USING TEST ATTRIBUTES TO DESCRIBE PATIENT PREFERENCES FOR COLON CANCER SCREENING TESTS

  

Purpose: 1) To describe patient-reported values for colorectal cancer screening test attributes and, using these values, to categorize the strength of patient preference for colonoscopy vs. fecal occult blood testing; 2) to explore variations in colorectal cancer screening test attribute values and modality preferences by gender, age and race.   

Method: Patients (N=485) completed a telephone survey prior to a scheduled appointment at an internal or family medicine practice in Southeast Michigan. Patients were insured, aged 50-80, and due for colorectal cancer screening. Survey respondents indicated how important various colorectal cancer screening test attributes were to them (e.g., test accuracy, complication risk, required preparation). Responses were used to categorize patients into screening modality preference groups: strong colonoscopy, weak colonoscopy, strong fecal occult blood testing, weak fecal occult blood testing, or an unclear preference. Differences in attribute importance and preference groups by age, gender and race were evaluated using chi square tests.    

Result: Test accuracy was the attribute most often reported to be most important (46%), followed by risk of complications (15%). Based on all attribute assessments, the most common screening modality preference was a weak preference for colonoscopy (39%). Other preference categorizations were less common: 7.8% strong colonoscopy, 10.5% strong fecal occult blood testing, and 20% weak fecal occult blood testing. Almost a quarter of participants’ responses were not consistent with a clear preference for either screening modality (23%). There were no significant differences in reported attribute importance or modality preference categorization by gender, race, or age, with the exception of those who weakly prefer fecal occult blood testing screening, where larger proportions are females (p=0.0448).   

Conclusion: Based on the test attributes patients indicated were most important, most patients did not have a strong preference for either colonoscopy or fecal occult blood testing. The most common categorization was a weak preference for colonoscopy, but many patients could not be linked to a preferred test using their most important attributes. How to make a colorectal cancer screening recommendation that is consistent with patients’ underlying preferences when those preferences do not clearly lean towards one screening modality presents a clinical challenge warranting further study.