Rapid Core-to-Peripheral Tissue Heat Transfer During Cutaneous Cooling Olga Plattner, MD*, Junyu Xiong, MD*, Daniel I. Sessler, MD*t, Harald Schmied, MD$,

Richard Christensen, BP, Minang Turakhia, BS*, Martha Dechert, BA+, and David Clough, MD*

*Department of Anesthesia, University of California, San Francisco, California, tDepartment of Anesthesia and Intensive Care, University of Vienna, Vienna, Austria, and *Department of Anesthesia and Intensive Care Medicine, Hospital of Amstetten, Amstetten, Austria

Perioperative thermal manipulations are usually directed at the skin surface because methods of directly warming the core are invasive or ineffective. However, inadequate heat flow between peripheral and core compartments will decrease the rate at which core temperature changes. We therefore determined whether core hypothermia is de- layed after initiation of surface cooling. Six volunteers were anesthetized with propofol and midazolam, and maintained under three layers of passive insulation for 2.5-4 h. Subsequently, the skin surface was cooled using forced air, 1000 L/min, at 10°C. Isoflurane was added as necessary to maintain arteriovenous shunt vasodilation. Overall heat balance was determined from the difference between cutaneous heat loss (thermal flux transducers) and metabolic heat production (oxygen consumption).

Average arm and leg (peripheral) tissue temperatures were determined from 19 intramuscular needle thermo- couples, 10 skin temperatures, and “deep” foot tempera- ture. Overall body heat content decreased -234 kcal dur- ing 2.5 h of active cooling. Core temperature, which was nearly constant before active cooling, decreased -1.3”C/h. There was no delay between initiation of active cooling and the decrease in core temperature. Further- more, peripheral (arm and leg) and core (trunk and head) tissue heat contents decreased at virtually the same rates: -50 kcal/h and -47 kcal/h, respectively. These data in- dicate that there is little restriction of heat flow between peripheral and core tissues in vasodilated, anesthetized subjects.

(Anesth Analg 1996;82:925-30)

S urgical patients are often actively warmed to treat inadvertent hypothermia; they may also be therapeutically cooled for protection against

brain (1) or spinal cord (2) ischemia. Effective methods of directly altering core temperature, such as peritoneal dialysis, are invasive. In contrast, noninvasive meth- ods, including airway heating and humidification, are of little value in adult patients (3). Consequently, peri- operative thermal manipulations are usually directed at the skin surface-by far the most important phys- iological heat exchanger in humans.

A limitation of adding or removing heat from the skin surface is that resulting changes in core temperature may be delayed, since heat must not only traverse the skin surface but must also be conveyed to core tissues.

Supported by National Institutes of Health Grant GM49670, Au- gustine Medical, Inc., the Max Kade Foundation, New York, NY, and the Joseph Drown Foundation, Los Angeles, CA.

Accepted for publication January 17, 1996. Address correspondence to Daniel I. Sessler, MD, Department of

Anesthesia, 374 Parnassus Ave., University of California, San Fran- cisco, CA 94143-0648. Address e-mail to sesslerQvaxine.ucsf.edu. World Wide Web: http://thermo.ucsf.edu.

01996 by the International Anesthesia Research Society 0003~2999/96/$5.00

Mechanisms by which heat can be transferred from skin to the core include: 1) direct conduction through tissues; 2) blood-borne convection from the skin sur- face; and 3) blood-borne convection from “deep” pe- ripheral tissues, especially muscle.

To the extent that conductive and convective heat flow between peripheral and core thermal compart- ments is inadequate, surface thermal manipulations will change core temperature less than mean body (weighted average tissue) temperature. Furthermore, if core temperature perturbations are restricted by tissue heat transfer rates, increasing device efficacy (e.g., cutaneous heat transfer) may prove only margin- ally helpful. We therefore determined whether core hypothermia is delayed after initiation of surface cool- ing, using an experimental design allowing us to quantify regional body heat distribution and flow.

Methods With approval from the Committee on Human Re- search at the University of California, San Francisco,

Anesth Analg 1996;82:925-30 925

926 PLATTNER ET AL. RAPID CORE-TO-PERIPHERAL HEAT TRANSFER

ANESTH ANALG 1996;82:925-30

and informed consent, we studied six male volunteers who were within 20% of ideal body weight. The vol- unteers’ morphometric characteristics included age 29 +- 6 yr, height 178 + 8 cm, and weight 72 ? 6 kg. Ambient temperature was maintained at 23.0 + 0.3”C and ambient relative humidity at 38% + 3% during the study period.

