Oncology Care Model: Progress along the learning curve
Click Here to Manage Email Alerts
Recognizing the importance of addressing value in cancer care, I have been an interested observer, until recently, of the progress of CMS’s Oncology Care Model, or OCM.
The change occurred earlier this year, when I joined one of the just over 190 institutions that participate in the OCM. Like so many aspects of taking up a new position, the learning curve has been steep.
In a spirit of transparency, I must confess that during my first few weeks here, I began to wonder whether the resources we are investing in OCM participation would pay off in the long term. I had certainly bought into the OCM philosophy of promoting value in cancer care, but I was not able to see any tangible benefits for our patients or our cancer programs after almost 3 years of participation.
With the benefit of 6 months of experience with the OCM and guided by the expertise of my colleagues, I have now begun to understand the importance of this initiative, the wealth of data it is collecting to help us understand and improve our practice, and the potential for practice transformation with broader adoption of bundled payments of this type.
I use the term “potential” deliberately, because although there are some early signs of quality improvement, it is too early to claim victory for the OCM as a scalable model of value-based care.
OCM basics
As a reminder, the OCM is piloted by the CMS Innovation Center to improve the effectiveness and efficiency of cancer care.
The 5-year program started in July 2016 with a goal of using aligned financial incentives to promote enhanced care coordination, appropriateness of care and access to services for patients receiving chemotherapy. The model includes Medicare beneficiaries undergoing 6-month chemotherapy episodes.
Participating groups are required to implement various practice redesign activities, including 24/7 clinician access with real-time access to medical records, use of data for quality improvement, a comprehensive care plan that includes the 13 Institute of Medicine-recommended components, meaningful use of electronic health records, patient navigation and use of nationally recognized clinical guidelines.
In the financial model, participating centers continue to receive fee-for-service payments for services provided to Medicare beneficiaries, as well as a monthly per-beneficiary payment for each patient on chemotherapy. Additionally, there are performance-based payments contingent upon how each center performs against quality improvement and financial targets, calculated by complex formulae that I do not pretend to fully understand.
Early positive signals
Early data from the first 6-month performance period, which began in July 2016, are now emerging.
A study by Brooks and colleagues, published last month in Journal of Oncology Practice, describes early findings in terms of practice impacts and changes in payments for episodes of care. The authors compared OCM participants with a comparator group of practices chosen by propensity matching — researchers adopted this approach because OCM participation is voluntary and, therefore, an analysis would be subject to the risk for substantial selection bias in terms of practice characteristics.
The high-level conclusions of this study show some small beneficial changes in acute care services. Specifically, compared with non-OCM participants, there were slightly fewer ICU admissions and ED visits, as well as fewer hospital and ICU admissions within 30 days of death. There were no apparent differences in hospice use, overall hospital admissions or readmissions, and no difference in total episode payments. So, although the practice-related differences were mostly favorable, they were small (1% to 3%), with no apparent financial gain.
These data reflect the first 6-month period of the model, and many of the practice redesign interventions had not yet taken place.
Additionally, the impact of redesign may take many months to manifest, so it may be too early to tell how effective these measures will be. The intrinsic lag in access to the OCM data means we may need to wait another year or 2 to really understand the impact of these later interventions. That said, there are some early positive signals.
Our own center’s experience in OCM so far has shown a similar pattern. Since participation began, we have seen modest improvements in several metrics, including hospital admissions, ED visits, hospice use, pain assessment and depression screening, as well as in patient experience. How much of this improvement represents the effect of center-wide quality initiatives vs. specific interventions related to OCM is difficult to determine, but the trend is in the right direction.
Although these improvements may be relatively small so far, I am learning that perhaps one of the major benefits of participation has been the access to claims data, which has given us a much better understanding of our practice, our costs and the way patients use our services, at least for the Medicare population.
The insights we have gained from these data are powerful stimuli for practice change. We have gained a more holistic view of what happens to our patients not only within our system, but when they access other systems for components of their care — the use of multiple EDs across the region is one example. We have been able to assess our overall rates of hospital admission for various cancers and ED use and are building acute care services that we hope will address part of this problem.
From a therapeutic perspective, we have detailed data on drug use and costs that will allow us to look for opportunities to optimize and standardize therapy. These data may, of course, reflect the nature of our practice as an academic tertiary and quaternary cancer center, but this information will be essential to inform future decision-making around appropriate performance targets.
Finally, the patient-specific data has given us very detailed insight into the treatment courses and subsequent management of individual patients. Although this anecdotal information may seem to be of limited value, it has been very informative in terms of understanding how to restructure some of our services, especially in the acute care setting.
Embracing the learning curve
Whether early output from the OCM is truly informative for practice redesign is unclear.
Because of the voluntary nature of participation, there are concerns that the practices included in the model are not truly representative of real-world practice.
A second study in Journal of Oncology Practice reports some potentially important differences between oncologists in participating and
Another commonly raised concern is that when such a high proportion of cancer care cost is attributable to drugs, our ability to control this cost may be limited, even with appropriate decision support tools in place to help inform treatment choices.
These are legitimate concerns that will need to be addressed in future models, and it may take several more years to evaluate the true impact of the OCM and design future models.
In the meantime, the data will continue to provide valuable insights into our own practices, expose deficiencies and identify opportunities to improve patient care — it is an important learning curve to be on.
References:
Brooks GA, et al. J Oncol Pract. 2019;doi:10.1200/JOP.19.00265.
Parikh RB, et al. J Oncol Pract. 2019;doi:10.1200/JOP.19.00047.
For more information:
John Sweetenham, MD, FRCP, FACP, is HemOnc Today’s Chief Medical Editor for Hematology. He also is associate director for clinical affairs at Harold C. Simmons Comprehensive Cancer Center at UT Southwestern Medical Center. He can be reached at john.sweetenham@utsouthwestern.edu.
Disclosure: Sweetenham reports no relevant financial disclosures.