A Report from Families USA
May 1998

Monitoring Medicare HMOs:
A Guide to Collecting and Interpreting Available Data

Table of Contents

Introduction

HCFA Performance Monitoring Reports

Disenrollment Data

Utilization Data

Medical Loss Ratio

HCFA Medicare Appeals Data

Sources of Comparative Information

NCQA Information
Report Cards
Medicare Managed Care Comparison Database

State Sources of Information

State monitoring reports
State Complaint Data
Medicaid Monitoring Reports
Penalty Information
Physician Information

Appendix

Tables and Exhibits

Exhibit A: Sample Format for Findings from Analysis of HCFA Performance Monitoring Reports

Table 2: Sample Format for Analyzing HCFA Performance Monitoring Report

Exhibit C: Sample Freedom of Information Act Request

Table 5: Medicare Encounters Per Enrollee: California Medicare HMOs in 1994

Exhibit D: Physician Encounters Per Enrollee

INTRODUCTION

BACKGROUND

Medicare beneficiaries are choosing to enroll in managed care plans over the traditional Medicare fee-for-service program in increasing numbers. They are drawn by added benefits, such as prescription drugs, and freedom from the costs of supplemental policies, Part A and Part B deductibles, and co-insurance. In 1985, 4 percent of Medicare beneficiaries were in managed care, but by 1998, that number had risen to 16 percent. This trend will likely accelerate as policymakers turn to managed care to contain soaring Medicare costs.*

While policymakers enthusiastically embrace Medicare managed care as the solution to financing problems, the truth is that no one really knows how the Medicare population will fare under managed care. Some evidence indicates that seniors do not do as well in managed care. For this reason, it is important that advocates monitor the quality of Medicare HMOs.

Managed care presents many quality concerns, largely due to its built-in incentive to underserve. Yet, in many ways, managed care is better suited for systematic quality data collection and analysis than the fee-for-service system. Managed care plans assume full responsibility for most of their members' care needs, so only one entity is accountable for providing quality care. In a highly fragmented fee-for-service environment, it is far more difficult to identify and remedy quality problems.

_______________
* The Balanced Budget Act of 1997 permits the Medicare program to contract with a range of managed care and insurance entities.

[Return to top]

PURPOSE OF THE GUIDE

Medicare beneficiaries considering a managed care plan need information to help them make a good choice. This guide will not tell you which HMOs are the best. However, it does show groups that work with Medicare beneficiaries how to obtain and analyze information that can bring beneficiaries a little closer to knowing which plans are the best. Collecting and analyzing this information serves three purposes: 1) making comparative information on HMOs available to seniors and younger Medicare beneficiaries with disabilities; 2) monitoring the quality of HMOs; and 3) pushing state and federal agencies that monitor Medicare HMOs to improve plan oversight.

AVAILABILITY

Consider yourself a bit of a pioneer. Health care quality assessment is still in its infancy, and the tools to truly measure the quality of HMOs are still being developed. Standardized quality measures like the Health Plan Employer Data and Information Set (HEDIS)—the most widely used quality assessment tool—examine only a limited number of factors important to assessing quality of care. Much of the information that is already collected is not yet made available to the average consumer.

The good news is that the field of quality assessment is changing rapidly. Medicare contractors are required to collect data for a growing range of measures, including a number of areas specific to the health status of seniors. Increasingly, this information is being released to consumers. In 1998, HCFA will make available its first comparative database of information on Medicare plans. It will feature comparative information on HMO benefits and on composition of physician networks. Eventually, this database will include HEDIS information (described later in this guide). And some states, like Maryland, are developing consumer report cards that rank HMOs by their performance in a variety of areas.

DATA COLLECTION HINTS

Obtaining large amounts of data from government agencies can be a challenge. However, it can be done. Remember, you are entitled to publicly available information from state and federal agencies. Here are some tips on data collection:

Know Your Rights
It is a good idea to familiarize yourself with state and federal laws regarding public information. Most states have some sort of freedom of information or public records legislation that regulates the distribution of public information. These laws will tell you the type of information that is available to you and the time within which agencies have to respond to your request. You can probably get a copy of these laws from your state's Attorney General's office or from the public information officer in the state or federal agency from which you are requesting data. Also, most state laws are available on the Internet and can be accessed through your state's official home page. If all else fails, head to the nearest law library!

Call Ahead
To minimize aggravation and wasted time that can result from requests lost in the void of some state bureaucracy, call before you write to request information. Find out if the information is available. Try to talk to the people who work with or generate the data you are requesting so you can make your written request as specific as possible. Find out the name of the person to whom you should address your request and the exact name of the report or statement you are requesting. Some agencies have public information officers who handle all requests; others require you to request information from individual departments.

Be Persistent
Don't be put off if you are at first told that the information you are requesting does not exist or is impossible to obtain. Contact different departments and speak with different people in a particular department before concluding that the information is truly unavailable. While many fine people work in government agencies, some go strictly by the book. Consider that a clerk in one state's department of insurance once held up a request for an annual financial statement because the requester mistakenly asked for an annual financial report. You should be able to obtain a good deal of the information described in the guide. However, you may find that the data in your state differs from what we've described or is called by a different name in your state.

Collect More than One Year's Worth of Data
Collect at least two years of information when possible. Because the data are only crude indicators of plan performance, it is important to establish trends over time to make reliable assessments of quality. By comparing more than one year's worth of data, you can observe improvement (or decline) in performance or dramatic changes between years that may signal reporting errors.

Collaborate with a University
Advocates might consider developing a partnership with a public health or public policy department of a nearby university for the project. Advocates could provide the data and help develop the qualitative analysis; the university could carry out the data analysis. Both partners could share credit for the report.

[Return to top]

HOW TO USE THIS GUIDE

The publicly available data discussed in this guide are tools for assessing the quality of health plans and the quality of oversight by state and federal agencies. This guide will show you how to obtain, analyze, and present these data. We have included examples of how other organizations used this data to educate consumers on the quality of plans, to push HMOs to improve care, or to encourage government agencies to improve plan oversight. A chapter is devoted to each of the five most significant types of data currently available—HCFA performance reviews, utilization data, medical loss ratios, appeals data, and disenrollment data. The remaining types of data are presented in a chapter on state data and a chapter on other data sources. Each chapter includes sections on how to obtain the data described in the chapter, what you can learn from it, how to analyze it, and what the limitations of the data are.

HCFA PERFORMANCE MONITORING REPORTS

WHAT IS A HCFA PERFORMANCE MONITORING REPORT?

Every two years the Health Care Financing Administration (HCFA) of the U.S. Department of Health and Human Services, which administers the Medicare program, examines the health plans participating in the Medicare program. The results of these on-site reviews, called performance monitoring reports, record each plan's compliance with federal requirements for HMOs that contract to serve Medicare beneficiaries. HCFA reviews do not actually evaluate the quality of care provided by plans. They merely examine whether a plan has appropriate systems in place to provide quality care. In addition, HCFA conducts focused reviews of an HMO when there are indications of problems in one or more specific areas.

Performance monitoring reports are accompanied by a cover letter to the HMO summarizing the major findings of the review. Performance monitoring reports are generally divided into two parts. Part I is a checklist of requirements. This part indicates whether each legal requirement is "met" or "not met"; for each HCFA policy requirement, the report notes "yes" or "no" to indicate a plan's compliance with regulations.* Part II contains statements of findings—detailed descriptions of areas in which a plan is not in compliance—and the corrective steps HCFA requires to bring the HMO into compliance, called corrective action requirements (CARs). HMOs must submit a plan for improvement, known as a corrective action plan (CAP), detailing the steps they will take to ensure future compliance in all deficient areas.

________________
* In addition to formal regulations, HCFA issues operational policy letters (OPL) that specify a range of plan requirements. For example, in 1997, HCFA issued OPLs on gag rules. HCFA uses OPLs because they can be issued more quickly than regulations.

WHERE DO YOU GET PERFORMANCE MONITORING REPORTS?

Performance monitoring reports are public documents. Request HCFA performance monitoring reports from your regional office of HCFA. If you have developed a good working relationship with your regional office, you may be able to obtain them through an informal request. However, you may have to file a formal request under the Freedom of Information Act (FOIA). Consult the Freedom of Information Officer in your HCFA regional office for specifics on how to make a request for performance monitoring reports and to get an estimate for the time required for fulfillment of the request. It can take a few months for a request to be honored, so it is best to submit it as early as possible. (See sample FOIA request letter.)

FOIA Requests

Make your request in writing to a Freedom of Information Officer in your HCFA region.

Provide a detailed description of the information you are requesting.

Include your telephone number as well as your address.

Request that fees be waived since the information will be used in the public's interest and will not serve a commercial interest.

HOW DO YOU USE PERFORMANCE MONITORING REPORTS?

Analyzing performance monitoring reports is a matter of organizing the information so consumers can easily compare report findings. The easiest way to do this is to group HCFA's requirements into broad categories, tally the number of deficiencies in each category, and display the information for each plan in a table or chart. This way, consumers can compare the number of deficiencies plans have in any category. It is important to remember that deficiencies in some areas are more critical than in others. For this reason, the table should be accompanied by a description of significant findings (as discussed below). Also, please note that in a number of categories, HCFA will find a requirement "met" if the plan meets the requirement 95 percent of the time. In your table and in your description, indicate those plans that have nearly met these requirements so you present differences in the severity of deficiencies among plans fairly.

STEP 1: GROUPING INFORMATION BY CATEGORIES AND TALLYING DEFICIENCIES

HCFA has already done most of the work of grouping its information by category. Part I of the HCFA performance monitoring report is organized by topic. Look at the list of topics and their descriptions below. You may decide to simplify HCFA's categories. Health Services and Quality Assurance could, for example, be combined under Quality of Care; Applications and Enrollment can be combined with Disenrollment under Enrollment/Disenrollment. Just be sure to describe the categories clearly so that consumers will know what they include.

The 13 sections of Part I of a performance monitoring report on a Medicare Managed Care plan are:

Administration and Management: This section indicates whether an HMO has the administrative capability to carry out its Medicare contract. Requirements listed include whether the plan has effective and adequate systems and staff to report information to the Secretary of the Department of Health and Human Services and to administer and manage plan activities, such as marketing, health services, and quality assurance.

