Note: This Guide, co-authored by Rachel Mandel, M.D., was first published as a series of blog posts in 2021. We are now re-publishing in consolidated form for state agencies and healthcare leaders considering the new CMS AHEAD model, which draws largely on Maryland's long experience with an all-payer system and total cost of care. For additional insights from Dr. Mandel, please see this Q&A.
The State of Maryland has been experimenting with healthcare payment models for decades. Since 1977, Maryland’s Health Services Cost Review Commission (HSCRC) has been prospectively setting hospital payment rates for all payers, with considerable success. According to information from the HSCRC, data from 1975 to 2005 shows that Maryland had the lowest cost increase per inpatient admission of any state in the US.
In 2014, Maryland introduced a new payment model, the “All-Payer Model” to expand payment control to cover all hospital-based services. In 2019, Maryland modified its model once again, adopting the “Total Cost of Care Model” in attempt to manage costs across the full continuum.
While Maryland’s experience could comprise a graduate course for an entire semester, the purpose of this guide is to understand the Maryland models, what has worked, and what lessons health systems in other states can learn to better prepare for widespread payment reform.
Note: As sourced throughout this guide, RTI International was tasked by CMS to evaluate and report on the All-Payer Model. The comprehensive report can be found here, and some key findings are included in this guide. This guide also includes the contributions of our colleague, Rachel Mandel, MD, who has lived the Maryland experiment as both a practicing clinician and healthcare executive.
Understanding Maryland's Payment History
What is the Medicare waiver?
In the 1970s, Maryland negotiated a waiver from the traditional Medicare payment system and related regulations in return for the State agreeing to actively manage healthcare costs by setting rates and cost control goals for hospitals. In this “all-payer” system, the State’s independent agency, the Health Services Cost Review Commission (HSCRC), sets hospital charges for each organization based on a complex algorithm. For each hospital, the HSCRC establishes a payment rate for each service that is the same for every payer, including Medicare, Medicaid, and private insurers. The state’s hospitals must meet certain collective goals or tests that are set by Medicare and the HSCRC to keep the waiver in place. According to the Maryland Hospital Association, a key advantage of the all-payer system is elimination of cost shifting between payers. All payers pay their fair share of costs, and each payer equitably shares the burden of uncompensated care.
The All-Payer Model
Although service-specific rates were fixed under the decades-old waiver, hospitals could still generate more revenue with increased volume. That is, until 2014. In that year, the State implemented a new model, the “Maryland All-Payer Medicare Model Contract.” Under the All-Payer Model (APM), most hospitals were paid using a Global Budget Revenue (GBR) methodology, whereby annual revenues were subject to a fixed cap. GBR hospitals were held accountable not only to a fixed revenue amount but also for quality performance, including rates for readmissions and hospital-acquired infections. Until 2019, however, the APM impacted only hospital-based inpatient and outpatient services, also referred to as “regulated services.” Any freestanding or “unregulated” services, including physician services, were not covered by the waiver and remained under a traditional fee-for-service, volume-driven model. From the outset, the APM was expected to be temporary before transition to a population-based model.
The Total Cost of Care Model
In January 2019, the APM model morphed into the Total Cost of Care (TCOC) model, which is set to run through 2028. Under TCOC, hospitals are responsible not only for the cost of care for regulated services via the GBR payment methodology, but also for their attributed Medicare patients’ cost along the entire continuum of care, including physician services, skilled nursing facilities and non-hospital related healthcare services.
The goal is to encourage navigation, prevention, education, and other initiatives that ultimately will avoid the need for expensive acute care, as well as prevent illness and unnecessary readmissions.TCOC brings more financial risk, whereby hospitals must find new and creative ways to engage patients and coordinate their care to reduce the cost of that care. Success of the TCOC model hinges on the ability of a hospital system to build partnerships and collaborate with community partners who have influence and credibility with the population, particularly those at risk. Hospitals are finding the need to align with private practitioners, skilled nursing facilities, assisted living facilities, government agencies and organizations that provide services that impact social determinants of health—a whole new world for most healthcare systems.
