What You'll Find in This Year's Report

For the fourth straight year, Michigan hospitals are voluntarily releasing valuable information to patients and purchasers of health care. Thousands of people have requested copies of our first three reports or accessed them through the Internet, so we know interest is high and that people want to learn more about health care. The MHA and its members are pleased that so many people are using the reports to make more informed health care decisions. This report affirms that Michigan hospitals continue to be accountable to patients and purchasers of health care.

Like the previous reports, this report shows if a hospital is performing as would be expected based on the hospital's unique patient population and compared to statewide averages. (See pages 23-152.) Measuring a hospital's performance against its expected performance is an excellent way to help the hospital improve. In addition, the report shows each hospital's mortality rates and lengths of stay for common medical and surgical cases, as well as important statistics from the Michigan Department of Community Health about the overall health of Michigan's citizens. This is important consumer information because people who smoke, are overweight, or have more health problems will respond differently to health care than people who are relatively more healthy. Compared to the rest of the nation, Michigan ranks relatively poorly in health status, which means more health care services are needed in our state.

Like the third report, the 1999 edition contains information on hospitals' community benefits - that is, how hospitals serve and better their local communities over and above providing traditional health care. (See pages 15-22.) In fact, measuring what hospitals do for local communities is one of the most important and far-reaching projects ever undertaken by the MHA. Finally, the report also includes financial information. The financial section lists what Medicare pays for the five most frequent reasons Medicare patients are hospitalized, along with other financial indicators. (See pages 153-170.) By including financial information, this report allows the public to begin learning the challenges Michigan hospitals must face in providing services to everyone in the community - including those who can't afford to pay.

About the Hospital Performance Data

The Michigan Inpatient Data Base (MIDB), which contains information from each of the state's acute-care hospitals, was used to create the hospital performance data section. Not all hospitals appear in all parts, either because they do not offer the services noted, have too few cases to be statistically valid, or have elected not to participate in this effort. This is a voluntary initiative focusing only on hospital inpatient data.

Important Notes on the Performance Section
Feedback from people who used the previous releases of the Michigan Hospital Report has been overwhelmingly positive. Suggestions for improvement were received from MHA members, consumers and others. Here are some of those improvements:

• Selected information on the general health status of Michigan citizens has been included in this report because it is valuable consumer information. It's also included because the relative health of a population helps determine how effective health care services will be. In addition, when appropriate, information on selected health status measures that may contribute to health problems (i.e., smoking and heart disease) precedes the appropriate indicator. This information was obtained from the Michigan Department of Community Health's Division for Vital Records and Health Statistics.
• The actual vs. expected range of data for 1997 is displayed graphically for easier use by readers.
• Hospitals are listed alphabetically by region for the whole state. A map of the regions in the state is located at the front of each section.
• Appendices with information about the growth of managed care and how to actively participate in choosing appropriate health care have been included.

How Performance was Determined

This report shows how hospitals performed on eight types of cases:

• groups of selected medical cases
• groups of selected surgical cases
• selected obstetrical rates
• nonsurgical heart care
• coronary artery bypass grafts (CABGs)
• heart valve repair
• total hip replacement
• total knee replacement

The report shows if each hospital performed as well as it should have performed, based on that hospital's patient population. For all measures (except obstetrics and joint replacements), hospital performance was measured on two factors - length of stay and mortality rates.

The selected obstetrical measures are cesarean section (C-section) rates and the rate of vaginal births after C-section (VBACs), which are measured against unadjusted statewide averages. For joint replacement, mortality is extremely rare, so only length of stay is analyzed.

Where the Data Came From
The data, voluntarily provided to the MHA, represent hospital patient discharges in 1997. From the data, one can determine how long a patient stayed in the hospital, the patient's age and gender, and medical conditions that may influence how well a patient responded to treatment. Patient confidentiality is protected.

Data Grouping
Some hospitals and health systems did not perform enough cases of any one procedure or diagnosis to be included in the report. By grouping the procedures or diagnoses, we were able to include information from most hospitals in the state and get a better picture of how hospitals performed in the selected areas.

SELECTION OF INDICATORS

The indicators in this report were chosen because they represent some of the most frequent reasons for being admitted to a hospital.

