SUMMARY. In order to estimate the importance of hypothermia in burn patients, bibliographic databases were searched for relative articles published in MEDLINE between the years 1966 and 2002. The search strategy was based on the terms “burns” and “hypothermia”. We retrieved any kind of article referring to these concepts “limited to text word”. The preliminary data obtained using these search criteria consisted of 108 references. A first analysis performed to apply the first inclusion criterion (referring to burn patients) showed that only 28 references were suitable for analysis. When the second inclusion criterion (implication of hypothermia) was considered only seven articles were found to be suitable for analysis. Data from these studies showed that from the pathophysiological point of view burn patients with a slow recovery from operative hypothermia exhibited impaired oxygen extraction and dependence of oxygen consumption on oxygen delivery over a wide range; the data also showed that patients unable to generate heat well in post-operative hypothermia were unable to produce the additional energy required to overcome sepsis. Considering the epidemiological aspects, about 2% of hypothermic patients had sustained burn wounds and 20% of deceased burn patients were hypothermic. Finally, an analysis of the therapeutic implications suggests that the infusion of hot crystalloids given via central venous access is safe and may be an acceptable adjuvant in attenuating hypothermia during operative procedures.
Whenever we read about the optimal assessment of burn victims in a pre-hospital setting, we must have a clear understanding of the pathophysiological changes occurring locally and systemically after injury. For the mobile emergency unit it is essential to possess an appropriate protocol for the prevention of hypovolaemia, hypothermia, and hypoxaemia. Hypothermia following the pre-hospital treatment of burn patients is a common risk with increasing lethality.1 We commonly read recommendations for the prevention of hypothermia, e.g. we should avoid applying cold water and protect the patient with blankets. But there are few articles about the exact incidence of hypothermia in burn patients, and many questions arise: when is hypothermia a key factor in the evolution of burn patients? What are the appropriate measures for treating (not preventing) hypothermia? And, finally, what is the prognosis of hypothermic burn patients?
In the light of these considerations, we decided to perform a meta-analysis of hypothermia in burn patients in order to try to answer these questions. First of all we tried to define what hypothermia is considered to be.
We found several classifications. Hypothermia can be divided into accidental, primary, or secondary, and it is also classified by severity as mild (32-35 °C), moderate (28-32 °C), or severe (< 28 °C). It may also be acute (minutes), sub-acute (hours), or chronic (days), depending on the time of development.
The pathophysiology of this process shows that hypothermia is caused by a disturbance in the net regulation of heat production and heat loss, weighted towards the latter. This can result from defective homeostatic regulation, reduced metabolism (including diminished cellular metabolism and shivering), increased loss from exposure to extreme cold, and impaired cardiovascular response, especially the loss of vasomotor tone.2 In burn patients, the loss of skin function places patients at risk for hypothermia, fluid and electrolyte imbalances, and systemic sepsis.3 Increased evaporative water loss following thermal injury sometimes results in electrolyte abnormalities, negative nitrogen balance, and hypothermia.4
To estimate the importance of hypothermia in burn patients, bibliographic databases were searched for articles published in MEDLINE between the years 1966 and 2002. The search strategy was based on the terms “burns” and “hypothermia”. We retrieved any kind of article referring to these concepts “limited to text word”.
Inclusion, exclusion, and evaluation criteria for studies
Studies were evaluated independently by four researchers on the basis of predefined validity criteria. Differences were resolved by consensus between the researchers.
The inclusion criteria used to select studies were as follows:
Non-compliance with either of the two criteria led to exclusion from the study.
After retrieving the articles we proceeded to an in-depth analysis of the contents in order to elaborate a document containing the most important conclusions.
Preliminary data obtained with the search criteria “burns” and “hypothermia” limited to “text word” from 1966 to 2002 in MEDLINE produced 108 references. A first analysis performed to apply the first inclusion criteria (referring to burn patients) showed that only 28 references were suitable for analysis. The second inclusion criteria (implication of hypothermia) yielded only seven articles suitable for analysis.
The most relevant data obtained after analysis of these seven articles were:
Physiopathology
Patients with burns who eventually succumbed to their injuries tended to recover more slowly from operative hypothermia than those who survived. Slower recovery was associated with lower post-operative oxygen consumption (VO2). Shiosaki et al.5 investigated whether this was due to impairment of oxygen delivery or extraction. Their study was performed in 13 adult patients with severe burns. One hundred and four measurements of VO2 by indirect calorimetry were made during recovery from 23 episodes of operative hypothermia in 11 patients. Sixty-six measurements of oxygen transport variables by balloon-tipped pulmonary artery catheter were made after 17 episodes of operative hypothermia in six patients. Body temperature was monitored in the urinary bladder. The results obtained in this study showed that the rate of temperature rise (T) showed a strong positive correlation with VO2 measured both by indirect calorimetry (r = 0.91, p < 0.001) and by balloon-tipped pulmonary artery catheter (r = 0.83, p < 0.001). Oxygen delivery (DO2) was above normal in nearly all patients. Oxygen extraction was low in patients recovering slowly (T < 1.0 °C/h)
and high in patients recovering quickly (T Ž 1.0 °C/h). During fast recovery VO2 (373 ± 77 ml/min/n2; mean ± SD) was approximately three times the normal value and was independent of DO2. In contrast, a strong linear relationship existed between VO2 and DO2 during slow recovery (r = 0.76, p < 0.001). Final conclusions indicated that patients suffering from burns with slow recovery
from operative hypothermia exhibited impaired oxygen extraction and dependence of VO2 on DO2 over a
wide range. This picture resembled that of critically ill patients.
