A Guide to Monitoring Medicaid Managed Care

Chapter 3:
 Enrollment and Disenrollment Data

Under managed care, Medicaid beneficiaries are generally offered a choice between at least two health plans. When enrollment in managed care is voluntary, Medicaid beneficiaries may choose either a health plan or Medicaid fee-for-service. When enrollment in managed care is mandatory, Medicaid beneficiaries must choose one of the health plans offered. In both voluntary and mandatory managed care programs, beneficiaries who do not make a choice within a given time frame will be assigned to a health plan by the state. 

Enrollments in Medicaid managed care plans turn over constantly. Enrollees leave and join plans as they lose and gain Medicaid eligibility and leave plans that do not meet their needs and join plans that promise better service or benefits. The rates at which individuals enroll (join a plan) or disenroll (leave a plan) can be useful indicators of consumer satisfaction. Enrollment and disenrollment statistics can also provide information about the success of a state's efforts to encourage Medicaid beneficiaries to choose their own managed care plans and about the impact of changing Medicaid eligibility on continuity of care. Each state categorizes enrollment and disenrollment statistics somewhat differently. The following are some factors that advocates and state officials might examine:

Voluntary enrollments: The number of people who choose to enroll in a particular managed care plan after being offered a choice between at least two plans in their geographic area.

Auto-assigment enrollments (also known as default assignments): The number of people whom the state or its contractor has assigned to a managed care plan. These beneficiaries have failed to indicate an enrollment choice within time frames established by the state.

Disenrollments: The number of people who leave a managed care plan, either to go into a different managed care plan or to go into fee-for-service Medicaid. Advocates need to consider four types of disenrollments-rapid, emergency, routine, and involuntary disenrollments.

• Rapid disenrollments are the disenrollments that occur within the first 90 days of plan membership. High rapid disenrollment rates usually indicate marketing problems as new enrollees try to use their plan's services and learn that the plan is not what they expected.
• Emergency disenrollments (also called "urgent"or "expedited" disenrollments) occur when someone is improperly placed in a health plan. For example, some states exempt certain people from mandatory managed care. If a state mistakenly assigns an exempt person to a plan, the state may permit the person to disenroll immediately so that he or she can continue seeing providers outside managed care plans' networks. Large numbers of emergency disenrollments suggest auto-assignment problems rather than marketing problems.
• Routine disenrollments occur for a variety of reasons throughout the year or during open enrollment periods. For example, enrollees may leave a plan because they want to use a provider in a different network.
• Involuntary disenrollments are disenrollments that occur for reasons other than the beneficiary's choice. Generally, involuntary disenrollments result from loss of Medicaid eligibility. These statistics might be useful, for example, if you want to analyze disruptions in care due to changing Medicaid eligibility. Sometimes, a managed care plan may disenroll a member whom they consider disruptive or noncompliant with treatment. Find out whether your state records disenrollment at the request of a managed care plan and disenrollment due to loss of Medicaid eligibility in separate categories.

Disenrolling from Medicaid Managed Care Plans

In some states, Medicaid managed care enrollees may disenroll from plans monthly. In other states, Medicaid managed care enrollees are "locked in" to plans for six months or nine months. When managed care enrollment is voluntary, enrollees may disenroll from a managed care plan and return to fee-for-service Medicaid or switch to a different plan; when managed care enrollment is mandatory, enrollees can switch plans. Federal law allows Medicaid managed care enrollees to disenroll from their plans for "cause" at any time. States can restrict disenrollments "without cause" to the first 90 days of a 12-month enrollment period. 1 Advocates should keep differences in enrollment periods in mind when comparing their state's disenrollment statistics with statistics from other states. 

