal.12 This reset fuels the hypermetabolism seen after burns and increases burn patients’ risk of hypothermia. In addition, the multiple surgical procedures required to debride necrotic tissue and close the resulting wounds repeatedly expose burn patients to the chilled environment of the operating room (OR). Vast expanses of normal and burned skin are bared, and this environmental chilling is augmented by the hypothalamic effects of general anaesthesia, such as decreased metabolic rate and the redistribution of blood flow, and the systemic administration of large volumes of chilled fluids.13 One study of peri-operative hypothermia in burn patients10 reported that 89% of patients with extensive burns who were undergoing extensive surgical debridements experienced core temperatures of less than 35 °C.

The purpose of this study was to determine the incidence, risk factors, and sequelae of peri-operative hypothermia in burn patients subjected to non-extensive surgical procedures. This is the commonest practice in our unit.

Material and methods

This prospective observational trial was conducted over a nine-month period (May 1998-February 1999). Acutely burned patients requiring surgical debridement in areas � 15% body surface area (BSA) under general anaesthesia were included. Patients requiring extensive debridement (> 15% BSA), amputations, or flaps were not included in the study. Patients with platelet counts below 100,000 platelets/mm3, prothrombin time > 16.7 sec (below 60% of normal activity), or partial thromboplastin time > 45 sec were excluded. Patients with both earlobes burned were also excluded.

The patients received both inhalation (isoflurane) and intravenous anaesthetic agents such as ketamine, propofol, fentanyl, and midazolan. It is our peri-operative routine to administer fluids and colloids at room temperature. Respiratory and heart rates, blood pressure, and rectal temperatures were monitored continuously intra-operatively. The patients were brought nude to the OR. Dressings were removed and the entire body was scrubbed with a povidone-iodine solution and then rinsed with warm water. The patients were then draped in sterile linen and/or disposable paper sheets, leaving only the operative site exposed to the air. Excised wounds were dressed in silver sulphadiazine and polyurethane. At the conclusion of the procedure, the patients were awakened and extubated prior to leaving the OR. They were then wrapped in warm blankets before being transported to their room to complete their recovery.

Room temperatures were measured before surgery (patient room), during surgery (operating theatre), and after surgery (patient room) using standard mercury thermometers (range, -25 to +50 °C). Core body temperature (rectal) was taken before surgery, during the procedure, and at 0, 60, and 180 min after surgery. Rectal temperatures were routinely measured using mercury thermometers (range, 35 to 42 °C) and read 2 min after insertion; however, if the temperature was � 35 °C, a second reading was obtained with a thermometer of increased range (range, -10 to 110 °C).

The patients’ cold sensation was measured using a 10-point cold sensation scale (CSS): 1 point was given to qualify “no feeling of cold” and 10 points to “the coldest possible”. Fifteen minutes post-operatively, the patients were asked to evaluate their perceived feelings of cold using the CSS. The occurrence of post-operative shivering was noted and its duration was timed.

Bleeding time was measured before, during, and after surgical procedures by incising the most dependent portion of the earlobe with a 0.5 x 16 mm needle and measuring the length of time before bleeding ceased.

Data analysis

A rectal temperature of less than 36 °C was defined as hypothermia. Since the patients were studied more than once, we analysed “procedures” in which hypothermia occurred and “patients” who presented hypothermia. The relevant data are presented in the text and tables as mean ± standard deviation (0 ± SD). Comparisons between the groups were performed by Student’s t-test, using a Bonferroni correction for multiple comparisons. Significance was accepted at p < 0.05, unless otherwise specified.

Results

Sixty-six procedures were performed on 14 patients, aged between 12 and 59 years of age. The excisions were performed by one of two surgeons, averaging 1.9 ± 3.3% BSA. Hypothermia occurred in 13 of the 66 procedures studied (19.7%) and in seven of the 14 studied subjects (50%). One patient was hypothermic before surgery owing to septic shock and the data of this procedure were therefore excluded from the analysis. Three of the seven patients (43%) who experienced hypothermia died, whereas all seven patients who did not have hypothermia survived (p < 0.006). There was a statistically significant correlation between open area (burn size) and the incidence of hypothermia (Pearson product correlation, p = 0.045).