The same authors previously published another article6 performed to clarify the cause of post-operative hypothermia in extensively burned patients. Factors affecting post-operative hypothermia were studied in 16 extensively burned adult patients (8 survivors and 8 non-survivors) with a burn index greater than 35. Body temperature was monitored continuously in the urinary bladder or rectum. Hypothermia of less than 35 °C occurred in 66 out of 74 (89%) of the total operations performed in the 16 patients. The rate of temperature rise (RTR) was significantly lower in non-survivors (0.4 ± 0.2 °C/h) than in survivors (1.7 ± 0.9 °C/h; p < 0.001). Continuous indirect calorimetry performed in seven patients (four survivors and three non-survivors) demonstrated that RTR was determined primarily by heat production. The measured energy expenditure reached only 1.7 ± 0.2 times the basal energy expenditure during rewarming in non-survivors, whereas it was 2.7 ± 0.9 times the basal energy expenditure in survivors (p < 0.01). Surprisingly, in non-survivors, RTR was significantly decreased even during the first two weeks. These findings suggest that patients who cannot generate heat well during post-operative hypothermia are unable to produce the additional energy required to overcome sepsis.
Epidemiological considerations on hypothermia in burn patients
Hajek et al.7 analysed a group of 194 deceased subjects (119 men and 75 women; mean age, 15.96 yr) who suffered thermic damage in the years 1981-1989. As regards cause, compared with cases in the period 1966-1974 there was a striking increase in the number of persons suffering burns due to propane-butane explosions. The main thanatological findings were pneumonia (49%), shock (28.3%), sepsis (26.8%), damage to the upper gastrointestinal tract (23.7%), and thrombotic embolism (18%) of the subjects who died. More than half died were normothermic, 25% were hyperpyrexic, and more than 20% were hypothermic.
During the period May 1992-April 1993 Gunning et al.8 evaluated 108 patients admitted to Liverpool Hospital with Injury Severity Scores (ISS) > 15. The temperatures of 100 of these had been recorded and, of these, 17 had a core temperature of less than 35 °C documented within 24 h of arrival. The hypothermic group presented with more severe injuries and contained a disproportionate number of females. Hypothermia was found to be more common in the winter months, but it was not associated with a delay in reaching hospital after the injury. When injuries were ranked by ISS, both hypothermic and normothermic patients were found to be equally likely to have received a blood transfusion; however, the mean number of units of packed cells transfused was greater in the hypothermic group with ISS < 41 than in the similarly injured normothermic group. Two patients in the hypothermic group had sustained burns, and both of these were hypothermic on arrival. All the hypothermic patients who required surgery developed hypothermia in the operating theatre.
Therapeutic implications
It has been reported8 that hypothermia following pre-hospital treatment of burn patients is a common risk with increasing lethality. Soon after admission to a burn unit, the body temperature of 212 adult patients with more than 5% total body surface area burned was documented. No influence was found of the time of pre-hospital care and cold-water treatment alone on body temperature. If the patients were not anaesthetized, the initial temperature was normal. Only anaesthetized and artificially ventilated patients were hypothermic. It was concluded that hypothermia was not a problem in non-anaesthetized and cold-water-treated patients. However, all anaesthetized patients must be carefully treated in order to avoid hypothermia, which is an important complication in pre-hospital management.
Gore and Beaston9 reported that hypothermia exacerbated coagulopathy and was thus a cause of potentially devastating morbidity during operative debridement of burn wounds. Current techniques for maintaining body temperature include warming intravenous fluids to 38 °C. The purpose of Gore and Beaston’s study was to assess the safety of infusing saline heated to 55-60 °C. These researchers, using a modified fluid-warmer, administered saline heated to 60 °C that was infused through a central venous access in eight adult patients undergoing debridement of burn wounds. The temperature of the saline actually entering the patient was measured by a thermocouple attached at the connection to the central line catheter. The results obtained showed that the actual infusate temperature was 54.0 ± 1.2 °C. In the first hour, 1,100 ml of hot saline was given, thus delivering 17.6 kcal more heat than fluid warmed with respect to the traditional 38 °C. Core temperature measured via oesophageal and Foley catheters showed an insignificant upward trend during the operative procedure. There was no evidence of intravascular haemolysis or coagulopathy. These researchers concluded that their pilot study suggested that the infusion of hot crystalloids given via central venous access is safe and may be an acceptable adjuvant in the attenuation of hypothermia during operative procedures.