Fiscal Solvency/Insolvency Protection Plan: This section indicates whether an HMO has adequate financial resources to meet its obligation to members. It includes requirements for handling insolvency, including mandatory reserves.

Utilization Management: This section indicates whether an HMO has systems in place to evaluate the appropriateness of the health care services it provides and to control costs of basic and supplemental health services to achieve utilization goals.

Incentive Arrangements: This section deals with payment arrangements that HMOs make with providers. Federal regulations are designed to keep plans from inducing physicians to deny members appropriate care. Reviewers indicate whether plans have been in compliance with regulations such as those that prohibit specific payment to physicians for withholding, limiting, or reducing services.

Health Services Delivery System: This section indicates whether HMOs have provided required Medicare services, met standards of availability and accessibility of services, maintained adequate medical records, and assured continuity of care.

Quality Assurance: This section indicates whether HMOs have met the requirements to provide ongoing quality assurance programs that stress health outcomes, include peer review programs, incorporate systematic data collection, and include written procedures for taking appropriate remedial action when poor service is provided.

Marketing: This section indicates whether HMOs have met the requirements for marketing activities and whether they have engaged in prohibited marketing activities such as door-to-door solicitation.

Applications and Enrollment: This section notes whether HMOs have met requirements for enrollment, such as holding open enrollment for at least 30 consecutive days during each contract year and notifying Medicare beneficiaries about these enrollment periods.

Membership: This section indicates whether HMOs have met requirements to inform members of their rules and charge members appropriate premiums and fees.

Disenrollment: This section indicates whether HMOs have complied with requirements governing voluntary and involuntary termination of enrollment of members including cases when there has been a failure to collect premiums, indication of fraud or abuse, a move by an enrollee out of a plan's geographic area, and loss of Medicare benefits.

Claims Processing: This section indicates whether HMOs are paying for emergency and urgently needed services that its Medicare members obtain outside the plan, whether they process claims in an appropriate and timely manner, and whether they handle denial of claims appropriately, including informing members of their rights to appeal.

Medicare Appeals: This section indicates whether an HMO properly identifies and processes appeals in a timely fashion. It also notes whether an HMO informs beneficiaries and contracting providers of the appeals process.

Internal Grievance Process: This section indicates whether an HMO has adequately handled complaints that are not subject to appeals, such as complaints about quality of service and waiting times.

STEP 2: PRESENTING FINDINGS

Table 1 (not included here) is one section from the determinations portion of a HCFA performance monitoring review from Region IX. The first column gives HCFA's code for each requirement. The second column contains a brief description of each statute, regulation, or policy guideline that the reviewers evaluate and a check mark indicating whether a regulation has been met or not met. The next two columns, titled "Yes" and "No," indicate whether a plan complies with policy requirements.

Table 2 shows how the same information can be presented in a format that is more reader-friendly. This table, while useful for comparison purposes, does not convey that some deficiencies are more serious than others. For this reason, you should add a narrative section to your analysis of performance reviews where you describe important findings and explain their significance for consumers.

PRESENTING YOUR FINDINGS

To help you see what your final product might look like, we present Sample Findings and a Sample Table from a hypothetical analysis of eight HMOs. The Sample Findings below show you how you might write up your analysis of HCFA performance monitoring reports. The Sample Table on the next page presents the data from two of the eight plans in our hypothetical study.

Exhibit A
SAMPLE FORMAT FOR FINDINGS
FROM ANALYSIS OF HCFA PERFORMANCE MONITORING REPORTS

General Findings

There were significant differences in compliance among plans: Plan B had a higher percentage of unmet requirements than other plans in 7 out of 11 categories. Plan A had the highest percentage of unmet requirements in only one area.

Noncompliance was more prevalent in certain categories across plans. Plans generally did poorly in the categories of Administration and Management, Medicare Appeals, and Applications and Enrollment. Six out of eight plans had deficiencies in this area. Three plans (Plan B, Plan D, and Plan H) failed to comply with 25 percent of the requirements in the area of Medicare Appeals.

Plans with especially high noncompliance: Plan B was noncompliant in seven areas of review. It had a noncompliance percentage of 25 percent or more in six of these areas.

Plans with high compliance: Plan C was short of full compliance in only two areas of review—Administration and Management, and Applications and Enrollment. Plan A was noncompliant in seven areas; however, it had the lowest overall percentage of unmet requirements.

Table 2
SAMPLE TABLE
Presenting Information from HCFA Performance Monitoring Reports

HCFA reviews indicate what the agency deems are especially important or "critical" requirements with an asterisk next to their codes. However, advocates may not agree with this assessment. For example, HCFA does not view any of the appeals requirements as critical. Advocates should not rely on HCFA's assessment in determining which requirements are the most significant. Instead, you should make your own best judgment about which categories are important and highlight those categories.

The reviewers' summary of findings, which makes up the second part of the report, is particularly useful for the narrative section of your analysis because it describes deficiencies in greater detail. These notes can provide a more complete picture of a plan's noncompliance. They reiterate the requirement and explain the reason for the negative determination. A corrective action requirement (CAR) details the steps the reviewers expect the plan to take to remedy the problem.

WHAT CAN HCFA PERFORMANCE MONITORING REPORTS TELL YOU?

HCFA performance monitoring reports tell you how well Medicare HMOs are doing in meeting federal requirements to provide services to Medicare beneficiaries. HCFA monitoring reports can tell you how well HCFA is doing overseeing Medicare HMOs. Advocates can learn whether HCFA is reviewing HMOs as often as required and whether they successfully compel HMOs to improve in deficient areas.

WHAT ARE THE PROBLEMS WITH PERFORMANCE MONITORING REPORTS?

Performance monitoring reports are limited by the scope of their review and the quality of the reviewers. Although HCFA's reviews are quite comprehensive, they are only designed to monitor a plan's compliance with federal requirements. Quality-of-care indicators that are not required by regulation or policy guidelines are, therefore, beyond the monitoring capability of these reviews.

Exhibit C
SAMPLE FREEDOM OF INFORMATION REQUEST

Date

Director
Freedom of Information and Privacy Office
Address of your HCFA Regional Office

To whom it may concern:

On behalf of [name of your group] , a nonprofit consumer advocacy group, we are writing pursuant to the Freedom of Information Act (FOIA), 5 U.S.C. §552, to request the two most recent HCFA performance monitoring reports for the following HMOs in [name of state] : [list all HMOs you want information about] . Please expedite this request. The data will be used for an analysis of Medicare HMOs in order to provide consumer-friendly comparative information on HMOs to Medicare beneficiaries.

Further, we request that copying and search fees be waived. Disclosure of this information meets the two fee waiver tests set forth in the FOIA regulation (53 FR 47697 §5.45 (November 25, 1988): (1) disclosure is in the public interest because it is likely to contribute significantly to public understanding of the operations or activities of the government, and (2) disclosure is not primarily in the commercial interest of the requester.

The requested performance monitoring reports will be used to inform the public about (1) government oversight activities of Medicare HMOs and (2) the quality of care provided by HMOs under direct contract with the Health Care Financing Administration. Medicare beneficiaries are interested in learning more about the quality of HMOs that participate in the Medicare program. Currently, Medicare managed care consumers make many decisions about their health care without any information about the quality of care provided by plans. As a consumer health care advocacy group, [name of your group] is committed to providing Medicare beneficiaries the information they need to make more informed choices. [Name of your group] will distribute its analysis of Medicare HMOs free of charge to organizations in [name of state] representing Medicare beneficiaries and will make the information available to the public in the form of free publications and on the Internet. The information would in no way further the commercial interest of [name of your group] .

Sincerely,

Name and Title

[Return to top]

DISENROLLMENT DATA

WHAT ARE DISENROLLMENT DATA?

Disenrollment data indicate the number of people who leave a plan in a given year. There are many reasons why people leave a plan. Some leave involuntarily (e.g., they die or lose their Medicare benefits), and some leave voluntarily (they to switch plans or return to fee-for-service). For the purpose of finding out how well plans do at keeping their members, we are interested in the number of people who voluntarily disenroll from an HMO. We will discuss three types of disenrollment information:

Voluntary Disenrollment: This refers to Medicare beneficiaries who voluntarily disenroll from a Medicare HMO during the year. High rates of voluntary disenrollment may be caused by competition in the market, marketing problems, or quality-of-care issues.

Rapid Disenrollment: This refers to Medicare beneficiaries who voluntarily leave a plan before enrollment is final or within three months of enrollment. When combined with high voluntary disenrollment rates, high rates of rapid disenrollment usually indicate marketing problems.

Disenrollment Destination: This measure provides information on where Medicare beneficiaries who disenroll from an HMO go: to traditional fee-for-service Medicare or to another Medicare HMO.

WHERE DO YOU GET DISENROLLMENT DATA?

The national office of the Health Care Financing Administration (HCFA), located in Baltimore, Maryland, collects plan disenrollment data on Medicare managed care plans. Although HCFA does not publish this data, you may request it. If your initial request is denied, you may have to file a Freedom of Informaiotn Act (FOIA) request to obtain it. (Region IX, serving Arizona, California, Hawaii, Guam, Nevada, and Samoa, tracks disenrollment data and publishes it—but not in a consumer-friendly format. Advocates in Region IX can request this information from the regional office.)

The data can be provided either on reel-to-reel computer tape or as a computer printout. Advocates may request this information directly from HCFA. Reel-to-reel tape is probably easier to analyze, but you will have to find a firm that can convert the tape to a floppy disk. (If you have arranged to do this analysis in cooperation with a local university, you may find that they have the capacity to work directly with a reel-to-reel tape without converting it to disk.)

WHAT DISENROLLMENT DATA ARE CURRENTLY AVAILABLE?

HCFA publishes two reports with disenrollment data:

Monthly Disenrollment Patterns Report includes monthly information on the total number of voluntary disenrollees. This report breaks down the total number of voluntary disenrollees by their months in the plan, so you can determine the number of voluntary disenrollees who disenrolled within the first three months of enrolling.

Disenrollment Rates Report includes the annual number of disenrollees from a plan who joined another managed care plan and the annual number of disenrollees who transfer to fee-for-service Medicare.

HOW DO YOU CALCULATE DISENROLLMENT RATES?

