COLORECTAL CANCER SCREENING RECOMMENDATIONS IN PRIMARY CARE

Purpose. Almost half of patients with a recent visit to primary care have not been screened for colorectal cancer (CRC). Little is known about the frequency, content or effectiveness of CRC screening recommendations in primary care.  Using direct observation and audio-recordings, we describe the frequency of, content of, and adherence to CRC screening recommendations care during annual “check-ups.”  Methods. Study eligible physicians are general or family medicine physicians practicing in a salaried medical group in southeast Michigan.  Eligible patients are insured under an affiliated health plan, aged 50-80 years with no evidence of fecal occult blood testing (FOBT) within a year, barium enema or flexible sigmoidoscopy within 5 years, or colonoscopy within 10 years.  Eligible patients scheduling a check up with a participating physician are approached for study enrollment.  Enrollment includes completion of a pre-visit telephone interview, audio-recording of the visit, and a post-visit survey.  Observational data are joined with automated health plan data to collect information on CRC screening procedure use in the 6 months following the observed visit.  Visit percentages are reported and precision of estimates is assessed taking into account clusters of patients seeing the same physician.  Results. 45% of eligible physicians (N=75) and 49% of contacted patients (N=257) have agreed to study participation.  Results from the first 201 observed/recorded visits indicate that CRC screening is discussed 95.0% of the time among patients due for screening and, when discussed, is generally first raised by the physician (84%).  Physician recommendations for screening are most often for colonoscopy (58.2%), FOBT (4.5%) or both (24.9%).  In 45.8% of visits there is evidence that the patient and physician have discussed CRC screening during at least one previous visit.  At the time of visit closure, 91.5% of patients recall discussing CRC screening during the visit, 87.5% report that their physician recommended CRC screening, and 31.4% indicate that the decision regarding CRC screening was shared.  At 6-months post-visit follow up, 53% of patients have completed recommended screening as evidenced in automated claims data.  Conclusions. Results are consistent with previous findings that physician recommendations alone are not sufficient to ensure receipt of recommended CRC screening.  Additional work is needed to determine the extent to which different physician recommendation processes and content result in improved adherence to recommended CRC screening.