The Impact of Perioperative Hypothermia on Plastic Surgery Outcomes: A Multivariate Logistic Regression of 1062 Cases

Research The Impact of Perioperative Hypothermia on Plastic Surgery Outcomes: A Multivariate Logistic Regression of 1062 Cases Ryan S. Constantine, BA; Matthew Kenkel, BA; Rachel E. Hein, BS; Roberto Cortez, BS; Kendall Anigian, BS; Kathryn E. Davis, PhD; and Jeffrey M. Kenkel, MD, FACS Abstract Background: Perioperative hypothermia has been associated with increased rates of infection, prolonged recovery time, and coagulopathy. Objectives: The authors assessed the impact of hypothermia on patient outcomes after plastic surgery and analyzed the impact of prewarming on postoperative outcomes. Methods: The medical charts of 1062 patients who underwent complex plastic surgery typically lasting at least 1 hour were reviewed. Hypothermia was defined as a temperature at or below 36 C. Postoperative complication data were collected for outcomes including infection, delayed wound healing, seroma, hematoma, dehiscence, deep venous thrombosis, and overall wound problems. Odds ratios (ORs) were estimated from 3 multivariate logistic regression models of hypothermia and one model of body contouring procedures that included prewarming as a parameter. Results: Perioperative hypothermia was not a significant predictor of wound problems (OR = 0.83; P =.28). In the stratified regression model, hypothermia did not significantly impact wound problems. The regression model measuring the interaction between hypothermia and operating time did not show a significantly increased risk of wound problems. Prewarming did not significantly affect perioperative hypothermia (P =.510), and in the model of body contouring procedures with prewarming as a categorical variable, massive weight loss was the most significant predictor of wound complications (OR = 2.57; P =.003). Prewarming did not significantly affect outcomes (OR = 1.49; P =.212). Conclusions: Based on univariate and multivariate models in our study, mild perioperative hypothermia appears to be independent of wound complications. Level of Evidence: 4 Accepted for publication May 22, 2014. All surgical patients have the potential to develop perioperative hypothermia, which has considerable implications for both patients and surgeons. It has been reported that temperatures falling below 36 C can lead to complications and negative outcomes such as infection, prolonged recovery time, and coagulopathy. 1-6 In plastic surgery, the potential for perioperative hypothermia is a cause for concern due to the effects of general anesthesia on human thermoregulatory thresholds (eg, vasoconstriction). In addition, a large body surface area is often exposed in plastic and reconstructive Aesthetic Surgery Journal 2015, Vol 35(1) 81 88 2015 The American Society for Aesthetic Plastic Surgery, Inc. Reprints and permission: journals.permissions@oup.com DOI: 10.1093/asj/sju022 www.aestheticsurgeryjournal.com Risk procedures, underscoring the need to pay particular attention to maintaining optimal internal body temperature. From the Department of Plastic and Reconstructive Surgery, University of Texas Southwestern Medical Center, Dallas, Texas. Corresponding Author: Dr Jeffrey M. Kenkel, Department of Plastic Surgery, University of Texas Southwestern Medical Center, 1801 Inwood Rd, Dallas, TX 75390-9132, USA. E-mail: jeffrey.kenkel@utsouthwestern.edu

82 Aesthetic Surgery Journal 35(1) Passive warming, which includes blankets and other insulating materials, and active warming, which includes forced air and warmed fluids, are aimed at maintaining normothermia during surgery and avoiding the side effects of perioperative hypothermia. 7-9 Much of the current literature, particularly in the field of plastic and reconstructive surgery, has focused on educating clinicians on the perils of not maintaining normothermia and on improving methods for maintaining normothermia, rather than on evaluating the efficacy of current techniques on patient outcomes. 