Use of family history to improve risk prediction in clinical care: Colorectal cancer as an example.

摘要

Family history has been called the "cornerstone of individualized disease prevention" but it is underutilized in clinical practice. In order to use it more effectively, its role in assessing risk for disease needs to be better quantified and understood. Family history has been identified as an important risk factor for colorectal cancer (CRC) and risk prediction in CRC is potentially worthwhile because of the possibility of preventing the disease through application of individualized screening programs tailored to risk. The overall project objective was to explore how family history can be better utilized to predict who will develop CRC. First, we used the Utah Population Database (UPDB) to define familial risk for CRC in more detail than has previously been reported. Second, we explored whether individuals at increased familial risk for CRC or at increased risk based on other risk factors such as a personal history of CRC or adenomatous polyps, are more compliant with screening and surveillance recommendations using colonoscopy than those who are at normal risk. Third, we measured how well family history can predict who will develop CRC over a period of 20 years, using family history by itself as a risk factor, and also in combination with the risk factor, age. We found that increased numbers of affected first-degree relatives influence risk much more than affected relatives from the second or third degrees. However, when combined with a positive first-degree family history, a positive second- and third-degree family history can significantly increase risk. Next, we found that colonoscopy rates were higher in those with risk factors, according to risk-specific guidelines, but improvements in compliance are still warranted. Lastly, it was determined that family history by itself is not a strong predictor of exactly who will acquire colorectal cancer within 20 years. However, stratification of risk using absolute risk probabilities may be more helpful in focusing screening on individuals who are more likely to develop the disease. Future work includes using these findings as a basis for a cost/benefit analysis to determine optimal screening recommendations and building tools to better capture and utilize family history data in an electronic health record system.