Exploring the All-Payer Model
The All-Payer Model was in place from January 2014 through December 2018. This model capped hospital revenue each year based on a complex algorithm that included value-based penalties and bonuses, commonly known as Global Budget Revenue (GBR), whereby annual revenues were subject to fixed cap. Hospitals’ “day-to-day” revenue was still paid based on the per-volume rates set by the Health Services Cost Review Commission, though the HSCRC allowed hospitals to modify their rates periodically to ensure that they would receive maximum revenue allowed under their global budget. While the prior waiver simply limited costs per admission, the APM, with its fixed revenue, created an incentive to limit both volume and cost of admissions.
APM was modified in 2017 to address one of the challenges for hospitals. With implementation of the Care Redesign Program (CRP), hospitals could make incentive payments (within specified parameters) to community providers, including physicians, who collaborate in redesigning care intended to drive improvement in quality. According to RTI’s analysis, hospitals participating in CRP performed better financially under APM than those who did not participate.
Hospital Strategies Under APM
Maryland hospitals deployed a number of strategies under APM. Only two resulted in both improved financial and patient care performance as measured by the RTI analysis: employment of physicians and patient-specific education and coaching. Hospitals employing more physicians reportedly did so to ensure adequate coverage in their community and/or to ensure that physicians were more aligned with the hospital’s quality performance goals.
By the end of the APM, most hospitals had 24/7 coverage for staffing strategies presumed to be foundational to operate under the model, including increased use of care coordination, care management, and discharge planning staff. In addition, many had added community health workers, particularly for complex patients and to expand primary care access.
According to RTI's analysis, more than 80% of Maryland hospitals created alternative sites to accommodate non-emergent ED visits.
Hospital executives also believed that the use of data analytics, frequently customized and including employees dedicated to the function, was another foundational tool for success with APM. At the same time, converting that data into useful knowledge was sometimes a challenge. Though a data analytics strategy did not result in improved financial performance, it did result in improved patient care performance.Strategic investments in social determinants of health were associated with improved patient care performance, but not financial performance. Examples included: identifying affordable housing, connecting with job training, and providing transportation to healthcare services.
Critical Success Factors Under APM
Hospitals reported three factors that promoted success under APM:
- APM strategies that were led by senior executives. RTI’s analysis showed that implementation led by the CEO or CFO was associated with better performance on patient care metrics.
- Grounding in continuous improvement. A culture steeped in improvement provided the built-in structure to continually adapt under APM, including innovative quality improvements.
- Systemic focus on waste reduction. An ongoing process designed to eliminate waste, rather than a haphazard approach, provided financial benefits as well as reinforcement of an improvement culture.
Two common obstacles to success were cited:
- Patient compliance, despite multi-level strategies to redirect and/or incentivize use of less costly settings.
- Insufficient community resources, particularly in rural communities and particularly for behavioral health services.
Factors that varied among hospitals included:
- Physician engagement. Some hospitals found ways to successfully align physicians with their APM strategies, but many were challenged. A key obstacle to alignment was physician payment that continued under a fee-for-service model, while hospitals were incentivized with fixed revenue.
- Clinical accountability. Using data analytics to drive improvements in care and quality was a critical success factor for some hospitals; others experienced the opposite, challenged by too much data (and perhaps not enough information) and physicians’ lack of acceptance.
- The model structure itself created variability among hospitals in ways that were antithetical to traditional payment. With revenue fixed, hospitals that experienced significant growth did not receive a revenue increase commensurate with that growth. On the other hand, revenue was more predictable for those hospitals with declining volume.
Hospital Results Under APM
For Maryland to maintain the waiver, it had to perform well on the five-year metrics set by the State and CMS. The Maryland Hospital Association reports the following results of the APM:
As shown, Maryland hospitals exceeded all the goals set by the State and CMS over the five years under this model, indicating modest success at a minimum and setting the foundation for progression to the TCOC model.
Despite success with the model’s target metrics, Maryland hospitals experienced mixed results that were not always considered positive.
- Notwithstanding a cap on revenues, Maryland hospitals’ operating margins increased under APM, from just over 1% in 2013 to range of 2.7% to 3.7% during the APM years.
- APM did not incentivize, and in fact penalized for some, appropriate shifts of care to less expensive, non-hospital settings. A decline in hospital-based volume led to a decline in revenue, leaving hospitals unable to successfully invest in community-based programs.