10-12 In this study, we evaluated the prevalence and effect of perioperative hypothermia on patient outcomes in a large sample size of patients treated in an academic medical setting. We also analyzed the impact of prewarming on the development of perioperative hypothermia and on postoperative outcomes in the body contouring population. METHODS At our academic medical center, we conducted a retrospective chart review of 1801 complex plastic surgery procedures occurring over a 4-year period from January 1, 2008, to January 31, 2012, at 4 separate surgical facilities. Data from patient charts and hospital records were extracted and compiled in a database. A broad range of complex procedures and some combined procedures were included, characterized by long operating times of typically at least one hour. Aesthetic procedures such as abdominoplasty, brachioplasty, thighplasty, breast mastopexy/reduction, lower body lift, circumferential body lift, buttock lift, facelift, and rhinoplasty, as well as reconstructive procedures such as breast reconstruction and various free-flap procedures were included in the analysis. All patients underwent general anesthesisa. Preoperative patient characteristics such as gender, age, and body mass index (BMI) and common comorbidities associated with postoperative complications were recorded. 13,14 Patient perioperative temperature was also included in the database and subsequent analyses. Patients that did not have perioperative temperature data were excluded from the analysis. Hypothermia was indicated if perioperative temperature decreased to below 36 C. Outcomes were measured first with univariate analysis and then with multivariate logistic regression for relevant patient demographics such as age, gender, and BMI as well as common comorbidities. Outcomes were defined as surgical site infection, dehiscence, erythema, necrosis, seroma, hematoma, delayed wound healing, and composite wound problems, which included all of these outcomes. These complications were then extracted from the chart review and included in the database. Patient demographics were summarized as means with standard deviations (SDs) for continuous variables and as percentages for categorical variables. We compared patient characteristics and comorbidities with univariate analyses, and relevant factors were selected for further analysis. Odds ratios (ORs) were estimated using 3 logistic regression models: (1) a base model including hypothermia status, (2) a model adjusting for stratified perioperative hypothermia, and (3) a model measuring the interaction between duration of surgery and hypothermia. The base model analysis included 3 separate regression models that characterized procedure type as body contouring procedures, general flap-based procedures, or reconstructive breast procedures. These 3 regression models all evaluated the common endpoint of composite wound problem. Massive weight loss was included as a comorbidity in the body contouring procedure analysis, because it has been shown to be a significant predictor of wound complications. 15 This first regression analysis was continued as an aggregate of all 3 proceduretype cohorts for all subsequent analyses. The second logistic regression model stratified hypothermic patients in 3 increments of 0.5 C, with the third stratification including all patients whose body temperatures decreased to below 35 C. Operating time was stratified in 2-hour increments (0 to 2hours,2to4hours,4to6hours,6to8hours,and8+ hours). A separate analysis was conducted to determine the impact of prewarming on developing hypothermia and patient outcomes in the body contouring population. Warming status was included as an endpoint in the overall database. Univariate analysis and multivariate logistic regression were conducted on all patients whose primary surgery included abdominoplasty, panniculectomy, suction-assisted lipectomy, or any other body contouring procedure. Statistical significance was defined as P <.05, and all tests were 2-sided. All data analyses and graphics were conducted and generated with Stata/SE Version 12.0 (StataCorp, Inc, College Station, TX) and GraphPad Prism, version 6.00 for Mac (GraphPad Software, La Jolla, CA) statistical software. 