We will show how to calculate voluntary disenrollment rates, rapid disenrollment rates, and the percentage of disenrollees who go to another HMO or to fee-for-service. Advocates will need to calculate these quantities on an annual basis, but HCFA presents the data on a monthly basis, which means getting reports for each month in the calendar year (or for a 12-month period) in order to calculate annual rates. The following examples show you how to extract information on membership, voluntary disenrollment, and rapid disenrollment from the monthly reports. We will use data from the Monthly Disenrollment Patterns report shown on Table 3 (not included here). We use H0150-Health Partners of Alabama for our example.

Step 1: Find the Number of Voluntary Disenrollments

The number of voluntary disen-rollments is calculated by subtracting involuntary disenrollments (for example, people who died) from total disenrollments. The Monthly Disenrollment Patterns report provides a voluntary disen-rollment figure that is calculated by subtracting out disenrollments due to death or Medicare ineligibility from total disenrollments.

NOTE: These reports do not subtract out disenrollments that occurred because people moved out of the area. While disenrollment due to a move out of the area is indeed a voluntary action, it does not indicate dissatisfaction with a plan. Although out-of-the-area moves are only a small percentage of total disenrollment, if you are able to obtain the number of disenrollments due to out-of-area moves, you should subtract this number from the number of voluntary disenrollments.

A Note on HCFA Data

In the near future, it may no longer be necessary for advocates to calculate disenrollment rates. HCFA plans to publish disenrollment data by late 1998 or early 1999 but has not publicized details of the format it will use. It is extremely difficult to obtain and analyze HCFA's disenrollment data. However, the concepts and formulas in this chapter will help you make use of HCFA's disenrollment data when it becomes available. Since it is not clear that HCFA will use methodology that will indicate "churning"—i.e., the amount of enrollee turnover a plan experiences—advocates may want to analyze disenrollment data themselves. The method described in this chapter is a practical way to analyze disenrollment data.

Note: Other methods of computing disen-rollment involve following a group or "cohort" of beneficiaries over time. However, that would require resources that only HCFA or a research institute with access to HCFA's databases would have.

Step 2: Calculate the Average Monthly Membership

To calculate plan size, you must calculate the average monthly membership. You must first calculate the average membership for each month. The Monthly Disenrollment Patterns report lists the beginning and end membership totals for each month. For each month, you should add these two figures and then divide by two.

One Month's Average Membership =
[ (column 3, first line) + (column 3, second line)] ÷ 2

EXAMPLE: Health Partners of Alabama

(6,506 + 6,682) ÷ 2 = 6,594

After you have computer each month's average membership, you then add up the 12 months' averages and divide by 12 to determine the average monthly membership of each plan.

Step 3: Calculate the Voluntary Disenrollment Rate

To calculate the voluntary disenrollment rate, divide the plan's total number of voluntary disenrollments during the year by its average monthly membership, and then multiply the result by 100 to obtain a percentage.

Voluntary Disenrollment Rate =
[( # of voluntary disenrollments (column 5) - (# who moved
out of area, if possible)) ÷ average monthly membership (from Step 2, above)] x 100

EXAMPLE: Health Partners of Alabama

(43 ÷6,594) x 100 = 0.65%

Note: The number in this example is low because it reflects only one month's disenrollments. This number will increase when you sum the voluntary disenrollments for an entire year.

Step 4: Calculate the Number of Rapid Disenrollments

The Monthly Disenrollment Patterns report breaks down the number of disenrollees by the length of time they have been in the plan. It is, therefore, possible find out the number of people who disenrolled from a plan within the first three months of enrolling—those who rapidly disenrolled from a plan. HCFA's report includes the number of people who canceled their enrollment before it became final (those who were in the plan for 0 months) and those who were in the plan for no more than three months. Sum the number people who canceled and the number of people who disenrolled within three months to determine the number of people who rapidly disenrolled from a plan.

Number of Rapid Disenrollments =
# of Cancellations (column 6) + # who disenrolled after being
in the plan less than three months (column 7)

EXAMPLE: Health Partners of Alabama

Number of Rapid Disenrollments = 7 + 8 = 15

Step 5: Calculate the Rapid Disenrollment Rate

Divide the number of Rapid Disenrollments (calculated in Step 4) by the number of Voluntary Disenrollments (Step 1) to determine the percentage of voluntary disenrollees who disenrolled within the first three months of enrollment.

Rapid Disenrollment Rate =
[# of Rapid Disenrollees (from Step 4, above) ÷ # of Voluntary Disenrollees (column 5)] x 100

EXAMPLE: Health Partners of Alabama

Rapid Disenrollment rate = (15 ÷43) x 100 = 35%

Step 6: Calculate the Percentage of Disenrollees Transferring to Another HMO or to Fee-for-Service Medicare

The Disenrollment Rates Report (Table 4, not included here) includes annual information on the number of disenrollees who transfer to fee-for-service and the number who join another HMO.

Add the number of people who transferred to fee-for-service and the number of people who joined another HMO. This becomes your denominator.

Percentage who transfer to fee-for-service =
[# of transfers to fee for service (column 7) ÷
(# of transfers to fee-for-service (column 7) +
# of transfers to other HMOs (column 8))] x 100

EXAMPLE: Health Partners of Alabama

% FFS = [29 ÷ (29 + 2)] x 100 = 94%

Percentage who transfer to other HMOs =
[# of transfers to other HMOs (column 8) ÷
(# of transfers to fee-for-service (column 7) +
# of transfers to other HMOs (column 8))] x 100

EXAMPLE: Health Partners of Alabama

% HMOs = [2 ÷(29 + 2)] x 100 = 6%

Note: HCFA's data do not currently indicate the destination of beneficiaries who disenroll through Social Security offices (column 9). Therefore, it is impossible to calculate the exact percentage of beneficiaries who disenroll to fee-for-service or other HMOs. In general, however, only a small number of beneficiaries disenroll through Social Security offices, so the figures calculated above will be fairly accurate in most instances.

WHAT CAN DISENROLLMENT INFORMATION TELL YOU?

Disenrollment information can give advocates hard data on how well particular plans are doing at keeping their members. Currently, this is the only hard data available that tells the number of Medicare enrollees who "vote with their feet" by leaving particular plans. High disenrollment rates may indicate that a plan has quality-of-care problems or that the plan's marketing agents have provided inaccurate or misleading information about the plans benefits and restrictions.

WHAT ARE THE LIMITATIONS OF DISENROLLMENT DATA?

Disenrollment data do not provide specific reasons why disenrollees leave. While high disenrollment rates may indicate that a plan has quality-of-care problems, people disenroll for many reasons. Some people disenroll to take advantage of lower premiums or special perks at a competing plan.

It is important to exercise caution when interpreting disenrollment data. The voluntary disenrollment rate of a small plan may be misleading because aberrations—e.g., a new plan enters the community, inducing members to leave their existing plan—can easily skew the results. Similarly, rapid disenrollment rates should be considered in relation to voluntary disenrollment rates. Plans with low voluntary disenrollment rates are likely to have high rapid disenrollment rates, but this does not necessarily signal a problem with the plan. Advocates should be concerned about plans with both high voluntary disenrollment rates and high rapid disenrollment rates.

You should concentrate on such significant problem indicators as:

plan has a high disenrollment rate for more than one year,

plan has a high disenrollment rate and a high rapid disenrollment rate, or

plan's disenrollment rate is significantly higher than that of other plans in the community.*

_______________

* Families USA. Comparing Medicare HMOs: Do They Keep Their Members? (Washington, DC: Families USA Foundation, December 1997), p. 6.
Table 3: Monthly Disenrollment Patterns
Table 4: Monthly Disenrollment Rates

[Return to top]

UTILIZATION DATA

WHAT ARE UTILIZATION DATA?

"Utilization data" is the kind of term your high school English teacher had in mind when she warned you not to use a ten-dollar word where a nickel word would suffice. Simply put, utilization data measure the amount of health care services—e.g, visits to doctors, nurse practitioners or laboratory technicians—used by a health plan's enrollees. Most states require HMOs to report at least the number of outpatient services used and the number of days members spent in the hospital (known as hospital patient days).

Managed care and fee-for-service differ in how they gather and record utilization data. In fee-for-service, patients are billed directly for the services that they use. Billing records are a rich source of data for quality assessment, providing specific information on the services provided. Because HMOs are not reimbursed for individual services provided but rather through flat fees that cover all necessary care, they do not generate billing records. As a substitute for billing data, many HMOs require providers to collect and report "encounter data," which record each service or group of services that enrollees receive. Encounter data may be used to monitor the kinds of services plans are providing and their frequency, members access to these services, and quality of care generally.

WHERE DO YOU GET UTILIZATION DATA?

Almost all states require plans to report utilization data in quarterly and yearly financial statements. These states generally report utilization data in a format suggested by the National Association of Insurance Commissioners (NAIC). These financial statements can be obtained from states' departments of insurance.* Generally, there is a per page copying fee for these statements—anywhere from $0.25 to $1.00. Fortunately, you will only need one or two tables from each statement, so the costs will be limited.

You can call or write the office of public information within your state's department of insurance (or whichever office in your state's department of insurance handles information requests) to request a waiver of charges. Some states and the District of Columbia have provisions within their Public Records Acts laws to waive fees for information requested for the public good. A public information officer will instruct you on how to apply to have fees waived, if it is possible to do so in your state.

Ask your state's department of insurance if it uses the NAIC format. If it does, request the table entitled Premiums, Enrollment and Utilization. This will provide you with the all the information necessary to calculate utilization rates. If your state doesn't use the NAIC format, find out what utilization data, if any, it does collect and request the appropriate documents. We will show you how to calculate the utilization rates for outpatient physician and non-physician encounters and hospital days. The same calculation methods can be used to examine other services, such as emergency room visits or primary and specialty care visits per Medicare enrollee, should your state collect these data.

_______________

* Some states require reporting to Department of Health. In California, the Department of Corporations collects utilization data.

WHAT UTILIZATION DATA ARE CURRENTLY AVAILABLE?