- Despite equal payment, changes in utilization (IP admissions, OP visits and ED visits), hospital service mix, expenditure reductions, and shifts between settings were inconsistent among payers. For example, ED visit rates for commercially insured and Medicaid patients decreased under the model, relative to the non-Maryland comparison groups, but rates for Medicare patients increased.
- The inherent nature of the model did not provide direct funding for medical education, innovation or capital investments, and cost savings were not a sufficiently viable source for some hospitals.
- An admitted limitation of APM was the lack of engagement with non-hospital providers. As result, the most significant changes under APM were those under the control of hospitals (e.g., hospital-spending growth and hospital-acquired infections), not those such as readmissions that require collaboration with non-hospital providers. As noted previously, CRP was implemented in the middle of the APM years to encourage collaboration with non-hospital providers.
- Some hospitals transformed away from inpatient care, including UMS Shore Medical Center at Dorchester and McCready Memorial Hospital in Crisfield, both of which were replaced or retooled as freestanding emergency rooms.
The Next Step in Maryland's Experiment
Not surprisingly, based on its name, the Total Cost of Care model focuses on the total cost of care for each Medicare beneficiary, rather than the narrower per capita hospital cost of the APM. TCOC calls attention to the health of the community as well as the individual. It incentivizes the healthcare system to partner across the care continuum while investing in community-based care and social determinants of health.
Under the TCOC model, hospitals take on more risk than they encountered under the APM. Medicare beneficiaries can still access care in a fee-for-service provider environment, with the freedom to see providers of their choosing, while costs accrued under Medicare Parts A and B are attributed to the patients in the hospital’s catchment area and ultimately the hospital. Currently, 1% of each hospital’s Medicare revenue is at risk for the cost category.
TCOC continues to offer incentives for improving hospital quality of care. Under this category, 7% of hospitals’ inpatient revenue, across all payers, is at risk. The motive behind setting these metrics is to end patient harm in health facilities, reduce potentially avoidable conditions, enhance coordination across care settings and attempt to engage patients in improving their own care and experience.
In addition to retaining CRP, TCOC added the Maryland Primary Care Program (MDPCP) as another tool to transform care delivery through investment in primary care. The Program pays advanced primary care providers a per beneficiary per month amount for care management services, as well as incentive payments associated with quality/experience outcomes and reduced hospital/ED utilization.
Among other requirements, participants in MDPCP must provide comprehensive primary care and promote access and continuity for patients. In fact, MDPCP requirements include all five elements found in Ascendient’s Primary Care Done Right:
- Same-day access
- Extended hours
- Virtual care
- Team-based care
- Comprehensive care
In the first year of the program, the percentage of practices offering same- or next-day appointments increased from 59.6% to 68.6%, and practices offering some version of an electronic visit increased from 47.4% to 54.3%, with 75% reporting so in the last quarter of the first year.
Ongoing Challenges
It all sounds well and good. One could make the case that the APM model was a success and that evolution into TCOC is rational and potentially inevitable. Policymakers are looking at the Maryland waiver and whether it is successful or not. If it is successful it should and could be adopted across the country. However, the implementation and execution was not without controversy, challenges and pain points. Identified below are some of the features of the waiver that have been the most challenging.
The single biggest challenge during the implementation of APM is the misalignment between what hospitals are being asked to do and what the healthcare payment model is paying for. Not unlike systems elsewhere still operating under PPS, this disconnect is painfully evident on multiple fronts as hospitals struggled to fund non-revenue producing infrastructure, build capacity and implement community-based programs.
Fee-for-Service vs. Value-Based Care
In the early days of the APM, hospitals were incentivized to focus on value-based healthcare under a fixed revenue budget, while community-based practitioners were still working entirely under the fee-for-service compensation model. At the most basic level, the incentives for hospitals and private practitioners were completely different. This disconnect created tension between hospitals and physicians because they had different goals. It took a great deal of communication and education for physicians to understand the rules of the APM and how it affected hospitals and their mutual patients.
Medical staff practice patterns and culture evolved more quickly with the advent of accountable care organizations (ACOs) and clinically integrated networks. As Gonzalez-Smith et al note in the October 2019 issue of Health Affairs, there is an increasing number of ACOs accepting downside risk and more that are successfully producing savings.