Average core body temperature (CBT) was significantly lower at the end of surgery than it was before (36.7 ± 1.0 °C vs 37.8 ± 0.7°C, respectively; p < 0.001) (Fig. 1). As expected, the patients gradually recovered after leaving the OR, with a CBT of 36.9 °C (± 0.8) at 60 post-operative min and 37.7 °C (± 0.7) at 180 min. The variations in CBT were greater in hypothermic patients and recovery time was longer in patients who died (Fig. 2).

There were statistically significant differences between the ages of the hypothermic and euthermic patients (42.8 ± 10.4 yr vs 28.6 ± 15.2 yr, respectively; p < 0.002). OR temperature in patients with hypothermia was significantly lower than in those who did not have hypothermia (22.6 ± 2.5 °C vs 25.6 ± 2.5 °C; p < 0.0003). With regard to the time spent in the OR, there was no statistical difference between the two groups (Table I).

Cold sensation was evaluated on the 10-point CSS in 46 for the 66 procedures (65%). Assessments were not made in patients who were so deeply sedated that they required continued mechanical ventilation or who were mentally incapable of responding. The mean CSS in patients with peri-operative hypothermia was 5.6 ± 3, compared with 2.6 ± 3 in patients without hypothermia (Table II).

We observed shivering after 21 of the 66 procedures (32%). There were five instances of shivering in hypothermic patients and 16 in non-hypothermic patients, i.e. 38% of hypothermic patients and 30% of non-hypothermic patients exhibited shivering.

In the two hours prior to surgery and up to three hours post-surgery, no red blood cells were administered. There was a significant difference in post-operative bleeding times between the groups. No cardiac event was observed during the period of study. Bleeding time was longer in hypothermic patients, both before and after surgery, but this achieved statistical significance only during the post-operative period (Table III). Bleeding time measured before and after surgery increased, decreased, and was not modified in similar percentages in both non-hypothermic and hypothermic procedures (Fig. 3).

Using standard regression techniques, immediate post-operative rectal temperature can be predicted by the formula:

In these series, the ambient temperature inside the OR was the single most significant predictor of post-operative hypothermia, being responsible for approximately 75% of the change in CBT. Age and the area of excision were also significant contributors - approximately 7% of the change was attributable to each of these variables in the prediction of post-operative hypothermia.

Discussion

Body temperature is usually higher than environmental temperature, and it is maintained with little variation by a system that regulates the production, loss, and maintenance of heat.

Burn patients have increased heat loss because of skin disruption.11 Normal skin acts as an effective coverage, reducing the temperature gradient between body temperature and the environment by means of vasoconstriction.

The administration of general anaesthetics to burn patients further reduces their ability to regulate their body temperature by altering the vasoconstrictive thresholds, the metabolic rate, and thermoregulatory set points. Fluid loss from burned tissue further increases cooling due to evaporation.

The heat loss rate is related to BSA and the temperature gradient between body and the environment. Wilmore et al.15 evaluated the effect of ambient temperature on heat production and heat loss in burn patients. They concluded that a higher environmental temperature decreased the metabolic rate only in patients with large thermal injuries (> 60% BSA) in whom the decrement in dry heat loss produced by the higher ambient temperature exceeded the increase of wet heat loss.

All these factors explain why the burn surface area is so important as a hypothermic risk factor, especially during surgery.

We observed a significant reduction in CBT during surgery and a slow recovery during the post-surgical period. Hypothermic patients had more extensive burned surface areas than non-hypothermic patients. Hypothermic patients were older than non-hypothermic patients. This could be related to differences in the capacity for warmth production. We did not observe any differences in surgical time between hypothermic and non-hypothermic patients. Extensive surgery was not however included in this study because we consider that other forms of management are required.

When the body isolation barrier is altered, as happens in burn patients, the increase in the body/environment temperature gradient increases the velocity of heat loss. We observed that the OR temperature was significantly different in hypothermic and non-hypothermic patients. Kelemen et al.16 observed that the metabolism of burn patients was dependent on BSA when the room temperature was less than 28 °C, but not with temperatures above

32 °C. In cases of trauma and burns, some researchers12,17 recommend an OR temperature of above 30 °C during surgery. However, raising the temperature in standard wards with no special airflow or temperature control facilities may cause patients to sweat, thus further increasing heat loss.

This level of environmental temperature is also uncomfortable for the surgical team and is probably

not necessary for all surgical procedures in all burn patients.