As mentioned above, most states require plans to report annual financial statements in the format of the National Association of Insurance Commissioners (NAIC). The NAIC format requires plans to report: 1) outpatient physician encounters; 2) outpatient non-physician encounters (e.g., visits to nurse practitioners); and 3) total hospital patient days (the accrued number of hospital patient days experienced by the total membership). This format also requires plans to break down utilization data for their total enrolled population as well as for their Medicare and Medicaid populations.

Some states—e.g., California and New York—require plans to report additional categories of encounter data. California collects average-length-of-stay data in its financial statements. New York State's Department of Health collects, for its Medicaid and commercial populations, inpatient and outpatient data on specific services (e.g., emergency room visits) as well as encounter data for selected procedures, such as hysterectomy.

HOW DO YOU CALCULATE UTILIZATION RATES?

Gross utilization data—that is, the total number of different services provided by a plan—do not provide useful information because they do not take into account the number of enrollees in a plan. If a plan with 2,000 members provides the same number of health care services in a year as a plan with 20,000 members, you know that one of these plans must be providing an inappropriate amount of care. For utilization data to be truly useful, they must take into account the number of people for whom a plan provides services. Thus, utilization data are expressed as rates per enrollee.

PART 1: CALCULATING AVERAGE ENCOUNTERS PER ENROLLEE

Note: Use Table 6 (not included here) NAIC Premiums, Enrollment and Utilization, for Chesapeake Health Plan in Maryland for the following step-by-step instructions. For the purpose of our example calculation, use the figures for Medicare Risk found in column 7.

Step 1: Calculate Average Monthly Medicare Membership

To calculate utilization rates, you have to know what the average number of enrollees is. Because people join and leave plans throughout the year, it would be inaccurate to use the total membership from one point of the year. Therefore states use something called "member months" to calculate the average number of members during the year.

Average Monthly Medicare Membership =
Current Year Member Months (Column 7, Line 6) ÷ 12

EXAMPLE: Chesapeake Health Plan

Average Monthly Medicare membership =
9,927 ÷12 = 827

The average monthly membership for Chesapeake Health Plan is 827 members.

Step 2: Calculate Physician Encounters Per Medicare Enrollee

Physician Encounters Per Medicare Enrollee =
Total Member Physician Encounters (Column 7, Line 7) ÷ Average Monthly Medicare Membership

EXAMPLE: Chesapeake Health Plan

Physician Encounters per Medicare Enrollee =
7,851 ÷ 827 = 9.49

According to the reported data, Chesapeake Health Plan's Medicare members had an average of 9.49 physician visits during the year.

Step 3: Calculate Non-Physician Encounters Per Medicare Enrollee

Non-Physician Encounters Per Medicare Enrollee =
Total Member Non-Physician Encounters (Column 7, Line 8) ÷ Average Monthly Medicare Membership

EXAMPLE: Chesapeake Health Plan

Non-Physician Encounters per Medicare Enrollee =
1,017 ÷ 827 = 1.23

Thus, Chesapeake Health Plan had an average of 1.23 non-physician encounters per Medicare enrollee.

Step 4: Calculate Hospital Patient Days Per Medicare Enrollee

Hospital Patient Days Per Medicare Enrollee =
Hospital Patient Days Incurred (Column 7, Line 10) ÷ Average Monthly Medicare Membership

EXAMPLE: Chesapeake Health Plan

Hospital Patient Days Per Medicare Enrollee =
1,348 ÷827 = 1.63

Thus, Chesapeake Health Plan had an average of 1.63 hospital patient days per enrollee.

Step 5: Calculate Hospital Patient Days Per Thousand Enrollees

Note: Hospital days are usually expressed per 1,000 enrollees.

Hospital Patient Days Per Thousand =
Hospital Patient Days Per Medicare Enrollee x 1,000

EXAMPLE: Chesapeake Health Plan

Hospital Days Per Thousand Medicare Enrollees =
1.63 x 1,000 = 1,630

This means that Chesapeake Health Plan covered 1,630 days in the hospital for each 1,000 Medicare enrollees.

PART 2: CALCULATING THE AVERAGE ENCOUNTERS PER ENROLLEE FOR ALL PLANS

To get a better sense of what the utilization rates mean, you will need to compute the averages for all plans in the community. Individual rates alone do not tell you which of these rates are more desirable: By calculating the average of all the plans you wish to examine, you establish a norm, or benchmark, against which the performance of individual plans can be better evaluated.

There are two types of averages you will need to calculate for this project: unweighted and weighted. Unweighted averages are the garden-variety averages you use all the time. To calculate them, simply add the utilization rates for all the plans you are examining and divide by the number of plans you are including (see below).

A weighted average, by contrast, takes into account the differences in relative size between the items being averaged. In our case, we are averaging utilization rates, so a weighted average would take into account the difference in the size of enrollment for each plan. An unweighted average gives a very small plan with only a few members the same weight as a large plan with many members. A weighted average offsets this bias by counting the utilization rate of small plan in proportion to its size relative to the other plans in the computation.

In the following examples we will calculate the weighted and unweighted averages for physician encounter data of several Medicare HMOs in California (Table 5)

Step 1: Calculate the Unweighted Average Number of Encounters

Unweighted Average =
Sum of Physician Encounters Per Enrollee (Column 2, line 8) ÷ (Number of Plans)

EXAMPLE: California Medicare HMOs

Unweighted Average = 54.03 ÷7 = 7.72

Table 5
Medicare HMOs in California in 1994
Medicare Encounters Per Enrollee

Source: Medi-Cal Managed Care Information Project. Utilization Rates for Enrollees in Medi-Cal Managed care Plans: Do Low Rates Reflect Inappropriate Care (Los Angeles: Center for Health Care Rights, September, 11, 1995).
*Numbers may not add due to rounding

Step 2: Calculate the Weighted Average

To find the weighted average, you will perform the following three calculations (use Contra Costa as example):

First: Calculate Each HMO's Share of Total Enrollment

Share of Total Enrollment (Column 4) =
Individual HMO's Enrollment (Column 3) ÷ Total Enrollment (Sum of Column 3)

EXAMPLE: Contra Costa

Share of Total Enrollment
1,157 ÷ 488,214 (total enrollment) = .0024

Second: Multiply Each HMO's Utilization Rate by Its Share of Total Enrollment

Column 5 = Physician Encounters Per Enrollee (Column 2) x Share of Total Enrollment (Column 4)

EXAMPLE: Contra Costa

Column 5 = 10.92 x .0024 = .03

Third: Calculate the Weighted Average

Weighted Average = Total of Column 5

EXAMPLE: California Medicare HMOs

Weighted Average = Total of Column 5 = 7.15

The weighted average is 7.15. It is lower than the unweighted average because of the relatively higher utilization rates of the smaller plans. This same process can be used to calculate the weighted average for other utilization rates (e.g., hospital days).

PART 3: INTERPRETING UTILIZATION DATA

HOW TWO GROUPS USED UTILIZATION DATA

Center for Health Care Rights

The Los Angeles-based Center for Health Care Rights (CHCR) analyzed utilization data from the 1994 financial statements of 18 Medicare HMO plans and found:

-utilization rates varied widely among plans. Physician visits per Medicare enrollee ranged from 4.11 to 26.43; non-physician visits per enrollee ranged from 0.038 to 16.44, and annual hospital days per 1,000 ranged from 40 to 2,732;

-plans reported and collected utilization data inconsistently. Twelve plans failed to include all required utilization data; and

-the Department of Corporations, which collects HMOs' financial statements, apparently did not check the accuracy of utilization data reported. One plan with unusually high hospital days per 1,000 told CHCR that the data they filed with the state were incorrect.

***

Champaign County Health Care Consumers

In response to a proposed premium hike by HMOs in Illinois, the Champaign County Health Care Consumers analyzed financial statements of two area Medicare HMO plans from 1984-1988 and found:

-while total utilization for both plans did increase, the average number of visits per patient actually declined;

-one HMO increased payments to physicians by 26 percent, while the average number of visits to physicians declined by 38 percent; and

-the other HMO increased payments to physicians by 23 percent, while the average number of visits to physicians declined by 5 percent.

WHAT CAN UTILIZATION DATA TELL YOU?

Differences in utilization rates between plans may indicate a number of things, including differences in the health of the enrolled population, differences or errors in data reporting, and over-utilization or under-utilization of services. Analysis of utilization data can help advocates monitor care in the following ways:

Indicate Plans That May Be Providing Too Little Care: Utilization data do not provide information on the appropriate level of patient care. High utilization rates are not necessarily good, and low utilization rates are not necessarily bad. A plan with high utilization may be inefficient, and a plan with low utilization may simply have healthier members. It is unusually high or low rates (the so-called outliers) that signal that a plan may be providing an inappropriate level of service. Given that all managed care organizations have an incentive to reduce care, advocates should monitor HMOs by watching out for plans with unusually low utilization rates (plans have every incentive to fix overly high utilization rates on their own).

Effective monitoring requires reliable data. Advocates can push agencies with oversight responsibilities to improve the quality of available data by publishing analyses that highlight data deficiencies. When plans are evaluated unfavorably, they inevitably challenge the integrity of the data. Too often they are right: The data are inaccurate. For example, in data analyzed by the Los Angeles-based Center for Health Care Rights (CHCR), the health plan Foundation Health Plan (FHP) reported 26.43 physician visits per Medicare beneficiary. For this to be true, Medicare enrollees would have had to have visited the doctor, on average, more than 26 times during the year—a figure that seems impossibly high. The rate was so large (twice the rate of the plan with the next highest rate) that CHCR was certain the data were flawed.The state apparently had not checked the data for accuracy.

If you feel that the data you are receiving are flawed, you can request data directly from the HMOs. However, plans are not required to comply with your requests and, therefore, may not cooperate. Where you suspect or can confirm that data are inaccurate, you should note this in your analysis. Advocates perform a public service when they point out these inaccuracies to HMOs, government agencies, and the public.

Indicate Appropriateness of Premium Adjustments: In 1990, an Illinois consumer group, the Champaign County Health Care Consumers (CCHCC), used utilization data to contest the claims of area HMOs that said they were increasing premiums because members were unnecessarily using care and because they were serving an increasingly sicker population. CCHCC's analysis showed that utilization rates had actually decreased relative to the previous year
(see box ).

WHAT ARE EFFECTIVE WAYS TO PRESENT UTILIZATION DATA?