Protocol

Studies began at approximately 9:00 AM and volun- teers fasted during the preceding 8 h. Throughout the study, minimally clothed volunteers reclined on an operating room table set in chaise-lounge position. An intravenous catheter was inserted, and =lO mL/kg lactated Ringer’s solution warmed to 37°C was in- fused immediately before induction of anesthesia. Subsequently, fluid was infused at a rate of -5 mL * kg-’ * h-i. To minimize redistribution hypothermia (4), the volunteers were prewarmed (5) for 2 h before induction of anesthesia using a disposable paper/ plastic full-body Bair Hugger@ cover (Augustine Med- ical, Inc., Eden Prairie, MN) attached to a Bair Hug- ger@ forced-air warming unit set on “high.” Two cot- ton hospital blankets were positioned above the forced-air cover.

Anesthesia was induced without any premeditation by administration of propofol 3-4 mg/kg and mida- zolam 2-4 mg. Vecuronium bromide, 10 mg, was ad- ministered intravenously, and the trachea intubated. Muscle relaxation was subsequently maintained with an infusion of vecuronium adjusted to maintain O-l twitches in response to supramaximal train-of-four electrical stimulation of the ulnar nerve at the wrist. Ventilation was controlled by an Ohmeda ventilator incorporated into a Modulus CD@ integrated anesthe- sia machine (Ohmeda, Inc., Madison, WI). The system was modified so that respiratory gases were not re- breathed. The volunteers’ lungs were ventilated with air at a rate and volume sufficient to maintain end- tidal Pco, near 35 mm Hg. Heat and moisture- exchanging filters were not used. Anesthesia was maintained with an infusion of propofol (7-10 pg. kg-’ . mini) and midazolam (1 mg/h); because we previously demonstrated that vasoconstriction de- lays core cooling (6), isoflurane was added when nec- essary to prevent peripheral vasoconstriction (see below).

After induction of anesthesia, a bladder catheter was inserted. During the first 2.5-4 h of anesthesia, the volunteers remained covered with the forced-air cover and two cotton blankets; however, the forced-air cover was uninflated. This amount of insulation was chosen because previous studies (7) indicated that three lay- ers of passive insulation would reduce heat loss to the level of metabolic heat production. Redistribution hy- pothermia (4) would still be expected to reduce core

temperature during the first ml.5 h after induction, but core temperature would then remain nearly con- stant. Subsequently, at elapsed time 0, the Bair Hug- ger@ cover was attached to a forced-air cooling unit (PolarAir@; Augustine Medical, Inc.). This device pro- vides -1000 L/min air at 10°C. The study concluded after 2.5 h of active cooling. The arms and legs were included in active cooling; however, two cotton towels protected the hands and feet from direct contact with cold air to avoid locally mediated vasoconstriction (8).

Heat Balance Monitoring

Heat flux from 15 skin-surface sites was measured using thermal flux transducers (Concept Engineering, Old Saybrook, CT) (4). As in a previous study, meas- ured cutaneous heat loss was augmented by 10% to account for insensible transcutaneous evaporative loss and 3% to compensate for the skin covered by the volunteers’ shorts (9). We further augmented cutane- ous loss by 10% of the metabolic rate (as determined from oxygen consumption) to account for respiratory loss (10). We defined flux as positive when heat tra- versed skin to the environment.

Oxygen consumption was measured using a meta- bolic monitor (DeltatracTM; SensorMedics Corp., Yorba Linda, CA). Oxygen consumption (mL/min) was con- verted to equivalent metabolic heat production (W) assuming the caloric value of oxygen to be 4.82 kcal/L (respiratory quotient = 0.82) (ll), and using a conver- sion of 1 kcal/h = 1.16 W.

Blood Flow Monitoring

Absolute right middle fingertip blood flow was quan- tified using venous-occlusion volume plethysmogra- phy at 5-min intervals (12). Arteriovenous shunt (13) vasoconstriction was also estimated using laser Dop- pler flowmetry (Periflux 3; Perimed Inc., Piscataway, NJ) with an integrating multiprobe (“wide-band” set- ting) positioned on the right fourth finger. Total digi- tal flow also was evaluated on the right second toe using the Perfusion Index which is derived from ab- sorption of two infrared wavelengths (14).

