Perioperative thermal insulation: minimal clinically important differences?

British Journal of Anaesthesia 92 (6): 836±40 (2004) DOI: 10.1093/bja/aeh156 Advance Access publication April 19, 2004 Perioperative thermal insulation: minimal clinically important differences? A. BraÈuer 1 *, T. Perl 1, Z. Uyanik 1, M. J. M. English 2, W. Weyland 3 and U. Braun 1 1 Department of Anesthesiology, Emergency and Intensive Care Medicine, University of GoÈttingen, Robert-Koch-Str. 40, D-37075 GoÈttingen, Germany. 2 Department of Anaesthesia, Montreal General Hospital and McGill University, Montreal, Canada. 3 Department of Anaesthesia and Intensive Care Medicine, Evangelisches Bethesda-Krankenhaus, Essen, Germany *Corresponding author. Email: abraeue@gwdg.de Background. Reduction of heat losses from the skin by thermal insulation is used to avoid perioperative hypothermia. However, there is little information about the physical properties of various insulating materials used in the operating room. Methods. The following insulation materials were tested using a validated manikin: cotton surgical drape tested in two and four layers; Allegiance drape; 3M Steri-Drape; metallized plastic sheet; ThermadrapeÔ; Barkey thermcare 1 tested in one and ; hospital duvet tested in one and. Heat loss from the surface of the manikin can be described as: QÇ=h DT A where QÇ is heat ux, h is the heat exchange coef cient, DT is the temperature gradient between the environment and surface and A is the area covered. The heat ux per unit area (QÇA ±1 ) and surface temperature were measured with nine calibrated heat- ux transducers. The environmental temperature was measured using a thermoanemometer. DT was varied and h was determined by linear regression analysis as the slope of DT vsqça ±1. The reciprocal of h de nes the insulation. Results. The insulation value of air was 0.61 Clo. The insulation values of the materials varied between 0.17 Clo ( of cotton surgical drapes) to 2.79 Clo ( of hospital duvet). Conclusions. There are relevant differences between various insulating materials. The best commercially available material designed for use in the operating room (Barkey thermcare 1) can reduce heat loss from the covered area by 45% when used in. Given the range of insulating materials available for outdoor activities, signi cant improvement in insulation of patients in the operating room is both possible and desirable. Br J Anaesth 2004; 92: 836±40 Keywords: complications, hypothermia; equipment, insulation; equipment, manikin; heat loss; measurement, heat ux Accepted for publication: February 1, 2004 Perioperative hypothermia is still a common problem during anaesthesia and is associated with various medical risks. These risks include coagulopathy, 12 increased blood loss, 3±6 morbid cardiac events, 7 prolonged postoperative recovery, 8 increased muscle protein breakdown, 9 wound infections, impaired wound healing and prolonged hospital stay. 4 Perioperative hypothermia is also associated with higher treatment costs. 4510 During long surgical operations, perioperative hypothermia can be avoided only if the patient's heat loss is offset by an equal heat gain, either from metabolic heat production or from an external heat source. It is usual to exploit a combination of measures to maintain normothermia. Heat gain from metabolic heat production can be augmented by the infusion of amino acids, 11 while external heat can be applied by conductive warming methods 12 or forced air warming. 4510 Heat losses from the airways can be reduced by the use of heat and moisture exchangers 13 and heat losses from skin that can not be warmed actively can be reduced by insulation. 14 15 Although insulation is invariably used to Ó The Board of Management and Trustees of the British Journal of Anaesthesia 2004