Thanks to the variety of easy-to-use computer software, you can present your findings graphically. Charts and tables can make data much easier for a reader to digest, particularly when you are trying to present comparative data. Look again at Table 5, at the very beginning of the section on calculating averages. Now look at the bar chart below. Notice how much easier it is to see the difference between HMOs' rates at a glance.

WHAT ARE THE PROBLEMS WITH UTILIZATION DATA?

Although utilization data are an important source of information, plans often collect and report data inconsistently. Because state and federal requirements for reporting are not yet standardized, differences in utilization rates among plans may reflect differences in reporting methodology, not quality.

Moreover, states often do not adequately monitor data collection. Although required by states, many plans do not report utilization data at all. Of the sampling of annual statements collected by Families USA in a 10-state survey of publicly available managed care data, the 1995 statement of Intergroup Prepaid Health Services of Arizona, Inc. did not include any ambulatory utilization data. In its 1995 study, CHCR found that several plans failed to report utilization data for all or some of their membership groups (i.e., Medicaid, Medicare, commercial). The CHCR study also uncovered instances where the California Department of Corporations failed to catch incorrectly reported data.

What's in a Name?

Part of the trouble with utilization data is that one plan's encounter is another plan's multiple encounters. NAIC defines an encounter as a "contact between an HMO member and a provider of health care services who exercises independent judgment in the care and provision of health service(s) to the member." Sounds easy enough, but in practice, plans end up reporting encounters very differently. Consider this: Physician X and Physician Y both performed the same number of obstetrics services in a given year.Physician X from HMO A reports the delivery but not each prenatal visit. Physician Y from HMO B reports each prenatal visit and the delivery. Though both physicians performed the same number of services, Physician Y appears to have performed many more.

This kind of data collection and reporting is a relatively new area, and both plans and regulators are finding their way. Plans must adapt their existing methods to new requirements, and regulators are fine-tuning their data requests. It therefore takes time to get plans to report this data consistently and accurately. For example, New York State, which has been systematically collecting Medicaid encounter data since July 1, 1996, has only recently made this data public because of inconsistency of plan reporting.

WHAT DATA WILL BE AVAILABLE IN THE FUTURE?

It is hoped that a broad range of standardized encounter data will be available in the not-too-distant future. The Health Insurance Portability and Accountability Act of 1996 (also called the Kassebaum-Kennedy Act) requires the U.S. Department of Health and Human Services to develop standard sets of data for health information, including encounter data, by mid-1998. The Balanced Budget Act of 1997 requires plans to report standardized encounter data for Medicare and Medicaid populations.

Exhibit D

Physician Encounters Per Enrollee

MEDICAL LOSS RATIO

WHAT IS A MEDICAL LOSS RATIO?

"Medical loss ratio" is insurance-speak for the percentage of each premium dollar that is spent on the provision of health care services. (To an insurer, after all, patient care costs are expenses that take away from profits and, therefore, represent loss.) Health care expenses include physician services, outside referrals to physicians, emergency room and out-of-area services, and costs associated with inpatient care. For the purpose of calculating medical loss ratios, health care expenses are distinguished from administrative costs, such as marketing, and profit.

WHERE DO YOU GET MEDICAL LOSS RATIO DATA?

The information needed to calculate medical loss ratios comes from the same place utilization data (discussed later in this guide) comes from: annual financial statements filed with your state's department of insurance. These financial statements can be obtained from states' departments of insurance. (Generally, there is a copying fee for these statements—anywhere from $0.25 to $1.00 per page.)

WHY DO YOU WANT IT?

Medical loss ratios are a very rough measure of how much a plan spends on patient care. They do not measure the quality of the care that an HMO provides. They can merely alert consumers to plans that may be spending too little on patient care to ensure appropriate quality of care. While medical loss ratios cannot determine the quality of care per se, plans with unusually low medical loss ratios may be skimping on care. Plans with very high loss ratios may be incorrectly reporting medical costs.

PART 1: CALCULATING AN HMO'S MEDICAL LOSS RATIO

Medical Loss Ratio =
[Amount Incurred for Provision of Health Care Services
(Line 17a) ÷ Premiums Earned (Line 14)] x 100

EXAMPLE: Chesapeake Health Plan

Medical Loss Ratio=
($76,933,541 ÷ $101,264,647) x 100 = 76%

The Medical Loss Ratio for Chesapeake Health Plan is 76 percent. This means that the plan spends 76 cents of every premium dollar on patient care.

PART 2: CALCULATING PLAN PROFIT

You can use medical loss data to analyze a plan's expenditures. Some administrative expenses are good for consumers. Quality assurance programs, for example, are administrative expenses that directly improve the quality of care for consumers; but other administrative costs, such as marketing and profits, can add to health care costs without providing benefits to consumers.

Step 1: Calculating plan income as a percentage of total revenues

Plan Income as a Percentage of Total Revenues =
Income (Line 29) ÷ Total Revenue (Line 7)

EXAMPLE: Chesapeake Health Plan

$7,773,884 ÷ $104,306,656 = 7.45%

Chesapeake Health Plan keeps 7.45 percent of its revenue as profit.

Note: Although the NAIC model financial statement contains Medicare-specific information on enrollment, it does not include specific information on the amount of income derived from Medicare. Therefore, it is only possible to calculate income for the plan's total enrollment and not per Medicare beneficiary.

Step 2: Calculate Income Per Enrollee

To find the income per enrollee, you will need to calculate average total membership using the Premiums, Enrollment and Utilization table (not included here). For the remaining calculations in this section, use the Statement of Revenue, Expenses and Net Worth (not included here). Because people join and leave plans throughout the year, it would be inaccurate to use the total membership from one point of the year. Therefore states use something called member months to calculate the average number of members during the year.

Average Total Membership =
Current Year Member Months (line 1, column 2) ÷12

EXAMPLE: Chesapeake Health Plan

786,430 ÷ 12 = 65,536

Income per Enrollee =
Income (Line 29) ÷ Average Total Membership
(calculated above)

EXAMPLE: Chesapeake Health Plan

$7,773,884 ÷ 65,536 = $119

This means that Chesapeake Health plan earned an average profit of $119 per enrollee per month before taxes.

PART 3: INTERPRETING MEDICAL LOSS DATA

How Two Groups Used Medical Loss Information
California Medical Association

Using data from annual reports filed with the California Department of Corporations, the California Medical Association (CMA) publishes an annual report comparing the medical loss ratios of California HMOs. In addition to the overall medical loss ratios of California HMOs, the report includes the percentage of revenue that goes to medical care, administrative costs, and profit. Remember, the higher the loss ratio, the better. For 1995-1996, CMA found:

-the portion of premiums spent on medical care ranged from 73 percent to 96.6 percent among commercial plans;

-70 percent of plans with the lowest medical loss ratios were for-profit plans;

-70 percent of plans with the highest medical loss ratios were nonprofit plans; and

-one plan with one of the lowest medical loss ratios had a profit ratio (plan income as a percentage of total revenues) of 10.5 percent compared to a profit ratio of 0.5 percent for the plan with the highest medical loss ratio.

***
Citizen Action

Citizen Action compared the premiums, profits, and medical loss ratios of the 10 largest HMOs in Illinois using data from 1996 annual financial reports filed with the Illinois Department of Insurance. Among the group's findings:

-premiums per enrollee were similar for all 10 plans;

-premiums per Medicare enrollee ranged from $5,490 to $5,845 (three times the average premium for non-Medicare coverage);

-one in eight premium dollars went to non-medical uses;

-nonmedical expenses as a percentage of premium dollars ranged from 5.8 percent to 24.5 percent; and

-combined profits for all plans averaged $46 per enrollee.

WHAT DO MEDICAL LOSS DATA TELL YOU?

Medical loss ratios are a controversial way of monitoring plan performance. As we've already mentioned, they do not measure quality of care directly. At best, they can indicate plans that may be skimping on care. One of the problems with medical loss data is that they suffer from some of the same standardization problems as utilization data (see section beginning on page 31). Though the NAIC format specifies categories for reporting the amount incurred for provision of health care services (e.g., physician services, other professional services, outside referrals), it is extremely difficult to codify expenses for plans with highly diverse organizational structures, product lines, and enrollment bases. For example, how does a plan with some centralized administrative functions accurately account for the percentage of premium dollars from various regions that contributed toward administrative costs? Companies that operate HMOs may also offer other products in the state, such as PPOs, which may be reported in the same financial statement. In such a case, it is impossible to extricate HMO-specific information for that plan. Advocates can push states and plans to devise more accurate methods of accounting for expenses and revenues. States could, for example, demand that plans report expenses for different products separately.

Some critics of medical loss data argue that the data will never be useful as a monitoring tool. They argue that direct quality indicators, such as patient satisfaction surveys, health status indicators, and outcomes measures, are much more effective ways of measuring plan performance. We couldn't agree more. Yet standardized, systematic quality measures are still in their infancy and are not yet universally published in comparative format. In the meantime, consumers have to rely on the data that are now available, including medical loss data. While advocates should avoid judging the quality of an HMO based solely on its medical loss ratio, they can serve the public by revealing plans with unusually low ratios and pushing state agencies to investigate these plans to ensure that patient care is not compromised.

HCFA MEDICARE APPEALS DATA

WHAT ARE APPEALS DATA?

Medicare enrollees may appeal whenever their HMO denies, terminates, or fails to pay for a Medicare-covered benefit. If a beneficiary is not satisfied by the outcome of an appeal handled by the HMO, he or she is entitled to have that decision reviewed by an independent party if the case involves a denial of a Medicare-covered benefit. Beneficiary complaints about the quality or timeliness of services, about administrative issues, and about non-Medicare covered benefits (e.g., prescription drugs) are handled by the HMO's grievance system, which is independent of the Medicare appeals process. HMOs are required to inform beneficiaries about both their own grievance process and the Medicare appeals process.

Appealing a decision under Medicare can involve four levels:

1. Initial Determination: HMOs must provide expedited review for denials or terminations of care that could jeopardize an enrollee's life, health, or ability to regain maximum function. Otherwise, HMOs must make an initial determination within 60 days of the enrollee's request for service or payment of services. Failure to provide a determination within the required time is treated as a decision that is unfavorable to the beneficiary.

