3114 — Medical Record Validation of the National Cancer Institute’s Colorectal Cancer Screening Status Questionnaire
Voils CI (Durham COE), Dudley TK
(Durham COE), Coffman CJ
(Durham COE), Grubber J
(Durham COE), Vernon SW
(University of Texas-Houston School of Public Health), Provenzale D
To assess comprehension and validityof a colorectal cancer (CRC) screening questionnaire developed by the National Cancer Institute (NCI).
We conducted cognitive interviews with 36 primary care veterans, aged > = 50 years, a VA facility. Based on these interviews, we made iterative changes to the questionnaire to improve comprehension. The revised questionnaire was administered in person to 201 primary care veterans, aged > = 50 years, at two VA facilities. Participants’ responses were compared to the CRC screening procedure (colonoscopy, flexible sigmoidoscopy, barium enema (BE), and fecal occult blood test (FOBT)) data abstracted from their VA and non-VA medical records. The primary outcome of the validation study was agreement between self-reported NCI questionnaire responses and medical records as to whether patients were up-to-date for CRC screenings by the published Multi-society guidelines We estimated sensitivity, specificity, positive and negative predictive values, concordance and report-to-records ratios.
Four rounds of cognitive interviews were conducted. Interview participants were 64% male and 53% Caucasian. Participants comprehended the descriptions and questions for FOBT, colonoscopy, and flexible sigmoidoscopy. The BE description, however, required three iterative revisions to resolve confusion. Most respondents had difficulty recalling procedure dates for any of the tests, but all who reported having undergone the test were able to choose among specified time intervals. Validation study participants were 74% male and 78% Caucasian. Eighty percent were up-to-date for at least one CRC test and 38% reported receipt of at least one CRC test at a non-VA facility. The sensitivity of the questionnaire for up-to-date CRC screening status was 94%, specificity was 63%, concordance was 88% and report-to-records ratio 1.04.
Self-report is a highly sensitive method to determine CRC screening status when administering the NCI questionnaire in person.
VA primary care patients frequently undergo CRC screening tests outside the VA. Given the difficulty in obtaining records from multiple facilities, self-report using this questionnaire is a reasonably accurate option in practice and research.