Monitoring Quality of Care and Improving Service Delivery

The quality of the nation's health care varies widely across populations, geographic areas, providers, and health insurance arrangements. The current system's inability to apply evidence-based medicine consistently and avoid medical mistakes contributes to hundreds of thousands of preventable illnesses and deaths and costs billions of dollars annually. Mathematica's quality-of-care research probes variations in quality in a variety of health care settings and professions and provides information to support health care improvement in the public and private sectors.

We have been a leader in studying pay for performance, public reporting of quality data, disease management, and use of technology to monitor and improve care—four of the most promising strategies available to policymakers and private health care leaders to encourage quality improvement. Sound quality measurement is central to these strategies, and we have worked on the development, analysis, and evaluation of quality measures, including claims-based, medical-chart-based, and patient-reported measures. Finally, we have a rich history of research aimed at assessing and improving quality for low-income populations.

We also design and implement quality-of-care studies, drawing from our rich toolbox of health research methods, including major random assignment studies, analysis of large claims databases, collection of clinical data from multiple sites, site visits, and many kinds of surveys.

Pay for Performance

As policymakers and health plans search for ways to align payment with quality, Mathematica has been conducting case studies and quick-turnaround discussions with providers and health plans to offer timely information about pay for performance in the Medicaid and Medicare programs. For example, a report for the Center for Health Care Strategies summarized lessons from Medicaid health plans' experiences with using financial and nonfinancial incentives with physicians to improve quality. In addition, we briefed Centers for Medicare & Medicaid Services (CMS) officials on experiences of 30 hospitals participating in the Premier Quality Incentive Demonstration and changes they made in their quality improvement programs. A project for the California HealthCare Foundation is studying the development and implementation of the largest pay-for-quality initiative implemented at this time, the pay-for-performance project run by the Integrated Healthcare Association Center for Health Care Strategies. It involves six health plans and over 35,000 physicians. Our major ongoing evaluation of a pay-for-performance demonstration involves an impact analysis as well as a qualitative analysis of implementation.

Public Reporting of Quality Data

Consumers and purchasers who can review information about the quality of care provided by various health plans may make better choices. This could, in turn, drive the entire marketplace toward improved quality. Hospitals recently began public reporting of quality data for the first time under the Medicare Modernization Act. Mathematica conducted some of the first research on this issue, based on discussions with hospitals about how they have reacted to this reporting—including the level of effort required and the changes they've made to their quality programs. Another project for the California HealthCare Foundation tackles the issue of public reporting of quality data across several different types of health care settings, including health plans and nursing homes. The National Healthcare Quality Report and National Healthcare Disparities Reports are critical national-level report cards on quality issued on a regular basis by the Agency for Healthcare Research and Quality (AHRQ). We assisted AHRQ in evaluating its process for developing these reports and thinking through future strategic plans.

Disease Management

Disease management has risen in prominence among health care leaders because it holds promise for increasing quality of care and improving outcomes while simultaneously reducing costs. Purchasers, health plans, providers, and consumers all hope to benefit, but evidence of the best techniques, target populations, effectiveness, and return on investment has lagged behind the concept's popularity. Mathematica is evaluating two major demonstrations of disease management interventions in the elderly—one designed to improve the health care and outcomes of Medicare beneficiaries with heart failure, coronary heart disease, or diabetes, and the other aimed at Medicare beneficiaries with chronic illnesses in the fee-for-service program. An issue brief based on this work uses lessons from Medicare to assist state Medicaid programs in designing effective disease management programs. In 2005, Mathematica began assisting CMS in implementing and developing performance measures for its Medicare Health Support (MHS) program. MHS is a population-based disease management program that is expected to become the national disease management component of Medicare. In addition, staff have authored papers on the evolution of health plans' disease management programs based on site visits to 12 communities with our research affiliate, the Center for Studying Health System Change, for its Community Tracking Study.

Health Information Technologies

Health information technologies hold promise for improving quality and consistency of care, but they have been adopted slowly and cautiously by practitioners. We are evaluating a major demonstration that provides financial incentives to small physician practices for adoption of these technologies to determine whether the incentives can drive use and positively affect patient outcomes for Medicare beneficiaries with certain chronic diseases. In another project, we are evaluating whether the use of telemedicine services for Medicare patients with diabetes randomly assigned to receive these services leads to improved outcomes.

Development, Analysis, and Evaluation of Quality Measures

Sound quality measurement is the foundation of quality improvement. Despite the rapid evolution of this area and increased availability of quality measures, quality measurement still faces formidable challenges. One important issue involves how to make the most of administrative information, such as claims data, since obtaining information from medical records is costly. For CMS, we developed quality and patient safety indicators based on administrative data to serve as a surveillance system for quality monitoring. Another challenge in quality measurement is helping funders effectively target their efforts. We are helping AHRQ and its collaborators in CAHPS^® (formerly the Consumer Assessment of Health Plans) to take stock of what they have accomplished, and to identify strengths and weaknesses, future priorities, and models for structuring a public-private partnership to finance and move toward priorities in the future. We are also working with our subcontractors, the American Medical Association and the National Committee for Quality Assurance, to develop performance measures for specialty physicians. CMS intends to use these measures as the basis for its pay-for-performance program for physicians.

Quality for Low-Income Populations

Our research has always focused on vulnerable populations, and our quality-of-care work is no exception. Assessing quality in Medicaid managed care has been an area of particular emphasis. In one recent study, we identified ways states can be doing more within federal regulations to assess the quality of preventive and developmental services for children, using their External Quality Improvement Organization contractors to better advantage. Another study examined the effectiveness of convening health plan workgroups to generate and sustain changes in quality.

Understanding what quality programs are in place and identifying patterns in health plan performance have also challenged policymakers. As part of a larger survey of Medicaid managed care plans, we asked plans about their condition-specific quality improvement programs and described the national picture, including differences between Medicaid-dominated and commercial plans. We also analyzed and updated how Medicaid HEDIS effectiveness-of-care and access scores vary by type of health plan. Our SCHIP evaluation is analyzing HEDIS data to assist states with better data collection and reporting of key measures.

Quality of care for safety net providers has been another important area of research. One of our recent studies found disease management by a clinical pharmacist to be a potentially effective tool for improving outcomes for community health center patients with diabetes.

We have also analyzed diabetes care and care for people with hypertension who visit community health centers, using the large-scale Community Health Center User/Visit Surveys along with the National Health Interview Survey and National Hospital Ambulatory Medical Care Surveys.