2. Reconsideration (step 1): The beneficiary may file a request for reconsideration within 60 days of the initial determination; plans have 60 days from receipt to handle the reconsideration. The reconsideration decision may not be made by the individual who makes the initial determination.

Reconsideration (step 2): If the plan's reconsideration determination is adverse to the enrollee, the plan must forward the appeal to the Center for Health Dispute Resolution (CHDR), with an explanation for the adverse determination. HCFA contracts with CHDR to handle Medicare appeals. CHDR is not required by law to make a decision on the appeal within a specified time period, but unwritten guidelines require a decision within 30 days.

3. Administrative Law Judge: An adverse decision in a reconsideration request by CHDR may be appealed within 60 days to an administrative law judge if at least $100 is at issue.

4. Appeals Council: Enrollees can appeal an administrative law judge's decision to an appeals council if at least $1,000 is at issue.

This chapter discusses data from the third level of the process—where CHDR reviews a complaint and makes a determination whether to uphold or overturn a plan's decision. CHDR publishes information on the number and the type of complaints received from each HMO, the percentage of reconsiderations from each plan that are overturned, and the length of time HMOs take to process appeals.

In addition, HCFA assesses whether HMOs are properly handling appeals and grievances in on-site performance reviews conducted every two years. These reviews note whether HMOs have met the agency's requirements for handling appeals before they get to CHDR. This information is important to analyzing appeals data, as we will discuss later.

WHERE DO YOU GET IT?

CHDR maintains statistics on the number, the type, and the outcome of the appeals (or reconsiderations) it receives as well as whether plans processed appeals in the time allotted by regulation (60 days for typical appeals, 72 hours for expedited appeals). Request appeals data directly from CHDR (you can get the phone number from your regional HCFA office). HCFA performance reviews can be obtained from HCFA Regional Offices (see section on monitoring reviews in this guide).

WHAT DO YOU DO WITH APPEALS DATA?

Because CHDR's data are already published in comparative form, analyzing appeals data is relatively easy. You simply need to decide which of the available data you wish to highlight and then present them in an effective manner. The three most important things you can learn from the data are 1) the appeals rate for individual HMOs, 2) the percentage of appeals partially or fully overturned by CHDR, and 3) the timeliness of plans' processing of appeals. Table 9 (not included here) is a sample of HCFA appeals data from Region I for 1995 and 1996. Moving from left to right on the page, you will notice that most of the columns are self-explanatory.

Columns 1 and 2: The numbers in the very first column preceded by the letter H are simply the contract numbers that HCFA assigns to each Medicare plan. Column 2 contains the name of each plan.

Columns 3 and 4: Enrollment figures are given for each plan for the years 1995 and 1996.

Column 5: RATE stands for the rate of reconsiderations per 1,000 members. The rate is an important indicator of the number of enrollees who appeal and are dissatisfied with a plan's internal appeal decisions. The rate is calculated by taking the number in column 6 (see below) and dividing by the sum of columns three and four.

Columns 6 and 7: Column 6 contains the number of reconsiderations that CHDR receives. Column 7 contains the number of reconsiderations that CHDR has received but has not yet decided. Alone, these raw numbers are not relevant, because they do not take into account the size of each plan's enrollment.

Columns 8 and 9: Column 8 contains the number of reconsiderations or appeals that CHDR upheld. In other words, this column indicates the number of appeals in which CHDR felt the plan made a correct decision. Column 9 contains the percentage of all decisions for each plan that CHDR upheld.

Columns 10 and 11: Column 10 contains the number of reconsiderations or appeals that CHDR overturned. These are important numbers because they indicate the number of cases in which CHDR ruled that plans made incorrect decisions in either denying, terminating, or refusing to pay for a Medicare-covered benefit. Column 11 is the percentage of total reconsiderations that CHDR overturned.

Column 12 and 13: Column 12 includes the number of appeals that are partially overturned. Let's say a beneficiary goes to an out-of-plan physician for emergency treatment and then continues to see this out-of-plan physician for non-emergency treatment. The HMO denies payment to the physician for both the emergency and non-emergency care. In this case, CHDR is likely to partially overturn the HMO's decision because, while HMOs are entitled to deny payment for out-of-plan non-emergency care, they are required to pay for services provided by out-of-plan physicians for emergencies. Column 13 shows the percentage of appeals that have been partially overturned.

Column 14 and 15: Column 14 contains retroactive disenrollments. Retroactive disenrollments are disenrollments approved by HCFA in special cases, such as when a beneficiary misunderstands key HMO provisions (e.g., that they must see providers within an HMO's network). In such cases, HCFA agrees to treat the enrollment as if it had never happened, and the traditional Medicare fee-for-service program picks up the tab for the contested benefit. The retroactive disenrollments presented here are only those for which the enrollee's appeal has been withdrawn because the enrollee retroactively disenrolled. Because the number of retroactive disenrollment appeals withdrawn is small, these are not critical data. Column 15 contains the percentage of cases that result in retroactive disenrollment.

Note: Use the second row, which contains appeals data on Humana Phoenix (H0307), for the appeals calculations.

Appeals Rate =
[# of Reconsiderations or Appeals (Column 6, Line 2) ÷ Enrollment (Column 3, Line 2 + Column 4, Line 2]) x 1000 = (Column 5, Line 2)

EXAMPLE: Humana Phoenix

Rate = [172 ÷ (17,118 + 18,484)] x 1000 = 4.8

Humana Phoenix has an appeals rate of 4.8 per 1,000 enrollees. Nearly 5 out of every 1,000 Medicare beneficiaries enrolled in Humana Phoenix has an appeal considered by CHDR.

Percentage Upheld =
[Reconsiderations Upheld (Column 8, Line 2) ÷ # of Reconsiderations (Column 6, Line 2 -Column 7, Line 2)] x 100

EXAMPLE: Humana Phoenix

Percent Upheld = (92 ÷ 170) x 100 = 54 percent

CHDR agreed with the decision of Humana Phoenix in over half of the appeals cases against the plan.

Percentage Overturned =
[Reconsideration Overturned (Column 10, Line 2) ÷ # of Reconsiderations (Column 6, Line 2 -Column 7, Line 2)] x 100

EXAMPLE: Humana Phoenix

Percent Overturned = [56 ÷ (172 - 2)] x 100 = 33%

CHDR did not agree with the decisions of Humana Phoenix in 33 percent of the cases reviewed.

Percentage Partially Overturned =
[# of Reconsiderations Overturned (Column 12, Line 2) ÷ # of Reconsiderations (Column 6, Line 2 - Column 7, Line 2)] x 100 = (Column 13, Line 2)

EXAMPLE: Humana Phoenix

Percent Partially Overturned =
[1 ÷ (172 - 2)] x 100 = 0.59%

Humana Phoenix had one appeal that was partially overturned. Because both partially and completely overturned appeals represent cases where plans made incorrect decisions (and thus cases where consumer dissatisfaction was valid), you may want to combine the two categories when you present your data. This puts the relevant data in one place, making it easier for people to understand.

Total Overturns =
[Overturns (Column 10, Line 2) +
Partial Overturns (Column 12, Line 2)] ÷ # of Reconsiderations (Column 6, Line 2 - Column 7, Line 2) x 100

EXAMPLE: Humana Phoenix

[(56 + 1) ÷ (172 - 2)] x 100 = 34%

APPEALS DATA FROM HCFA HMO PERFORMANCE REVIEWS

The information on HMOs' appeals processes contained in HCFA's performance reviews complements the CHDR's appeals data. While CHDR's data describe the outcome of a small number of individual complaints, the appeals information in HCFA's review describe how successful HMOs are in handling complaints generally.

To see why appeals information from HCFA performance reviews are important, take a look at Intergroup Prepaid Health (H0351) and Humana Phoenix (H0307) on our sample of CHDR appeals data (Table 8). With a combined two-year enrollment of 35,000 Humana has one-half the enrollment of Intergroup. Yet, Humana has 31 percent more Reconsiderations than Intergroup. Alone, this number suggests that Intergroup is more successful handling appeals internally or receives fewer complaints, but this data says nothing about how well each plan handles its appeals before they get to CHDR. It is possible that Intergroup is merely less conscientious about informing beneficiaries of their right to external review than Humana or that Intergroup incorrectly categorizes appeals, which entitle beneficiaries to external review, as grievances. For these reasons, information about how well Medicare HMOs handle the appeals process contained in HCFA's performance monitoring reviews can be helpful in interpreting what high or low appeals rates mean.

WHAT ARE THE DRAWBACKS TO USING APPEALS DATA?

Appeals data draw only a partial picture of consumer dissatisfaction. Appeals data do not take into account the beneficiaries who will never even make an initial complaint to their HMO and will choose, instead, to vote with their feet and disenroll, or worse, just accept an HMO decision they feel is unjust. As we've already mentioned, appeals data do not include consumer complaints about the quality of services they received or record complaints that were resolved at the plan level. Appeals data really account for only a fraction of complaints and only those involving a Medicare-covered benefit.

Because these number are so small, you have to be careful about drawing sweeping conclusions about plans on the basis of appeals data. Still, these data can point out problems with individual HMOs' reconsideration processes that lead to inappropriate terminations or denials of payment or services for Medicare-covered benefits. Consider the plans from Region IX in Table 9 on page 63. Even though the number of appeals that reached CHDR is relatively small, some plans had a much higher rate than others. For example, Foundation Health Plan (H0557) and Cigna Health Plan (H0354) had rates considerably higher than other HMOs in the region. Maricopa County Health Plan (H0350) had such a small enrollment compared to the other plans in the region that its rate of 10.6 is suspect.

Still, sometimes there is strength in even a few numbers. If you consider that by the time these cases arrive at the CHDR, they have already been through an initial decision and a plan reconsideration, we should find that the plans have made the correct decision in the majority of cases. In fact, CHDR reports that it upholds the plan's decision in over 60 percent of cases. Advocates should, therefore, be concerned by HMOs that have a high rate of appeals and/or number of decisions overturned through the appeals process.

These same small numbers suggest that beneficiaries either may not know how to use the appeals process or may find the process too cumbersome to use. Advocates should push HCFA and plans to improve beneficiary awareness of and access to the appeals process.