The question we had posed before beginning this trial was: How common is peri-operative hypothermia in burn patients? There is not much in the literature on this topic. One trial10 showed a very high incidence of hypothermia (89%). This incidence was higher than the incidence we found in the present study, in which hypothermia developed in less than 20% of the operations. Moreover, all the hypothermic patients in the Shiozaki study had a temperature of less than 35 °C, while in our study only one patient had less than 35 °C and this was found in a pre-surgical hypothermic patient with sepsis. Various factors, such as differences in the severity of thermal injuries and the types of operation studied, could explain these results. The OR temperature was not recorded.

The recovery time after hypothermia was another interesting finding. We observed that the patients who died had had a slower recovery time than those who survived.

Shiozaki et al.19 showed that oxygen consumption was greater in patients who recovered from hypothermia faster, probably because of a better haemodynamic reserve.

The haemodynamic reserve is related to the capacity to increase oxygen delivery, consumption, and extraction in response to injury, and is therefore related to the outcome.20, 21 The prognostic value of recovery time can be evaluated as early as the first two weeks.19

The increased risk of peri-operative bleeding is one of the most important complications related to hypothermia. It has been demonstrated that even when hypothermia

is as little as 0.5 °C (i.e., body temperature less than

36.5 °C), it significantly slows blood clotting.3 We evaluated this alteration with a simple method - bleeding time - and we found significant differences between hypothermic and non-hypothermic patients. However, after surgery, we observed in both groups some patients who prolonged bleeding time and others who reduced it. We cannot rule out methodology problems, because bleeding time leaves much to be desired in terms of reproducibility.22 No two skin areas are exactly the same and it is impossible to produce a truly standard wound. Moreover, vasoconstriction mediated by hypothermia can reduce bleeding time although the coagulation system is altered. Finally, the liberation of thromboplastin during the surgical trauma can exceed the anticoagulant effect of mild hypothermia. However, bleeding time changes had no clinical importance because blood transfusion requirements did not increase in hypothermic patients.

Post-surgical cool sensation and discomfort were greater in hypothermic patients than in non-hypothermic patients, although these differences did not achieve statistical significance. This was possibly due to the small size of the sample, which was the result of the impossibility of evaluating a larger proportion of the patient population.

The incidence and duration of shivering were not different in hypothermic and non-hypothermic patients.

To sum up, peri-operative hypothermia was not very frequent and in our routine practice we did not find any clinically important complications related to it. An OR temperature close to 25 °C should therefore be safe for most operations in burn patients. This trial was not however extensive enough to deny all peri-operative hypothermic importance and the negative consequences described by others. Patients with extensive burn surface areas and aggressive surgical debridement in prolonged surgery, and especially those who require large amounts of fluids, may need other measures to prevent hypothermia. These measures include warmed fluids for intravenous and local administration, humidified and warmed air for mechanical ventilation, convective air blankets to cover non-surgical body areas, and - probably - a higher surgical room temperature than we routinely use in most surgical operations.

Conclusion

In our cases, surgical procedures in burn patients induced a significant reduction in central body temperature. However, less than 20% of the patients subjected to surgery developed hypothermia, and this was mild in all cases and without clinical relevance. The OR temperature, age, and open BSA were determined to be significant contributors to the development of post-operative hypothermia.

RESUME. Les patients atteints de brûlure sont particulièrement susceptibles à l’hypothermie. Les Auteurs de cette étude prospective et d’observation, conduite entre mai 1998 et février 1999, ont évalué la fréquence, les facteurs de risque et les séquelles de l’hypothermie intra-opératoire dans leurs patients brûlés. Pendant la période d’observation quatorze patients brûlés (surface corporelle brûlée [SCB], 10-64%) ont subi 66 débridements chirurgicaux (1-15% SCB). Les patients ont manifesté l’hypothermie (température, ± 36 °C) en 13 des 66 procédures (19,7%). L’hypothermie était légère en tous les cas (35,2 ± 0,3 °C; moyenne ± DS). La SCB, l’âge et la température de la salle opératoire ont été identifiés comme facteurs de risque pour l’hypothermie, mais non la durée de l’intervention chirurgicale. Le temps de saignement était plus long dans les patients hypothermiques, mais sans aucune manifestation de conséquences cliniques.