General Indicators: Group of Selected Medical and Surgical Cases
These indicators provide a general picture of each hospital's performance on a group of selected medical and surgical cases.

• The group of selected medical cases includes stroke, pneumonia, chronic obstructive pulmonary disease, and gastrointestinal bleeding.
• The group of selected surgical cases includes lung surgery, lower bowel surgery, vascular surgery, spine surgery, prostate surgery, and hysterectomy.

The report provides a picture of each hospital's performance on the group of selected medical and surgical cases in the aggregate. For example, the length of stay for the group of medical cases does not reflect how successful a hospital was in treating stroke, pneumonia, or any other individual diagnosis. In the same respect, the mortality rate for the group of surgical cases does not necessarily reflect the mortality rate for any individual procedure, such as prostate surgery.

Selected Obstetrics: C-section and VBAC Rates
This section examines selected procedures performed during the delivery of babies, including C-sections, which involve surgery, and VBACs, which are normal deliveries after a previous C-section. The rates for these procedures are generally accepted performance measures in obstetrics. However, as with other measures, these rates alone cannot be used to judge the overall quality of care. Generally, a lower C-section rate is preferred, because fewer mothers are exposed to surgical procedures that may put them at higher risk for complications. Generally, a higher VBAC rate is preferred because fewer mothers are exposed to complications that may be associated with surgery.

The obstetrics level of service is also reported in this section. Hospitals are classified into three levels.

• Level 1 hospitals have all the capabilities for normal births and births with minor complications.
• Level 2 hospitals have additional equipment and staff to deal with more complicated deliveries.
• Level 3 hospitals have equipment and staff to handle very complicated cases, such as premature births requiring intensive care.

Cardiac Care
This section examines the performance of hospitals that provide heart care. It evaluates indicators related to nonsurgical heart care (angina, heart attacks, and heart failure and shock) and open-heart surgery, which includes coronary artery bypass grafts and valve repairs.

The indicator for nonsurgical heart care includes patients who were admitted for heart problems but did not need heart surgery. Not all hospitals have data in the CABG and valve repair sections because open-heart surgery is performed at relatively few hospitals.

Joint Replacements: Total Hip and Total Knee Replacements
This section shows if hospitals performed as expected in the area of joint replacement. The data reflect patients who were discharged for either total hip or total knee replacement after deciding with their physicians that this procedure would benefit them. These are nonemergency elective operations. Patients who have had previous joint replacements and needed a second operation are not included. Patients who required emergency joint replacements are also not included.

HOW AND WHY THE INFORMATION IS ADJUSTED

As noted before, the data in this report are risk and severity adjusted. This is a standard statistical method to help researchers more accurately examine different populations. This method enables the report to take into account each hospital's specific patient population. It follows that mortality rates should be higher and lengths of stay longer in those hospitals that tend to treat sicker patients.

What is an Expected Range?
The expected range is a prediction of the actual performance, based on statewide averages. These ranges involve complex statistical calculations and were computed using a proprietary software product from the Sachs Group called The Quality Planner™. For this report, the data were severity adjusted using 3M's™ All Patient Refined Diagnosis Related Groups.

What is Severity of Illness?
The same care, given by the same physician, in the same hospital, might have very different effects on a patient who is healthier than another patient. For example, a patient requiring a heart valve repair who also has an infection resistant to antibiotics is more severely ill entering surgery than a valve repair patient who is otherwise healthy. The severely ill patient may not respond as well to treatment or surgery and, therefore, may have to stay in the hospital longer or may not recover at all.

What is Risk?
Risk includes health-related circumstances specific to an individual that affect how the patient responds to treatment or care. For example, a patient who has cancer may not recover from pneumonia, regardless of the treatment provided; or a woman with severe diabetes may have a higher-risk pregnancy requiring a C-section rather than a normal delivery.

Where Does the Risk and Severity Information Come From?
The medical record created for a patient who enters the hospital contains information on all of the patient's conditions. The information is used to assess how ill the patient was or what risk factors influenced the patient's ability to respond to treatment.

