We retrospectively analyzed prospectively collected data for all 1214 patients who underwent primary elective TKA at our institution from January 1998 to December 2005. Prosthetic joint infections were defined using the US Centers for Disease Control classifications for surgical site infections [15
]. All patients who incurred an acute deep infection up to 1 year after TKA at our institution were included in the analysis (Table ). The median age of the patients was 72 years (interquartile range, 65–77 years); 765 patients were females and 449 were males. There were 564 left and 650 right knees replaced. The median body mass index (BMI) of the group was 30.7 kg/m2
(interquartile range, 27.1–35.3 kg/m2
). Reasons for patients undergoing TKA were osteoarthritis (n = 1114), rheumatoid arthritis (n = 95), osteonecrosis (n = 3), and trauma (n = 2). The major comorbidities were cardiovascular (n = 874), respiratory (n = 211), and diabetes mellitus (DM) (n = 206). Seventy-two percent of patients had multiple comorbidities (two or more). No patients were lost to followup during the first 12 months after their procedure, but there were 11 deaths (0.9%) during the first 12 months after index surgery. No patients were recalled explicitly for this study and all data came from the records.
Obesity was defined according to the US Centers for Disease Control guidelines [3
]. A BMI less than 30 kg/m2
was classified as nonobese, 30 kg/m2
to 39 kg/m2
as obese, and 40 kg/m2
or greater as morbidly obese. Of the 1214 patients who underwent TKA, 59% were either obese (n = 592) or morbidly obese (n = 123). Substantially more women (63% [n = 498]) were obese or morbidly obese than men (48% [n = 217]). The number of patients with multiple comorbidities was similar between groups: obese, 73% (n = 431); morbidly obese, 77% (n = 95); and nonobese, 70% (n = 349). The median age at the time of surgery was younger (p < 0.001) for the obese and morbidly obese groups than for the nonobese group (71 years [interquartile range, 64–76 years] and 67 years [interquartile range, 62–73 years] versus 74 years [interquartile range, 68–78 years], respectively) (Table ).
TKAs were performed in a positive-pressure operating room. Patients received perioperative prophylactic antibiotics, which consisted of 1 g intravenous cefazolin on induction and continued for 24 hours after the procedure. When a preexisting allergy to cephalosporins was identified, prophylaxis was substituted with 1 g vancomycin twice a day for 24 hours with the first dose administered on induction. The dosage and timing of antibiotic prophylaxis were as per protocol in all 18 patients diagnosed with deep infection, with one of the 18 patients receiving vancomycin. All patients had a urinary catheter inserted in the operating room just before surgery, which remained in situ for 48 hours. Gentamicin was given just before insertion and removal of the urinary catheter. Regional anesthesia was used in 1083 patients and general anesthesia in 131 patients. A totally cemented prosthesis was used in all patients. Antibiotic-impregnated cement was introduced at our institution in March 2002 and was used in 621 patients. In the majority of patients (n = 1165), a medial parapatellar approach was used; 44 patients underwent surgery through a midvastus approach and five underwent surgery through a lateral parapatellar approach. No cases of infection occurred using the latter two approaches.
Postoperative care for all patients was standardized through the use of a clinical pathway for TKA introduced at our institution in 1995 [8
]. Low-pressure suction drains were used in 1109 patients and remained in situ for 24 to 48 hours. The median postoperative hemoglobin level was 102 g/dL (interquartile range, 90–112 g/dL), and 300 patients received an allogeneic blood transfusion postoperatively. Autologous blood transfusions were not used during the study.
We reviewed the medical records for all patients in this group, including inpatient data, discharge summaries, and all outpatient followup notes. Data collected included patient demographics, comorbidities, operative time, length of stay, discharge destination, complications, and readmissions that occurred within the first 12 months of the index surgery. Once data collection was complete, patients were separated into obese and nonobese groups according to their BMI classification as described previously.
Differences in infection rates between obese and nonobese patients were determined using the chi square test. Multiple logistic regression was performed to determine whether any comorbidities commonly linked with obesity (cardiovascular disease or diabetes) might be associated with an increased risk of periprosthetic infection and whether patient (gender, etiology, or smoking) and surgical factors (transfusion, drain tube, or antibiotic-impregnated cement) influenced infection rates. The OR and 95% confidence intervals were tabled for each variable tested. Comorbidities associated with an increased rate of prosthetic infection were further analyzed in combination with obesity using a chi square test. Data were maintained and analyzed using Microsoft® Excel® (Microsoft Corp, Redmond, WA) and SigmaStat® for Windows® Version 3.0.1 (SPSS Inc, Chicago, IL).