Conventional wisdom holds that urban academic medical centers are ill-suited for value-based care initiatives such as the Medicare Shared Savings Program (MSSP). Nevertheless, NewYork Quality Care, the MSSP Accountable Care Organization (ACO) established in 2015 as a joint venture between NewYork-Presbyterian, Columbia University Vagelos College of Physicians and Surgeons, and Weill Cornell Medicine, achieved savings in each of its 3 performance years by implementing a range of operational, financial, and analytical initiatives.
The success of NewYork Quality Care is the result of a close collaboration between all three of our participating organizations. Along the way, we gained a number of insights that will inform our future value-based care efforts and that can help other organizations hoping to follow a similar path:
Provide Data and Enable Transparency: Care teams need detailed, up-to-date patient-level data to effectively target and manage patients. Providing transparency on performance across entities and at all levels encourages collaboration and accountability.
Upend Existing Practices: While existing care practices may be convenient for providers, addressing difficult problems often involves redesigning processes. In some cases, the incorporation of new technologies such as telehealth can help to achieve change.
Engage All Stakeholders: NewYork Quality Care was deliberately designed with an ownership and governance structure that was shared equally among the three member organizations. By giving all participants a voice in the design and operations of NewYork Quality Care, we built trust and shared learnings that helped everyone to improve their performance.
Align Physician Incentives to Achieve Engagement: Participating in an ACO requires physicians to change practices and provide care in ways that are not typically reimbursed in a fee-for-service context, such as spending additional time ensuring that quality measure gaps are filled. Consequently, it is essential to implement an incentive program that specifically encourages desired behaviors. To date, we have found this program to be most effective for primary care physicians.
NewYork-Presbyterian, Columbia University Vagelos College of Physicians and Surgeons (Columbia), and Weill Cornell Medicine (Weill Cornell) together form one of the nation’s leading academic centers. Following the passage of the Affordable Care Act in 2010, our three institutions came together to consider how to adapt to the changing health care landscape. While delivering cutting-edge specialty care was our forte, we did not have experience in jointly participating in value-based care programs.
However, we did have experience in managing the health of a population outside of a value-based care arrangement: In 2009, NewYork-Presbyterian and Columbia launched the successful Regional Health Collaborative to improve care for residents of the Washington Heights/Inwood neighborhood and succeeded in reducing emergency department visits for ambulatory care-sensitive conditions by 9.2% and hospitalizations by 5.8% within the first 6 months of the program.
Our three organizations ultimately decided to jointly launch an MSSP ACO that we named NewYork Quality Care. Success was not assured, given the complexity of our three organizations, our predominant experience with fee-for-service tertiary and quaternary care, and the challenge of serving in an urban context with multiple competing hospital systems only blocks away. For NewYork Quality Care to deliver higher-quality and lower-cost care while also providing crucial learnings, our three organizations would have to manage the care of patients across multiple clinical settings in complex organizations while also balancing the academic mission to advance research and education.
The fact that all three institutions used different electronic medical record platforms added another level of complexity, meaning that we would have to build new tools to share data and provide real-time insights. Finally, there was the challenge of precedent: MSSP ACOs that include hospitals have generally been less successful in lowering costs than physician group–led MSSP ACOs.
With that backdrop in mind, NewYork-Presbyterian, Weill Cornell, and Columbia jointly established five goals for NewYork Quality Care:
- Improve the quality of care and reduce costs for attributed Medicare fee-for-service beneficiaries.
- Address the fragmentation of care in the current health care environment.
- Build and share data-analytics capabilities across the enterprise.
- Share expertise and learnings across our institutions and care settings.
- Advance our collective experience in population health.
In 2014, cross-functional teams from across NewYork-Presbyterian, Weill Cornell, and Columbia came together to develop and implement a plan for what ultimately became NewYork Quality Care. That process established a Board of Managers consisting of physician and operational leaders, with equal representation from each organization. Agreeing to an equally shared governance structure helped to achieve alignment on other issues, such as how operating expenses — along with any shared savings received — would be equally allocated among the three institutions. The Board of Managers then hired an Executive Director and created a three-member Executive Committee; those four individuals were charged with establishing and monitoring the ongoing priorities of NewYork Quality Care.
Today, daily operations of NewYork Quality Care are managed by the Executive Director. The Executive Director works in close collaboration with the chief medical officers who represent the four entities that have attributed Medicare beneficiaries: the Weill Cornell and Columbia faculty practices (the Weill Cornell Physician Organization and ColumbiaDoctors), NewYork-Presbyterian’s Ambulatory Care Network, and NewYork-Presbyterian’s employed medical groups in Westchester County (NYPMG-Westchester). NewYork Quality Care also built a team of seven analysts, including a director, to supply reporting to all four groups.
With >31,000 initially attributed beneficiaries, NewYork Quality Care’s first step was to build an analytics infrastructure to increase our understanding of the care patterns of attributed beneficiaries and to provide actionable data to clinicians. This step involved the creation of shared dashboards that were populated with the monthly claims-level data files provided by Medicare.
These dashboards offer insight on chronic conditions, cost of care, quality measures, and care provided to attributed beneficiaries both inside and outside of the three entities that constitute NewYork Quality Care. The dashboards provide views at the ACO, entity, practice, and individual clinician levels. All our clinicians with attributed beneficiaries are provided with access to these dashboards, enabling them to both evaluate their individual patients’ needs and understand the performance of NewYork Quality Care as a whole.
