PURPOSE: There is increasing pressure to reduce the length of stay (LOS) in hospital after colectomy. To assess the impact of unintended variations on LOS, actual discharge management was compared with the expectations expressed by general surgeons. METHODS: We retrospectively examined 262 patients who underwent elective open colectomy. The effects of patient demographics, surgical variables, and functional recovery on postoperative LOS were assessed in univariate and multivariate models. We also used a structured questionnaire to assess the expectations of surgeons regarding criteria affecting patient discharge. RESULTS: The median LOS was 11 days, which was univariately correlated with morbidity, bowel movement, oral feeding, mobilization, surgeon, rectal procedures, comorbidity, type of analgesia, blood transfusion, and colorectal cancer. According to the multivariate analysis, morbidity (P < 0.001), colorectal cancer (P = 0.026), rectal procedure (P = 0.037), and analgesia (P = 0.04) were correlated with LOS. The surgeons we interviewed believed that discharge should be based on the return of bowel movement and the absence of morbidity. By discharging our patients on postoperative day 5 or 6, these criteria would have been respected in 84% and 88%, respectively. CONCLUSION: Prolonged hospitalization after colectomy persists when critical pathways are not observed and there is a lack of institutional pressure to reduce to the LOS. Many factors affect the delay in patient discharge, possibly reflecting unwanted variation rather than the surgeons' expectations.
PURPOSE: There is increasing pressure to reduce the length of stay (LOS) in hospital after colectomy. To assess the impact of unintended variations on LOS, actual discharge management was compared with the expectations expressed by general surgeons. METHODS: We retrospectively examined 262 patients who underwent elective open colectomy. The effects of patient demographics, surgical variables, and functional recovery on postoperative LOS were assessed in univariate and multivariate models. We also used a structured questionnaire to assess the expectations of surgeons regarding criteria affecting patient discharge. RESULTS: The median LOS was 11 days, which was univariately correlated with morbidity, bowel movement, oral feeding, mobilization, surgeon, rectal procedures, comorbidity, type of analgesia, blood transfusion, and colorectal cancer. According to the multivariate analysis, morbidity (P < 0.001), colorectal cancer (P = 0.026), rectal procedure (P = 0.037), and analgesia (P = 0.04) were correlated with LOS. The surgeons we interviewed believed that discharge should be based on the return of bowel movement and the absence of morbidity. By discharging our patients on postoperative day 5 or 6, these criteria would have been respected in 84% and 88%, respectively. CONCLUSION: Prolonged hospitalization after colectomy persists when critical pathways are not observed and there is a lack of institutional pressure to reduce to the LOS. Many factors affect the delay in patient discharge, possibly reflecting unwanted variation rather than the surgeons' expectations.