Involuntary disenrollment rates vary among states, depending on the length of a state's Medicaid eligibility periods. Generally, states must redetermine a beneficiary's Medicaid eligibility every 12 months. During that year, beneficiaries must report changes in their income or other circumstances that would affect eligibility, and the state must "promptly" redetermine Medicaid eligibility. However, states can opt to guarantee all Medicaid managed care enrollees six months of enrollment in a managed care plan, even if the beneficiary would otherwise lose Medicaid eligibility during the 6-month time period. States can also opt to guarantee children 12 months of continuous eligibility for Medicaid, even if the family income increases. (For adults, 12-month eligibility can only be provided through a federal waiver.)

Usefulness of Enrollment/Disenrollment Data

The U.S. General Accounting Office (GAO) wrote about the importance of disenrollment data for monitoring Medicaid managed care:

Collecting and analyzing data on disenrollments can provide important insights into plan performance. . . . More analyses of these disenrollment data-even if the rate at which beneficiaries leave or switch plans is low-could reveal significant problems. Disenrollments concentrated in an area or among people having similar needs, such as people with AIDS, may indicate a potential problem in a plan. Also, any plan having higher disenrollment rates than other plans may merit scrutiny to determine the reason. 2

In addition, the Office of the Inspector General in the federal Department of Health and Human Services found that disenrollment data are a meaningful performance indicator for Medicare enrollees, another group of publicly insured individuals-a finding with implications for Medicaid beneficiaries. The Office of the Inspector General coupled information from surveys of Medicare beneficiaries 3 with disenrollment data. Beneficiaries who reported certain problems on surveys 4 and those with disabilities were more likely to disenroll from managed care plans than other Medicare beneficiaries. 5

All other factors being equal, voluntary disenrollment rates among Medicaid managed care plans within a state should be roughly comparable if Medicaid enrollees are equally satisfied with their plans. Particularly high disenrollment rates may indicate problems. Perhaps a plan misled beneficiaries during marketing, and people disenrolled in high numbers when they discovered a plan could not live up to its promises. Perhaps people had difficulty obtaining care or felt that their plan provided poor quality care.

Federal Requirements for Enrollment/Disenrollment Data

The Balanced Budget Act of 1997 sets forth some general rules about enrollment and disenrollment in Medicaid managed care plans. States must follow these rules unless HCFA has approved their request for a "waiver" of a specific provision. For example, the state must provide prospective enrollees with comparative information to use in making managed care choices, and the auto-assigment system must attempt to maintain existing provider-patient relationships or relationships with traditional providers. If this is not possible, auto-assignments must be distributed evenly among managed care plans until the plans reach their enrollment capacities.

As of Spring 2000, regulations implementing the Medicaid managed care provisions of the Balanced Budget Act have not been finalized. Federal regulations that predate the Balanced Budget Act and that are still in effect require state Medicaid agencies to monitor the enrollment and disenrollment practices of managed care organizations to ensure that plans do not discriminate based on health status.   State Medicaid agencies must also arrange for an annual medical audit of each contracting health plan. As part of their review, auditors examine data on reasons for enrollment and disenrollment from a health plan. Health plans that contract with Medicaid agencies must have internal systems for monitoring and improving the quality of their care. HCFA suggests that plans consider measures of enrollee satisfaction, such as information on disenrollments and requests to change providers, as one source of information for quality improvement projects.

HCFA's regional offices monitor the states' implementation of Medicaid managed care and oversight of managed care contractors. HCFA has drafted a monitoring guide that serves as a resource tool for the regional offices. It comprises a set of questions from which reviewers can pick and choose in their monitoring efforts. One set of questions addresses enrollment/disenrollment and auto-assignment. Among other questions, the HCFA guide asks:

• What percentage of beneficiaries are auto-assigned? Has this percent decreased since the inception of the program? since the last monitoring period? If not, why not? What steps has the state taken to minimize the percent of beneficiaries that must be assigned to managed care organizations?
• What monitoring does the state do in regard to plan disenrollment and requests for changes of primary care providers (either using its own resources or by asking managed care organizations to provide figures)?
• What is the state's mechanism for handling plans with high rates of disenrollment and changes in primary care providers?

