The Impact of Diabetes on Perioperative Patient Outcomes After Total Hip and Total Knee Arthroplasty in the United States Michael P. Bolognesi, MD,* Milford H. Marchant Jr., MD,* Nicholas A. Viens, MD,* Chad Cook, PT, PhD, MBA, y Ricardo Pietrobon, MD, PhD, MBA, y and Thomas Parker Vail, MD z Abstract: The purpose of this study was to determine whether patients with diabetes mellitus (DM) have a higher likelihood of immediate, inpatient complications following primary and revision total hip (THA) and total knee arthroplasty (TKA) than patients without DM. From 1988 to 2003, the Nationwide Inpatient Sample identified 751340 primary or revision THA or TKA patients. 64262 (8.55%) had DM. Comparisons of specific outcome measures between diabetic and nondiabetic cohorts were performed using bivariate and multivariate analyses with logistic regression modeling. Diabetic patients had fewer routine discharges and higher inflation- adjusted hospital charges for all procedures. Although complications were not uniformly increased, diabetic patients had significantly increased odds of pneumonia, stroke, and transfusion (P b .001) after primary arthroplasty. This analysis of a large patient database indicates clinically relevant information for patients and surgeons, suggesting that patients undergoing THA and TKA demonstrate more complications and utilize more resources if they have the comorbidity of DM level II evidence. Key words: diabetes, total knee arthroplasty, total hip arthroplasty. © 2008 Published by Elsevier Inc. As the average age of the population in the United States continues to increase, there has been progressive growth in the incidence and prevalence of not only diabetes mellitus (DM) but also arthritis [1-4]. The rate of joint arthroplasties continues to increase substantially [5]. Because the number of patients with diabetes and arthritis is increasing, the number of patients with diabetes electing to undergo joint arthroplasty is likely to increase as well. Several studies within the orthopedic literature suggest the presence of comorbidities, particularly diabetes, may have a negative effect on surgical outcomes; however, limited research currently exists that focuses on total joint replacement patients [6-13]. Analysis of small case series from single institutions have suggested that the complication rates in patients undergoing these procedures increased when the patients carried the comorbidity of diabetes [6-9]. Evaluation of patients who underwent total hip arthroplasty (THA) showed variable increases in the rates of postoperative infection but consistent increases in the overall complica- tion rates experienced by patients with diabetes [6,8]. Similar studies, specific to total knee arthroplasty (TKA) patients, found significantly decreased subjective outcome scores for patients with diabetes and increased likelihoods of postoperative infections, wound complications, and revisions [7,9]. Because of variation in operative protocols and outcome measurement tools used in studying orthope- dic patients, it has been difficult to draw conclusions regarding the overall effect of diabetes on perioperative morbidity and mortality for all orthopedic patients, and From the *Division of Orthopaedic Surgery, Duke University Medical Center, Box 3269, Durham, North Carolina; yCenter for Excellence in Surgical Outcomes, Duke University Medical Center, Box 3094, Durham, North Carolina; and zDepartment of Orthopaedic Surgery, University of California, San Francisco, Box 0728, San Francisco, California. Submitted September 14, 2007; accepted May 11, 2008. No benefits of funds were received in support of the study. Reprint requests: Michael P. Bolognesi, MD, Division of Orthopaedic Surgery, Duke University Medical Center, Box 3269, Durham, NC 27710. © 2008 Published by Elsevier Inc. 0883-5403/08/2306-0017$34.00/0 doi:10.1016/j.arth.2008.05.012 92 The Journal of Arthroplasty Vol. 23 No. 6 Suppl. 1 2008