Reader's Guide

Beginning in November of 1992, the Pennsylvania Health Care Cost Containment Council has released a series of public reports about coronary artery bypass graft (CABG) surgery. These reports contained information about the risk-adjusted patient mortality statistics for Pennsylvania's cardiac surgeons (who perform CABG surgery) and the Pennsylvania hospitals where the procedures are performed. This report, which includes data from 1994 and 1995, is the latest in the series.

What's New in This Report?

This report contains two new additions: Information about the length of hospitalization after surgery and information about CABG patients according to selected health plans.

While Pennsylvania's CABG report has historically contained information about the risk-adjusted mortality numbers for CABG patients specific to Pennsylvania hospitals and cardiac surgeons, this report marks the first time that similar data for individual health plans is included.

Why Has Information About Health Plans Been Added to This Report?

The health care industry is experiencing enormous change. Part of this movement involves a shift in traditional roles, especially as it relates to the management of health care. Larger hospital and health care systems are emerging. The relationships among hospitals, physicians and insurers are changing.

Payors have evolved from the traditional approach of financing the delivery of health care to one of influencing, on an increasing basis, the organization of the delivery system. This takes the form of quality improvement efforts, re-certification, utilization management, promulgation of physician practice guidelines, development of select physician and hospital networks, and financial incentives - the increasing "management" of care. While it is important to remember that patients are not treated by payors, rather by doctors and hospitals, it is increasingly the case that payors influence, directly or indirectly, the delivery of care in today's market.

In late 1995, the Pennsylvania Health Care Cost Containment Council, through a series of strategic planning sessions, identified as its primary future role the development of information about the influence of managed care on health care cost and quality issues.

This report is a first step in the direction of measuring health plans in that it includes for the first time in a Council report risk-adjusted mortality and length of hospital stay data for enrollees in selected health plans that had CABG surgery. In doing so, it builds upon previous Council CABG reports that have included data about Pennsylvania hospitals where CABG operations have been performed, and the cardiac surgeons who have performed them. (This report also contains the hospital and surgeon-related risk-adjusted mortality and length of hospital stay data.)

As with any first-time initiative, these health plan data should be interpreted with caution. The data in this report should not be the only factor in any health care decision, particularly with regard to choosing a health plan. Many other factors should be weighed before making such a decision.

A Limited View of Managed Care

What Do We Mean by Managed Care?

It is important to note that this first step is a limited one. Managed care is a general term that means different things to different people. In its most restrictive sense the term can be applied to Health Maintenance Organizations (HMOs), which are included in this report. There is a broad range of health insurance plans in Pennsylvania, including many different types of employer or employee-determined managed care arrangements and levels of health care management, some of which will be found within traditional fee-for-service plans, others as options or variations within licensed HMOs.

For the purposes of this report, however, the Council reported only on those managed care plans that are licensed by the Pennsylvania Department of Health as Health Maintenance Organizations (HMOs). In addition, the report includes traditional fee for service plans, Medicare and Medicaid. Beyond these broad categories, the Council did not differentiate among the different insurance products available in Pennsylvania. To do so would involve significant additional data collection and verification among the hospitals, health plans, and the Council.

CABG Surgery Constitutes a Small Portion of Overall Health Plan Admissions

While the Council's ultimate goal is to provide an increasingly comprehensive picture of the system of care, this report focuses on only one procedure. Although a high-volume, high-cost procedure, CABG surgery generally represents a small portion of the overall health care utilization of health insurance plan enrollees. Therefore, a plan's results in this report should not be used to generalize about the overall quality of a health plan.

Times Have Changed - and so Has the Health Care Market

The data is this report are from 1994 and 1995, the most recent data available. The marketplace, especially with the growth of managed care plans, has changed dramatically since that time. The same categories examined today might show very different results. While managed care covers only a small percentage of CABG patients in 1994 and 1995, the Council will use this time period as a baseline for comparison with later years when the market penetration of managed care has grown.

Measuring Quality

The risk-adjusted mortality and risk-adjusted length of stay statistics included in this report are important measures of quality as well as resource utilization, but are not and cannot be considered the only measures of importance. Additional measures, such as whether CABG surgery was the appropriate treatment option, readmission rates, or the quality of patients' lives after receiving surgery are not captured here.

The measurement of quality is highly complex and the information used to capture such measures is limited. Hospital deaths are frequently an unavoidable consequence of a patient's medical condition. Hospitals and physicians may do everything right and still the patient can die. In addition, marked differences in health plan populations in terms of social, economic, and behavioral characteristics might put some groups at higher risk of mortality - factors not completely captured by the Council's risk-adjustment model.

