Effect of Personalized Risk Messages on Uptake of Colorectal Cancer Screening : A Randomized Controlled Trial.

Background: Providing personalized risk information to patients and their providers could improve colorectal cancer (CRC) screening.

Objective: To determine whether providing information on patient risk for advanced colorectal neoplasia (ACN; which includes CRC and advanced precancerous lesions) to patients and providers affects screening uptake, and to identify effect moderators.

Design: Randomized controlled trial (2 × 2 factorial design). (ClinicalTrials.gov: NCT04683731).

Setting: Primary care clinics in 2 health care systems.

Participants: 214 providers and 1084 average-risk patients due for screening.

Intervention: Participants were randomly assigned to view a CRC screening decision aid with or without a personalized message about ACN risk. Providers were randomly assigned to receive notifications that the patient was due for screening, with or without a personalized message about the patient's ACN risk.

Measurements: Screening test completion by 6 months. Logistic regression was used to estimate intervention effects.

Results: Overall, there were no differences in screening uptake or test completion for the provider notification (predicted probabilities, 41.5% vs. 36.4% for personalized vs. generic; difference, 5.1 [95% CI, -1.6 to 11.8] percentage points) or decision aid (predicted probabilities, 36.8% vs. 41.0% for personalized vs. generic; difference, -4.1 [CI, -10.2 to 1.9] percentage points) interventions. Health system was an effect moderator for stool testing. For one health system, the stool testing rate was higher for personalized versus generic provider notification (predicted probabilities, 21.1% vs. 7.9%; difference, 13.2 [CI, 1.6 to 24.8] percentage points) when the decision aid was generic. The stool testing rate was higher for the personalized versus the generic decision aid (predicted probabilities, 21.4% vs. 7.9%; difference, 13.5 [CI, 2.4 to 24.5] percentage points) when the provider notification was generic.

Limitations: Few participants had high-average ACN risk. Non-English-speaking patients were excluded.

Conclusion: Although including personalized risk for ACN in a decision aid or provider notification had no overall effect, it increased uptake of stool testing in one health system.

Primary Funding Source: Patient-Centered Outcomes Research Institute (PCORI).