Perioperative thermal insulation reduce heat loss from the skin, there is little information about the physical properties of various insulating materials used in the operating room. Therefore the following study compared the ef cacy of seven insulating materials using a validated copper manikin to simulate the human body. Methods Heat exchange between a dry surface and the environment is caused by radiation and convection. This heat exchange process can be described as: Q Ç =h DT A where Q Ç is heat ux (W), h is the heat exchange coef cient (W m ±2 C ±1 ), DT is the temperature gradient between the environment and surface ( C) and A is the area covered (m 2 ). Q Ç per unit area (Q Ç A ±1 ) can be measured directly with heat- ux transducers, and temperatures can be measured with standard thermometric techniques. From these data, h, which de nes the ef ciency of heat exchange, can be calculated. Covering a surface with insulation decreases heat ow to the environment, therefore lowering h. The reciprocal of h de nes the resistance to heat exchange, or the insulation. Insulation values can be expressed in SI units ( Cm ±2 W ±1 ), `togs' (0.1 Cm ±2 W ±1 ) or Clo units. One Clo unit is equivalent to the insulation required to keep a seated subject comfortable at an air temperature of 21 C in an air movement of 0.1 m s ±1. Such insulation is provided by an ordinary suit, with shirt, trousers etc. 16 One Clo is equivalent to 0.155 C m ±2 W ±1. The following insulation materials were tested: (i) cotton surgical drape TB 202 B (160 cm 3 140 cm) (Karl Dieckhoff GmbH & Co. KG, Wuppertal, Germany), tested in two and four layers, because surgical cotton drapes are rarely used in one layer; (ii) Allegiance beach chair shoulder drape (262 cm 3 411 cm) (Allegiance Healthcare Corporation, McGaw Park, IL, USA); (iii) 3M Steri-Drape adhesive split sheet No. 9045 (228 cm 3 260 cm) (3M Health Care, St Paul, MN, USA); (iv) metallized plastic sheet (140 cm 3 220 cm) (VauDe, Norderstedt, Germany), tested with the silver side facing the manikin; (v) ThermadrapeÔ Blanket T2000 (120 cm 3 120 cm) (OR Concepts, Roanoke, TX, USA); (vi) Barkey thermcare 1 whole body blanket for adults (220 cm 3 140 cm) (Barkey GmbH & Co. KG, LeopoldshoÈhe, Germany) which consists of cotton, polyester and polyester with polyurethane coating; it was tested in one and ; (vii) hospital duvet (188 cm 3 122 cm) (Brinkhaus GmbH & Co. KG, Warendorf, Germany), lled with Trevira (100% polyester). The hospital duvet was tested together with its covering (140 cm 3 200 cm) (Karl Dieckhoff GmbH & Co. KG, Wuppertal, Germany) made of 50% polyester and 50% cotton and was tested as one and two duvets (one and two layers). The manikin The manikin consists of six copper tubes painted matt black. Two tubes serve as arms, two as legs, one as the head and one as the trunk. The total surface area of all tubes is 1.98 m 2. In order to set surface temperature and achieve steady-state conditions, water mattresses (Maxi-Thermâ, Cincinnati Sub-Zero Products Inc., Cincinnati, OH, USA) are bonded to the inner surface of the copper tubes. The circulating water is warmed and cooled by a hypohyperthermia system (Hico-Variotherm 530, Hirtz & Co. Hospitalwerk, Cologne, Germany). Measurement of environmental conditions Air humidity and velocity were measured using a gauged thermoanemometer (Velocicalc plus TSI â Model 8388-M-D, TSI Incorporated, St Paul, MN, USA). Measurement of heat exchange at the manikin We measured Q Ç A ±1 between the environment and the manikin with nine calibrated heat- ux transducers (Heat Flow Sensor Model FR-025-TH44033-F16, Concept Engineering, Old Saybrook, CT, USA) distributed equally over the trunk of the manikin. Measurement of temperature gradient The temperature gradient was de ned as the difference between the environmental temperature and the surface temperature of the manikin underneath the heat- ux transducer. The environmental temperature was measured in the middle of the room and near the wall using the thermoanemometer. The surface temperature of the manikin was measured with calibrated thermistors incorporated into the heat- ux transducers. Data sampling Heat- ux signals were measured and digitized using a Dash TC AD converter (Keithley Instruments Inc., Taunton, MA, USA). The thermistors incorporated into the heat- ux transducers for measurement of the manikin surface temperature were connected to Hellige Servomed 236039 monitors (Hellige, Freiburg, Germany). The signal of these monitors was digitized on a Dash 1402 A/D board (Keithley Instruments Inc., Taunton, MA, USA). All data were sampled synchronously in 10 s intervals on a computer, averaged over 1 min and written to a hard disk. Determination of the heat exchange coef cient The trunk of the manikin was completely covered with the insulation material, which was smoothed at to exclude any obvious trapped air. To determine h, Q Ç A ±1 and DT were measured simultaneously over a range of temperature 837

