1 British Journal of Anaesthesia 1998; 80: Effects of preinduction and intraoperative warming during major laparotomy M. BOCK, J. MÜLLER, A. BACH, H. BÖHRER, E. MARTIN AND J. MOTSCH Summary We have investigated the influence of active warming before and during operation on blood loss, transfusion requirements, duration of stay in the post-anaesthesia care unit (PACU) and perioperative costs in 40 patients undergoing major abdominal surgery. Patients were allocated randomly to one of two groups: in the study group (n 20), patients were actively warmed using forced air for 30 min before induction of general anaesthesia and during anaesthesia. Passive protection against heat loss consisted of circulating water mattresses, blankets and fluid warming devices, and was used both in the active warming group and in the control group (n 20). At the end of surgery the change in core temperature was significantly less in the group of actively warmed patients (0.5 (SD 0.8) C vs 1.5 (0.8) C; P 0.01). Blood loss and transfusion requirements were less in the actively warmed patients, who had a shorter duration of stay in the PACU (94 (SD 42) min vs 217 (169) min; P 0.01) and a 24 % reduction in total anaesthetic costs. (Br. J. Anaesth. 1998; 80: ) Keywords: surgery, abdominal; hypothermia; temperature, effects; temperature, monitoring; equipment, warming devices Central body temperature decreases after induction of gen eral an aesthesia as a result of reduced metabolic heat production and redistribution of heat from the central to peripheral compartments. 1 During laparotomy, heat loss may also be increased because the large surface of the intestine is exposed to the relatively cold environment of the operating room. These findings are relevant because intraoperative hypothermia is associated with an increased rate of postoperative myocardial ischaemia, 2 increased plasma concentrations of catecholamines 3 and an increased rate of postoperative wound infection. 4 Moreover, in patients undergoing total hip replacement, blood loss and transfusion requirements were reduced when central body temperature was kept normal. 5 During hip arthroplasty, intraoperative hypothermia was prevented by preoperative warming with an electric warming blanket for 90 min. 6 Heat loss in the central body compartment after induction of general anaesthesia may be minimized by active warming of the arms and legs for 30 min. 7 Conflicting results on the effect of central body temperature on duration of stay in the post-anaesthesia care unit (PACU) have been reported. 8 9 As these findings imply that hypothermia may be associated with increased perioperative costs, we investigated if warming for 30 min before induction of anaesthesia combined with intraoperative warming with forced air prevents hypothermia during major abdominal surgery. Additionally, intraoperative blood loss, time to discharge from the PACU and total costs of perioperative management were compared with a control group who received conventional treatment of hypothermia, consisting of fluid warming devices, circulating water mattresses, cotton blankets and postoperative radiant warming. Patients and methods After obtaining approval from the local Ethics Committee and written informed consent, we studied 40 patients undergoing major abdominal surgery, allocated randomly to receive either passive isolation against hypothermia with fluid warming devices, circulating water mattresses, two layers of blankets and postoperative radiant warming, or additional active warming of the upper body using forced air, beginning 30 min before induction of general anaesthesia and during anaesthesia. During a 1-yr period from November 1995 until October 1996, ASA I III patients undergoing major abdominal surgery for cancer or inflammatory bowel disease were included. In one additional patient, a gastrojejunostomy was performed because of annular pancreas (table 1). All surgical procedures were performed via a median laparotomy, and surgery lasted for more than 2 h. Patients undergoing surgery of the biliary tract, liver, pancreatic surgery or gastrectomy were excluded, as were those who underwent preoperative radiotherapy or suffered from recurrent rectal cancer because of the greater risk of blood loss. No patient showed signs of bowel obstruction or acute onset of inflammatory bowel disease. Patients did not suffer from diabetes mellitus, thyroid disease, autonomic dysfunction, Cushing s syndrome, peripheral M. BOCK, MD, J. MÜLLER, A. BACH, MD, H. BÖHRER, MD, E. MARTIN, MD, FANZCA, J. MOTSCH, MD, Department of Anaesthesiology, University of Heidelberg, Im Neuenheimer Feld 110, D Heidelberg, Germany. Accepted for publication: September 17, Correspondence to M. B. Presented in part at the annual meeting of the American Society of Anaesthesiologists, New Orleans, LA, USA, October 19 23, 1996, and at the annual meeting of the German Society of Anaesthesiologists, Hamburg, Germany, April 23 26, 1997.
