A study co-authored by MIT economist Josh Angrist reveals that colon cancer screenings are significantly more effective than previously thought. Analyzing data from five trials across various countries, the study found that screenings reduce colon cancer rates by about 0.5 percent, which is double the impact estimated by earlier studies. This new finding is based on the actual screening effect, rather than just the effect of being invited for screening.
The study highlights a common issue in clinical trials: nonadherence to the intended treatment, especially in trials involving uncomfortable procedures like colon cancer screenings. Many participants offered screenings through colonoscopy or sigmoidoscopy often choose not to undergo them. Previous studies did not adequately account for this nonadherence.
Angrist and co-author Peter Hull, a professor at Brown University, used an econometric method called “instrumental variables” to adjust for the actual number of people screened, rather than just those invited. This method provided a more accurate measure of the screening’s effectiveness.
Their analysis reconciled the variability in findings across different trials, consistently showing a 0.5 percentage point decrease in cancer incidence among those actually screened. The study, published in the Proceedings of the National Academy of Sciences, calls for the routine inclusion of instrumental variable analysis in clinical research to obtain more accurate results.
Key Insights
- Importance of Actual Screening vs. Invitation: The study underscores the distinction between the effects of being invited for screening and the actual impact of undergoing the screening. This distinction is crucial for understanding the true effectiveness of cancer screenings.
- Challenges in Clinical Trials: Nonadherence is a significant challenge in clinical trials, particularly for procedures perceived as unpleasant. This study demonstrates the necessity of accounting for nonadherence to accurately assess the effectiveness of medical interventions.
- Advancements in Econometric Methods: The use of instrumental variables in this study highlights the evolving nature of econometric methods and their application in various fields, including medical research. This approach could lead to more precise evaluations of treatments and interventions in healthcare.