The risk prediction model and risk index for colorectal cancer

Objectives: Various prediction models or indices for cancer risk have been developed in Western countries, but no such instruments are available in Taiwan. Colorectal cancer is the third leading cause of cancer deaths in Taiwan, thus it is essential to establish a prediction model or risk index for the prevention of colorectal cancer. Methods: The risk prediction model was established by a case-control study using multivariate logistic regression. Individual odds ratios (OR) for each risk factor were transformed into a risk point. An overall risk point was calculated as an index by summing all risk points for the related risk factors. The risk index was divided into seven categories according to the distribution of the total risk points. Results: In the risk prediction models, ten and nine significant risk factors were identified for males and females, respectively. With the exception of alcohol consumption in females, nine risk factors were identical in both genders, including race, occupation, physical activity, coffee intake, and the consumption of white meat, seafood, vegetables, and fruits, and the method of cooking of meat. After checking the strata for the selected risk factors, individuals can acquire their total risk points using the sum of the risk points for each risk factor. Referring to the level of the risk index by matched total risk points, individuals can recognize their risk level for the development of colorectal cancer. Conclusions: This study has established a risk prediction model and risk index for colorectal cancer in Taiwan. It is a simple and convenient method to assess colorectal cancer risk for the population in Taiwan. A prospective cohort study, however, is needed to help validate the accuracy of this risk prediction model and risk index.