Recalibration of Gail’s Model for Risk Assessment of Breast Cancer Among Female Adult Population in Iran: a case-control Study

Introduction & Aim: Since the increasing incidence and mortality rate of breast cancer (BC), in order to prevent the disease and adopt risk reduction strategies for high-risk women in clinical practice, it is substantial to evaluate the individualized risk of developing the disease among Iranian population. There are several models for estimating the risk of developing BC. Gail is the most widely utilized.Methods: The subjects included in the current matched case-control study consists of 1000 female incident BC in the ages of 20-80 referred to Imam-Khomeini hospital. The same sample size was considered for the control group (n=1000). We applied the logistic regression to our data and re-estimated coefficients for Iranian female population and estimated individualized long-term risk of developing BC. The discriminatory accuracy of new model was compared with Gail’s estimation of coefficients using the Hosmer–Lemeshow test, and Receiver Operating Characteristic (ROC) curve analysis. All statistical analysis were performed in R version 3.4.1.Results: In terms of calibration, the difference in AUCs was not statistically significant (an AUC of 63.6% for the new model, compared with an AUC of 62.7% for the Gail s). The Hosmer-Lemeshow goodness-of-fit test is in consistent with the above result as well.Conclusion: According to the current results, it can be suggested that since application of the current version of the Gail model will underestimate the risk of BC development in Iranian female population, further studies adding other principal variables is required to optimize the applicability of the model in Iranian female population.