Left calf and forearm blood flows were quantified using “extensometer” plethysmography (15,16). We did not isolate arteriovenous shunts in the hand or foot with an arterial tourniquet because we were in- terested in total extremity blood flow.

Tissue Temperature and Heat Content

Core temperature was measured in the distal esopha- gus. Skin-surface temperatures were recorded from thermocouples incorporated into thermal flux trans- ducers. Area-weighted mean skin temperature was calculated from 15 sites as previously described (4).

ANESTH ANALG 1996;82:925-30

PLATIYNER ET AL. 927 RAPID CORE-TO-PERIPHERAL HEAT TRANSFER

Arm and leg tissue temperatures were determined as previously described (4). Briefly, the length of the thigh (groin to midpatella) and lower leg (midpatella to ankle) were measured in centimeters. Circumfer- ence was measured at the mid upper thigh, mid lower thigh, mid upper calf, and mid lower calf. At each circumference, right leg muscle temperatures were recorded using, B-, 18-, and 38-mm-long, 21-gauge needle thermocouples (Mallinckrodt Anesthesiology Products, Inc., St. Louis, MO) inserted perpendicular to the skin surface. Skin-surface temperatures were recorded immediately adjacent to each set of needles, and directly posterior to each set. Subcutaneous tem- perature was measured on the ball of the foot using a Coretemp@ (Terumo Medical Corp., Tokyo, Japan) “deep tissue” thermometer (17).

The lengths of the right arm (axilla to elbow) and forearm (elbow to wrist) were measured. The circumfer- ence was measured at the midpoint of each segment. As in the leg, needle thermocouples were inserted into each segment and skin-surface temperatures were recorded immediately adjacent to each set. Additionally, an B-mm-long needle thermocouple was inserted directly into the adductor pollicis.

The leg was divided into five segments: upper thigh, lower thigh, upper calf, lower calf, and foot. Each thigh and calf segment was further divided into an anterior and posterior section, with one third of the estimated mass considered to be posterior. Anterior leg tissue temperature distributions were calculated using fourth-order regression. In contrast, arm and posterior leg segment temperatures were evaluated using parabolic regression. In each case, inspection of the raw data confirmed adequate correlation. Limb heat content was estimated from these temperatures, as previously described (41, by integrating over vol- ume and multiplying by the specific heat of muscle. Posterior segment heat content was similarly calcu- lated, using the assumption that radial temperature distribution in the posterior leg segments would also be parabolic. “Deep temperature,” measured on the ball of the foot, was assumed to represent the entire foot. Foot heat content thus was calculated by multi- plying foot temperature by the mass of the foot and the specific heat of muscle. We further assumed that tissue temperature in the contralateral limb was similar.

Arm tissue temperature and heat content were de- termined similarly. Adductor pollicis temperature was assumed to represent that of the entire hand. Hand heat content thus was calculated by multiplying adductor pollicis temperature by the mass of the hand and the specific heat of muscle. As in the leg, average temperatures of the arm and forearm (forearm and hand) were calculated by weighting values from each of the three segments in proportion to their estimated masses. We have previously described the derivation

of these equations, and their limitations (18). Changes in trunk and head heat content were modeled simply by multiplying the weight of the trunk and head by the change in core temperature and the average spe- cific heat of human tissues.

Statistical Analysis

Results from the first 1.5-2 h after induction of anes- thesia were discarded because redistribution is the most important cause of core hypothermia during this period (4). Subsequent results were indexed to initia- tion of forced-air cooling (elapsed time 0).

Changes in mean body temperature were deter- mined using two separate methods: “Overall,” by in- tegrating the difference between heat production, and “Extremities & Core,” by adding the change in ex- tremity heat content to the change in core temperature multiplied by the mass of the core and the specific heat of human tissue (0.83 cal * “C-i * g-l) (19).

The effects of surface cooling were evaluated by comparing peripheral and core tissue temperatures and heat contents. Time-dependent changes were evaluated using repeated-measures/analysis of vari- ance; values were compared with those recorded at elapsed Time 0 with Dunnett’s test. Peripheral blood flows were compared with those during the precool- ing period (-1-O h) using repeated-measures analysis of variance and Dunnett’s tests. Results are expressed as mean + SD; differences were considered statistically significant when P < 0.01.

