Should lymph node retrieval be a surgical quality indicator in colon cancer?
648 Background: Adequate lymph node harvest is easily obtained and is necessary for proper staging of colorectal cancer, making it an attractive measure of surgical quality for policy makers. However, achieving an adequate lymph node harvest requires a multidisciplinary effort. The purpose of this study was to determine if it is appropriate to use this measure as a surgical quality indicator for individual surgeons. Methods: The study was undertaken at a high volume center with standardized colon cancer specimen processes. The charts of 1,138 consecutive segmental colon cancer surgeries performed between 2002 and 2008 were retrospectively analyzed. The primary outcome was inadequate lymph node retrieval for colon cancer surgery defined by fewer than 12 lymph nodes on pathology. Predictor variables were based on patient, surgeon, pathology and tumor related factors. Univariate analysis was performed on all potential predictor variables, followed by multivariate logistic regression. Results: 841 cases (69.0%) achieved adequate lymph node harvest, while 377 (31.0%) were inadequate. Factors on univariate analysis associated with inadequate lymph node harvest were specimen length (p<0.0001), tumor location (p<0.0001), and T-stage (0.0015), all of which remained significant multivariate logistic regression. The average specimen length differed by 3.6 cm between non-adequate and adequate specimens. When broken down by procedure, resection length did not vary significantly between high and low volume surgeons or between colorectal and non-colorectal surgeons. Furthermore when surgeons were ranked according to their success rate of >12 LN retrieval, there was no difference between surgeons in mean specimen length. Conclusions: In a high-volume, tertiary care centre that uses standardized practices for specimen processing, 31% of cases yielded fewer than 12 lymph nodes. Factors relating to the patient and tumor were the primary predictors of a successful outcome and there was no association between surgeon-related factor and adequate LN retrieval. Caution should be used when considering LN harvest as a surgical quality indicator for individual surgeons.