The results of the second year of CMS’s voluntary bundled payment experiment – Bundled Payment for Care Improvement (BPCI) – are now in, and the results for CV patients are decidedly a “mixed bag.” A 248-page evaluation report was issued in August and is available for review. We thought that this evaluation could possibly yield some good insights into CMS’s bundling initiatives in general, given the proposed mandatory bundling for select cardiac patients announced in July to begin next year. As you will remember, the original bundling initiative commenced in 2013 on a voluntary basis and covered a number of patient groups, including cardiac surgery under Model 2. This model included anchor hospitalization; 30-, 60-, or 90-days of care post discharge; and fee-for-service (Medicare Parts A and B) payments. We will focus on the results of this part of the initiative (refer to pages 134 through 137 of the report, E. Impact of BPCI on Cardiovascular Surgery).
Participants and Characteristics
There were 30 hospitals participating, with 2,859 CV surgery episodes.1 About 3% were cardiac valve MS-DRGs, and 28% were CABG MS-DRGs. These 30 participating hospitals averaged 463 beds and were disproportionately teaching hospitals. Beneficiaries treated in the bundling model were similar to all Medicare beneficiaries treated in the same MS-DRG – similar age, gender, etc.
There were statistically significant changes in utilization of post-acute care (PAC) providers in the BPCI episodes relative to comparison groups:
- The number of discharges to a PAC provider decreased from 55% in the baseline to 44% in the initiative, but referrals to home health actually increased from 15.8 visits to 16.6 visits – an average increase of 1.5 visits.
- About 13% of CV surgery episodes were 30-day episodes; 87% were 90-day episodes.
- The Medicare payment for 30-day episodes did have a statistically significant decrease in payment in relation to a comparison group.
- The Medicare payment for 90-day episodes did not have a statistically significant decrease in payment in relation to a comparison group.
Quality of Care
- The emergency department visit rate during the 30-day episode increased significantly from 9.3% to 12.0%. The 90-day rate change was not significant.
- The change in readmission rates was not significant.
- Mortality rates caused some technical issues with comparison groups. Subsequent analysis showed no statistically significant change.
On a preliminary basis, it appears that the bundled pricing model yielded neutral results for CV patients – no significant decrease in payments and no increase or decrease in quality outcomes. Realistically, the jury is still out. In a blog post, CMS acting Principal Deputy Administrator and Chief Medical Officer Dr. Patrick Conway acknowledged that it will take more data to fully evaluate the impact on cost and quality, although he characterized the results (for all specialties covered, especially orthopedics) as “encouraging.” The sample size is small, the hospitals are uncharacteristic of the average hospitals in the country (larger, urban, academic, and serving higher-income populations), and the CV procedures studied are a very mixed bag of surgical and procedural patients. The first year or two of any new program is always one of trial-and-error. Dr. Conway characterized the results this way:
“Cardiovascular surgery episodes under Model 2 did not show any savings yet but quality of care was preserved. Over the next year, we will have significantly more data available, enabling us to better estimate effects on costs and quality.”
Are there lessons to be learned that can be applied to the new, mandatory cardiac bundles announced in July? So far, probably not many. Care coordination focusing on post-discharge remains a challenge and a potential significant downside. Managing readmissions, emergency department visits, PAC provider utilization and care coordination remain critical for ultimate success. CFA remains confident that costs in acute hospitalization can be managed appropriately if appropriate resources are focused, but if readmissions are not addressed, ED visits occur, home health not aggressively managed and other, potential PAC issues not coordinated, then any gains made on the hospital side will be negated or certainly reduced.
Anyone contemplating the new mandatory cardiac bundles should read the appropriate portions of this report, as there are statistics and conclusions that may be helpful in your hospital’s preparation – at least for CABG surgery. No matter what conclusions reports like this produce, Medicare is not giving up on the bundled payment model and appropriate preparation and planning remain mandatory.
If you are interested in learning more about developing a bundled-payment program for cardiovascular services, please contact CFA at (949) 443-4005 or by e-mail at CFA@charlesfrancassociates.com.
1 Episodes included: AICD generator or lead, cardiac defibrillator, cardiac valve, coronary artery bypass graft, major cardiovascular procedure, other vascular surgery, pacemaker, pacemaker device replacement or revision, and PCI.