16 RESULTS Patient demographics and complication rates are summarized in Table 1. One thousand sixty-two patients met inclusion criteria for analysis. Of these, 873 were women and 189 were men, with a mean (± SD) age of 50.4 ± 13.4 years (range, 15-85 years), and BMI of 27.8 ± 7.1 kg/m 2.The overall complication rate was 25.7%, with a mean follow-up time of 6 months. The majority of patients were hypothermic: 820 of 1062 (77.2%) had perioperative body temperatures at or below the 36 C hypothermia threshold (Figure 1). Hypothermic patients had a mean (± SD) body temperature of 35.6 C ± 0.71 C (range, 32.6 C to 39.7 C). All patients received perioperative warming, and there were similar rates of hypothermia at all 4 facilities (Table 2). Mean operating time was 4.4 ± 3.0 hours (range, 0.62 18.32 hours), reflecting the complexity of the operations typically included in the analysis (Table 3). In the

Constantine et al 83 univariate analysis, there was no significant difference in outcomes between hypothermic and normothermic patients for any patient characteristic other than gender, hypertension status, or cancer status (Table 4). Multivariate Table 1. Patient Demographics (N = 1062) Characteristic N % Male gender 189 17.8 Age >45 y (mean ± SD, 50.4 ± 13.4 y) 720 67.8 BMI 30 kg/m 2 (mean ± SD, 27.8 ± 7.1) 310 29.5 Smoker 111 10.4 Diabetes mellitus 125 11.8 Hypertension 345 32.5 CAD 26 2.4 MI 20 1.9 Other cardiac disease 229 21.6 COPD 30 2.8 Other pulmonary disease 215 20.2 Renal diseases 92 8.7 Cancer 307 28.9 HIV/AIDS 6 0.6 DVT 17 1.6 Hypothermic body temperature (mean ± SD, 35.6 C ± 7.1) 820 77.2 Wound problem a 273 25.7 Infection 77 7.6 Dehiscence 54 5.1 Erythema 74 7.0 Necrosis 22 2.1 Seroma 75 7.1 Hematoma 16 1.5 Delayed wound healing 38 3.6 CAD, coronary artery disease; COPD, chronic obstructive pulmonary disease; DVT, previous history of deep vein thrombosis; HIV/AIDS, human immunodeficiency virus/acquired immunodeficiency syndrome; MI, previous myocardial infarction; SD, standard deviation. a Wound problem includes any type of complication (infection, dehiscence, erythema, necrosis, seroma, hematoma, and delayed wound healing). logistic regression analyses by procedure type indicated that hypothermia status was not a statistically significant indicator of composite wound problems (body contouring procedures: OR, 1.80; P=.077; flap-based procedures: OR, 0.69; P=.148; and reconstructive breast procedures: OR, 0.77; P=.451; Tables 5-7). In the aggregate base model, hypothermia was not a statistically significant predictor of composite wound problems (OR, 0.83; P =.277; Table 8). In the stratified model based on body temperature, all 3 temperature categories (36.0 C-35.5 C, 35.5 C-35.0 C, and <35.0 C) were not statistically significant indicators of wound problems. In the interaction model between surgery time and level of hypothermia, none of the interaction terms were statistically significant predictors for composite wound complications. In the investigation of the impact of prewarming on body contouring outcomes, 343 body contouring procedures were included in the univariate and multivariate regression analyses. The Fisher exact test for significance indicated that prewarming did not significantly affect Figure 1. (A) Percentage of patients in the overall patient cohort (N=1062) who experienced perioperative hypothermia (n=820) and who were normothermic (n=242). (B) Patient cohort stratified by perioperative temperature: N=340 (for 35.5 C < X < 36.0 C); N=317 (for 35.0 C < x 35.5 C); N=163 for x 35.0 C Table 2. Perioperative Hypothermia by Surgical Facility Cases of Perioperative Hypothermia Facility 1 (n = 540) Facility 2 (n = 40) Facility 3 (n = 383) Facility 4 (n = 99) Total (N = 1062) n (%) 375 (69.6) 31 (77.5) 329 (85.9) 84 (84.8) 1062 Hypothermia was defined as a minimum temperature of 36.0 C.