[Return to top]

SOURCES OF COMPARATIVE INFORMATION

NCQA INFORMATION

What is NCQA?
The National Committee for Quality Assurance (NCQA) is a private nonprofit organization that accredits HMOs. NCQA also develops and collects the most commonly used form of standardized performance measures for HMOs, known as HEDIS (short for Health Plan Employer Data and Information Set). HEDIS was developed to provide employers with comparative information by which to judge the quality of health plans. Currently, HCFA requires HMOs with Medicare contracts to report selected measures for the latest version of HEDIS (see a list of measures required for reporting Medicare HMOs below).

Why do you want NCQA information?
Quite frankly, NCQA is pretty much the only source of comprehensive comparative information on HMO quality. A number of large employers require contracting HMOs to obtain NCQA accreditation and to provide HEDIS data. Right now, HEDIS is the most widely used set of standardized performance measures. Every risk HMO with a Medicare contract is required to collect and report HEDIS informa-tion. Some of this HEDIS information directly pertains to seniors (see list of measures currently required, below).

Two types of NCQA information are available: 1) accreditation status of HMOs and 2) HEDIS information.

1) Accreditation standards for HMOs

Plans are reviewed by a team of physicians and managed care experts. A national committee of physicians reviews the team's findings and determines the appropriate level of accreditation. Currently, NCQA uses 50 accreditation standards. The standards are designed to determine whether the HMO plan has good systems in place to evaluate the quality of care given to its members, to assess physician credentials, to inform members of their rights and responsibilities, to encourage and deliver preventive services, to decide what services are appropriate, and to maintain medical records sufficient to meet NCQA's standards for quality care. The five possible accreditation levels are explained below:

Full Accreditation is awarded to plans that meet NCQA's standards and is effective for three years.

One-Year Accreditation is awarded to plans that meet most of NCQA's standards but not enough to obtain full accreditation. After the year, plans are reassessed to see whether they meet requirements for full accreditation.

Provisional Accreditation is granted for one year to plans that meet some of NCQA's requirements and have adequate quality improvement programs. Plans must demonstrate progress before they can be accredited at a higher level.

Denial is given to plans that fail to meet enough of the accreditation standards.

Under Review applies to plans that have received an initial accreditation determination but have requested to move to the next accreditation level.

2) HEDIS Information Required from Medicare HMOs

1. Breast Cancer Screening estimates the percentage of female members between the ages of 52 and 69 who had at least one mammogram during the past two years.

2. Beta Blocker Treatment after a Heart Attack estimates the number of members discharged from the hospital after a heart attack who were given a prescription for beta blockers (only considers those who did not show evidence that beta blockers might have adverse affects for them).

3. Eye Exams for People with Diabetes estimates the percentage of diabetic plan members who received an eye exam in the past years.

4. Follow-up after Hospitalization for Mental Illness

5. The Health of Seniors assesses the quality of life of seniors in a plan by using a survey asking them to rate whether their ability to function has lessened or improved over time.

6. Adults Access to Prevention/Ambulatory Health Services indicates whether members are getting preventive and ambulatory services.

7. Availability of Primary Care Providers indicates how many primary care providers are accepting new patients.

8. Availability of Mental Health/Chemical Dependency Providers indicates the number of mental health and chemical dependency providers accepting new patients.

9. Availability of Language Interpretation Services indicates the number of health care providers and member services personnel who speak languages other than English. Also measures availability of out-of-plan interpreters.

10-13. Health Plan Stability (10 separate measures) includes information about the plan's disenrollment rate, physician turnover rate, and the plan's financial stability.

14. High-Occurrence/ High-Cost Diagnostic Related Groups assesses the average cost of treatment of common diseases.

15. Rate Trends measures plan's actual expenses per member per month compared to plan's prediction about cost of delivering care.

16. Provider Compensation describes the payment arrangements that a plan has with its physicians (e.g., capitation, fee-for-service, salary).

17. Total Enrollment indicates what proportion of health plan members are enrolled through Medicaid, commercial, and Medicare risk contracts.

18. Enrollment by Payer indicates age and sex profiles of persons covered by payer.

19. Frequency of Selected Procedures indicates the frequency of several commonly performed procedures.

20. Utilization:
* Use of Ambulatory Care
* Inpatient Utilization - Non-acute Care
* Mental Health Utilization - Length of Stay
* Mental Health Utilization - Percentage of Patients Receiving Ambulatory Service
* Readmission for Specified Mental Disorders
* Chemical Dependency Utilization - Length of Stay

Where do you get NCQA accreditation information?
If you have access to the Internet, you can download the accreditation status of HMOs at www.ncqa.org. The site it is easy to use, simply indicating whether a plan has been accredited. If you do not have access to the Internet, call NCQA for a listing of accreditation status.

Where do you get HEDIS information?
You can purchase HEDIS information for commercial plans from NCQA for $800. This buys you access to Quality Compass, NCQA's database of comparable performance and accreditation information. However, HMOs submit information for inclusion in Quality Compass voluntarily, and those that participate are free to submit some measures and not others.

HCFA will make some HEDIS information available on the Internet in late 1999 (see section on Medicare Plans Comparative Database below). Also, some states— Maryland, for example—are producing report cards comparing and ranking HMOs using HEDIS data.

What are the drawbacks to using NCQA information?
As with all private accrediting organizations, NCQA is not publicly accountable for its practices. The organization can—and does—charge hefty fees for the information it collects. NCQA also does not require HMOs to report or share this information with the public. HMO participation is voluntary for HMOs that do not contract with HCFA for Medicare or Medicaid. Further, NCQA is supported in part by the HMO industry through accreditation fees. While this doesn't mean that NCQA isn't performing a useful service, it does mean that its organizational perspective is not purely consumer-oriented. From a consumer perspective, accreditation standards are not especially stringent, although the organization is in the process of improving its standards to conform with some of the more recent state HMO consumer legislation. As a result, the overwhelming majority of plans that apply for accreditation receive one-year or full accreditation. Only 10 percent of plans are denied accreditation of any sort. As of this writing, an HMO can be accredited without having to report HEDIS data, although this is likely to change in the near future.

NCQA accreditation does not always mean than an HMO is without serious problems: New York Oxford health plan faced near bankruptcy, despite a 3-year accreditation by NCQA. In 1998, the North Carolina Department of Insurance fined Blue Cross/Blue Shield $300,000 for failure to meet state licensing requirements, although the plan had just received a 3-year accreditation from NCQA.

HEDIS data have been criticized for focusing more on process measures—e.g., does the plan have adequate systems in place to monitor and improve quality of care—than outcomes measures—e.g., what is the survival rate of heart patients; how well do surviving patients function after treatment. If you look at the list of measures required for reporting by Medicare HMOs, you will notice that some of the measures do not directly measure the quality of care. Instead, they measure the adequacy of the systems that plans have in place to deliver care. For example, measures 7 and 8 pertain to the availability of providers, surely important information for Medicare beneficiaries; however, these measures do not tell you whether beneficiaries are accessing mental health providers or whether they receive appropriate care when they do.

[Return to top]

REPORT CARDS

WHAT ARE REPORT CARDS?

One of the few sensibly named terms you're likely to encounter in health care-speak, "report cards" grade or score HMOs based on their performance on consumer surveys and/or on measures designed to evaluate quality of care. Report cards come in all shapes and sizes and vary in quality and nature. In fact, you may decide to use the information from this guide to make a report card of your own.

WHY DO YOU WANT REPORT CARDS?

Report cards may provide information on the quality of care of health plans that cannot be obtained from the types of data described in this guide. Although report cards don't provide consumers with everything they might want to know, they are a good place to start assessing the care provided by HMOs.

WHERE DO YOU GET REPORT CARDS

Report cards can be obtained from a number of sources.

Purchasers: Some large employers, such as Xerox Corporation, regularly produce comparative information about plans for their employees. The information from these sources is often referred to as purchaser information because it is information produced by organizations that purchase insurance on behalf of other employees, members, etc. These companies may or may not provide this information to the general public.

Nonprofit Alliances: Some nonprofit groups, such as the Pacific Business Group on Health in California, produce report cards. Some of these groups are coalitions of area businesses that use their collective bargaining power to negotiate with plans about premium rates and health care quality issues.

Magazines: Health Pages features consumer information about health plans across the nation. Health Pages publishes reports cards in regional issues for the following locations: Atlanta, Denver, Phoenix, Los Angeles, Pittsburgh, South Florida, St. Louis, and in combined issues for New Hampshire-Vermont and Cincinnati-Columbus-Dayton.

Large circulation magazines such as U.S. News & World Report and Consumer Reports also publish HMO comparison data based on HEDIS and other information.

State Agencies: States are starting to produce report cards as well. For example, The Health Care Access and Cost Commission of Maryland recently released its first report card on commercial plans, using HEDIS data and results from a consumer survey. The New York State Department of Health publishes comparative HEDIS for Medicaid and commercial plans.

HMOs: Many HMOs publish their own report cards for consumers. However, because HMOs use self-reported and self-selected measures for these reports, consider the information with caution. HMOs are not likely to publish measures in which they have done poorly. As a result, HMO report cards are often just slick marketing tools.

WHAT ARE THE DRAWBACKS TO USING REPORT CARDS?

Perhaps the biggest limitation of report cards in some parts of the country is that they do not provide information on subcontracting medical groups. In Los Angeles, for example, HMO enrollees obtain their care from large subcontracting medical groups that often contract with all HMOs in the community. These groups make the decisions about referrals and often run the utilization review and appeals processes. Consequently, the question is not which HMO is best, but which medical group is best.

Report cards often rely on consumer satisfaction surveys. While consumers can judge the administrative quality of HMOs (e.g., how easy is it to schedule an appointment), they often cannot judge the quality of the care they receive because they are not professionals. A consumer who experiences a negative outcome—perhaps he or she does not recover quickly—may wrongly attribute it to poor care, even if he or she has received the best care available. On the other hand, a patient may be receiving poor care but might wrongly attribute a lack of progress to the nature of the illness or condition. Nevertheless, information such as how enrollees rate the travel time needed to obtain care or the ease of obtaining specialist referrals is critical information for prospective enrollees and purchasers.