How Your State Should Keep Enrollment/Disenrollment Statistics

To be useful as an indicator of satisfaction, states must separately record voluntary enrollment, auto-assignments, voluntary disenrollments, and disenrollments due to loss of Medicaid eligibility or loss of Medicaid managed care eligibility. States may also wish to record disenrollment statistics for vulnerable populations.  For example, California maintains statistics for various ethnic and linguistic populations.

Some states ask each member why he or she wishes to disenroll and report this data in categories such as dissatisfaction with quality of care, access to service, marketing practices, distance to services/transportation problems, prior relationship with provider not in a plan, provider changed plans, enrollee moved out of area, etc. Advocates should urge their states to record reasons for disenrollment or, at a minimum, to survey a sample of disenrollees from each plan. Many states require managed care plans to survey enrollees' satisfaction annually. One standardized instrument for doing this is the Consumer Assessment of Health Plans (CAHPS®), developed by Harvard Medical School, RAND, and the Research Triangle Institute for the Agency for Health Care Research and Quality. Advocates should urge states and plans to survey disenrollees using similar standardized surveys. 

When states begin new Medicaid managed care programs or renegotiate managed care contracts, advocates may be able to influence the enrollment information their states collect. Advocates might recommend: 

• public reporting of enrollment and disenrollment information in specific categories, 
• state surveys of both enrollees and disenrollees on satisfaction with managed care plans, 
• examination by plans of reasons for disenrollment as part of their quality improvement activities, 
• examination by external organizations of disenrollment rates as part of their annual review of managed care plans' quality performance, 
• state-provided separate enrollment and disenrollment data for vulnerable populations such as persons with disabilities or persons whose primary language is not English.

What You Can Learn From Enrollment/Disenrollment Data

Advocates may want to use enrollment and disenrollment data to answer the following questions:

1) Do most people voluntarily select their plans, so that auto-assigment rates are minimal? Are auto-assigment rates particularly high among special populations, such as those whose primary language is not English? How do the monthly proportions of people auto-assigned to each plan compare with the proportions that voluntarily select each plan? Is the state assigning people to the plans consumers most often select themselves, or do auto?assignments support unpopular plans? For answers, examine auto-assignment rates.

2) Are the proportions of enrollees who voluntarily leave some health plans higher than the proportions that voluntarily leave other health plans? Do high numbers of people disenroll rapidly from some plans, indicating perhaps that they misunderstood enrollment choices or that the state auto-assigned people to plans that did not meet their needs? For answers, examine voluntary disenrollment rates and rapid disenrollment rates.

3) What do we know about reasons for disenrollment? Do people tend to leave certain plans because of dissatisfaction with service or inability to see the doctor of their choice? For answers, examine disenrollments by reason.

Sources for Enrollment/Disenrollment Statistics

Enrollment and disenrollment statistics may be collected by the state Medicaid agency itself or its enrollment contractor or by contracting health plans. Data on reasons for disenrollment will probably be more reliable if they are collected by the state or its enrollment contractor rather than by health plans. Since states may penalize plans for quality and access problems, plans have an incentive to code disenrollments related to perceived quality and access in another category instead (such as "no reason given" or "other").

Besides examining disenrollment data, you may want to compare plans' disenrollment procedures. Do plans clearly inform members of their rights to disenroll? You can directly review managed care plan member handbooks to see how clearly they inform beneficiaries of their rights. You can also find out whether your state has already reviewed disenrollment procedures as part of a managed care contract compliance review or an external quality review. For example, in Maryland and the District of Columbia, external reviewers determine whether plans have clear written enrollment and disenrollment procedures that are explained in member handbooks and whether plans monitor voluntary disenrollments and work to resolve any systemic problems members identify as their reasons for disenrollment. Portions of these external reviews of managed care plans' performance are available to the public.

Contact your state Medicaid agency to find out who keeps reports on enrollments and disenrollments in your state and what information is recorded about enrollments and disenrollments.

Step-by-Step Analysis

This section will guide you through three general aspects of the enrollment/disenrollment process for which the numerical data can provide information:

• Auto-assignment Rates 
• Disenrollment Rates 
• Disenrollments by Reason

a. How to analyze auto-assignment statistics

Basically, the auto-assigment rate is the number of people assigned to a plan divided by the number of people who had an opportunity to choose a plan. This statistic is probably computed by your Medicaid agency. You may want to look at auto-assigment rates within your state over time to gauge the effectiveness of enrollment outreach and education, or you may be interested in comparing auto-assigment rates in your state to those in other states. To do this, you will need to know how your state counts auto-assignments. Besides counting the number of enrollees who did not choose a plan after they were given the opportunity, some states include the following groups in their auto-assignment rates:

• newborns who are automatically assigned to their mother's health plans; 
• family members who are automatically assigned to the health plan chosen by the head of household; and
• re-enrollment of beneficiaries who are automatically assigned to their former health plan following temporary lapses in Medicaid eligibility.

Before comparing your state's auto-assignment rates to another state's rates, make sure each defines auto-assignees the same way. 

States and consumer organizations want Medicaid beneficiaries to choose their own health plans so that they can find the provider networks and add-on benefits that best meet their needs. In states where managed care enrollment is mandatory, states assign Medicaid beneficiaries to plans if they do not make enrollment choices within a given amount of time. Many states contract with enrollment brokers to educate Medicaid beneficiaries about their managed care choices and to enroll them in plans. Some states have established goals for voluntary enrollment and have increased outreach efforts to achieve these goals. Some states use voluntary enrollment and auto-assignment statistics as a performance measure for enrollment brokers. They pay bonuses to enrollment brokers who keep auto-assignment rates below certain percentages. 

Advocates can monitor monthly auto-assignment rates to spot problems in a state's efforts to educate Medicaid beneficiaries about their enrollment choices. Advocates might urge states to suspend auto-assignments and increase outreach if auto-assignments are excessive. Some states set 20 percent as a maximum auto-assignment rate.

b. How to analyze disenrollment statistics

Look for variations among plans' disenrollment statistics. On a monthly basis, percentages of the membership that disenroll will likely be small. Plans might show, for instance, 1 to 3 percent of their membership voluntarily disenrolling. You will see more variation if you compute annual voluntary disenrollments as a percentage of average annual membership. We recommend the following method:

1) Add the monthly membership totals in a plan and then divide by 12. This will give you the plan's average yearly membership.

2) Add the number of voluntary disenrollments in each month of the year. (Do not include people who disenrolled from a plan due to loss of Medicaid eligibility.)

3) Divide the plan's total number of voluntary disenrollments during the year by its average annual membership and then multiply the result by 100 to obtain a percentage.

4) Similarly calculate disenrollment rates for any particular areas of study, such as emergency disenrollments; rapid disenrollments; or disenrollment rates by linguistic group, age, or disability status.

If your state reports monthly disenrollment rates (voluntary disenrollments divided by the month's enrollment), you can add these monthly figures to get annual voluntary disenrollment rates.

c. How to analyze reasons for disenrollment

Many states collect data on the reasons that members disenroll from health plans. Look at your state's overall data to determine the most common reasons for disenrollment. These reasons may indicate systemic problems. For example, in Florida, about 110,000 Medicaid HMO members voluntarily disenrolled from a plan between 1992 and 1994. The Florida Sun Sentinel used this statistic, along with a survey on reasons for disenrollment (people wanted their own private doctor, quit because the plan was not what they were told it would be, or quit because they did not like the plan) and anecdotal information, to call attention to widespread marketing and quality problems.

Compare disenrollment rates by reason among the health plans. Is the proportion of people who disenroll for a certain reason higher in one plan than in the other plans? If so, this may indicate a problem with a particular aspect of that plan. For example, there may be widespread discontent with a particular plan's member services department. 

In any particular category, monthly disenrollment figures may be quite small. You might need to group similar categories (e.g., health plan did not meet needs, beneficiary preference, and doctor did not meet needs might all be part of one "dissatisfaction" grouping) in order to see trends.

Cautions on interpreting enrollment and disenrollment data

Disenrollment data should be used with caution. A number of factors, including those listed below, could account for high rates.

• Even a small number of disenrollees will result in a high disenrollment rate in plans with few members. You may therefore want to compare only plans with average membership of at least 1,000 Medicaid beneficiaries annually. New plans first enrolling Medicaid beneficiaries may have high disenrollment rates because the plan has not achieved the same type of consumer loyalty that older plans have achieved. High disenrollment rates for more than one year are an especially significant problem indicator.
• Market conditions may result in high disenrollment rates. For example, Medicaid enrollees may disenroll from one plan in high numbers if a second plan offering better or additional benefits (such as adult dental care or vision care) begins marketing in the same community.

While advocates should avoid judging the quality of a managed care plan based solely on a single measure, they can serve the public by revealing plans with unusually high disenrollment ratios and by encouraging state agencies to investigate these plans to ensure that patient care is not compromised.

Endnotes

1  42 U.S.C. § 1396a-2.

2  Medicaid Managed Care: Challenge of Holding Plans Accountable Requires Greater State Effort (Washington, DC: General Accounting Office, GAO/HEHS-97-86, May 1997).

3  Office of Inspector General, Disenrollment Rates as Performance Indicators of Risk HMOs (Washington, DC: DHHS Office of the Inspector General 06-96-00432, October 1995).

4  Problems included patients' reports that their health was declining due to HMO care, long waits in their primary care doctors' offices, and the perception that their HMO seemed to place more importance on holding down costs than providing the best possible medical care.

5  Office of Inspector General, op cit.

6  Balanced Budget Act of 1997 (P.L. 105-33), Subtitle H codified at 42 U.S.C. § 1396a-2.

7   42 C.F.R. §§ 434.25-434.27 and § 42 C.F.R. 434.63.

8  42 C.F.R. § 434.53. Proposed regulations published in December 1999 would eliminate required medical audits.

9  Quality Improvement System for Managed Care (QISMC) (Baltimore, MD: U.S. Department of Health and Human Services, Health Care Financing Administration, September 28, 1999).

10  Health Care Financing Administration, Monitoring Guide: Medicaid Managed Care Initiatives (Baltimore, MD: U.S. Department of Health and Human Services, Health Care Financing Administration, Working Draft, January 1997). Though never finalized, this draft is still in use as of 2000. HCFA is working on an update.

11   The U.S. General Accounting Office has suggested that states require disenrollment data by eligibility category and that individuals requesting disenrollment be interviewed to gather more in-depth information about the care they received. Medicaid Managed Care: Serving the Disabled Challenges State Programs (Washington, DC: General Accounting Office, GAO/HEHS-96-136, July 1996).

12  Families USA Foundation, The Quality of Maryland and District of Columbia Medicaid Managed Care Plans: External Reviews (Washington, DC: Families USA, 1998).

13  Mary Kenneson, Medicaid Managed Care Enrollment Study: Report of Findings from the Survey of State Medicaid Managed Care Programs (Princeton, NJ: Center for Health Care Strategies, 1997).

14  Families USA Foundation, A Guide to Marketing and Enrollment in Medicaid Managed Care (Washington, DC: Families USA Foundation, June 1997) and Medi-Cal Community Assistance Project, Promoting Consumer Choice in Medi-Cal Managed Care (San Francisco, CA: Medi-Cal Community Assistance Project, October 1997). [Available from Health Access, San Francisco, CA, 415-395-7955 

15   Fred Schulte, and Jenni Bergal, "Profits from Pain: Florida's Medicaid HMOs," Sun Sentinel, Fort Lauderdale, Florida, December 11 and 12, 1994.