2 160 British Journal of Anaesthesia Table 1 Surgical procedures in the two groups (n) vascular disease, platelet dysfunction or any other coagulopathy, or had a history of smoking more than 20 cigarettes per day for more than 5 yr or treatment for arterial hypertension. Patients weights were within 20% of ideal body weight. On arrival in the preoperative holding area, tympanic membrane temperature was measured, and patients with a temperature greater than 37.5 C or less than 36.5 C were excluded. STUDY DESIGN Active warming Major colonic resection Intestinal resection 1 2 Pancreaticojejunostomy 1 0 Duodenojejunostomy 1 0 Retroperitoneal lymphadenectomy 0 1 Control Passive prevention of hypothermia included covering the patient s body and extremities with two layers of blankets and the use of circulating water mattresses (temperature set at 39 C) and fluid warming devices for infusions. Active warming was provided by the Warm Touch System (Mallinckrodt Medical GmbH. Hennef/Sieg, Germany) set at the highest flow rate and in the temperature range C. When the system switched to the ran ge below this, the temperature of the forced air was reset to C. Ambient room temperature was maintained at approximately 22 C during the preinduction period and during anaesthesia. On the evening before surgery, the bowel was cleaned with a standard electrolyte solution. Ten minutes before arrival in the holding area, patients were premedicated with midazolam 7.5 mg orally. After haemodynamic and temperature monitoring had been established, the patient s abdomen and legs were covered with two layers of blankets. The arms and chest of those patients receiving preoperative active warming were warmed using forced air, whereas in the control group the arms and chest were covered with blankets. Thirty minutes later, general anaesthesia was induced with fentanyl 2 4 g kg 1, thiopental 3 5 mg kg 1 and atracurium 0.5 mg kg 1 or cisatracurium 0.1 mg kg 1. The trachea was intubated, a central venous catheter was inserted into the right internal jugular vein and a 14-gauge cannula into a forearm vein. Warmed fluids were provided via this cannula. The central venous catheter was used for drug administration and for measurement of central venous pressure. When considered necessary, a 20-gauge cannula for continuous measurement of arterial pressure was inserted into the right radial artery. Anaesthesia was maintained using fentanyl, isoflurane (maximum end-tidal volume 1.0 MAC) and nitrous oxide in 70% oxygen. Fresh gas flow was kept constant at 3.0 litre min 1. Fluid replacement was guided by central venous pressure. A core temperature of less than 35 C or significant peripheral vasoconstriction for more than 2 h were judged as criteria for maintaining mechanical ventilation after operation. The criteria for tracheal extubation after mechanical ventilation in the PACU were: central body temperature 36.0 C, absence of significant peripheral vasoconstriction, and heart rate and arterial pressure within 20% of baseline. In the postoperative phase, radiant heat was used to rewarm hypothermic patients. MEASUREMENTS Arterial pressure was measured non-invasively via oscillometry; in a few patients it was also recorded via invasive arterial pressure measurement when indicated medically. An electronic thermocouple (Mallinckrodt Medical) was used to record tympanic membrane temperature, which was considered to reflect central body temperature. 1 Skin surface temperatures were measured by thermocouples placed on the middle of the forearm and on the tip of the ipsilateral middle finger; temperatures were obtained from the arm that was not used for indirect measurement of arterial pressure. A difference between central and peripheral temperatures of more than 4.0 C was considered to indicate significant vasoconstriction. During operation, arterial pressure, heart rate, peripheral oxygen saturation, end-tidal carbon dioxide and isoflurane concentrations, and temperatures were recorded at 15-min intervals. Central venous pressure was measured at 30-min intervals, and central venous or, if available, arterial blood-gas samples were obtained every 60 min. After operation, temperatures were recorded at 30-min intervals until normothermia was achieved. On arrival in the PACU, blood samples were obtained for coagulation tests and blood cell counts, and to measure blood-gas tensions and serum concentrations of electrolytes, creatinine and urea. Laboratory analyses were repeated on the first postoperative day. For analysis of arterial and central venous blood-gas tensions, the Ciba Corning 288 (Ciba Corning, Fernwald, Germany) analyser was used. Blood loss was estimated by an independent anaesthetist who was not involved in the study and who took care of the patient during surgery or during the patient s stay in the PACU. The indication for transfusion of packed red blood cells or fresh frozen plasma was confirmed by the supervising staff anaesthetist who was also not involved in the study. Thus the anaesthetist in charge was always blinded to the fact that the amount of blood loss and transfusion requirements were part of the study. The presence or absence of shivering was determined by the anaesthetist in charge on the PACU who was blinded to the study. The time the patient was ready to be discharged from the PACU was set retrospectively by a blinded observer using a modified Aldrete and Kroulik scoring system. 