Results Estimated mass of the thighs and lower legs (includ- ing feet) were 12 5 1 kg and 8 + 1 kg, respectively. Similarly, estimated mass of the arms and forearms (including hands) were 5 + 1 kg and 3 + 1 kg, respec- tively. Consequently, the arms represented -13% of our volunteers’ mass and the legs represented -28% of our volunteers’ mass, for a total of 41%. Per study protocol, arteriovenous shunt vasodilation was main- tained throughout the study. Shunt dilation was doc- umented by absolute finger blood flow near 2.5 mL/min, laser Doppler perfusion near 150 U, and a perfusion index near 2.0 U. (Typical vasoconstricted values for the three measures might be cl.0 mL/min, ~50 U, and ~0.5 U.) Similarly, whole leg and whole arm blood flows remained increased throughout the study (Table 1).

Cutaneous heat loss increased from =60 kcal/h while the volunteers were covered with two blankets and an uninflated paper/plastic cover to -150 kcal/h during forced-air cooling. In contrast, metabolic heat production which was =60 kcal/h before cooling, de- creased only slightly during the study period. As a result, overall body heat content-calculated as the

928 PLATI’NER ET AL. RAPID CORE-TO-PERIPHERAL HEAT TRANSFER

ANESTH ANALG 1996;82:925-30

Table 1. Peripheral Blood Flows

Precooling (-1-O h) Cooling (O-1 h) Cooling (1.5-2.5 h)

Finger plethysmography (mL/min) 2.6 t 1.2 2.7 t 1.8 2.3 2 1.0 Finger laser Doppler (U) 176 + 2 166 + 2 152 + 3” Toe perfusion index (U) 2.0 rt 0.4 2.1 t 0.4 2.4 t 0.6 Arm extensometer 5.4 2 2.8 5.3 + 3.0 4.6 t 2.8

(mL * min-l * 100 g-‘1 Leg extensometer 5.9 2 1.2 5.4 It 1.5 4.1 2 1.1

(mL * min-* * 100 8-l)

Values are expressed as mean i: SD. *Statistically significant, but clinically unimportant, difference from the precooling period. The decrease in leg flow during the 1.5 to 2.5-h period was nearly

statistically significant (P = 0.02), but probably clinically unimportant.

time integral of metabolic heat production minus cu- taneous heat loss-decreased 234 kcal during 2.5 h of forced-air cooling. Changes in body heat content were virtually identical when independently calculated as the sum of the change in extremity heat content and the change in core temperature multiplied by the spe- cific heat of human tissue and the weight of the trunk and head (Fig. 1).

Mean skin temperature decreased more than 2°C during the first half hour of forced-air cooling, and subsequently decreased at a rate of =1.6”C/h. Core temperature remained nearly constant during the hour before active cooling started. Subsequently, the core cooled at a rate of -1.3”C/h (? = 0.99). There was no delay between initiation of active cooling and the decrease in core temperature. Furthermore, pe- ripheral (arm and leg) and core (trunk and head) tissue heat contents decreased at virtually identical rates: -50 kcal/h (? = 0.99) and -47 kcal/h (? = 0.99), respectively (Fig. 2).

Discussion Per study protocol, arteriovenous shunts remained dilated throughout the protocol, far exceeding typical vasoconstricted values (20). Blood flow to the entire arm and leg was also increased throughout the study. Arm and leg flows were only slightly less than values reported previously in anesthetized, vasodilated sub- jects, and approximately three-fold greater than in unanesthetized, vasoconstricted subjects (4).

Cutaneous heat loss during the precooling period (-1 to 0 elapsed hours) was -60 kcal/h, which is similar to that reported previously for subjects cov- ered by three layers of passive insulation (7). Appli- cation of forced-air cooling markedly increased loss to -150 kcal/h and resulted in a strongly negative over- all heat balance of -95 kcal/h. Mean body tempera- ture therefore decreased at a rate of 1.6”C/h.

