Center for Studying Health System Change

Introduction

he U.S. health care system is changing rapidly. Purchasers’ drive to control health care costs, new technologies for managing care, policy changes and competitive forces have led to significant changes in the health care market. Consolidation, vertical integration, the growth of managed care and the expansion of national, for-profit health care companies are central features of the changes. Although anecdotal information abounds, little systematic information on changes in the health system and their effects on people is available.

Because these changes are occurring locally, market by market, obtaining information requires studying individual local markets. The case studies of 12 communities in this volume are intended to provide an in-depth understanding of and insight into the organization and functioning of the health system in each area based on interviews with local leaders in the health system. The primary questions of interest are:

• What local forces are driving change in the health system? What roles do public policy, purchaser pressure and community culture and leadership play in driving health system change?
• How are insurers and health plans, physicians and hospitals responding to these changes? How are the ships between plans and providers changing? How are the relationships between hospitals and physicians being affected? What is the impact on providers of care for the poor?

The Community Tracking Study

The case studies are part of a larger longitudinal study, the Community Tracking Study.¹ The goal of the study is to provide an information and analytic base to monitor and understand the changes taking place in the health system and the effects of those changes on people. The study focuses on changes in the health care system in 60 sites that are representative of the nation. An initial round of surveys of households, physicians and employers was conducted in all 60 sites in 1996-97, and subsequent rounds are planned at two-year intervals. Twelve of the 60 communities are being studied more intensively through site visits and larger sample sizes in the surveys to produce site-specific information. The analysis of all 60 sites will permit generalization to the nation as a whole. Combining the findings from the case studies in the 12 markets with data from surveys will provide a unique and comprehensive view of health system change and how those changes are affecting people.

In addition to documenting and analyzing health system change in each community, the case studies play an important role in the broader study. First, they provide information for comparing and contrasting the 12 communities to analyze how and why health systems are changing. Cross-site analyses will examine the forces driving health system change, such as the role that purchasers, public policy and community culture play in how local health systems evolve, and the response of the health system to those forces, including how insurers and providers are reacting and adapting, and the effects on providers that traditionally serve the poor. Second, the case studies provide the baseline for future tracking of health system change in these markets. Follow-up site visits at two-year intervals will track the dynamics of change.

The Communities Studied

Much of the existing information on health system change is available on few specific markets, such as Minneapolis/St. Paul or areas of California. In order to obtain a broader picture of health system change in the United States generally, communities were selected randomly for this study. By doing so, we can be confident that, taken together, the 12 case studies are broadly representative of health system change in the nation.²

The resulting set of communities, where site visits were conducted between May 1996 and April 1997, are:

• Boston -- September 9-12, 1996
• Cleveland -- June 24-28, 1996
• Greenville, S.C. -- January 27-29, 1997
• Indianapolis -- November 4-7, 1996
• Lansing -- October 1-3, 1996
• Little Rock -- October 9-11, 1996
• Miami -- December 9-13, 1996
• Newark, N.J. -- October 28-31, 1996, and April 7-10, 1997
• Orange County, Calif. -- January 21-24, 1997
• Phoenix -- December 9-13, 1996
• Seattle -- September 30-October 4, 1996
• Syracuse -- May 20-24, 1996

Recent events were identified through follow-up phone calls to the sites.

Site Visit Methods

Site visits were conducted by four-person teams, each of which conducted at least two visits. Visits lasted three days for smaller sites (less than one million population) and four days for the larger sites (greater than one million population). This allowed approximately 36-48 interviews in smaller markets and 48-60 interviews in larger ones. The majority of interviews were conducted on-site, but some interviews may have been conducted after the site visit to contact individuals or organizations identified during the course of the actual visit.

Interviews were conducted with insurers and health plans, hospitals, physician groups, employers and other purchasers, community health centers, health departments, community agencies, policy groups, consumer advocates, community leaders and other actors important to the local market. To select specific respondents, the Center developed overall allocation targets to guide the number of interviews to be conducted within each category of respondents, as follows: 25 percent insurers/health plans; 20 percent providers; 10 percent providers traditionally serving the poor; 10 percent providers in the outlying areas of the community; 10 percent employers and other purchasers; 10 percent community leaders and consumer groups; 10 percent public health and other public agencies; and 5 percent miscellaneous. In insurers/health plans and provider organizations, attempts were made to interview an administrative and a clinical leader. Criteria also were developed to guide selection from among multiple respondents within a single category (e.g., to get a mix of large and small organizations, not-for-profit and for-profit organizations, free-standing organizations and health systems, etc.).

Prior to the visit, a selected number of telephone interviews were conducted with some of the more complex organizations that were to be interviewed while on site to obtain basic information about the organizational structure and services or product offerings. This allowed the interviewer to prepare adequately for the meeting and to make best use of the available time for the interview. A briefing book was developed for each site for use by the team prior to the visit. This briefing book included available data on the site to orient the team to the geography of the area, population characteristics and health status, major employers and employment patterns, local health plans and providers and relevant state policy.

The interview protocol for the visits was developed to reflect analytic priorities. Selected topics for analyses were identified in advance and provided guidance on the questions to be asked of respondents. Interview modules were developed that typically contained 5-10 questions on a specific topic. The topics covered included purchaser decisions, influence of policy, social responsibility and care of the poor, community decision making, influence of information, organizational change and competition, financial and administrative integration, clinical integration and care management, health plan design and rationale and the perceived impact of health system change on people.

For each respondent, an appropriate number and mix of modules was selected to guide the interview. This approach permitted a level of standardization across many different interviewers that strengthens the reliability of the information obtained, but still permits flexibility for each individual respondent. The information obtained during the interviews was recorded directly onto the module immediately after the visit and was entered into a text database. The database provided a reference for writing the case study reports and the cross-site analyses and also will provide the starting point for subsequent site visits.

The Case Studies in This Volume

The case studies describing each local market are collected in this volume. The first part of each study describes the roles of public policy, employers and other purchasers and community leadership and decision making and analyzes how these forces influence the pace and direction of health system change. The second part describes how the organizations in the health system -- health plans, hospitals and physician organizations -- are responding to those forces through horizontal and vertical integration and whether those organizational changes appear to be having an effect on clinical practice and the delivery of care. Special attention is given to that part of the health system that traditionally serves the poor. Each report concludes with a set of issues to be tracked and analyzed over time.

The case studies will form the basis of a systematic analysis of the similarities and differences across communities. It is clear from even a casual reading of the case studies, however, that although communities face many common forces for change, the effects of those forces vary based on local conditions and the history of the local health system. Whether or not the markets converge over time and the local health systems become more similar or continue to exhibit great variation remains to be seen. Tracking these changes in the future is the long-term goal of the Community Tracking Study.

NOTES

1. A complete description of the study design may be found in P. Kemper et al., "The Design of the Community Tracking Study: A Longitudinal Study of Health System Changes and Its Effects on People," Inquiry 33:195-206 (Summer 1996).

2. Technically, the study sites were selected by stratifying all areas of the continental United States by region, population size and whether they were metropolitan or non-metropolitan. Sixty sites were randomly selected with probability in proportion to the population. The 12 high-intensity sites were then selected randomly from those of the 60 sites that were metropolitan areas with more than 200,000 people.