The insights provided by this analytics infrastructure enabled NewYork Quality Care to design programs to meet the specific needs of our attributed beneficiaries. Given the large number of attributed beneficiaries, we needed to identify which patients to prioritize for care management. In some cases, this process would require strengthening existing care-coordination programs, whereas in other cases it would require building new programs. After reviewing the data, the medical leadership of NewYork Quality Care agreed to focus on attributed beneficiaries who had been discharged from the hospital, with particular attention being paid to the approximately 20% of attributed beneficiaries with high-risk clinical conditions such as heart failure and chronic obstructive pulmonary disease. We believed that this focus provided the best opportunity for us to have a meaningful impact on the quality and cost of care.
Our next step was to design the interventions most likely to improve care and reduce cost. Claims data helped us to identify patients who lacked a timely office visit following hospital discharge. Such visits represent a critical opportunity to assess both clinical and nonclinical needs and to coordinate access to appropriate services. Given the challenges of connecting with post-discharge patients in metropolitan New York, we opted to bring these follow-up visits to our beneficiaries by leveraging the consortium’s existing telehealth infrastructure. Recognizing that many beneficiaries had difficulty with technology, we enlisted a home health agency to facilitate the visit within 48 hours after discharge. This innovation had the added benefit of enabling our care managers to evaluate the home environment.
We also identified an opportunity to redesign care for our >500 patients with end-stage renal disease (ESRD), who typically have the highest per capita cost of all MSSP beneficiaries. In the past, ESRD patients who missed dialysis and presented to the emergency department were admitted to the hospital. By routing these patients directly to the inpatient dialysis unit without admitting them, we reduced unnecessary hospitalizations, contributing to savings of $3.1 million on attributed ESRD patients alone during the 2017 performance year.
While these projects addressed specific groups of beneficiaries, we also built a clinician incentive program to encourage our participating physicians to improve quality metrics across NewYork Quality Care. The first step was to provide the relevant information in a dashboard built on electronic medical record data. We then offered physicians a per-attributed-beneficiary-per-month payment based on a specified percentage improvement in quality measures. Given our belief in the importance of this incentive to our success, we allocated approximately 10% of NewYork Quality Care’s budget to this program. We have seen strong adoption to date: Currently, 78% of eligible clinicians are participating in this program.
While these and other interventions (e.g., care management, data transparency, etc.) contributed to the achievement of our goal of receiving shared savings, the experience of building and managing NewYork Quality Care had the secondary benefit of deepening collaborative relationships across all three of our organizations. Equal participation and the sharing of learnings from both successes and failures strengthened relationships and increased trust. This increased sense of collaboration also paid dividends beyond NewYork Quality Care by helping to set the stage for the 2018 launch of a multiyear transition to a unified electronic medical record.
Challenges and Next Steps
NewYork Quality Care was built through a collaboration between three large, complex organizations. As a result, one of the most significant challenges that we faced was efficiently and effectively achieving alignment around care redesign projects across all practice settings. We believed that the best way to effect such change would be through face-to-face meetings with frontline clinicians, but scheduling these meetings proved to be difficult.
We also faced the related challenge of translating the output of our analytics team into actionable insights that our clinicians could apply to their daily patient encounters. Meeting with clinicians to review their attributed patients’ cost and quality-of-care data and to answer any questions about the operations and goals of NewYork Quality Care ended up becoming a significant part of our clinician-engagement efforts.
The creation of NewYork Quality Care has established a blueprint for collaboration that has increased alignment among our three institutions. Looking ahead, our plan is to potentially expand NewYork Quality Care to include additional clinicians and their attributed beneficiaries. We also plan to leverage the care redesign and analytics capabilities that we have developed to inform other such efforts across our three institutions.
Metrics and Performance
Medicare evaluates the performance of all MSSP ACOs on the basis of a number of quality metrics, including measures of patient experience, care coordination, and preventive health. In the initial year, ACOs are only evaluated on their ability to report these metrics, whereas in later years they are evaluated on the basis of their actual performance on the measures. (For example, in Year 1 , NewYork Quality Care’s score was 100% because we received full credit for reporting on all measures, whereas in Year 3 , it was 80.56% because 24 of the 31 quality measures were performance-based).
Those quality metrics are then combined with the amount of savings achieved by the ACO relative to the benchmark to determine how much (if any) savings should be shared with the ACO. Savings are calculated through a detailed methodology that involves (1) establishing an expected risk-adjusted per-beneficiary Medicare Part A and B spend figure that is multiplied by the number of attributed beneficiaries and then (2) comparing that amount to the actual Part A and B claims for those beneficiaries over a given year.
NewYork Quality Care designed, implemented, and refined its programs over a period of 3 years, and its performance reflects this evolution. In its first year of participation (2015), NewYork Quality Care had 31,201 attributed beneficiaries and saved $2.9 million compared with the historical benchmark. This figure was, however, below the minimum savings threshold necessary to earn shared savings. In 2016, with 31,008 attributed beneficiaries, NewYork Quality Care saved $1.7 million compared to the historical benchmark — again, below the minimum savings threshold. With the addition of the NYPMG-Westchester clinicians in 2017, NewYork Quality Care increased to 38,033 attributed beneficiaries, saved $17.8 million, and earned shared savings of $7.1 million.