84 Aesthetic Surgery Journal 35(1) Table 3. Type and Frequency of Procedures, N = 1062 Primary Procedure Type n % Myocutaneous flap Head or neck 18 1.7 Trunk 224 21.0 Upper extremity 13 1.2 Lower extremity 30 2.8 Free muscle flap with microvascular anastomosis 24 2.6 Free skin flap with microvascular anastomosis 11 1.0 Free facial flap with microvascular anastomosis 3 0.3 SMAS flap 23 2.2 Excision Excessive skin and subcutaneous tissue (abdomen, infraumbilical); panniculectomy 95 8.9 Thigh 14 1.3 Leg 2 0.2 Hip 11 1.0 Buttock 2 0.2 Arm 24 2.3 Submental fat pad 1 0.1 Other areas 13 1.2 Excessive skin and subcutaneous tissue of abdomen 21 2.0 Suction-assisted lipectomy Trunk 72 6.8 Upper extremity 34 3.2 Lower extremity 58 5.5 Delayed insertion of breast prosthesis after mastopexy, mastectomy 121 11.4 Breast reconstruction Immediate or delayed, with tissue expander 110 10.3 Latissimus dorsi flap, without prosthetic implant 48 4.5 Other technique 58 5.5 TRAM flap 26 2.4 SMAS, superficial musculoaponeurotic system; TRAM, transverse rectus abdominis myocutaneous. Table 4. Univariate Analysis (N = 1062) Variable Percentage Normothermic (n = 242) Percentage Hypothermic a (n = 820) P Value Male gender 22.7% 16.3%.023 Age >45 y 64.1% 68.9%.156 BMI 30 kg/m 2 29.8% 29.3%.875 Smoker 8.3% 11.0%.224 Diabetes mellitus 12.8% 11.5%.568 Hypertension 38.0% 30.9%.037 CAD 2.9% 2.3%.611 MI 3.3% 1.5%.064 Other cardiac disease 19.4% 22.2%.357 COPD 3.7% 2.6%.339 Other pulmonary disease 20.2% 20.2%.999 Renal disease 8.3% 8.8%.802 Cancer 19.8% 31.6% <.000 HIV/AIDS 0.5% 0.8%.537 DVT 1.2% 1.7%.611 Wound problem b 28.5% 24.9%.256 Infection 9.1% 6.7%.209 Dehiscence 5.0% 5.1%.919 Erythema 6.6% 7.1%.805 Necrosis 1.7% 2.2%.603 Seroma 7.0% 7.1%.979 Hematoma 1.2% 1.6%.698 Delayed wound healing 5.8% 2.9%.035 BMI, body mass index; CAD, coronary artery disease; COPD, chronic obstructive pulmonary disease; DVT, previous history of deep vein thrombosis; HIV/AIDS, human immunodeficiency virus/acquired immunodeficiency syndrome; MI, previous myocardial infarction. a Hypothermia was defined as a minimum temperature of 36.0 C. b Wound problem includes any type of complication (infection, dehiscence, erythema, necrosis, seroma, hematoma, and delayed wound healing). complications (OR, 2.57; P=.003), whereas prewarming did not significantly impact outcomes (OR, 1.49; P =.212). perioperative hypothermia, as 73% of patients with prewarming became hypothermic and 69% without prewarming became hypothermic (P =.510). Multivariate logistic regression (Table 9) indicated that massive weight loss status was the most significant indicator for patient DISCUSSION Human thermoregulation is a complex and variable process that is known to be disrupted by modern anesthetic procedures and techniques. 7,17 The thermoregulatory

Constantine et al 85 Table 5. Multivariate Logistic Regression: Hypothermia Status in Body Contouring Procedures (N = 343) Male gender 0.55 0.19-1.61.273 Age 45 y 1.38 0.74-2.60.312 BMI 30 kg/m 2 1.18 0.60-2.33.625 Diabetes mellitus 0.48 0.13-1.80.274 Smoking 0.10 0.01-0.82.032 Hypertension 1.95 0.97-3.94.061 Cardiac disease other than MI 0.32 0.11-0.93.036 Pulmonary disease other than COPD 0.31 0.09-1.03.056 Renal disease 2.19 0.48-10.1.315 Cancer 0.95 0.29-3.09.933 MWL 3.58 1.92-6.67 >.000 Hypothermia 1.80 0.94-3.45.077 CI, confidence interval; COPD, chronic obstructive pulmonary disease; MI, previous myocardial infarction; MWL, massive weight loss; OR, odds ratio. Table 6. Multivariate Logistic Regression: Hypothermia Status in Flap-Based Procedures (N = 344) Male gender 1.06 0.66-1.72.791 Age 45 y 1.17 0.66-2.08.579 BMI 30 kg/m 2 0.94 0.58-1.54.817 Diabetes mellitus 1.67 0.88-3.16.116 Smoking 1.88 1.03-3.43.039 Hypertension 0.79 0.45-1.36.389 CAD 1.70 0.53-5.46.372 MI 3.51 0.97-12.7.056 Other cardiac disease 0.69 0.37-1.26.227 COPD 0.55 0.17-1.77.317 Other pulmonary disease 1.03 0.58-1.84.922 Renal disease 1.50 0.79-2.84.218 Cancer 1.19 0.64-2.22.582 