This progressive alignment with hospitals, population health and value-based care moves everyone in the same direction under the same strategy. Unfortunately, there are not enough ACOs or clinically integrated networks to have a significant impact on national trends. The State of Maryland has attempted to expedite this alignment by offering CRP — initiated under APM and continuing under TCOC — that incentivize community physicians to collaborate with acute care facilities in chronic disease management.
Strategy and payment alignment is not only important in the acute care setting, but throughout the entire continuum of care, especially under TCOC. Maryland hospitals have begun to generate preferred provider networks for nursing homes, assisted living and home health partners. They are working more closely with those who understand the goals of APM and TCOC. These community care facilities are encouraged to focus on decreasing complications and readmissions, which complements the hospital and provider network strategies.
Non-Revenue Producing Investments
Maryland’s Medicare waiver models have focused on decreasing cost and improving quality, but not enough attention has been paid to financing non-revenue producing infrastructure investments that support the move towards population health. Hospitals have increasing requirements for data sharing and analyses to support their initiatives, the majority of which are not supported directly by waiver monies. Not unlike their counterparts in other states, healthcare entities in Maryland are investing in technology without knowing if they can clearly document a return on investment that justifies the expense.
Maryland has responded to the data sharing challenge at the statewide level by launching CRISP, the Chesapeake Regional Information System for our Patients, a health information exchange. CRISP users have access to reports and analytics based on comprehensive, Maryland-wide Parts A, B and D Medicare claims. These reports, which are additions to the case management and prescription drug monitoring programs, provide timely population-level data to all hospitals and patient-level data to state policymakers and hospitals participating in Maryland’s care redesign programs.
Non-Traditional Community Care Services and Initiatives
Despite payment uncertainties, many healthcare organizations have moved ahead in the last five years with innovative population health programs. One such program is known as Community Aging in Place – Advancing Better Living for Elders, or CAPABLE. This program is a great example of a fresh approach to an identified need in the community. It keeps seniors safely at home, where they prefer to be, while controlling attendant healthcare costs. Unfortunately, it is not yet supported by Medicare, though it is currently under consideration for reimbursement. In the meantime, the program is funded by a mosaic of payers, which include ACOs, state senior agencies, pilot program monies, home health agencies and non-profits in the healthcare and housing industries.
Other hospitals are taking on non-traditional roles to tackle population health issues without traditional reimbursement options. For example, one Maryland hospital opened a dental clinic to serve adults who were either uninsured or underinsured. Historically, this patient population would present to the Emergency Room with their non-emergent but costly dental complaints. The clinic was subsidized by the hospital, grants and philanthropic gifts, and is staffed by volunteers and supervised dental students. The clinic utilizes a sliding payment scale. It has become a part of the system’s efforts to provide care in the right location for the right reasons at the right cost. Since the clinic’s opening, emergency room visits for dental complaints have dropped significantly, yielding a ROI in the form of expense reduction.
Conclusion
As healthcare morphs into a system based on population health and value-based care, there will be an increasing emphasis on non-traditional solutions to complex patient problems. The appropriate reforms to healthcare payment models must accompany this evolution if the innovations are to be encouraged and sustained. Slow or absent payment reform by stakeholders will stifle the change and limit success, such as that experienced by Maryland healthcare providers under the waiver during the APM years. Simply put, if the emphasis in healthcare is on population health, value-based care and social risks, then the payment model must address the same activities in a timely fashion.
Payment reform truly is the key. We need rational compensation models that advance innovation and community-based solutions to include community health workers, community paramedics, navigators, community clinics and home-based services. Other novel programs being deployed include food pharmacies, transportation solutions and school educational interventions which are, again, not traditionally reimbursable. Let’s not forget telehealth. Policy reform, now accelerated by COVID-19, should consistently reimburse the use of technological solutions that can mitigate limited access to care, geographic challenges and provider shortages. There is so much opportunity here.
If the rest of the nation is to follow in Maryland’s footsteps, payment policies will most certainly need to keep pace with these strategies and goals. In the meantime, other states and their healthcare providers should review Maryland’s experience and work with their government officials, insurers and partners to find the right way to avoid the pitfalls and embrace the successes of the Maryland models. The path may not be straight, but the importance of transforming to value-based and cost-effective quality care is quite clear.
