The Dartmouth Atlas of Health Care in Michigan is the sixteenth volume in the Dartmouth Atlas series developed by John E. Wennberg, MD, MPH. Dr. Wennberg has an international reputation for his work in studying regional variation in the use of health care services. Previous atlas volumes have been based on Medicare data, and therefore focused on inpatient services for the elderly population. BCBSM’s claims data base is one of the largest in the nation, and we have made a commitment to turn our raw data into useful information to improve the quality of health care throughout the state. Development of the atlas provides an unprecedented opportunity to examine variation in the use of a wide range of inpatient, outpatient and pharmacy services by the population under age 65.

The atlas provides a picture of where health care services are used in Michigan and gives us a basis for asking questions about how and why particular patterns of services exist. BCBSM views The Dartmouth Atlas of Health Care in Michigan as the first step in an ongoing dialogue with providers and communities about how to enhance the quality of health care throughout the state. Back to Top

The Dartmouth Atlas uses BCBSM membership and claims as its primary data source. BCBSM provided Dr. Wennberg with membership and claims data for one calendar year (1997). All claims and membership information was provided in a manner that protects individual confidentiality while allowing for analysis of the data.

In some cases, BCBSM data did not provide enough information for a thorough analysis of a particular service. Dr. Wennberg used Medicare data, either instead of or in addition to, BCBSM data in those cases. Chapters dealing with the supply of health care resources draw from other data sources, including the American Hospital Association, American Medical Association and American Osteopathic Association. The atlas includes an Appendix on Methods, which describes the data sources in considerable detail. These chapters will be added to the web site during the course of the summer, and will be included in the complete print version. Back to Top

BCBSM is strongly committed to protecting the confidentiality of our members’ health care data. All data transfers were structured to protect individual privacy in several ways:

For inpatient and outpatient procedures, the atlas presents rates of service use per thousand BCBSM members. Certain sections of the atlas specify a more targeted population, generally a rate per thousand adult or child members. The rates refer to services per thousand individuals during calendar year 1997.

The section on prescription drugs presents information in a somewhat different format. The denominator includes only BCBSM members with prescription drug coverage. Also, because members often receive more than one prescription for a particular drug, rates refer to members (per thousand members with drug coverage) who received at least one prescription during 1997. Back to Top

Yes. The data were adjusted for age and sex. Back to Top

Hospital Service Areas (HSAs) and Hospital Referral Regions (HRRs) are the geographic units of analysis used in the atlas. They represent local markets for health care. Each HSA includes at least one acute care hospital and all of the zip codes in which a plurality of residents obtains most of their care at that hospital. Similarly, each HRR includes at least one hospital that performs major cardiovascular procedures and neurosurgery and all of the HSAs in which most residents obtain these services at that hospital. It is important to note that rates of health service use presented in the atlas are based on the HSAs or HRRs in which people live, not those in which they obtain care. The Appendix on Methods in the atlas will provide more detail about HSAs and HRRs. Back to Top

The rates presented in the atlas are based on the HSA or HRR in which a patient lives, not where he or she obtains health care services. Most people obtain health care services close to home, and the use of HSAs and HRRs (instead of cities or counties, for example) gives an indication of how these local markets operate. However, the atlas provides an analysis of how health care use varies among populations, rather than among health care providers. It does not provide information about individual hospitals or physicians. Back to Top

The Dartmouth Atlas shows that some services, such as treatment for hip fractures, show little regional variation, while others, such as rates of hysterectomy or Ritalin use vary widely from place to place. There is nothing inherently wrong with variation in the use of health care services. In cases where patients choose among several treatments with similar outcomes, variation is a positive reflection of those choices. In other cases, however, variation indicates shortcomings within the current health care system – for example, problems in access to services, a lack of scientific understanding about treatment options, or inadequate education of providers and patients.

The Dartmouth Atlas of Health Care in Michigan does not explain why variation exists, and it does not differentiate between "good" and "bad" variation. But by pointing out where health care use varies, it raises important questions about health care in Michigan communities. Back to Top

The atlas measures use rates against an average rate for the state as a whole. This statewide average is not a benchmark rate. Benchmarks do exist for some populations and services (e.g., use of beta blockers after heart attacks) and the atlas refers to them where applicable. There are no current benchmarks for most services examined in the atlas, however. BCBSM believes that these "right rates" will have to be determined through further study and discussion at the community level. Back to Top

All health care services show some degree of variation, but the extent of variation differs greatly among health care services. In some cases – treatment of hip fracture, for example – diagnosis is straightforward and treatment options are limited. Variation is typically narrow for such services. Variation is wider in situations with a wider range of options. Diagnosis may be difficult or time consuming, there may not be a clear point at which treatment should begin, or several treatment options may provide similar outcomes. The atlas includes services that fall into all of these categories. Back to Top

Virtually all HSAs and HRRs in Michigan have use rates well above or below state averages for certain services. For some categories of services – e.g., hospitalizations for medical conditions – HSAs with above-average rates in one service tend to have above-average rates in others, and vice versa. For other categories, notably surgical procedures, each HSA has its own "signature" that includes higher-than-average rates for some services and lower-than-average rates for others. Back to Top

The atlas will be released in sections at bcbsm.com/atlas, beginning in April 2000 and continuing throughout the summer. The final electronic installments will be available in early autumn, at which time a complete print version of the atlas will be available for purchase. Back to Top

The Blues commissioned the Dartmouth Atlas of Health Care in Michigan as a way to open a statewide discussion about the quality of health care. The atlas findings will be shared with communities throughout the state in hopes of developing a better understanding of variation in the use of health care services and working with communities to enhance the quality of health care. The Blues plan to accomplish this over the next several months through use of this Web site and personal presentations.

These initial discussions will be followed by in-depth dialogue and improvement activities in several Michigan communities. The Blues will work collaboratively with physicians, hospitals, employers, labor unions and consumers in this ongoing effort. Back to Top