ABSTRACT
Surgery is the primary treatment for colorectal cancer. A prolonged Length of Stay (pLOS) can increase risk of complications and physical inactivity, leading to a decline in physical function. While promising results were seen from preoperative exercise training and post-operative functional recovery, the predictive potential of preoperative physical function has not yet been investigated. The objective of this study is to determine if preoperative physical function can predict pLOS in patients with colorectal cancer. A total of 459 patients from 7 cohorts were analyzed. Logistic regression was used to determine risk of pLOS (>3 days), and ROC curve was plotted to establish sensitivity/specificity. Selected variables included age, sex, BMI, comorbidity, ASA status, tumor site, surgical approach, handgrip strength, Timed-Up and Go, 30-s Sit-to-Stand, 30-s Arm Curl Test, 6-Minute-Walking Test (6MWT), CHAMPS Physical Activity Questionnaire for Older Adult and the 36-Item Short Form Survey. The results showed that patients with rectal tumor are 2.7x more at risk to be in the pLOS group compared to those with colon tumor (O.R. 2.7; C.I. 1.3-5.7, p=0.01). For every increment of 20 m in 6MWT, there is a decreased risk of 9% of being in pLOS group (C.I. 1.03-1.17, p=0.00). A cut-off of 431 m can predict 70% of patients in pLOS group (AUC 0.71 C.I 0.63-0.78, p=0.00). Tumor site (rectal) and 6MWT were significant predictors of pLOS. Using the 6MWT as a screening tool for pLOS with cut-off of 431 m should be implemented in the preoperative surgical pathway.