DVT 0.63 0.14-2.81.541 Hypothermia 0.69 0.42-1.14.148 CAD, coronary artery disease; CI, confidence interval; COPD, chronic obstructive pulmonary disease; DVT, previous history of deep vein thrombosis; MI, previous myocardial infarction; OR, odds ratio. Table 7. Multivariate Logistic Regression: Hypothermia Status in Reconstructive Breast Procedures (N = 375) Age 45 y 1.53 0.88-1.71.151 BMI 30 kg/m 2 2.12 0.94-1.75.010 Diabetes mellitus 0.35 0.71-1.81.077 Smoking 1.46 0.83-2.02.403 Hypertension 0.81 0.78-1.53.488 CAD 4.99 0.76-5.67.297 MI 0.41 0.35-3.29.679 Other cardiac disease 1.02 0.52-1.11.950 COPD 0.73 0.30-1.83.696 Other pulmonary disease 1.64 0.70-1.48.092 Renal disease 1.23 0.78-2.06.674 Cancer 0.98 0.86-1.63.919 DVT 4.27 0.24-2.40.250 Hypothermia 0.77 0.60-1.18.451 CAD, coronary artery disease; CI, confidence interval; COPD, chronic obstructive pulmonary disease; DVT, previous history of deep vein thrombosis; MI, previous myocardial infarction; OR, odds ratio. interthreshold range is typically 0.2 C; however, when general anesthesia is induced, this range goes up to 4.0 C. This means that there is no thermoregulatory response until the patient s temperature is 4.0 C (higher or lower) different than normal core body temperature as opposed to a 0.2 C change inducing a thermoregulatory response. 2,18,19 Patients are susceptible to 3 phases of hypothermia during anesthesia, including a body temperature drop of 1.0 C to 1.5 C during the first 60 minutes, a linear decline in temperature from 1 to 3 hours postinduction of anesthesia, and a plateau in core temperature from 2 to 4 hours of anesthesia. 18 However, these 3 phases can be counteracted with preoperative warming, forced air, warmed fluids, and/or passive warming through insulation. 20 Despite these warming techniques, perioperative hypothermia has been reported in the literature at rates from ranging from 50% to 90% of all patients undergoing major and minor surgical procedures with general anesthesia. 21-24 Our study population fell within this range, with 77.2% of patients having body temperatures at or below the 36 C threshold during the operating period. This rate of hypothermia occurred despite the fact that all patients received active and passive warming perioperatively. Mild perioperative hypothermia has been associated with various complications, including wound infection, morbid

86 Aesthetic Surgery Journal 35(1) Table 8. Multivariate Logistic Regression: Hypothermia Status (N = 1062) Sex (male) 1.17 0.80-1.70.426 Age 45 y 1.23 0.88-1.71.218 BMI 30 y 1.28 0.94-1.75.114 Diabetes mellitus 1.13 0.71-1.81.596 Smoking 1.29 0.83-2.02.260 Hypertension 1.09 0.78-1.53.609 CAD 2.08 0.76-5.67.152 MI 1.07 0.35-3.29.900 Other cardiac disease 0.76 0.52-1.11.151 COPD 0.74 0.30-1.83.515 Other pulmonary disease 1.02 0.70-1.48.937 Renal disease 1.27 0.78-2.06.344 Cancer 1.18 0.86-1.63.299 HIV/AIDS 2.60 0.50-13.60.258 DVT 0.75 0.24-2.40.634 Hypothermia 0.83 0.60-1.18.277 CAD, coronary artery disease; CI, confidence interval; COPD, chronic obstructive pulmonary disease; DVT, previous history of deep vein thrombosis; HIV/AIDS, human immunodeficiency virus/acquired immunodeficiency syndrome; MI, previous myocardial infarction; OR, odds ratio. cardiac events, perioperative blood loss, and longer duration of hospitalization. 1,25-28 Among the most commonly cited articles evaluating perioperative hypothermia and its effects on wound infection rates is a randomized trial by Kurz et al 1 in which a standard warming procedure was followed in half of the patients and not in the other half. This trial has served as the foundation for several subsequent studies and patient safety continuing medical education modules. However, the study design itself prompted multiple researchers and clinicians to cite significant study limitations, such as ignoring best clinical practices in 50% of patients by avoiding active warming, relatively small sample size (n = 200) prone to statistical fluctuation, and infection rates that were artificially inflated by failing to use standard temperature-regulating techniques. 