Guiding Principles to Address the Impact of Algorithm Bias on Racial and Ethnic Disparities in Health and Health Care.

Importance: Health care algorithms are used for diagnosis, treatment, prognosis, risk stratification, and allocation of resources. Bias in the development and use of algorithms can lead to worse outcomes for racial and ethnic minoritized groups and other historically marginalized populations such as individuals with lower income. Objective: To provide a conceptual framework and guiding principles for mitigating and preventing bias in health care algorithms to promote health and health care equity. Evidence Review: The Agency for Healthcare Research and Quality and the National Institute for Minority Health and Health Disparities convened a diverse panel of experts to review evidence, hear from stakeholders, and receive community feedback. Findings: The panel developed a conceptual framework to apply guiding principles across an algorithm's life cycle, centering health and health care equity for patients and communities as the goal, within the wider context of structural racism and discrimination. Multiple stakeholders can mitigate and prevent bias at each phase of the algorithm life cycle, including problem formulation (phase 1); data selection, assessment, and management (phase 2); algorithm development, training, and validation (phase 3); deployment and integration of algorithms in intended settings (phase 4); and algorithm monitoring, maintenance, updating, or deimplementation (phase 5). Five principles should guide these efforts: (1) promote health and health care equity during all phases of the health care algorithm life cycle; (2) ensure health care algorithms and their use are transparent and explainable; (3) authentically engage patients and communities during all phases of the health care algorithm life cycle and earn trustworthiness; (4) explicitly identify health care algorithmic fairness issues and trade-offs; and (5) establish accountability for equity and fairness in outcomes from health care algorithms. Conclusions and Relevance: Multiple stakeholders must partner to create systems, processes, regulations, incentives, standards, and policies to mitigate and prevent algorithmic bias. Reforms should implement guiding principles that support promotion of health and health care equity in all phases of the algorithm life cycle as well as transparency and explainability, authentic community engagement and ethical partnerships, explicit identification of fairness issues and trade-offs, and accountability for equity and fairness.

Innovative Home Visit Models Associated With Reductions In Costs, Hospitalizations, And Emergency Department Use.

While studies of home-based care delivered by teams led by primary care providers have shown cost savings, little is known about outcomes when practice-extender teams-that is, teams led by registered nurses or lay health workers-provide home visits with similar components (for example, care coordination and education). We evaluated findings from five models funded by Health Care Innovation Awards of the Centers for Medicare and Medicaid Services. Each model used a mix of different components to strengthen connections to primary care among fee-for-service Medicare beneficiaries with multiple chronic conditions; these connections included practice-extender home visits. Two models achieved significant reductions in Medicare expenditures, and three models reduced utilization in the form of emergency department visits, hospitalizations, or both for beneficiaries relative to comparators. These findings present a strong case for the potential value of home visits by practice-extender teams to reduce Medicare expenditures and service use in a particularly vulnerable and costly segment of the Medicare population.

Physician Engagement Strategies in Care Coordination: Findings from the Centers for Medicare & Medicaid Services' Health Care Innovation Awards Program.

OBJECTIVE: To identify roles physicians assumed as part of new health care delivery models and related strategies that facilitated physician engagement across 21 Health Care Innovation Award (HCIA) programs. DATA SOURCES: Site-level in-depth interviews, conducted from 2014 to 2015 (N = 672) with program staff, leadership, and partners (including 95 physicians) and direct observations. STUDY DESIGN: NORC conducted a mixed-method evaluation, including two rounds of qualitative data collected via site visits and telephone interviews. DATA COLLECTION/EXTRACTION METHODS: We used qualitative thematic coding for data from 21 programs actively engaging physicians as part of HCIA interventions. PRINCIPAL FINDINGS: Establishing physician champions and ensuring an innovation-values fit between physicians and programs, including the strategies programs employed, facilitated engagement. Among engagement practices identified in this study, tailoring team working styles to meet physician preferences and conducting physician outreach and education were the most common successful approaches. CONCLUSIONS: We describe engagement strategies derived from a diverse range of programs. Successful programs considered physicians' values and engagement as components of process and policy, rather than viewing them as exogenous factors affecting innovation adoption. These types of approaches enabled programs to accelerate acceptance of innovations within organizations.

Innovative Models for High-Risk Patients Use Care Coordination and Palliative Supports to Reduce End-of-life Utilization and Spending.

BACKGROUND AND OBJECTIVES: Care coordination and palliative care supports are associated with reduced anxiety, fewer hospital admissions, and improved quality of life for patients and their families. Early palliative care can result in savings in the end-of-life period, but there is limited evidence that larger-scale models can improve both utilization and the cost of care. Three models that received Health Care Innovation Awards from the Centers for Medicare & Medicaid Services aimed to improve quality of care and reduce cost through the use of innovative care coordination models. This study explores the total cost of care and selected utilization outcomes at the end-of-life for these innovative models, each of which enrolled adults with multiple chronic conditions and featured care coordination with advance care planning as a component of palliative care. These included a comprehensive at-home supportive care model for persons predicted to die within a year and two models offering advance care planning in nursing facilities and during care transitions. RESEARCH DESIGN AND METHODS: We used regression models to assess model impacts on costs and utilization for high-risk Medicare beneficiaries participating in the comprehensive supportive care model (N = 3,339) and the two care transition models (N = 587 and N = 277) who died during the study period (2013-2016), relative to a set of matched comparison patients. RESULTS: Comparing participants in each model who died during the study period to matched comparators, two of the three models were associated with significantly lower costs in the last 90 days of life ($2,122 and $4,606 per person), and the third model showed nonsignificant differences. Two of the three models encouraged early hospice entry in the last 30 days of life. For the comprehensive at-home supportive care model, we observed aggregate savings of nearly $19 million over the study period. One care transition model showed aggregate savings of over $500,000 during the same period. Potential drivers of these cost savings include improved patient safety, timeliness of care, and caregiver support. DISCUSSION AND IMPLICATIONS: Two of the three models achieved significant lower Medicare costs than a comparison group and the same two models also sustained their models beyond the Centers for Medicare & Medicaid Services award period. These findings show promise for achieving palliative care goals as part of care coordination innovation.