ABSTRACT
BACKGROUND: Length of stay following elective colorectal surgery is being reported as a quality measure in surgical outcome registries, such as the National Surgical Quality Improvement Program. Regional referral centers with large geographic catchment areas attract patients from significant distances. OBJECTIVE: The aim of this study was to examine the effect of patient distance traveled, from primary residence to a tertiary care hospital, on length of stay in elective colorectal surgery patients. DESIGN: Retrospective population-based cohort study uses data obtained from the National Surgical Quality Improvement Program database. SETTINGS: This study was conducted at a tertiary referral hospital. PATIENTS: Data on 866 patients undergoing elective colorectal surgery from May 2003 to April 2011 were reviewed. MAIN OUTCOME MEASURES: Demographics, surgery-related variables, and distance traveled were analyzed relative to the length of stay. RESULTS: Of the 866 patients, 54% were men, mean age was 57 years, mean distance traveled was 145 miles (range, 2-2984 miles), and mean length of stay was 8.8 days. Univariate analysis showed a significant increase in length of stay with increased distance traveled (p = 0.02). Linear regression analysis revealed a significant association between increased length of stay and male sex (p = 0.006), increasing ASA score (p = 0.000), living alone (p = 0.009), and increased distance traveled (p = 0.028). For each incremental increase in log distance traveled, the length of stay increases by 2.5%. LIMITATIONS: This is a retrospective review that uses National Surgical Quality Improvement Program data. It is not known how many patients left the hospital and did not return to their primary residence. CONCLUSIONS: In a model that controlled for variables, increased travel distance from a patient's residence to the surgical hospital was associated with an increase in length of stay. If length of stay is a reportable quality measure in patients undergoing colorectal surgery, significant travel distance should be accounted for in the risk adjustment model calculations.
