Heckman, Michael

Introduction Surgical site infections (SSI) continue to be a risk associated with surgery. The Centers for Disease Control (CDC) uses surgical wound class (SWC) to predict SSI postoperatively. Prior studies have shown that the current wound class definitions are unreliable for surgical subspecialties. Therefore, an alternative surgical wound classification (ASWC) was created to more accurately reflect risk of SSI following orthopedic procedures. The purpose of this study was to compare the two systems. Methods Seventy patients who developed 90-day SSI following total joint arthroplasty (TJA) were matched 1:1 to patients without postoperative SSI. The SWC was recorded from operative reports. Wounds were then retrospectively reclassified according to the ASWC based on preoperative history from the medical record. The degree of agreement and degree of association with occurrence of SSI was compared between the two systems. Results The proportion of agreement between the systems was poor (41.4%). The current SWC was associated with risk of SSI in both unadjusted (p=0.045) and multivariable analysis (p=0.050). Comparatively, the ASWC demonstrated a stronger association with SSI in both unadjusted (p=0.001) and multivariable analysis (p=0.001). Conclusion The results of this study suggest that the ASWC is more predictive for SSI following TJA than the current SWC definitions for total joint arthroplasty.

238 Background: To compare post-transplant recurrence rates of hepatocellular carcinoma (HCC) and differences in percent tumor necrosis in patients with pre-operative superselective chemoembolization (CE) vs. non-superselective CE. Methods: Retrospective review of 147 patients with liver transplant (LT) for HCC treated with pre-operative CE. Those transplanted less than 24 hours after CE and those who had other concomitant treatments were excluded. The angiographic studies and written reports were reviewed to categorize patients into superselective and non-superselective groups. CE was done with an emulsion of lipiodol, cisplatin, doxorubicin, mitomycin C, and contrast, with or without embolization particles. Superselective CE was defined as treatment directly into the vessel(s) feeding the tumor. Treatment from more proximal vessels was considered non-superselective. Data gathered included age, gender, date of CE, date of LT, tumor size and number, tumor stage, MELD score, Child-Pugh class, alpha fetoprotein, percent tumor necrosis at explant, date of recurrence, date and cause of death, and date of last follow-up. Patient and disease characteristics and percent tumor necrosis were analyzed with a Wilcoxon rank sum test or Fisher’s exact test. The Kaplan-Meier method was used to estimate the cumulative incidences of death due to recurrent disease after LT. Results: 99 patients were included, 53 had superselective CE and 46 non-superselective. HCC recurrence rates at 1, 3, 5, and 8 years were all significantly less for superselective CE compared to non-superselective CE (p = 0.039 single, p = 0.011 multivariable). Using multivariable analysis, superselective CE had a lower risk of death due to recurrent HCC at 1, 3, 5, and 8 years (p = 0.046). There was a trend toward more percent tumor necrosis after superselective CE, but it did not reach significance (p = 0.057). Conclusions: This study provides evidence that HCC recurrence after LT and death due to recurrent HCC is lower after pre-LT superselective CE versus non-superselective CE. The strong trend for greater percent tumor necrosis with superselective CE may explain the differences in recurrence rate.