MCPR 2000 - Measuring the Quality of Pennsylvania's HMOs - About The Data


Sources of Data

Inpatient hospital data used in the analysis of treatment measures were submitted to PHC4 by Pennsylvania hospitals. Information included in the Plan Profiles was provided by the Pennsylvania Department of Health. The Pennsylvania Insurance Department provided financial information used to calculate the financial indicators for each HMO. The National Committee for Quality Assurance (NCQA), an independent organization that reports information about managed care plans, was the source of the Health Plan Employer Data and Information Set ® (HEDIS). HEDIS provides the specific prevention measures included in this report. The member satisfaction measures were taken from the Consumer Assessment of Health Plans Survey ® (CAHPS).

Limitations of the Data

This report is not intended to be a sole source of information in making choices about HMO plans since the measures included are important, but limited, indicators of quality. Hospital admissions, complications and rehospitalizations are sometimes unavoidable consequences of a patient's medical condition. Hospitals, physicians and health insurance plans may do everything right and still the patient may experience other problems.

In addition, an HMO's success in helping members to manage health problems depends in part upon members' willingness and ability to comply with their providers' treatment decisions. While HMOs play an important role in the delivery of care, it is hospitals and doctors who ultimately provide health care for patients.

This report may not provide exact comparisons for several reasons. Benefit plan designs differ among and within HMOs. Enrollment in HMOs is constantly changing. Furthermore, since this report includes data from only one year, it is only a snapshot of what occurred during a limited period of time. Finally, the Council's risk-adjustment model may not completely capture some groups at higher risk due to social, economic, and behavioral differences.

All HMOs included in this report verified that they were the primary insurer for the hospitalizations analyzed in this report.

Because the methods to compare health plans are not yet well developed, this report addresses a limited number of indicators that are not intended to represent an HMO's overall performance. As with any new initiative, these data should be interpreted with caution.

PHC4 would like to emphasize that this report is about helping people make more informed choices and stimulating a quality improvement process where differences in important health care measures are identified and appropriate questions are raised and answered.

Accounting for Differences in Illness Level, Age and Sex Across HMOs

PHC4 compiles "expected" rates for many of the measures in this report based on a complex mathematical formula that assesses the degree of illness or risk for patients. In other words, HMOs that have sicker members or a higher percentage of high-risk members are given "credit" in the formula; more patients can be expected to be admitted to the hospital, have longer lengths of stay, or have greater potential for complications because they are more seriously ill or at greater risk.

Age and sex adjustments are also applied. For example, a particular HMO that has a higher proportion of older patients in comparison to other HMOs will also have a higher expected hospitalization rate. PHC4's system "expects" more health problems in HMOs with older populations and makes adjustments for that expectation to allow for fair comparisons across the HMOs.

A comprehensive description of these and other issues can be found in the Technical Report available in hard copy and on the PHC4 website http://www.phc4.org.