However, after taking the significant patient risk factors available to the Council into account, differences with respect to patient mortality and lengths of stay do exist among hospitals, cardiac surgeons, and health plans.

Other Measures of Quality

HEDIS (Health Plan Employer Data and Information Set), a system for measuring health plan effectiveness developed by the National Committee for Quality Assurance, uses a number of measures related to utilization rates per 1,000 covered lives. Additionally, these rates include only enrollees with a sufficient record of enrollment, normally 12 to 36 months of continuous enrollment. This report does not account for differences in the length of enrollment among plans. In order to do so in future reports, the Council would need the cooperation of health insurers in providing enrollment data.

Overall patient satisfaction can be an important component in assessing quality. Several states such as Utah, Maryland and New Jersey have developed and published patient/consumer satisfaction surveys as measures of the quality of care received within health plans.

The Report is Not All-Inclusive

This report includes only inpatient hospital information, and only information about hospitals, surgeons and health plans with 30 or more cases in 1994 and 1995. While taking into account such issues as the race, ethnicity and urban/rural status of CABG patients, the report may not fully capture the impact of socioeconomic status on CABG patients' risk of mortality. It does not allow for trend analysis since the data cover only two years.

The Verification of Data for Health Insurers

The Council formed a new Payor Advisory Group to assist in developing ways to approach various issues related to the development of this report, in particular the verification of plan data. The verification of the plan data, which originates with the hospitals, was a two-step process.

First, the hospital where the procedure was performed verified the plan that paid for the surgery.

Secondly, all plans were given an opportunity to verify that the hospital identification of the CABG payor was correct. Many plans chose to do so and the Council would like to acknowledge those plans for their participation. However, some plans chose not to verify the hospital identification of the payor and in those instances, the hospital-only data is reported.

In some instances the plan with financial responsibility for a CABG case was not necessarily the plan that managed or oversaw the treatment of the same case. In these situations, the Council assigned the case to the plan that paid for the case. In situations where more than one plan was financially responsible for a CABG patient, the plan listed as the primary payor was assigned the case.

An Important Beginning

It's important to recognize that efforts to compare health plans are still in their infancy. This report is just a starting point; useful as a basis for identifying both similarities and differences among health plans, asking why they exist, and as a basis for further study.

The treatment of patients is a varied and complex process, one that involves many players. The Council would like to emphasize that this report is about pointing out differences in patient outcomes. It is about asking why those differences exist. It is about stimulating a dialogue among purchasers, consumers, providers and payors, and a quality improvement dynamic that will attempt to raise appropriate questions.

So, while this report represents a limited view of managed care, it is nevertheless an important step and can serve as a baseline for future reporting. Future reports can better serve the public with the inclusion of enrollment information, data which can only be provided by the health plans themselves and which can serve to overcome some of the limitations of this project. The report is a start towards helping Pennsylvanians examine the system of care involved in treating CABG surgery patients. It is a start in helping the Council examine additional areas of health care delivery and the growing interrelationships among hospitals, physicians, and payors in delivering quality outcomes for the patients and enrollees for which they provide, pay and manage care.

Heart Disease and Coronary Artery Bypass Graft Surgery in Pennsylvania

About 57 million Americans have some form of heart-related disease. It is the leading cause of death in the United States. This booklet is designed to provide the public withinformation on the surgical procedure used to treat one type of heart disease known as atherosclerotic coronary artery disease.

This booklet provides comparisons among hospitals and cardiac surgeons should surgical treatment for coronary artery disease be required. It provides information that can be used as a basis for asking questions and to make more informed choices when selecting a hospital or surgeon for coronary artery bypass graft (CABG) surgery. This information should be used in conjunction with a doctor, hospital and health plan representative.

The charts show the number of CABG surgery cases and the risk-adjusted patient mortality percentages for each hospital, health plan and cardiac surgeon in Pennsylvania where at least 30 coronary artery bypass graft operations were performed in 1994-1995. Thirty is considered by the Council to be a minimum number in order for the information to be statistically meaningful.

What is Atherosclerotic Coronary Artery Disease?

Atherosclerotic coronary artery disease occurs when the arteries which supply blood to the heart muscle become lined with fatty deposits, harden, and become partially blocked. The amount of blood reaching the heart is reduced. This reduced flow of blood can cause chest pain (angina) or a heart attack.

Which Methods Are Used to Treat Heart Disease?

It is important to discuss this with a physician. Depending on a patient's condition and the doctor's recommendation, the following are among treatment methods that might be used: changes in lifestyle habits such as diet or smoking, medication, balloon angioplasty, laser angioplasty, and coronary artery bypass graft surgery.

