In Honor of Martin Luther King Jr. : Seeking Social Justice in Value-Based Payment
Today we celebrate the vision and memory of Dr. Martin Luther King Jr., Baptist Pastor and Leader of the American Civil Rights Movement. Dr. King’s life and legacy paved the way for equality, hope, and acceptance for all without concern for discriminations of race, religion, creed or other social difference. After being imprisoned for participating in a Birmingham, Alabama civil rights demonstration, Dr. King penned a letter to fellow ministers in which he exclaimed, “Injustice anywhere is a threat to justice everywhere.” He espoused that the acts of identifying and addressing injustice are the first steps in righting discovered inequalities. Here is one such example of how advancing health programming policy will remediate an unjust bias in health care delivery.
In 2019, The National Institute on Minority Health and Health Disparities (NIMHD) and the National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS) studied the Comprehensive Care for Joint Replacement (CJR) Model in hospital patients having hip and knee replacement in dual eligible populations (meaning patients with both Medicare and Medicaid coverage). Hospitals with high-dual volumes have increased disadvantaged populations with increased proportions of medically complex and social risk patients. The study explored care outcomes, incentivization, and model implementation costs in high- and low-dual CJR hospitals. The results revealed low-dual hospitals (i.e., less medically complex and lower social risk patients) were more likely to receive model incentive bonuses than high-dual hospitals (i.e., higher medically complex and higher social risk patients). Researchers also found increases in imposed model penalties on high-duals than low-duals hospitals. Further, to receive annual performance bonuses, high-dual per episode spending reductions needed to decrease by $887 to $2231, compared to $89 to $215 for low-dual groups. Overall, high- and low-duals decreased healthcare spending for model implementation, and discharges to post-acute settings and readmissions did not change in group.
Hip and knee joint replacements or lower extremity joint replacements (LEJR) are the most common inpatient surgery for Medicare beneficiaries, with over 400,000 performed annually. In the Centers for Medicare & Medicaid Services (CMS) CJR value-based payment model, LEJR are the focus for the nation’s first mandated bundled payment model. Starting in 2016, 67 randomly selected metropolitan statistical areas (MSAs) hospitals were mandated to participate, except for those participating in the Bundled Payment for Care Improvement Initiative (BPCI). The investigators studied outcomes in 1165 participating hospitals (291 high-duals and 874 low-duals) and 768,224 patients (68.3% women) in 2016-2017. Using a similar classification process to other CMS’ value-based payment systems, the study categorizes hospitals into quartiles, with the high disadvantaged hospital populations in the high-dual category and the remaining quartiles in the low-dual category. In the CJR model, participating hospitals are accountable for per episode spending and quality of care for patients receiving LEJR surgical episodes during hospitalization and in the 90-day post discharge timeframe. Within the model, when spending is below benchmarks and care quality expectations are met, hospitals receive a bonus. Hospitals exceeding quality-adjusted spending and/or not meeting quality benchmarks are penalized.
Both NIMHD and NIAMS sought to understand how care delivery was modified for the two populations based on model implementation. in three performed analyses:
1. Primary Outcomes
a. Episode Spending - During the index hospitalization and 90 days post discharge (except durable medical equipment [DME] and hospice)
b. Health Service Use – Discharge to institutional post-acute care settings (long-term care hospitals, inpatient rehabilitation facilities, skilled nursing facilities, or swing beds)
c. Quality of Care – CJR relevant readmissions within 90 days post index hospital discharge
d. Secondary Outcomes – Spending at each care setting, discharge to home and home health care, days for institutional post-acute care, index hospitalization stays, complication rates, emergency department visits, mortality rates, discharges to a skilled nursing facility with a 4- or 5-star rating, and physical therapy within 0 to 2 days of hospital discharge
2. 2016-2017 CJR Reconciliation Payments – Model bonuses for quality care and spending stratified as positive, zero, and negative
3. Mean Spending Reductions for Bonus – Reductions in per episode spending needed to attain annual model bonus
For the primary outcomes, care delivered during hospitalization and 90-day post-discharge period, episode spending decreases were noted at both high-dual hospitals by $851(–$1556 to –$146) and low-dual hospitals by $567(–$933 to –$202). Discharges to institutional post-acute care settings and readmission did not change among both high- and low-dual hospitals. In CJR reconciliation payments, high-duals were less likely to receive model bonuses compared with low-duals (40.3% vs 59.1% in 2016 and 56.9% vs 76.0% in 2017), although both dual groups demonstrated statistically increased likelihood from the previous year. To receive an annual bonus, high-dual hospitals needed to reduce their per episode spending from $887 to $2231, while low dual eligible hospitals needed to reduce per episode spending by $89 to $215. Both dual hospitals made changes in care during the model, although the magnitude of spending changes between these 2 groups did not differ.
Within the CJR model, current spending benchmarks are based on a mix of hospital and regional historical spending without adjustment for preexisting social and medical complexity. In the future, care episodes spending will eventually be calculated completely on regional historical spending. A preponderance of evidence demonstrates that patients with higher medical and social and economic care barriers require additional medical spending and practice interventions to ameliorate the impacts of these vulnerabilities. In the recent CMS CRJ Year 2 Evaluation Report, CMS notes specific population characteristics impact model outcomes, including patient age, sex, race, Medicaid eligibility, prior healthcare use, and disabilities. They also state that a portion of episode spending benchmarks “may be due to a healthier mix of patients at CJR participating hospitals”, which, by definition, penalizes participating hospitals high-dual patient populations.
While traditional risk calculations limit risk estimates to medical case mix and other similar risk adjustment mechanisms, both CMS and the National Quality Forum (NQF) have worked diligently to understand the impacts of socioeconomic risk factors on health and cost outcomes. It is widely accepted that advancements in existing risk adjustment methodologies are required in both quality measures and payment innovation to account for disparities among diverse populations. The CMS’ Meaningful Measures Framework “identifies the highest priorities for quality measurement and improvement” naming the elimination of health disparities as a top priority. NQF’s A National Call to Action: Quality and Payment Innovation in Social Determinants of Health (SDOH) states, the focus of “quality and payment innovation in SDOH are to improve health and well-being, reduce disparities, and achieve health equity across the nation”. HealthyPeople.gov reports that SDOH may account for nearly 60% of all health outcomes. Both entities set a high bar for equitable healthcare outcomes in all populations. This would include remediation of policies and program implementation strategies when disparities are identified. By mandating participation in the CJR model, both medical and social risk adjustments representative of population variations must be integrated into both payment and quality measure risk models, especially for highly disadvantaged populations.
By continuing to risk adjust the CJR model for participating hospitals by regional historical spending and not including participant hospital spending patterns and population characteristics, a known pattern of care delivery inequality would persist. With an aim to improve model participation, and normalize risk-sharing and incentives, the CMS Innovation Center Episode Payment Models January 2020 publication finds improvements episode payment models could include capture participant’s historical costs, as well as risk and peer performance adjustments, including medical complexity and social risk. By incorporating and normalizing episodes risks including robust population characteristics (historic episode spending, medical complexity, and social risk) and setting more accurate target prices, CMS is more likely to realize equitable model successes and savings to the Medicare Trust Fund. Here, CMS identifies an unintended consequence of model implementation, and will presumably begin addressing and righting this acknowledged outcomes disparities.
- By Sharon M. Hibay, RN, DNP
CEO, Advanced Health Outcomes
Accountability & Measures Authority