Increasing Colon Cancer Screening Uptake
Personalized behavioral messages to increase adherence to colorectal cancer screening tests
Partners
Meuhedet Health Services, GistMD, funded by Ministry of Health award
Collaborators
Judith Zamir, Dr. Ronen Bareket , Dr. Siven Berger-Achituv and Giuseppe A. Veltri
The Problem
Colorectal Cancer (CRC) is the second most commonly diagnosed malignancy in Israel and the second most common cause of cancer death. IScreening tests like, the Fecal Occult Blood Testing (FOBT) reduce the odds of late-stage CRC diagnosis. However, the FOBT adherence rates in Israel are suboptimal and many people fail to take regular screening tests for early detection of CRC. Remote care and digital-based interventions aimed at promoting healthy behaviors in large populations (e.g., national prevention programs) typically use a “one-size-fits-all” approach, limiting their effectiveness. Personalized interventions, that have the potential to significantly improve the impact and efficiency of large-scale interventional programs, are rarely tested or implemented in the field.
Our Approach
This study is designed to evaluate the efficacy of personalized interventions in improving adherence to screening tests for early detection of CRC.
Behavioral Solutions
The study is being carried out among Meuhedet HMO members aged 50-75, who have been referred for annual FOBT. It includes two large-scale experiments evaluating the efficacy of personalized vs. one-size-fits-all interventions in improving FOBT adherence. The first, exploratory experiment examined how patients with different levels of individual traits react to different interventions. Using a machine-learning causal forest approach, we calculated the individual-level treatment effect for each participant and determined the optimal message for each participant. The second, validation experiment will test whether the interaction model developed in the first experiment improves FOBT adherence more than non-personalized interventions.
Results and Impact
Improved adherence through personalization will directly lead to reduced CRC morbidity and mortality. The model can then be adapted to other screening test adherence challenges.