Date Published: September 9, 2015
Publisher: Springer US
Author(s): Bruce Pyenson, Perry J. Pickhardt, Tia Goss Sawhney, Michele Berrios.
Purpose: To compare the Medicare population cost of colorectal cancer (CRC) screening of average risk individuals by CT colonography (CTC) vs. optical colonoscopy (OC). Methods: The authors used Medicare claims data, fee schedules, established protocols, and other sources to estimate CTC and OC per-screen costs, including the costs of OC referrals for a subset of CTC patients. They then modeled and compared the Medicare costs of patients who complied with CTC and OC screening recommendations and tested alternative scenarios. Results: CTC is 29% less expensive than OC for the Medicare population in the base scenario. Although the CTC cost advantage is increased or reduced under alternative scenarios, it is always positive. Conclusion: CTC is a cost-effective CRC screening option for the Medicare population and will likely reduce Medicare expenditures for CRC screening.
We estimated the per-screen costs of OC and CTC, the frequency of colonic and extra-colonic screening findings and the resulting rescreen times, the size and demographic mix of the Medicare population, and built a simulation model to produce Medicare population-level cost comparisons of the two screening methods. In addition, we tested several alternative scenarios. Throughout this paper, “costs” refers to Medicare allowed amounts, which include the Medicare payment and the enrollee cost sharing payment. For bowel preparation agents, the allowed amounts are those administered by the Medicare Part D insurer.
We identified 127,175 Medicare colonoscopies performed in 2013 (Table 2) and classified 56,578 (44%) as screening for purposes of calculating average costs. 46% of the colonoscopies were excluded because the coding indicated that they were diagnostic, 7% because they were performed on the same-day as an upper endoscopy, 2% because they were incomplete, and 1% because the Medicare enrollee was under age 50.Table 22013 Medicare colonoscopiesColonoscopy days identifiedNumberPercentageTotal127,175100Exclusions: Diagnostic58,20646 Incomplete24412 With same-day upper endoscopy91397 Enrollee under age 508111Net remaining: screening colonoscopy days56,57844Source: Authors’ analysis of 2013 Medicare 5% sample data
To understand the model sensitivity to certain assumptions and potential changes to OC and CTC standards and practices, we tested several alternative scenarios. Table 10 describes the scenarios (with more detail provided in the Appendix in supplementary material) and explains why we selected them. Table 11 summarizes the results of the scenarios.Table 10Alternative scenariosDescriptionAlternative model inputsExplanation1. Fewer large and small polypsApply a 0.80 adjustment factor to the probability of large and small polyps and increase the probability of diminutive size polyps by the same amount so that the probability of finding a polyp of any size remains unchangedOur base scenario, using data centered at approximately 2005, produces an approximately 18% aggregate probability of a large or small polyp finding. We have found smaller studies that indicate an aggregate probability in the 13–15% range [47–49]. Since the probability of finding a polyp of any size, however, uses recent data we left the total probability unchanged2. Add costs for OC and CTC complications and CTC extra-colonic findingsOC costs: add $20 and $96 to OC without and with biopsy costs, respectively, for OC complicationsWe used published literature , trended to 2015, to estimate the costs of complications and the follow-up diagnosis costs of CTC extra-colonic findings. OC has more complications than CTC and within OC, OCs with biopsies have substantially more complications than OCs without biopsies [47, 49, 51]. See Appendix in supplementary material for more details. The ACR currently recommends the reporting of potentially significant extra-colonic findings CTC costs: add $131 to CTC for CTC complications and extra-colonic findings3. Increase anesthesia use for OCAssume that 80% of OCs will have separately billed anesthesia, a 40% increase in use and costs from the 57% base scenario assumptionIn 2013 separately billed anesthesia was subject to Medicare cost sharing; as of January 1, 2015 it is no longer subject to cost sharing . Anesthesia use may therefore increase over the next couple of years4. Add costs for CTC shared decision makingAdd a $20 cost to all CTCs and another $20 cost to CTCs with small polypsMedicare covers CT lung cancer screening with the provision that the first screening must include a documented shared decision making consultation . If Medicare adopts a similar approach for CTC screening, two consultations may be required: the first for the screening and the second for the decision for follow-up OC if the patient has small polyps. See the Appendix in supplementary material for more details5. Decrease maximum screening ageDecrease maximum screening age from age 84 to age 74UPSTF recommends CRC screening until age 75. Medicare, however, currently pays for screening for all ages 50 and over6. Decrease OC follow-up rate for CTCs with small polyp findingsDecrease OC follow-up rate for CTCs with small polyp findings from 50% to 25%No one knows how many patients with small polyps will opt for OC polypectomy vs. CTC surveillance. The percentage will likely vary substantially by clinic and physician7. Increase OC follow-up rate for CTCs with small polyp findingsIncrease OC follow-up rate for CTCs with small polyp findings from 50% to 75%8. Decrease rescreen years for both OC and CTC for screenings with small polypsRescreen in 3 years instead of the 6 years for OC and 5 for CTCLiterature indicates that many OC patients rescreen sooner than recommended by guidelines 9. Increase rescreen years for CTC to match OCRescreen in 3, 6, 7, and 10 years for large, small, diminutive, and no polyps, respectivelyBy removing the rescreen time differential, this scenario compares the per-screen costs of CTC and OCTable 11Alternative scenario simulation results2015 cost per Medicare enrollee per monthScreening pathCTC savings (%)ScenarioOCCTC0. Base$9.34$6.59291. Fewer large and small polyps9.316.09352. Add costs for OC and CTC complications and CTC extra-colonic findings9.898.74123. Increase anesthesia use for OC9.686.67314. Add costs for CTC shared decision making9.347.00255. Decrease maximum screening age6.664.46336. Decrease OC follow-up rate for CTCs with small polyp findings9.346.13347. Increase OC follow-up rate for CTCs with small polyp findings9.347.03258. Decrease rescreen years for both OC and CTC for screenings with small polyps9.656.80309. Increase rescreen years for CTC to match OC9.343.9758Source: Authors’ simulation. Assumes all Medicare enrollees have perfectly adhered to an OC or CTC screening path since age 50
For a Medicare population, when compared to OC, CTC satisfies the third goal of the triple aim: reducing the per capita cost of health care. For our base scenario, CTC is 29% less costly than OC per Medicare enrollee. Although the CTC cost advantage is quite variable under the alternative scenarios, the cost advantage is positive for all alternative scenarios (range: 12–58%).