Mortality Rates and Quality of Care
Experts in quality measurement have strong reservations about using severity-adjusted mortality rates as an absolute gauge of the quality of care provided by a hospital, because many other factors also must be considered. For example, a hospital treating more elderly patients might have more patients with orders not to resuscitate and would have more deaths. Other factors include patient risk factors, lifestyle choices and socioeconomic issues. Experts say mortality is just one indicator of hospital performance. It cannot be used as the primary measure of the overall quality of care provided by a hospital.

Average Length of Stay
The average time a patient needs to stay in the hospital is best represented by an expected range. This report shows whether a hospital's average length of stay is within an expected range, given the hospital's specific patient population. Length of stay is considered an indirect indicator of efficiency.

A Note About Patient Volume
The number of cases treated is not an indicator of the quality of care you might receive in a hospital. Volume does, however, help you understand the size of the program in the hospital. The greater the number of cases a hospital performs, the narrower its expected range. For example, if you flipped a coin 10 times, you would expect heads four, five, or six times - that's an expected range of 40 percent to 60 percent. However, if you flipped the coin 10,000 times, it would be extremely rare to get 4,000 or 6,000 heads. A more reasonable expected range might be 49 percent to 51 percent.

Hospital Comments
Because performance is affected by many variables, some hospitals have submitted comments to help consumers and purchasers better understand their performance. Hospital comments, indicated by footnotes, are found on the same pages where the data appear.

Why Trends are Important
Because statistical processes are not perfect, and because hospitals and health systems are always changing, consumers and purchasers should not make judgments based only on the results in this report. Information in this report will be most useful when considered over time and when analyzed for trends in subsequent years. If a hospital performs consistently above or below its expected range, this is an indication that there might be something unique about this facility. For example, it may reflect a difference in philosophy in the way patients are treated or be dependent on the level of specialists. What About Other Procedures and Treatments at the Hospital? This report offers no conclusions about the performance of hospitals other than on the eight types of cases specifically measured.

ABOUT HEALTH STATUS AND HEALTH CARE

The health of an individual is influenced by many factors. According to the US Surgeon General, 20 percent of health is determined by genetics, 20 percent by the environment, 10 percent by the medical system, and 50 percent by lifestyle choices such as physical activity, nutrition, tobacco use, etc. For any given person, when all of these factors are examined together, health professionals can get a fairly good picture of the overall health of the individual and if the person is at risk for developing future health problems. This is called an individual's "health status." In the same way, when a group's or population's health status is examined collectively, health professionals can identify the potential health problems of that group or population.

An individual's lifestyle choices can influence the number and types of health problems encountered throughout life. A person's health status also offers clues about his or her ability to recover from illness following surgery. For instance, an overweight person who smokes might be more predisposed to developing heart disease and have a more difficult time recovering from coronary artery bypass graft surgery. Thus, a person's health status indirectly influences how long he or she may need to be hospitalized (length of stay) or if they recover at all (mortality rate).

Each year, the Michigan Department of Community Health (MDCH) publishes data on the overall health of Michigan citizens. Health indicators measured by the MDCH disclose important facts about the health of Michigan's population, and there are many areas where improvements must be made. The MDCH data show there is a gap in health outcomes for minority groups as compared to the population as a whole.

African Americans in Michigan have a lower survival experience for heart disease, cancer, stroke, and diabetes than the general population. A high percentage of Michigan's population is overweight. Smoking is the leading cause of preventable deaths; yet nearly 26 percent of our state's population still smokes.

All of these health status factors can influence how often a person seeks medical care, requires surgery, and how he or she recovers to good health.

When appropriate, health status information is included at the front of each indicator section in this report to give consumers and purchasers a better idea of how the overall health of the population might influence how hospitals perform.

WHERE FROM HERE?

The Michigan Health & Hospital Association is committed to providing consumers and purchasers with information to help them make better-informed health care decisions. Reports produced in future years will continue to benefit the public by tracking how hospitals have performed and improved.

But you can't improve the entire health care system by improving just one of its parts. Data from physicians, insurance companies, employers, and the rest of the health care system would help achieve the very significant goal of improving health care delivery for all Michigan citizens. The MHA encourages other health care providers to include their data in future or separate reports, as Michigan nursing homes did earlier this year.