BraÈuer et al. differences. Six tests were created by setting six different surface temperatures of the manikin (22, 26, 30, 34, 38 and 42 C). Each test consisted of a 60 min preparation period to achieve steady-state conditions followed by a 20 min measurement period. The collected data were averaged for the single measurement period. Each test was repeated three times. There were nine sites, six tests and three repetitions, so that h was calculated from 162 results for Q Ç A ±1 and the corresponding temperature gradients. h was calculated by linear regression analysis as the slope of Q Ç A ±1 as a function of the temperature gradient. Heat ux from the manikin to the environment was called heat loss and was assigned a negative value. Calculation of the insulation values of the tested materials The insulation of the trunk of the manikin when covered with an insulation material represents the total insulation provided by the insulation material and the insulation of air. Therefore the insulation of air was determined by exposing the manikin, using only air as the insulating material. Subtracting the insulation of air from the total insulation gave the insulation value of the tested material. Results The mean ambient temperature for all trials was 22.6 (SD 0.3) C, the relative humidity was 40.5 (6.3)% and the air velocity was below 0.2 m s ±1 with no relevant difference between the single measurement series. Insulation value of air h for the trunk of the manikin was 10.6 W m ±2 C ±1 (Fig. 1) The reciprocal of h de nes the resistance to heat exchange, or the insulation. This resistance is 1/h=1/10.6 W m ±2 C ±1 = 0.09 C m ±2 W ±1 or 0.9 tog or 0.61 Clo. Insulation value of the materials Values of h for the trunk of the manikin and the insulation materials varied between 8.3 and 1.9 W m ±2 C ±1. This corresponds to total insulation values of 0.12±0.53 C m ±2 W ±1 or 1.2±5.3 tog or 0.78±3.40 Clo for the material and air together. Subtracting the insulation of air gave insulation values for the materials of 0.3±4.4 tog or 0.17±2.79 Clo (Table 1, Fig. 2). Discussion Thermal insulation reduces heat loss from the surface of the manikin to the environment by decreasing the radiative and convective heat exchange. This decrease in heat loss is shown by an equivalent decrease in h, provided that the same temperature gradient exists. When total insulation is Fig 1 Determination of the heat exchange coef cient (h) of air. The slope of the temperature (T) gradient between the environment and the surface of the manikin vs the measured heat ux per area (QÇ A ±1 ). The data show the regression line and 95% con dence intervals. Table 1 Heat exchange coef cient (h) of the heat exchange between the manikin and the environment, corresponding total insulation (insulation of the material and air) and insulation of the material Insulation material h (W m ±2 C ±1 ) Total insulation (Clo) Air 10.6 0.61 Cotton surgical drape four layers 8.3 6.5 0.78 0.99 0.17 0.38 Allegiance 7.8 0.83 0.22 3M Steri-Drape 7.8 0.83 0.22 Metallized plastic sheet 4.9 1.32 0.71 Thermadrape 5.5 1.17 0.56 Barkey thermcare 1 one layer Hospital duvet one layer 6.1 4.8 3.4 1.9 1.06 1.34 1.90 3.40 expressed in SI units, h is inversely related to total insulation since insulation is 1/h. When total insulation is expressed in Clo units, then h=6.45/x, since 1 Clo=0.155 Cm ±2 W ±1 and 1/0.155 = 6.45 (Fig. 2). This relationship implies that adding a little insulation to an uninsulated surface can decrease heat loss in a relevant way (e.g. insulating an exposed surface with 0.17 Clo reduces heat loss by 28%) but adding more insulation to an already well insulated surface will have only a small effect on heat loss (e.g. increasing insulation from 2.6 to 2.8 Clo will only decrease h from 2.48 to 2.30 W m ±2 C ±1, a reduction in heat loss of only 9%). Different types of thermal insulators Insulation of the material (Clo) 0.45 0.73 1.29 2.79 Thermal insulators can be divided into two different types. The majority consist of mass insulators (cotton surgical 838

