OBJECTIVE: To construct a predictive model of postoperative colorectal neoplasm development using a nomogram. BACKGROUND: Although patients with colorectal cancer (CRC) are known to be at high risk of developing metachronous adenoma or CRC, no statistical model for predicting the incidence of postoperative colorectal lesions has been reported. METHODS: A total of 309 CRC patients who underwent surgical resection received regular endoscopic follow-up to detect the development of metachronous adenoma or adenocarcinoma. The patients were divided into the derivation set (n = 209) and the validation set (n = 100). The nomogram to predict the 3- and 5-year adenoma-free survival rates was constructed using the derivation set, and a calibration plot and concordance index (c-index) were calculated. The predictive utility of the nomogram was validated in the validation set. RESULTS: Sex, age, and number of synchronous lesions at the time of surgery for primary CRC were adopted as variables for the nomogram. The nomogram showed moderate calibration, with a c-index of 0.709 in the derivation set and 0.712 in the validation set. CONCLUSIONS: A nomogram based on sex, age, and number of synchronous lesions at the time of surgery has the ability to predict postoperative adenoma-free survival.
OBJECTIVE: To construct a predictive model of postoperative colorectal neoplasm development using a nomogram. BACKGROUND: Although patients with colorectal cancer (CRC) are known to be at high risk of developing metachronous adenoma or CRC, no statistical model for predicting the incidence of postoperative colorectal lesions has been reported. METHODS: A total of 309 CRC patients who underwent surgical resection received regular endoscopic follow-up to detect the development of metachronous adenoma or adenocarcinoma. The patients were divided into the derivation set (n = 209) and the validation set (n = 100). The nomogram to predict the 3- and 5-year adenoma-free survival rates was constructed using the derivation set, and a calibration plot and concordance index (c-index) were calculated. The predictive utility of the nomogram was validated in the validation set. RESULTS: Sex, age, and number of synchronous lesions at the time of surgery for primary CRC were adopted as variables for the nomogram. The nomogram showed moderate calibration, with a c-index of 0.709 in the derivation set and 0.712 in the validation set. CONCLUSIONS: A nomogram based on sex, age, and number of synchronous lesions at the time of surgery has the ability to predict postoperative adenoma-free survival.
Authors: Tarik Sammour; Andrew Macleod; Tim J Chittleborough; Raaj Chandra; Susan M Shedda; Ian A Hastie; Ian T Jones; Ian P Hayes Journal: Int J Colorectal Dis Date: 2016-03-16 Impact factor: 2.571