[Return to top]

HCFA MEDICARE MANAGED CARE PLANS COMPARISON DATABASE

WHAT IS THE MEDICARE MANAGED CARE PLANS COMPARISON DATABASE?

The Medicare Managed Care Plans Comparison Database is an electronic database of Medicare HMOs. The database contains information on each HMO, including toll free telephone numbers, website addresses, the service areas, and the benefits packages the plan offers. The database includes detailed information on copayments, deductibles, and benefits.

WHY DO YOU WANT INFORMATION FROM THE MEDICARE HMO DATABASE?

You can use the database to find comparative information on the costs and the benefits packages of Medicare HMOs in your area and publish this information for consumers. The information from this database is especially valuable because it describes HMOs' costs and benefits in a way that is comparable across plans. This is important because plans often describe their benefits and cost-sharing requirements quite differently in their own marketing materials. For example, one plan may describe the cost of a specific benefit in terms of its annual cost to a beneficiary, and another plan will describe the same benefit in terms of its monthly cost to a beneficiary. When you consider the number of benefits each plan may have and the growing number of Medicare HMOs among which beneficiaries can choose, it is easy to see why comparative information is critical for consumers.

WHERE DO YOU FIND THE DATABASE?

The database is available on the Internet at www.hcfa.gov.

[Return to top]

STATE SOURCES OF INFORMATION

In addition to the utilization and medical loss data described in this guide, you can collect and analyze other information on managed care. Because this information varies by state, we cannot tell you in detail how to analyze the different data. The sections below will tell you what types of data may be available in your state, why you may consider collecting them, and, where possible, what you may be able to do with this information.

[Return to top]

STATE MONITORING REPORTS

WHAT ARE STATE MONITORING REPORTS?

In most states, HMOs are monitored by the state's health department or the insurance department or both. Some states do actual on-site reviews of HMOs, monitoring HMOs' compliance with state regulations and performance in any number of review areas pertaining to quality of care. In addition, almost all states conduct some kind of financial review of an HMO, often—but not always—called market conduct examinations. Market conduct examinations generally are financial reviews but sometimes include quality indicators such as a review of complaints against an HMO or a review of an HMO's compliance with quality standards.

WHERE DO YOU GET STATE MONITORING REVIEWS?

It depends on which agency has oversight responsibility for HMOs. See the Appendix for the name of the appropriate agency in your state.

WHY DO YOU WANT STATE MONITORING REPORTS?

State data on HMOs can provide additional information about the quality of HMOs serving Medicare beneficiaries. Remember, you want to provide Medicare beneficiaries with the most complete information possible on the performance of Medicare HMOs in the community. While state information is often not specific to Medicare, it addresses the general quality of care provided by HMOs.

WHAT DO YOU DO WITH STATE MONITORING REPORTS?

Because state monitoring reports vary, we can't tell you exactly how to analyze state data. As with HCFA monitoring reports, you should present the findings from state reports in a format that allows consumers to understand how plans compare.

WHAT CAN STATE MONITORING REPORTS TELL YOU?

How helpful state monitoring reviews are will vary from state to state. In states where agencies with oversight of HMOs monitor quality of care, these reports may be a good indicator of a plan's performance. In other states, these reviews may not tell you anything of value about the quality of HMOs. By collecting and analyzing state monitoring reports, advocates can judge the quality of state oversight and push states to improve supervision of HMOs.

WHAT ARE THE PROBLEMS WITH STATE MONITORING REPORTS?

Some states protect the trade interests of HMOs, so state reports may be unavailable or difficult to get. In general, these reports are available through a public records act request (see Introduction).

[Return to top]

STATE COMPLAINT DATA

WHAT ARE STATE COMPLAINT DATA?

State complaint data come from complaints filed with state regulators. As with the other state information in this chapter, it varies considerably in content and quality across states. Oregon, for example, publishes a yearly comparative guide on consumer complaints against all types of insurers, including HMOs. The annual guide lists the name of each HMO, the number of complaints received for each HMO by the Oregon Insurance Division, the amount of premium dollars earned by the HMO in a year, and the ranking of each HMO based on its number of complaints in relation to the amount of premium dollars earned. By contrast, Arizona does not regularly publish complaint information at all, although the state will, upon request, tell consumers over the phone how many complaints a specific HMO has had, and actual complaint records can be reviewed.

WHERE DO YOU GET STATE COMPLAINT DATA?

You will usually get state complaint data from the department of insurance. It may also be collected by the health department or, in the case of California, the Department of Corporations. Usually, some sort of beneficiary hotline or consumer relations division handles complaints.

WHAT CAN YOU DO WITH STATE COMPLAINT DATA?

Obviously, what you can do with complaint data depends upon what complaint data your state makes available. If possible, present the number of complaints per enrollee so that you can account for differences in the size of plans.

WHAT ARE THE PROBLEMS WITH COMPLAINT DATA?

States often do a poor job of compiling the complaints that they do get. Often, complaints are not grouped by category, and there is no indication of what the complaint is about; or complaints are grouped by categories that are so broad, it is difficult to tell what type of complaint was made. In addition, very few states make the outcome of the complaints available, so state complaint data generally do not distinguish between valid and invalid complaints.

MEDICAID MONITORING REPORTS

MEDICAID INFORMATION MAY BE USEFUL

Most states now, or soon will, require some or all of their Medicaid populations to enroll in HMOs. Some of these Medicaid HMOs also serve the Medicare population, and to the extent they do, Medicaid data may provide additional information about the quality of HMOs serving Medicare beneficiaries. Because Medicaid is administered and monitored by state agencies, the quality of the information varies substantially from state to state. For this reason, we cannot tell you in detail how to analyze the different data. This section and the section below on penalty information will tell you the kinds of data that may be available in your state, why you may consider collecting them, and where possible, what you may be able to do with this information.

WHAT ARE MEDICAID MONITORING REPORTS?

Under federal law (42 C.F.R. §434.53), state Medicaid agencies are required to contract with independent agencies to conduct annual medical audits of managed care plans to ensure quality and access. The scope and the quality of these reviews vary from state to state. The written reports about these audits should be publicly available.

WHY DO YOU WANT MEDICAID MONITORING REPORTS?

Medicaid monitoring reports can tell you how well Medicaid HMOs that also serve Medicare beneficiaries perform in the areas reviewed by the Medicaid agency. The more comprehensive the review in areas that may concern seniors, the more the monitoring reports can tell you. While these reports address areas of concerns for women of childbearing age and children (e.g., child immunization rates, prenatal visits), some of the information they contain can speak to how well HMOs treat their populations at large. For example, the Medicaid agency in Maryland looks at a sampling of patient records and reviews the adequacy of medical records documentation. How well HMOs document patient care is a valuable piece of information for a Medicare beneficiary choosing a plan.

WHERE DO YOU GET MEDICAID MONITORING REPORTS?

Request Medicaid monitoring reports from the state Medicaid agency in your state. Ask for the division that monitors Medicaid HMOs. Often that will lead you exactly where you need to go. Don't let anyone tell you that monitoring reports do not exist. They are required under federal law and all state Medicaid agencies must contract with independent agencies to conduct these reviews. In your request to the state Medicaid agency, be clear that you are not interested in any patient-specific information and that you would like to receive the reports minus any privileged information. See the Appendix for the name of the agency that oversees Medicaid in your state.

WHAT ARE THE PROBLEMS WITH MEDICAID MONITORING REPORTS?

Medicaid monitoring reports are designed to monitor the care Medicaid beneficiaries receive from HMOs. Because the needs of the majority of Medicare enrollees are very different from those of the typical Medicaid beneficiary in an HMO, much of the information in these monitoring reports will not be useful for our purposes.

PENALTY INFORMATION

WHERE DO YOU GET PENALTY INFORMATION?

You can obtain information about what penalties the state has assessed against HMOs from a number of sources: the state's Attorney General's office, the state's Department of Insurance (for commercial plans), the state health department, and the state Medicaid agency.

WHY DO YOU WANT PENALTY INFORMATION?

Penalty information can tell you two things: 1) whether HMOs have failed to comply with the laws and what the nature of the infraction was; and 2) what penalties the state applies to plans for violations of the law. You should look back several years to be sure that HMOs don't repeat offenses often and that penalties fit the offense. Advocates should note overly lenient penalties and urge the state to penalize HMOs that break the law or otherwise perform poorly.

WHAT ARE THE DRAWBACKS TO USING PENALTY INFORMATION?

Penalty information is not made available in every state. Arizona, for example, does not release information about HMO penalties.

PHYSICIAN INFORMATION

WHAT KINDS OF PHYSICIAN INFORMATION ARE AVAILABLE?

In most states, HMOs are required to establish a credentialling system to assess and reassess the qualifications of its providers. The National Committee on Quality Assurance (NCQA) also requires HMOs to have a credentialling system as part of its accrediting standards. Many HMOs include information about the backgrounds of physicians with their provider lists. Some states require this, and others do not. Request this information from individual plans. If they do not cooperate, check with your state's department of insurance to see whether plans are required to release this information.

Usually the information that you can obtain about physicians pertains to board certification. Physicians may apply to receive certification from 24 different specialty boards approved by the American Board of Medical Specialties. These specialty boards determine whether candidates have received sufficient educational preparation to meet established standards, and they evaluate candidates for certification through rigorous examinations. HMOs may include information on physician board certification in their provider lists.

WHAT ARE THE DRAWBACKS TO USING PHYSICIAN INFORMATION?

None of the types of physician information discussed above directly measure the quality of care that physicians give. Board certification simply tells you whether a physician has passed a comprehensive test in his/her specialty. Malpractice information is not a direct indicator of quality. Sometimes people sue for malpractice and win, even when a doctor is not to blame. And of course, many doctors who have been negligent or reckless are never sued.

WHAT DO YOU DO WITH PHYSICIAN INFORMATION?

Advocates can provide a service to the public by simply publishing the information and distributing it to consumers.

[Return to top]

APPENDIX

Example of Presentation of Medicare HMO Reconsideration Data
Prepared by Families USA

STATE GOVERNMENT AGENCIES
WITH AUTHORITY TO REGULATE HMOS

This Guide contains tables and exhibits which are not posted here. To obtain a print copy of the guide, including tables and exhibits, contact Families USA at 202-628-3030.

[Return to top]