10 A core temperature of more than 36.0 C and a score of 14 points was regarded as a criterion for discharge (table 2). The costs of perioperative management were calculated with respect to the salaries and costs for the University of Heidelberg in August For preinduction warming, we calculated a nurse-to-patient ratio of 1:4; the costs of treating patients in the control group before induction of anaesthesia were calculated as zero. During surgery, an anaesthetist-topatient ratio of 1:1 and a nurse-to-patient ratio of 1:2 was assumed. In the PACU, the anaesthetist-topatient ratio was 1:10. For a patient who required
3 Preinduction warming and perioperative costs 161 Table 2 Criteria for discharge from the PACU. Central body temperature had to be 36.0 ºC Table 3 Patients characteristics (mean (SD or range) or number). No difference between groups Zero point One point Two points Active warming Control Activity mechanical ventilation in the PACU, the nurse-topatient ratio was regarded as 1:1, and for a spontaneously breathing patient in the PACU it was 1:3. A cost of sterling was calculated for an upper body layer of the Warm Touch System. No costs were calculated for the ventilator of the Warm Touch System or for the radiant heater because both devices were already present in our department. The costs calculated for each patient comprised the salaries for the anaesthetist and nurse assisting him and the costs for anaesthetic equipment, such as venous catheters, tracheal tubes, etc., and the medical treatment the patient required. The costs for isoflurane and nitrous oxide were not taken into consideration. An exchange rate of 1 to 2.8 German marks was assumed for calculation of perioperative costs. DATA ANALYSIS No No purposeful Respiration Apnoea Dyspnoea or limited breathing Correct on command Breathes deeply and coughs freely Supplementary O 2 require to maintain Sa O2 95% (litre min 1 ) Consciousness Unresponsive Arouses to verbal stimuli Fully awake Arterial pressure AP or 50% of baseline AP 20 50% of baseline Heart rate (beat min 1 ) 45 or or Gastrointestinal tract Severe Vomiting vomiting within 30 min Renal function Anuria 0.3 ml kg 1 h 1 AP 0 20% of baseline Little or no vomiting 0.3 ml kg 1 h 1 Time-dependent results were evaluated using analysis of variance (ANOVA) with post hoc Scheffé s test. Results in the two treatment groups were compared by unpaired, two-tailed t tests. Data are presented as mean (SD). Results Patient characteristics, duration of surgery and anaesthetics are given in table 3. There were no differences between the two groups. After operation, one patient in the control group was admitted to the intensive care unit which was not controlled by anaesthetists. Consequently, this patient was excluded from the calculation of perioperative costs. Duration of stay in the PACU was not investigated in this patient because postoperative treatment differed. Before operation, central body temperature was not altered in either group. During the preinduction period, central body temperature did not differ significantly between the two groups. From 15 min after intubation until the end of surgery, central body temperature was higher in the prewarmed group (P 0.01) (fig. 1) and remained higher in this group for Age (yr) 43 (19 67) 49 (27 78) Sex (F/M) 10/10 9/11 Height (m) 175 (8) 171 (9) Weight (kg) 73 (14) 72 (14) Length of operation (h) 4.0 (1.4) 4.3 (1.6) Fentanyl (µg kg 1 h 1 ) 3.0 (1.0) 3.6 (1.5) Intraoperative volume (ml) 982 (220) 1224 (273) the first 180 min after arrival in the PACU (P 0.01 for the first 150 min; P 0.05 for the following 30 min). At the end of surgery and during the first hour in the PACU, 12 of 20 patients in the control group showed significant vasoconstriction, whereas vasoconstriction was detectable in only four of 20 patients in the prewarmed group. On arrival in the PACU, seven patients in the control group showed signs of shivering, and six additional patients underwent mechanical ventilation because of hypothermia. In all patients in the prewarmed group, the trachea was extubated immediately after surgery. Only one of these patients showed shivering in the PACU. Intraoperative blood loss was significantly less in the normothermic group (635 (507) ml vs 1070 (803) ml; P 0.05). During operation, only one patient in the study group received 2 u. of packed red blood cells, while