Perioperative thermal insulation The correct determination of insulation values for different insulating materials is complicated by the fact that, while the air trapped within a mass insulator determines that insulator's characteristics, the variable amount of air trapped beneath the insulator will increase its apparent insulation effectiveness. For this reason we excluded any obvious trapped air between the insulating material and the trunk of the manikin by smoothing at all the test materials. In clinical practice there will be more trapped air under the insulation material and therefore the practical insulation of the materials will be slightly higher. Fig 2 Heat exchange coef cient (h) values vs total insulation values of the insulation materials. h is reciprocally related to total insulation. A reduction in h equals a reduction in heat loss if the temperature gradient between the environment and the surface is unchanged. drape, Allegiance drape, 3M Steri-Drape, Barkey thermcare 1 whole body blanket, hospital duvet). These insulators entrap air within a bre matrix. This entrapped air does not move and is called `still air'. Still air is a very effective insulator, with an insulation value of 1.8 Clo cm ±1. 17 Therefore the insulation value of these insulators is proportional to the thickness of the still air enclosed. The 16 18 kind of bre used to trap the air is of little importance. The second type of insulator is the radiant insulator (ThermadrapeÔ blanket, metallized plastic sheet) which re ects radiant heat back to the radiating surface and emits little radiant heat to the exterior. To provide a signi cant effect, the radiant insulator should have a distance of about 1 cm from the radiating surface. 17 This distance to the radiating surface should consist of air. If this distance is lled by loose material of low bulk density, the effect of the radiant insulator is reduced. 17 Simulation of heat loss and the in uence of insulation by the manikin Thermal manikins are used extensively in environmental physiology 16 19 and are a useful and valuable complement to direct experiments with human volunteers. The main application areas of thermal manikins are relevant simulation of human whole body and local heat exchange. Clothing insulation in particular has been extensively studied in heated thermal manikins and this work forms the basis of American and European standards. 19 The heat exchanging properties of our manikin have been validated. 20 h for the whole manikin is 11 W m ±2 C ±1.In this study, we used only the trunk and found a heat exchange coef cient of 10.6 W m ±2 C ±1. This corresponds very well with the heat exchange coef cient of 10.8 W m ±2 C ±1 we found in human volunteers. 20 Insulation values of the insulation materials The insulation value of air was 0.61 Clo, which means that air is a better insulator than the materials found in an operating room. A value of 0.61 Clo compares well with insulation values of air given by Burton and Edholm. 17 The insulation materials had insulation values between 0.17 Clo and 2.79 Clo. This result is different from the results of a study by Sessler and colleagues, 14 who concluded that there were only minor important differences among the thermal barriers. The reason for this is that we included effective insulating materials that are not commonly used in the operating room (e.g. hospital duvet). However, if we compare similar materials in both studies we nd very similar results. We have also found that disposable covers are more effective than a cloth surgical drape, but they are less effective than a re ective material (e.g. Thermadrape). Adding additional layers of the insulating material increases the ef cacy. This result is also comparable to a volunteer study. 15 However, the exact in uence of more layers on the reduction of heat loss is still to be determined. In contrast to many clinical studies, 21±23 the results of the radiant insulators were better than many other insulating materials. Possibly the ef cacy of these materials is lowered in clinical practice by placing additional sterile drapes on them. This consideration is con rmed by studies that have found no improvement of thermal insulation by adding 24 25 radiant insulators sandwiched into insulating materials. However, this problem requires further detailed analysis. Conclusion There are relevant differences between various insulating materials. Heat loss can be reduced in a relevant way by insulation materials, which should be applied to those areas of the body surface that cannot be warmed actively. The best commercially available material designed for use in the operating room (Barkey thermcare 1) reduces heat loss from the covered area by about 45% when used in. However, with better insulating materials (e.g. of a hospital duvet) heat loss can be reduced to about 80%. It should be possible to manufacture specially designed insulating materials for the operating room with insulation values of 2±2.5 Clo, as materials like this are used for 839