29-33 Furthermore, the authors of this study investigated outcomes of colorectal surgery, which is likely susceptible to a unique bacterial fauna when compared with the typical bacteria encountered in plastic or reconstructive surgery. Despite the limitations of the study by Kurz et al, 1 our results are in agreement with their conclusion that routine perioperative warming, by both passive and active methods, appears to mitigate complications from hypothermia. With Table 9. Multivariate Logistic Regression: Impact of Prewarming on Body Contouring Outcomes (N = 343) Male gender 0.78 0.30-2.02.608 Age 45 y 1.40 0.75-2.58.289 BMI 30 y 1.38 0.72-2.64.333 MWL 2.57 1.37-4.83.003 Diabetes mellitus 0.93 0.31-2.83.899 Smoker 0.11 0.01-0.86.035 Hypertension 1.91 0.97-3.74.061 Cardiac disease other than MI 0.46 0.18-1.18.104 Pulmonary disease other than COPD 0.28 0.09-0.96.042 Renal disease 1.62 0.37-7.20.525 Cancer 0.91 0.28-2.91.867 Prewarming 1.49 0.80-2.78.212 CAD, coronary artery disease; CI, confidence interval; COPD, chronic obstructive pulmonary disease; MI, myocardial infarction; MWL, massive weight loss; OR, odds ratio. both univariate and multivariate logistic regression models, we demonstrated that wound infection and wound complications are not significantly higher when patients become hypothermic perioperatively. This finding suggests that when patients are cared for under best practice standards for perioperative warming, hypothermia is independent of wound complications and infections. Prewarming before surgery has been proposed as a means of mitigating the effects of hypothermia induced by anesthesia and potentially as a means of preventing perioperative hypothermia from occurring in the first place. 34-36 In a small study population (n = 16), prewarming with an electric blanket for at least 90 minutes reduced postinduction hypothermia in patients who had receiving prewarming when compared with patients who did not. 37 However, to our knowledge, the impact of prewarming on individual patient outcomes has not been investigated. Our study indicates that prewarming appears to have no statistically significant effect on reducing the rate of hypothermia in plastic surgery, and that prewarming does not significantly impact patient outcomes. This finding is important given that patients who undergo plastic or reconstructive surgery often have a large body area exposed during the operation. Although prewarming should reduce hypothermia after induction of anesthesia, heat loss due to skin preparation, heat evaporation from the operating field, and infusion of intravenous fluids can negate these effects. 37 As physician and hospital quality assurance measures are increasingly implemented for reimbursement purposes

Constantine et al 87 through the Physician Quality Reporting System and Hospital Quality Initiative, further studies should address the definition and impact of perioperative hypothermia. The Agency for Healthcare Research and Quality (AHRQ) and the National Quality Forum (NQF) indicate that a single body temperature measurement of >36.0 C within 30 minutes before or 15 minutes after anesthesia is an adequate marker of normothermia. 38 The results of our study, however, indicate that this threshold may need to be reconsidered, particularly in conjunction with results of the foundational study that showed infection risk from perioperative hypothermia in a patient population with a mean temperature of 34.7 C ± 0.6 C well below the accepted 36.0 C threshold. 1,38 In our study, perioperative blood loss was not included as an endpoint because the authors believed that perioperative notes regarding estimated blood loss were both imprecise and arbitrary in many cases. Clinical outcomes studies analyzing the effect of hypothermia in plastic surgery have been very limited; however, one study noted a lower minimum temperature associated with both seroma and transfusion in patients who underwent body contouring. 