As expected, forced-air cooling rapidly reduced mean skin temperature. However, peripheral tissue and core temperatures decreased nearly as fast. There

Heat (kcal/h)

42 (kcal)

-100

-200

-300 J, -1

&psed’Time;h) 3

Figure 1. Cutaneous heat loss increased from -60 kcal/h while the volunteers were covered with a single blanket to -150 kcal/h during forced-air cooling. In contrast, metabolic heat production, which was -60 W before cooling, decreased only slightly during the study period. As a result, body heat content-calculated as the time integral of metabolic heat production minus cutaneous heat loss (Overall)-decreased 234 kcal during 2.5 h of forced-air cooling. Changes in body heat content were virtually identical when inde- pendently calculated as the sum of the change in extremity heat content and the change in core temperature multiplied by the spe- cific heat of human tissue and the weight of the trunk and head. Changes in heat content are presented cumulatively, as mean t SD, referenced to initiation of forced-air cooling at elapsed Time 0. *Values differing significantly from those at Time 0.

was thus little restriction of heat flow between periph- eral and core tissues, and minimal regional accumu- lation of heat. With cutaneous vasodilation, changes in core heat content and temperature were, therefore, limited by device efficacy, not by internal restrictions on heat flow. These data suggest that augmenting device efficacy (i.e., increasing cutaneous heat flux) will be matched by comparable increases in the rate of core temperature change.

Forced-air warming rapidly increases intraopera- tive core temperature (21), but reportedly has little effect postoperatively (22). The major difference be- tween these patient populations is that surgery was accompanied by general anesthesia, which in turn prevents thermoregulatory vasoconstriction (23). In

ANESTH ANALG l’LA7-I-NER ET AL. 929 1996,82 925,10 RAPID CORC-TO~I’EIZII’HERAL IHEAT TRANSFER

I I -1 0 1 2 3

Elapsed Time (h)

Figure 2. Core temperature remained nearly constant during the hour before active cooling started. Subsequently, the core cooled %1.3”C/li; there was no delay between initiation of active cooling and the decrease in core temperature. Peripheral (arm and leg) and core (trunk and head) tissue heat contents decreased at virtually the same rates: -50 kcal/h and -47 kcal/h, respectively. Values are presented as mean ? SD, referenced to initiation of forced-air cool- ing at elapsed Time 0. All values after 0 elapsed hours differed significantly from those at Time 0.

contrast, arteriovenous shunt vasoconstriction is uni- versal in hypothermic postoperative patients (24). Contrasting effectiveness of forced-air warming dur- ing and after surgery suggests that peripheral-to-core heat transfer is impeded by thermoregulatory vaso- constriction and/or facilitated by nonthermoregula- tory anesthetic-induced vasodilation. Consistent with this theory, vasoconstriction reduces the rate at which forced-air cooling decreases core temperature (6). Arteriovenous shunt vasodilation was maintained throughout our current study and thus represents a “best case” evaluation of intercompartmental heat transfer. It remains likely that heat flow between pe- ripheral and core compartments is impeded in un- anesthetized and/or vasoconstricted subjects in whom extremity blood flow is markedly reduced.

Rather than directly evaluate distribution of tissue heat within the trunk and head, we considered these tissues to be uniform, and applied changes in core temperature to the entire mass. To the extent that trunk tissue temperature is inhomogeneous, this as- sumption will be invalid. Our use of heat flux trans- ducers to estimate overall heat loss is fraught with potential errors: 1) transducers are only certified to i5%, and the calibration process is tricky (25); 2) flux across individual transducers was extrapolated to the entire region assigned to each; and 3) we augmented flux to account for insensible cutaneous and respira- tory heat losses without directly measuring either.

Similarly, the potential error in estimating metabolic heat production from oxygen consumption is proba- bly near 10%. And finally, our estimates of tissue heat

content involve many assumptions, including: 1) min- imal conduction along the stainless steel shaft of the needle thermocouples; 2) radial and axial symmetry within extremity segments; and 3) homogeneous tis- sue specific heats. Despite these potential sources of error, changes in body heat content were similar when determined using two independent methods (integrat- ed net heat loss and measured changes in extremity and core heat contents), suggesting that our results are generally correct.

In summary, there was no delay between initiation of active cooling and the resulting decrease in core temperature. Furthermore, average peripheral (arm and leg) temperature and core (trunk and head) tem- perature decreased at similar rates. These data indi- cate that there is little restriction of heat flow between peripheral and core tissues in vasodilated, anesthe- tized subjects.

Ohmeda, Inc., Salt Lake City, UT, loaned the Modulus@ CD inte- grated anesthesia machine and Mallinckrodt Anesthesiology Prod- ucts, Inc., St. Louis, MO, donated the thermocouples that were used. Datex Medical Instruments (Tewksbury, MA) kindly loaned us a Capnomac Ultimam.