BraÈuer et al. outdoor activities 18 and army uniforms. 17 The effects of multiple layers of insulation and the effects of radiant insulators require further investigation. Acknowledgement Thomas Schulze is thanked for programming the data acquisition program and excellent technical help. References 1 Valeri CR, Khabbatz K, Khuri SF, et al. Effect of skin temperature on platelet function in patients undergoing extracorporal bypass. J Thorac Cardiovasc Surg 1992; 104: 108±16 2 Rohrer MJ, Natale AM. Effect of hypothermia on the coagulation cascade. Crit Care Med 1992; 20: 1402±5 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±92 4 Kurz A, Sessler DI, Lenhard R, and the Study of Wound Infection and Temperature Group. Perioperative normothermia to reduce the incidence of surgical-wound infection and shorten hospitalization. N Engl J Med 1996; 334: 1209±15 5 Bock M, MuÈller J, Bach A, BoÈhrer H, Martin E, Motsch J. Effects of preinduction and intraoperative warming during major laparotomy. Br J Anaesth 1998; 80: 159±63 6 Winkler M, Akca O, Birkenberg B, et al. Aggressive warming reduces blood loss during hip arthroplasty. Anesth Analg 2000; 91: 978±84 7 Frank SM, Fleischer LA, Breslow MJ, et al. Perioperative maintenance of normothermia reduces the incidence of morbid cardiac events. A randomized clinical trial. JAMA 1997; 277: 1127±34 8 Lenhardt R, Marker E, Goll V, et al. Mild intraoperative hypothermia prolongs postanesthetic recovery. Anesthesiology 1997; 87: 1318±23 9 Carli F, Itiaba K. Effect of heat conservation during and after major abdominal surgery on muscle protein breakdown in elderly patients. Br J Anaesth 1986; 58: 502±7 10 Ng S, Oo C, Loh K, Lim P, Chan Y, Ong B. A comparative study of three warming interventions to determine the most effective in maintaining perioperative normothermia. Anesth Analg 2003; 96: 171±6 11 SelldeÂn E, BraÈnstroÈm R, Brundin T. Preoperative infusion of amino acids prevents postoperative hypothermia. Br J Anaesth 1996; 76: 227±34 12 Matsuzaki Y, Matsukawa T, Ohki K, Yamamoto Y, Nakamura M, Oshibuchi T. Warming by resistive heating maintains perioperative normothermia as well as forced air heating. Br J Anaesth 2003; 90: 689±91 13 Bickler PE, Sessler DI. Ef ciency of airway heat and moisture exchangers in anesthetized humans. Anesth Analg 1990; 71: 415± 18 14 Sessler DI, McGuire J, Sessler AM. Perioperative thermal insulation. Anesthesiology 1991; 74: 875±9 15 Sessler DI, Schroeder M. Heat loss in humans covered with cotton hospital blankets. Anesth Analg 1993; 77: 73±7 16 Clark RP, Edholm OG. Man and his Thermal Environment. London: Edward Arnold Publishers, 1985 17 Burton AC, Edholm OG. Man in a Cold Environment. London: Edward Arnold Publishers, 1955 18 Kaufman, WC, Bothe D, Meyer SD. Thermal insulating capabilities of outdoor clothing materials. Science 1982; 215: 690±1 19 HolmeÂr I. Thermal manikins in research and standards. In: Nilsson HO, HolmeÂr I, eds. Proceedings of the Third International Meeting on Thermal Manikin Testing, 3IMM, at the National Institute for Working Life, October 12±13, 1999. Stockholm: National Institute for Working Life, 2000; 1±7 20 BraÈuer A, English MJM, Sander H, Timmermann A, Braun U, Weyland W. Construction and evaluation of a manikin for perioperative heat exchange. Acta Anaesthesiol Scand 2002; 46: 43±50 21 Bennett J, Ramachandra V, Webster J, Carli F. Prevention of hypothermia during hip surgery: effect of passive compared with active skin surface warming. Br J Anaesth 1994; 73: 180±3 22 Berti M, Casati A, Torri G, Aldegheri G, Lugani D, Fanelli G. Active warming, not passive heat retention, maintains normothermia during combined epidural-general anesthesia for hip and knee arthroplasty. J Clin Anesth 1997; 9: 482±6 23 Simmons M, Phillips P, Doctor U, Liehr P. The effect of two intraoperative heat-conserving methods on orthopedic patients receiving regional and general anesthesia. J Post Anesth Nurs 1992; 7: 170±5 24 Kaufman WC, Bothe DJ. Thermal insulation of materials with possible aerospace application. Aviat Space Environ Med 1986; 57: 993±6 25 Light IM, Norman JN. The thermal properties of a survival bag incorporating metallised plastic sheeting. Aviat Space Environ Med 1980; 51: 367±70 840