39 Potential limitations of this study may be best addressed with further research utilizing a prospective database that would capture a greater number of variables for analysis, including operating room temperatures. Another limitation of our study is that our dataset did not include duration of hypothermia ostensibly, an important measure in investigating the relationship between hypothermia and surgical outcomes. As with all retrospective outcome reviews, confounding is a possibility in our study, and sample sizes for each procedure type limited our ability to further investigate outcomes by procedure type. CONCLUSIONS Our results suggest that mild perioperative hypothermia appears to be independent of wound complications in patients who undergo plastic surgery. This finding does not suggest a diminished importance of perioperative hypothermia and thermoregulation in surgery. Indeed, this study appears to reinforce the importance of perioperative warming. As plastic surgery enters the age of outcomes measurement, the definition of perioperative hypothermia and its effect on patient outcomes deserves further study. Plastic and reconstructive surgeons should remain vigilant regarding perioperative patient body temperature. Disclosures The authors declared no potential conflicts of interest with respect to the research, authorship, and publication of this article. Funding The authors received no financial support for the research, authorship, and publication of this article. Previous Presentation Presented at the American Society of Aesthetic Plastic Surgery Aesthetic Meeting, April 28, 2014; San Francisco, CA. REFERENCES 1. Kurz A, Sessler DI, Lenhardt R, et al. Perioperative normothermia to reduce the incidence of surgical-wound infection and shorten hospitalization. NEnglJMed. 1996; 334:1209-1215. 2. Sessler DI. Mild perioperative hypothermia. N Engl J Med. 1997;336:1730-1737. 3. Sessler DI. Complications and treatment of mild hypothermia. Anesthesiology. 2001;95:531-543. 4. Lenhardt R, Marker E, Goll V, et al. Mild intraoperative hypothermia prolongs postanesthetic recovery. Anesthesiology. 1997;87:1318-1323. 5. Reynolds L, Beckmann J, Kurz A. Perioperative complications of hypothermia. Best Pract Res Clin Anaesthesiol. 2008;22:645-657. 6. Rajagopalan S, Mascha E, Na J, et al. The effects of mild perioperative hypothermia on blood loss and transfusion requirement: a meta-analysis. Anesthesiology. 2008;108: 71-77. 7. Young VL, Watson ME. Prevention of perioperative hypothermia in plastic surgery. Aesthetic Surg J. 2006;26:551-571. 8. Lenhardt R. Monitoring and thermal management. Best Pract Res Clin Anaesthesiol. 2003;17:569-581. 9. Lista F, Doherty CD, Backstein RM, Ahmad J. The impact of perioperative warming in an outpatient aesthetic surgery setting. Aesthetic Surg J. 2012;32:613-620. 10. Young VL, Watson ME. Prevention of perioperative hypothermia in plastic surgery. Aesthetic Surg J. 2006;26: 551-571. 11. Collins JB, Verheyden CN, Mahabir RC. Core measures: implications for plastic surgery. Plast Reconstr Surg. 2013;131:1266-1271. 12. Hernandez M, Cutter TW, Apfelbaum JL. Hypothermia and hyperthermia in the ambulatory surgical patient. Clin Plastic Surg. 2013;40:429-438. 13. Romano PS, Roos LL, Jollis JG. Adapting a clinical comorbidity index for use with ICD-9-CM administrative data: differing perspectives. J Clin Epidemiol. 1993;46(10): 1075-1079. 14. Elixhauser A, Steiner C, Harris DR, Coffey RM. Comorbidity measures for use with administrative data. Med Care. 1998;36:8-27. 15. Constantine RS, Davis KE, Kenkel JM. The effect of massive weight loss status, amount of weight loss, and method of weight loss on body contouring outcomes. Aesthet Surg J. 2014;34(4):578-583. 16. StataCorp. 2011. Stata Statistical Software: Release 12. College Station, TX: StataCorp LP.