Bibliography

1. Frank S., Beattie C., Christopherson R., Norris E., Perler B., Williams G. et al.: Unintentional hypothermia is associated with post-operative myocardial ischemia. Anesthesiol., 78: 468-76, 1993.
2. Schmied H., Kurz A., Sessler D., Kozek S., Reiter A.: Mild hypothermia increases blood loss and transfusion requirements during total hip arthroplasty. Lancet, 347: 289-92, 1996.
3. Valeri R., Feingold H., Cassidy G., Ragno G., Khuri S., Attschule M.: Hypothermia-induced reversible platelet dysfunction. Ann. Surg., 205: 175-81, 1987.
4. Watts D., Trask A., Soeken K., Perdue P., Dols S., Kaufman C. et al.: Hypothermic coagulopathy in trauma: Effect of varying levels of hypothermia on enzyme speed, platelet function, and fibrinolytic activity. J. Trauma, 44: 846-54, 1998.
5. Frank S., Fleisher L., Breslow M., Higgins M., Olson K., Kelly S. et al.: Peri-operative maintenance of normothermia reduces the incidence of morbid cardiac events: A randomized clinical trial. JAMA, 277: 1127-34, 1997.
6. Lenhardt R., Marker E., Goll V., Tschernich H., Kurz A., Sessler D. et al.: Mild intra-operative hypothermia prolongs post-anaesthetic recovery. Anesthesiol., 87: 1318-23, 1997.
7. Kurz A., Sessler D., Lenhardt R.: Peri-operative normothermia to reduce the incidence of surgical-wound infection and shorten hospitalization. N. Engl. J. Med., 19: 1209-15, 1996.
8. Cheney F.: Should normothermia be maintained during major surgery? JAMA, 277: 1165-6, 1997.
9. Peng R., Bongard F.: Hypothermia in trauma patients. J. Am. Coll. Surg., 188: 685-96, 1999.
10. Shiozaki T., Kishikawa M., Hiraide A., Shimazu T., Sugimoto N., Shoshiaka T. et al.: Recovery from post-operative hypothermia predicts survival in extensively burned patients. Am. J. Surg., 165: 326-30, 1993.
11. Cone J.B., Wallace B.H., Caldwell F.T., jr: The effect of staged burn wound closure on the rates of heat production and heat loss of burned children and young adults. J. Trauma, 28: 968-72, 1988.
12. Herndon D. (ed.): “Total Burn Care”, 148-58, W.B. Saunders, London, 1996.
13. Ikeda T., Sessler D., Marder D., Xiong J.: Influence of thermoregulatory vasomotor and ambient temperature variation on the accuracy of core-temperature estimates by cutaneous liquid-crystal thermometers. Anesthesiol., 86: 603-12, 1997.
14. Schwartz M.M., Shires G.T., Spencer (eds): “Principles of Surgery”, 95-118, McGraw-Hill, Inc., New York, 1994.
15. Wilmore D., Mason A., Johnson D., Pruitt B.: Effect of ambient temperature on heat production and heat loss in burned patients. J. Appl. Physiol., 38: 593-7, 1975.
16. Kelemen J. III, Cioffi W., Mason A., Mozingo D., McManus W., Pruitt B., jr: Effect of ambient temperature on metabolic rate after thermal injury. Ann. Surg., 223: 406-12, 1996.
17. Ayres S., Grenvik A., Holbrook P., Shoemaker W. (eds): “Textbook of Critical Care”, Philadelphia, Pa, 1995; Editorial Medica Panamericana, 1402-15, 1996.
18. Martin C.J., Ferguson J.C., Rayner C.: Environmental conditions for treatment of burned patients by the exposure method. Burns, 18: 273-82, 1992.
19. Shiozaki T., Hiraide A., Shimazu T., Ohnishi M., Tasaki O., Shoshiaka T. et al.: Differences in body temperature changes during dressing change in surviving and non-surviving burned patients. Brit. J. Surg., 82: 784-6, 1995.
20. Shoemaker W., Appel P., Kram H.: Prospective trial of supranormal values of survivors as therapeutic goals in high-risk surgical patients. Chest, 94: 1176-86, 1988.
21. Schiller W., Bay C., Mclachlan J., Sagraves S.: Survival in major burn injuries is predicted by early response to Swan-Ganz guided resuscitation. Am. J. Surg., 170: 696-700, 1995.
22. Lee R., Bithell T., Foerster J., Atlens J., Lukens J. (eds): “Wintrobe’s Clinical Hematology”, 1301-24, Lea & Febger, 1993.