This report deals with coronary artery bypass graft operations, which are performed by a cardiac surgeon. However, when seeking treatment for heart disease, a cardiologist is usually involved in the diagnosis of heart disease. In general, it is the cardiologist who will diagnose the problem and refer the patient to a cardiac surgeon if surgery is being considered or recommended. This report can be used in conjunction with the advice of a cardiologist in selecting a cardiac surgeon.

What is Coronary Artery Bypass Graft Surgery?

An alternate path for blood, or "bypass," is made around the narrowed or blocked part of the coronary artery that has been identified during cardiac catheterization. The surgeon removes or diverts a section of blood vessel from another part of the body and attaches it around the blockage in the coronary artery. The blood flows through the inserted section, bypassing the blockage, to restore the blood flow to the heart muscle.

A Focus on Patient Outcomes

The primary focus of this report is on the outcome or result of CABG surgery. Although there are other ways to view a successful result of CABG surgery (lack of complications, improved quality of life, and recovery time), this report focuses on in-hospital mortality statistics and risk-adjusted lengths of hospitalization as measures of the outcome of CABG surgery. To be fair, the mortality data and length of stay data are adjusted to account for significant patient risk and severity of illness factors. The first section of this report includes risk-adjusted mortality statistics for CABG patients in 1994-1995 according to the health plan or program they belonged to, the hospital they were admitted to, and the cardiac surgeon of record who performed the coronary artery bypass graft surgery. The statistics are calculated and displayed in the same way for health plans, hospitals and cardiac surgeons. The health plans, hospitals and surgeons in this report were offered the opportunity to review the data applicable to them and attest to its accuracy.

About Figure 1 - Health Plans - Risk-Adjusted In-Hospital Mortality

There are 34 health plans (including Medicare and Medicaid) covered in this report. These include fee for service plans (Blue Cross and Commercial Insurance Companies), Pennsylvania-licensed Health Maintenance Organizations (HMOs), as well as the Medicare and Medicaid programs. However, only plans that had 30 or more CABG cases in 1994-1995 after exclusions are reported. (For a detailed list of exclusion criteria, see the Council's Research Methods and Results.

A risk-adjusted mortality percentage has been calculated for each plan and is presented in this section. The health plans are grouped according to the type of plan: Commercial fee for service plans together, Blue Cross-related fee for service plans together, Licensed HMOs, Medicare, and Medicaid.

About Figures 2A, 2B, 2C - Hospitals - Risk-Adjusted In-Hospital Mortality

Forty-four hospitals in Pennsylvania were approved to perform CABG surgery during 1994-1995. This report provides risk-adjusted mortality data, risk-adjusted lengths of hospital stay, and hospital charges for the 43 of those hospitals that had 30 or more cases. Passavant Hospital began its open-heart program in late 1995, and as such did not have enough cases to be reported. There are additional hospitals that have been set up to perform CABG surgery since 1995 but those hospitals do not appear in this report. In addition, hospital names may have changed since 1994-1995.

About Figures 3A, 3B, 3C - Cardiac Surgeons - Risk-Adjusted In-Hospital Mortality

Risk-adjusted mortality statistics are also reported for the CABG patients of Pennsylvania cardiac surgeons who performed CABG surgery during 1994-1995. There are surgeons who practiced at more than one hospital and that is noted in the report.

Figure 3 lists all the cardiac surgeons who practiced at a given hospital under that hospital name. The hospitals are sorted by region, then listed alphabetically, with each surgeon who practiced in that hospital also listed alphabetically. Many surgeons practiced at more than one hospital so they will be listed more than once. Only surgeons who treated 30 or more cases in 1994-1995 were rated; others are listed with the number of cases they performed.

About Figures 4 and 5A, 5B, 5C - Risk-Adjusted Post-Surgical Average Length of Stay According to Health Plans and Hospitals

In this report, the length of stay is post-surgical (hospital days prior to surgery are not counted) and has also been adjusted to take patient risk factors into account. The statistics are developed in the same way as the patient outcomes (mortality) section. An expected length of hospitalization is calculated and can be compared to the actual length of stay. These figures are expressed in number of days in the hospital. An asterisk (*) next to the hospital (or plan) name means that a hospital's actual length of stay was significantly greater than expected. An open bullet () next to the hospital (or plan) name means the length of stay was significantly less than expected. The absence of a symbol means that the actual length of hospital stay was within the expected range.

Length of stay data for each cardiac surgeon is reported in a separate Technical Appendix available from the Council upon request.