88 Aesthetic Surgery Journal 35(1) 17. Doufas AG. Consequence of inadvertent perioperative hypothermia. Best Practice & Research Clinical Anaesthesiology. 2003;17(4):535-549. 18. Sessler DI. Perioperative heat balance. Anesthesiology. 2000;92:578-596. 19. Sessler DI. Perioperative thermoregulation and heat balance. Ann NY Acad Sci. 1997;813:757-777. 20. Kenkel JM, Lipschitz AH, Luby M, et al. Hemodynamic Physiology and Thermoregulation in Liposuction. Plast Reconstr Surg. 2004;114:503-513. 21. Barone CP, et al. Postanesthetic care in the critical care unit. Crit Care Nurse. 2004;24(1):38-45. 22. Cooper S. The effect of preoperative warming on patients postoperative temperatures. AORN J. 2006;83(5): 1073-1084. 23. Kumar S, et al. Effects of perioperative hypothermia and warming in surgical practice. Int Wound J. 2005;2(3):193-204. 24. Benson EE, McMillan DE, Ong B. The effects of active warming on patient temperature and pain after total knee arthroplasty. AJN. 2012;112(5):26-33. 25. Frank SM, Fleisher LA, Breslow MJ, et al. Perioperative maintenance of normothermia reduces incidence of morbid cardiac events. A randomized clinical trial. JAMA. 1997;277:1127-1134. 26. Johansson T, Lisander B, Ivarsson I. Mild hypothermia does not increase blood loss during total hip arthroplasty. Acta Anaesthesiologica Scandinavica. 1999;43:1005-1010. 27. Winkler M, Akça O, Birkenberg B, et al. Aggressive warming reduces blood loss during hip arthroplasty. Anesthesia and Analgesia. 2000;91:978-984. 28. Widman J, Hammarqvist F, Sellden E. Amino acid infusion induces thermogenesis and reduces blood loss during hip arthroplasty under spinal anesthesia. Anesthesia and Analgesia. 2002;95:1757-1762. 29. Benzer A, Sparr HJ. Perioperative normothermia and surgical-wound infection [letter]. N Engl J Med. 1996;335 (10):747. 30. Kempen PM. Perioperative normothermia and surgicalwound infection [letter]. N Engl J Med. 1996;335(10): 747-748. 31. Osler T, Rogers F, Fletcher D. Perioperative normothermia and surgical-wound infection [letter]. N Engl J Med. 1996;335(10):748. 32. Schein M, Rucinski J, Wise L. Perioperative normothermia and surgical-wound infection [letter]. N Engl J Med. 1996;335(10):748-749. 33. Winfree CJ, Baker KZ, Connolly ES Jr. Perioperative normothermia and surgical-wound infection [letter]. N Engl J Med. 1996;335(10):749. 34. Delva E, Camus Y, Just B, Ollivier M, Lienhart A. Prevention of hypothermia by skin-surface warming during abdominal surgery in lithotomy position (abstract). Anesthesiology. 1991;75:A1011. 35. Moayeri A, Hynson JM, Sessler DI, McGuire J. Preinduction skin-surface warming prevents redistribution hypothermia (abstract). Anesthesiology. 1991;75:A1004. 36. Sessler DI, Schroeder M, Merrifield B, Matsukawa T, Cheng C. Optimal duration and temperature of prewarming. Anesthesiology. 1995;82:674-681. 37. Just B, Trévien V, Delva E, Lienhart A. Prevention of intraoperative hypothermia by preoperative skin-surface warming. Anesthesiology. 1993;79:214-218. 38. American Society of Anesthesiologists, Physician Consortium for Performance Improvement. Anesthesiology and critical care performance measurement set. American Medical Association. 2008. 39. Coon D, Michaels J, Gusenoff JA, et al. Hypothermia and complications in postbariatric body contouring. Plast Reconstr Surg. 2012;130:443-448.