After analysing the entire bibliography found in the MEDLINE search referring to hypothermia in burn patients, we first concluded that there was an extremely short list of relevant references related to this item. Our understanding is that hypothermia in burn patients is not misunderstood but underestimated. Every time we attend a “Burns Meeting on First Aid” we hear of measures to prevent hypothermia. But nobody talks about the detection of hypothermia in these patients when they are admitted to a burns unit or a general hospital.
First of all we have to consider that normal temperature is not necessarily correlated with a normothermic state. We may find that a patient who should be hyperpyretic owing to the presence of various injuries is surprisingly normothermic. We regard these patients as “functional hypothermic patients”.
The second aspect of the problem relates to when there is a high risk of the development of hypothermia in such patients. To this regard we found three precise moments: first, as is obvious, when patients are exposed to a cold atmosphere or freezing liquids in the acute phase - blankets and measures tending to prevent avoid cold exposure are well defined in “first-aid protocols for burn patients”; the second most important moment is that of surgery - when patients are being debrided, or simply exposed to a cold atmosphere in the operating room, they run a high risk of developing hypothermia; and the third moment is the post-surgical phase, owing to alterations in thermoregulation due to anaesthetic procedures.
The question that then arises is why hypothermia is underestimated. If we analyse the pathophysiological patterns of hypothermia in patients we find the following conditions: defective homeostatic regulation (both the intrinsic and the extrinsic coagulation systems are affected), ineffective platelet function due to inhibition of thromboxane B2, and increased fibrinolytic activity. A heparin-like substance is released, and it is also common to find a reduction in the enzyme activities necessary for initiating and maintaining platelet-fibrin clots, which results in a net increase in bleeding tendency. These features produce a disseminated intravascular coagulation-like syndrome, but with a marked haemorrhagic tendency. This can be aggravated in the hypothermic traumatized patient, who may require massive transfusions owing to blood loss. Other features are reduced metabolism (including diminished cellular metabolism and shivering), or increased loss due to exposure to extreme cold, or impaired cardiovascular response, especially loss of vasomotor tone (both in conditions of hypothermia.2 All these alterations can obviously be found in extensively burned patients even in the absence of hypothermia. That is why we consider that hypothermia is underestimated, which makes it difficult to establish a differential diagnosis.
Another consideration is the influence of hypothermia on burn patient mortality rates. There is no study that analyses this aspect, but when we extrapolated data from mortality rates associated with general hypothermia we found that overall mortality associated with hypothermia was about 17%, involving all ages, aetiologies, and classifications of hypothermia. The influence is thus considerable, and obviously underestimated.
We may say in conclusion that the therapeutic approach to clinically recognized hypothermia is not completely established. Preventive measures include the avoidance of exposure to a cold atmosphere. Perfusion with hot crystalloids is another way to control these alterations. However, throughout our entire meta-analysis, we did not find any complete protocol of detection and treatment of hypothermia.
There are few articles in the literature dealing with hypothermia in burn patients, and this reflects an underestimation of the prevalence of this condition. It would appear that the influence of hypothermia on mortality rates is even higher than our bibliographical review may suggest. For this reason we conclude that further studies should be performed in order to investigate the real incidence and prevalence of hypothermia, its pathophysiological features, its differential diagnosis, and the correct therapeutic approach.
RESUME. Pour évaluer l’importance de l’hypothermie dans les patients brûlés, les Auteurs ont effectué une recherche des bases de données bibliographiques pour trouver les articles relatifs publiés en MEDLINE dans la période 1966-2002. La stratégie de recherche se basait sur les termes “brûlures” et “hypothermie”. Tous les articles qui contenaient des références à ces termes ont été extraits. Les données préliminaires obtenues utilisant ces critères de recherche consistaient en 108 références. Une première analyse effectuée pour appliquer le premier critère d’inclusion (patients brûlés) indiquait que seulement 28 références étaient susceptibles d’analyse. Pour ce qui concerne le deuxième critère d’inclusion (hypothermie), seulement sept articles ont été retenus appropriés pour l’analyse. Les données qui émergeaient de ces études indiquaient que du point de vue pathophysiologique les patients brûlés qui présentaient une guérison lente après l’hypothermie opératoire démontraient une extraction de l’oxygène altérée et la dépendance de la consommation de l’oxygène à l’égard de la provision d’oxygène sur une grande échelle; elles indiquaient en outre que les patients incapables de générer bien la chaleur dans l’hypothermie post-opératoire ne pouvaient pas produire l’énergie additionnelle nécessaire pour vaincre le sepsis. Quant aux aspects épidémiologiques, environ 2% des patients hypothermiques étaient atteints de brûlures et 20% des patients brûlés décédés étaient hypothermiques. Les Auteurs concluent en affirmant que l’analyse des implications thérapeutiques suggère que l’infusion de cristalloïdes chauds administrés par voie veineuse centrale ne présente pas de risques et peut être utile pour atténuer l’hypothermie pendant les procédures opératoires.