References 1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

Baker KZ, Young WL, Stone GJ, et al. Deliberate mild intraop- erative hypothermia for craniotomy. Anesthesiology 1994;81: 361-7. Vacanti RX, Ames A III. Mild hypothermia and Mg’l protect against irreversible damage during CNS ischemia. Stroke 1984; 15:695-g. Deriaz H, Fiez N, Lienhart A. Jnfluence d’un filtre hygrophobe ou d’un humidificateur-rechauffeur sur l’hypothermie perop- eratoire. Ann Fr Anesth Reanim 1992;11:145-9. Matsukawa 7, Sessler DI, Sessler AM, et al. Heat flow and distribution during induction of general anesthesia. Anesthesi- ology 1995;82:662-73. Hynson JM, Sessler DI, Moayeri A, et al. The effects of pre- induction warming on temperature and blood pressure during propofol/nitrous oxide anesthesia. Anesthesiology 1993;79: 219-28. Kurz A, Sessler DI, Birnbauer F, et al. Thermoregulatory vaso- constriction impairs active core cooling. Anesthesiology 1995; 82:870&6. Sessler DI, Schroeder M. Heat loss in humans covered with cotton hospital blankets. Anesth Analg 1993;77:73-7. Wurster RD, McCook RD. Influence of rate of change in skin temperature on sweating. J Appl Physiol 1969;27:237-40. Hammarlund K, Sedin G. Transepidermal water loss in new- born infants. III. Relation to gestational age. Acta Paediatr Stand 1979;68:795-801. Bickler I’, Sessler Dl. Efficiency of airway heat and moisture exchangers in anesthetized humans. Anesth Analg 1990;71: 415-S. Pike RL, Brown ML. Nutrition, an integrated approach. 3rd ed. New York: John Wiley, 1984:76556. Rubinstein EH, Sessler DI. Skin-surface temperature gradients correlate with fingertip blood flow in humans. Anesthesiology 1990;73:541-5. Hales JRS. Skin arteriovenous anastomoses, their control and role in thermoregulation. In: Johansen K, Burggren W, eds. Cardiovascular shunts: phylogenetic, ontogenetic and clinical aspects. Copenhagen: Munksgaard, 1985:433-51.

930 PLATTNER ET AL. RAPID CORE-TO-PERIPHERAL HEAT TRANSFER

14. Ozaki M, Sessler DI, Lopez M, et al. Pulse oximeter-based flow index correlates well with fingertip volume plethysmography [abstract]. Anesthesiology 1993;79:A542.

15. Whitney RJ. The measurement of volume changes in human limbs. J Physiol (Land) 1953;121:1-27.

16. Brimacombe JR, Macfie AG, McCrirrick A. The extensometer: potential applications in anaesthesia and intensive care. Anaes- thesia 1991;46:756-61.

17. Kobayashi T, Nemoto T, Kamiya A, et al. Improvement of deep body thermometer for man. Ann Biomed Eng 1975;3:181-8.

18. Belani K, Sessler DI, Sessler AM, et al. Leg heat content contin- ues to decrease during the core temperature plateau in humans. Anesthesiology 1993;78:856-63.

19. Mekjavic IB, Bligh J. The increased oxygen uptake upon immersion: the raised external pressure could be a causative factor. Eur J Appl Physiol 1989;58:556-62.

20. Sessler DI, Moayeri A, Stsen R, et al. Thermoregulatory vaso- constriction decreases cutaneous heat loss. Anesthesiology 1990; 73~656-60.

ANESTH ANALG 1996;82:925-30

21. Kurz A, Kurz M, Poeschl G, et al. Forced-air warming maintains intraoperative normothermia better than circulating-water mat- tresses. Anesth Analg 1993;77:89-95.

22. Ereth MH, Lennon R, Sessler DI. Isolation of peripheral and central thermal compartments in vasoconstricted patients. Aviat Space Environ Med 1992;63:1065-9.

23. Matsukawa T, Kurz A, Sessler DI, et al. Propofol linearly re- duces the vasoconstriction and shivering thresholds. Anesthe- siology 1995;82:1169-80.

24. Kurz A, Sessler DI, Narzt E, et al. Postoperative hemodynamic and thermoregulatory consequences of intraoperative core hy- pothermia. J Clin Anesth 1995;7:359-66.

25. Ducharme MB, Frim J, Tikuisis I’. Errors in heat flux meas- urements due to the thermal resistance of heat flux disks. J Appl Physiol 1990;69:776-84.