in the control group six patients received blood products. By the morning after surgery, packed red blood cells or fresh frozen plasma had been transfused to two study patients. However, a total of nine control patients received blood products by the first morning after surgery. Expressed in millilitre per patient, warmed patients were given significantly less blood products on admission to the PACU (30 (134) vs 240 (432) ml/patient; P 0.05) and by the morning after surgery (95 (293) vs 502 (754) ml/patient; P 0.05). There were no significant differences in partial thromboplastin time or prothrombin time between the two groups before operation, on admission to the PACU, 12 h after operation or on the first morning after operation. Among warmed patients, platelet count decreased from ( ) l 1 before operation and ( ) l 1 on admission to the PACU, to (60 000) l 1 12 h after the end of operation. On the morning after surgery, platelet count was (80 000) l 1 in this group. In the control group, platelet count decreased from (83 000) l 1 before operation and (77 000) l 1 to (54 000) l 1 12 h after operation and remained constant for the next 24 h. There was no significant difference between the two groups, but using a two-tailed paired t test, the level of significance was at 12 h after operation. After 94 (42) min, prewarmed patients were ready to be discharged from the PACU compared with 217 (169) min in the control group (P 0.01). The costs of treatment from the beginning of preinduction warming until discharge from the PACU were 408 (105) for an aggressively warmed patient and 534 (250) for a control patient (P 0.05). Thus 24% of the costs for the anaesthetic treatment could be saved by maintenance of normothermia. Preinduction warming contributed to 4.6 (1.3)% of total costs.
4 162 British Journal of Anaesthesia Figure 1 Change in core temperature ( T) during the preinduction period and anaesthesia in the actively warmed patients (closed symbols) and in the control group (open symbols). Time zero induction of general anaesthesia. Data are mean (SD) (95% confidence interval of the mean). P 0.01 from 0.25 h to 6 h. Discussion Our major finding was that maintenance of normothermia reduced the total costs for anaesthetic treatment overall by 24% during major abdominal surgery. This was mainly because of a shortened stay in the PACU, lower incidence of postoperative mechanical ventilation and reduced perioperative blood loss, resulting in fewer transfusion requirements in normothermic patients. Our study supports the findings of Schmied and co-workers on the effect of core hypothermia on blood loss and transfusion requirements during total hip arthroplasty. 5 We did not include such a detailed haemodilution and transfusion procedure in our study because much greater blood loss could be tolerated by the younger patients undergoing abdominal surgery. In common with Schmied and co-workers, 5 we did not observe any significant differences in prothrombin time or partial thromboplastin time between groups. One reason might be that the laboratory investigations were performed at a temperature of 37 C and a reversible coagulopathy induced by hypothermia may only be detected when clotting tests are performed at the patient s central body temperature In contrast, some patients received fresh frozen plasma in the control group when coagulopathy was observed clinically, whereas among normothermic patients fresh frozen plasma was not given. The assumption that hypothermia produces a clinically relevant coagulopathy in abdominal surgery was also supported by a study on the effects of hypothermia on wound infection and duration of hospitalization. 4 Normothermic patients required less units of red blood cells than hypothermic patients during abdominal surgery lasting approximately 3 h. In our study, normothermic patients had a slight, but non-significant, higher platelet count 12 h after operation. This may also indicate a clinically relevant coagulopathy in hypothermic patients. In contrast with our results, hypothermia was found to have no influence on duration of stay in the PACU in children recoverin g from peripheral surgery. 8 However, that study was not randomized, and patients and surgical procedures were different from those in our investigation. The assumption that hypothermia prolongs post-anaesthetic recovery has recently been proved in a randomized prospective study 9 which concurs with our data. Unlike this investigation, we included central body temperature as a criterion for fitness of discharge from the PACU, as some hypothermic patients underwent mechanical ventilation when they arrived in the PACU. In contrast with the study by Lehnhardt and co-workers 9 however, hypothermia was treated after operation in our study. The estimated salary of a nurse taking care of four patients in a holding area and the cost of the upper body layer of the