Development of a risk index for colorectal cancer screening

1. From a cross sectional study, this derived index can obtain measures to stratify risk for advanced colorectal neoplasia.

Evidence Rating Level: 2 (Good)

Study Rundown: Colorectal cancer screening has been demonstrated to be both beneficial and cost-effective. The U.S. Preventive Services Task Force recommends several tests and approaches to screening, but overall inefficiency in these has led to high costs. This study sought to develop a decision-making index to risk stratify average-risk persons and distinguish those individuals that may be managed with less invasive testing. Using common risk factors, including age, sex, waist circumference, cigarette smoking, and family history, the study authors created a scoring model to categorize patients into very low, low, intermediate, and high risk strata. Using pathology reports and this classification system, the authors calculated that the majority of patients in the very low and low groups with advanced neoplasm could have been identified with sigmoidoscopy, with success likely increased if fecal blood testing was added, sparing the cost and invasiveness of colonoscopy. Strengths to the model include the use of 5 easily obtained screening factors and an even distribution across risk groups that is improved compared to previous models. Limitations include a high percentage of white patients (94%), a lack of comparison between less-invasive approaches, and an inability to distinguish between proximal and distal disease. Overall, this risk index may inform and improve future colorectal screening.

Click to read the study, published today in Annals of Internal Medicine

Relevant Reading: USPSTF Colorectal Cancer Screening

In-Depth [cross-sectional study]: A total of 4500 participants presenting for first screening colonoscopy were enrolled. Two-thirds of the study group were used as a derivation set and the remaining third as the validation group. Ideal formats for risk factors (i.e., cigarette smoking, history, body fat, age and sex) were determined using individual logistics regressions, and a multivariable regression model was created. The model was well-calibrated (p = 0.42) and demonstrated good discrimination (c-statistic = 0.72). A score of 0-13 was calculated as the sum of points given for age (0 if <55, 1 if 55-60, 2 if 60-65, 3 if 65-70, and 4 if >70 years), sex (0 if F, 1 if M), first-degree relative (1 if one or more), smoking history (0 if none, 2 if 0-30, 4 if >30 pack-years), and waist circumference (0 = small, 1 = medium, 2 = large). Risk groups were defined as very low (score 0), low (score 1-3), intermediate (score 4-6), and high (score > 6), and the risk of neoplasia was 1.92%, 4.88%, 9.93% and 24.9% in these strata, respectively. A total of 51/70 (73%; 95%CI 61-83%) of the advanced neoplasms in the derivation set in the very low and low groups, and 21/24 (87.5%; 95%CI 68%-97%) of the advanced neoplasms in the validation very low and low groups, would have been detected on sigmoidoscopy.

Image: CC/Wiki

©2015 2 Minute Medicine, Inc. All rights reserved. No works may be reproduced without expressed written consent from 2 Minute Medicine, Inc. Inquire about licensing here. No article should be construed as medical advice and is not intended as such by the authors or by 2 Minute Medicine, Inc.