About Figure 6 - Average Hospital Charges for CABG Surgery

This column shows the average amount a hospital charged for coronary artery bypass graft surgery in 1994-1995. The charges do not include physician fees, and are usually more than actual payments received by hospitals from the payor (such as your insurance company). For example, for Medicare patients, the actual payment to a hospital is based on a different formula and may be lower. Charges are, however, a reasonable and consistent basis for comparison.

Charges vs. Revenues

The amount a hospital bills for a patient's care is known as the charge. What the hospital actually receives is known as revenue. This report lists the average charges billed by hospitals for a CABG surgery. The charges are derived from hospital billing forms, which list the actual charges for each patient. However, hospitals generally do not receive full reimbursement of their charges. Hospitals frequently negotiate discounts with insurance companies or other large purchasers of health care services. The amount collected by the hospital may differ substantially from the amount billed.

An analogy can be made to the purchase of an automobile. Each automobile has a manufacturer's suggested list price (the charge). But the amount the buyer actually pays depends upon his or her ability to negotiate a discount from that charge. Purchasers of fleet vehicles have greater clout in negotiating discounts than do the buyers of a single vehicle. In the same way, large group purchasers have greater purchasing power when buying insurance or negotiating health care discounts than do privately or self-insured individuals.

About Table 1

Table 1 contains for each plan a listing of the hospitals that account for approximately 75% of the plan's CABG patients and the number of cases for each plan in each of the hospitals.

About Figures 7A, 7B - The Issue of Volume

The number of procedures performed in hospitals or by cardiac surgeons is often considered an important factor in deciding upon treatment options. Figure 6A provides a view of the total number of bypass procedures as well as the number of total open-heart procedures performed by Pennsylvania hospitals. Figure 6B details the average number of open-heart procedures per surgeon in each Pennsylvania hospital.

Interpreting the Data Actual to Expected Patient Mortality (Death)

Under the expert guidance of its Technical Advisory Group, a committee of physicians and health researchers, the Council develops a complex methodology to measure expected mortality. First, the Council identifies a list of health factors that have an impact upon patients' risk of dying during or following (prior to hospital discharge) CABG surgery. In compiling this list, the Council examines the scientific literature, and solicits feedback from hospital staff and physicians.

The next step is to determine which risk factors (of those available to the Council) had a significant impact on the in-hospital mortality of those patients hospitalized for a coronary artery bypass in 1994-1995. The rating system gives a certain weight (or importance) to key health factors that influence in-hospital mortality for each patient receiving a coronary artery bypass operation in 1994-1995. These risk factors are taken into consideration to create a risk profile for each patient.

By looking at all the individual patient data together, the Council is able to calculate an expected mortality percentage for each hospital, health plan and cardiac surgeon. The statistics are adjusted for the higher or lower risk of the patients of each provider and health plan. This provides a fair basis for comparison. By adjusting for risk, hospitals, health plans and cardiac surgeons are given extra credit for having treated "sicker" patients or patients with more risk factors. The higher the risk, the more deaths to be expected.

The figures allow you to compare the actual number of mortalities with the expected number of mortalities. These are expressed as percentage points. The expected mortality is expressed as a range of percentages representing the lowest number of mortalities you could expect to the highest. The expected range is based on a calculation that takes into account

range for that calculation. The length of the bar is based on a combination of patient volume and diversity of patient risk. Generally, the more CABG patients or the more similar patient risk is across hospital, surgeon, or health plan, the smaller the bar will be.

How to Interpret the Graphs

If the point falls within the bar, it means that the difference between the actual mortality and the expected mortality was not statistically significant. If the point falls to the left of the bar, the actual mortality was significantly lower statistically than what was expected. This is highlighted by an open bullet () next to the hospital, health plan, or surgeon's name. If the point falls to the right, the actual mortality was significantly higher than the expected. This is highlighted by a single asterisk (*) next to the hospital, health plan or surgeon's name. A point that is statistically significant will always fall clearly outside the bar. Tables that provide the numbers that correspond to the graphs in the mortality section are contained in the Technical Appendix.

What Do We Mean by Statistical Significance?

Scientists use the term "statistical significance" to indicate when a measurement or calculation is certain enough to be caused by something other than chance or random variation. If the actual percent mortality falls outside the expected bar, we can conclude with 95% certainty that the difference between what was expected and what actually occurred was not because of chance or random variation. If the actual percent mortality falls inside the bar, the difference may have been due to chance or random variation.

Technical Appendix

The actual figures related to the graphs presented in this report are contained in a separate Technical Appendix. The appendix also contains surgeon length of hospital stay data and a list of patient risk factors.