Warm Touch System contributed to 4.6% of the costs of total perioperative anaesthetic management during the day of surgery. Considering all aspects, however, the costs of perioperative management for each patient were significantly lower in the actively prewarmed group. Thus the 24% reduction in costs for each actively warmed patient represents a large economic impact compared with the cost of approximately 1430 for the ventilator of an active warming device. These results are new and have been confirmed only indirectly by other studies which showed complications associated with mild intra- and postoperative hypothermia. Thus maintenance of intraoperative normothermia has been shown indirectly to reduce the costs of hospitalization by incurring a lower incidence of wound infections and a shorter duration of hospitalization, 4 shorter duration in the PACU, 9 reduced transfusion requirements 5 and a lower incidence of myocardial ischemia. 2 However, our study is the first to have directly evaluated the economic benefits of intraoperative normothermia for an anaesthesia department. The reduction in total costs of perioperative treatment of a single patient produces an economic benefit for the anaesthetic department as a whole. This would appear to contradict a computer simulation of economic management of a PACU. 13 In that investigation, the distribution of admissions to PACU was found to be the major determinant of PACU costs, whereas the time of fitness for discharge was of minor economic importance. However, in that analysis, the PACU investigated was solely for outpatients, that is those recovering from minor surgery. Our PACU, however, provides care for patients recovering from outpatient surgery and for those undergoing major abdominal or vascular surgery, including the opportunity for postoperative mechanical ventilation. The costs of supplies, medication and transfusion products are thus likely to be higher in our PACU than in a PACU for outpatient surgery where they contribute to only 2% of the total costs. 13 However, the costs of loans are also the major determinant for the total costs in our PACU. Contrary to the study by Dexter and Tinker, 13 our PACU is also used during the night. Thus a smaller number of nurses have to work during the night shift in our PACU if patients are ready to be discharged earlier in the evening after maintenance of perioperative normothermia. In summary, maintenance of intraoperative core normothermia helped to reduce blood loss and save blood products. Furthermore, duration of stay in the PACU was prolonged in hypothermic patients, even though hypothermia was treated in the PACU. The costs of perioperative management were lower in
5 Preinduction warming and perioperative costs 163 patients who remained normothermic although they had to be supervised during the preinduction warming period. Acknowledgements We thank H. Bauer, MD, for statistical analysis and D. I. Sessler, MD, University of California, San Francisco, USA, for critical review of the manuscript. This work was part of the MD thesis of J. Müller, University of Heidelberg, Faculty of Medicine. References 1. Sessler DI. Perianesthetic thermoregulation and heat balance in humans. FASEB Journal 1993; 7: Frank SM, Beattie C, Christopherson R, Norris EJ, Perler BA, Williams GM, Gottlieb SO. Unintentional hypothermia is associated with postoperative myocardial ischemia. Anesthesiology 1993; 78: Frank SM, Higgins MS, Breslow MJ, Fleisher LA, Gorman RB, Sitzmann JV, Raff H, Beattie C. The catecholamine, cortisol, and hemodynamic responses to mild perioperative hypothermia. Anesthesiology 1995; 82: Kurz A, Sessler DI, Lenhardt R. Perioperative normothermia to reduce the incidence of surgical wound infection and shorten hospitalisation. New England Journal of Medicine 1996; 334: 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: Just B, Trévíen V, Delva E, Lienhart A. Prevention of intraoperative hypothermia by preoperative skin-surface warming. Anesthesiology 1993; 79: Sessler DI, Schroeder M, Merrifield B, Matsukawa T, Cheng C. Optimal duration and temperature of prewarming. Anesthesiology 1995; 82: Bissonnette B, Sessler DI. Mild hypothermia does not impair postanesthetic recovery in infants and children. Anesthesia and Analgesia 1993; 76: Lehnhardt R, Kurz A, Marker E, Sessler DI, Narzt E, Lackner F. Intraoperative hypothermia prolongs duration of postoperative recovery. Anesthesiology 1995; 83: A Aldrete JA, Kroulik D. A postanesthetic recovery score. Anesthesia and Analgesia 1970; 49: Rohrer MJ, Natale AM. Effect of hypothermia on the coagulation cascade. Critical Care Medicine 1992; 20: Valeri CR, Cassidy G, Khuri S, Feingold H, Ragno G, Altschule MD. Hypothermia-induced reversible platelet dysfunction. Annals of Surgery 1987; 205: Dexter F, Tinker JH. Analysis of strategies to decrease postanaesthesia care unit costs. Anesthesiology 1995; 82: