Connecticut's Medicaid Managed Care Program
An Interim Report
Medicaid expenditures are the largest single item in Connecticut's budget, representing in excess of 20 percent of the general fund. Administered by the Department of Social Services (DSS), this federal-state program provides health related services to over 315,000 state residents.
In 1995, DSS began implementing a managed care program, fundamentally transforming the way health care is delivered to a significant portion of the state's Medicaid eligible population. The first group transitioned to managed care comprised the Aid to Families With Dependent Children (AFDC) and AFDC related population. Enrollment of this group in managed care, approximately 235,000 eligibles, is scheduled for completion by January 1997.
In contrast to existing historical overviews, this report is intended as the second in a series of analyses designed to highlight key elements of Connecticut's Medicaid managed care program as they evolve.
Access to Health Care
In order to depart from traditional Medicaid fee-for-service (FFS) health care, offering in its place a managed care program, DSS sought, and in July 1995 obtained, a program waiver from the Department of Health and Human Services. This waiver, under 1915(b) of the Social Security Act, was a prerequisite to constructing a health care delivery system based upon managed care networks.
Under FFS, access to service is available from any health care provider that accepts Medicaid patients. In contrast, the program waiver permits DSS to contract with several Managed Care Organizations (MCOs) to offer comprehensive health coverage through provider networks.
In an effort to insure that each MCO has sufficient resources to serve its enrollees adequately, DSS has developed a methodology for gauging network adequacy. Using calendar year 1994 as its base period under the FFS system, DSS established the following ratios by provider type.(See Endnote 1).
|Provider Type||Number of Eligibles
|Adult Primary Care||387|
|Children's Primary Care||301|
|Obstetrics and Gynecology||835|
Each network is evaluated on a county-by-county basis. The above ratios are multiplied by the plans' number of providers in a given county to establish the maximum number of eligibles that MCOs are permitted to enroll. By way of illustration, the number of eligibles allowed per adult primary care physician is 387. If an MCO has 75 adult primary care physicians in a certain county its enrollment ceiling for that provider type in that county is 29,025 (75 x 387).
The minimum ceiling of the five provider types within a county determines a plan's overall enrollment ceiling. For example, if an MCO has in one county many providers in every category except children's primary care, for which it has only one provider, the plan's county ceiling would be limited to 301 enrollees.
In order to insure full enrollment of the targeted population in managed care, two designated providers were chosen to serve as default plans for those Medicaid eligibles who did not voluntarily select an MCO during their scheduled initial enrollment period. Oxford Health Plan was selected as the designated provider for western Connecticut; BlueCare Family Plan -- a Blue Cross product serving the Medicaid population -- was chosen for eastern Connecticut. Both designated providers were required to stay within the above-described ceilings in each county for which they served as the default plan.
Suspension of Enrollment
On March 11, 1996, DSS suspended BlueCare's designated provider status; BlueCare was also suspended from voluntarily enrolling eligibles in Tolland county which lacked any network dentists. The suspension prompted Senators Harp and Prague, Chair and Vice-Chair, respectively, of the Medicaid Managed Care Council, to request a review by the State Comptroller of the managed care program.
To address the specific issue of BlueCare's network capacity, and related matters, staff from the Comptroller's Office have had meetings and follow-up telephone interviews with key DSS personnel. In addition, this office has reviewed correspondence between DSS and BlueCare, and other pertinent documentation, relating to that MCO's provider network.
The review team has concluded that BlueCare made a good faith effort to expand its provider network as its enrollment grew. DSS, in turn, granted BlueCare appropriate leeway as the MCO attempted to correct network deficiencies. However, its analysis in March of BlueCare's network deficiencies in the context of rapid and continuing enrollment led DSS to suspend BlueCare's designated provider status for April and May 1996.
Specifically, the limited dental network resulted in excess enrollment in Hartford, Fairfield, New London and Tolland counties. In Windham county, excess enrollment was due to the limited number of providers serving children. Ultimately, BlueCare did eliminate the above-referenced deficiencies in its network and was reinstated as the designated provider for eastern Connecticut, effective June 1, 1996.
Limited Value of Current Enrollment Ceilings
In the absence of reliable measures of access to care, DSS opted to use provider ratios as a preliminary benchmark of appropriate levels of plan capacity. However, there are limitations to the ratios' value as measurements of adequate health care access and delivery. First, the ceilings are based on provider/enrollee ratios within each plan. They do not indicate a provider's total patient count. DSS has maintained a maximum ratio under FFS of 1 primary care physician (PCP) to 1200 patients when such physician serves only the Medicaid population. Under managed care, the enrollment ceiling for primary care physicians is 387 for adult PCPs and 301 for pediatricians. However, these limits are per plan; a physician may be under the ceiling for each MCO in which he or she is a member but in the aggregate treat more than the previously established 1200 patient limit. Additionally, the ceilings do not take into consideration the non-Medicaid population seen by providers.
DSS acknowledges these concerns noting, however, that providers are often unable to estimate their capacity. The agency also concedes that it lacks the resources to count the total number of managed care patients per provider across plan lines.
A second measure of plan performance has been a series of member satisfaction surveys conducted by MAXIMUS, Inc. under the auspices of the Children's Health Council and funded from DSS' budget. (See Endnote 2).Three enrollment surveys have been conducted to date: in December 1995 for eligibles enrolled in August-October 1995; in April and May 1996 for eligibles enrolled in November 1995; and, in June 1996 for eligibles enrolled in February 1996.
The survey instrument has undergone revision after each poll. However, all three surveys reported similar experiences for new enrollees in the program. Noteworthy findings include the following:
The findings suggest that plan dissatisfaction is not a significant issue at this time. However, the ability to change plans every month appears to diminish enrollee efforts at resolving issues by complaint or through the grievance process. That process should be encouraged as frequent plan changes have an adverse effect on continuity of care.
Provider ratios attempt to measure plan access. Membership surveys gauge plan satisfaction, an important but subjective appraisal tool. However, in order to achieve an objective, comprehensive analysis of plan performance, each MCO is required to submit a series of reports to DSS.
As one of the initial steps in implementing Medicaid managed care, in February 1995, DSS issued a Request for Application (RFA). The RFA, produced jointly by DSS and its consultant, Lewin-VHI, contained a comprehensive set of reporting requirements for MCOs.
One set of state requirements specifically involving Medicaid eligibles under age 21 is designed to report compliance with a federally mandated program for early and periodic screening, diagnosis and treatment (EPSDT), known in Connecticut as HealthTrack.
Additional reporting requirements set forth in the RFA involve inpatient utilization; mental health and substance abuse service overview (measuring inpatient, day treatment and outpatient treatment); readmission rates for both behavioral health categories; other services (including dental, prescription, vision and hearing screenings); immunization overview; and, maternal and prenatal care reports. Through a series of meetings between DSS staff and the MCOs, some modification to the original reporting formats has been agreed upon.
On March 18, 1996, the Association of Connecticut HMOs wrote to Joyce Thomas, Commissioner of DSS, raising several concerns with both the time frame for submission of the initial reporting series and purported difficulties associated with collecting the requisite data for specific reports.
In her response, Commissioner Thomas noted that the MCOs knew of the reporting requirements at the time they reviewed and responded to the RFA and that certain minor adjustments were permitted through the ongoing meetings between DSS and the MCOs.
DSS did agree to the delay of two reports. The first, a section of the Annual Immunization Overview, calls for MCOs to report vaccines received by children both during and prior to enrollment in their health plan. Submission of this report will be held in abeyance pending the establishment of a statewide immunization registry (discussed below). The Maternal and Prenatal Care Report requires information that can only be obtained from the Department of Public Health (DPH). DPH, DSS, the Connecticut Hospital Association and the MCOs are engaged in developing procedures for sharing the requisite data between the parties. Outstanding issues involve funding and confidentiality of records.
The initial set of reports was due May 1, 1996, reflecting plan experience for the last quarter of 1995, or from a plan's initial participation date, whichever was earlier. The second set of reports was submitted on June 30th of this year.
Several reports were submitted late and many were incomplete. Individual reports contained data which were either implausible or clearly incorrect. Nevertheless, sufficient data were presented to raise concern with perhaps the most significant performance measurement for the AFDC population as a whole -- the EPSDT report.
One component of the report, the participation ratio, measures the extent to which eligibles who should be screened during the year receive at least one initial or periodic screening service. The contract calls for an improvement of 15 percentage points above the participation ratio levels reported for Federal Fiscal Year 1995. Under this formula, the participation target levels for 1996 are 77 percent for children under one; 56 percent for children between one and five; 55 percent for children between six and fourteen; and, 38 percent for the fifteen to twenty year old population. However, for these four age brackets, the median participation ratios of all plans during the first quarter of 1996 were 39, 51, 35 and 34 percent, respectively.(See Endnote 3).
The RFA explicitly obligates the MCOs to conduct "...outreach to the families of Medicaid children to bring them into the HealthTrack [EPSDT] program."(See Endnote 4). This contract provision illustrates a critical distinction between Medicaid FFS and the managed care program. Under managed care, there is an affirmative obligation on the part of the MCOs to do just that -- effectively manage the delivery of health care to enrollees. Having sufficient capacity to serve its membership is only part of an MCO's responsibility. MCOs must determine why enrolled children are not participating in the program at acceptable rates and then take appropriate corrective measures.
Among the factors impeding the collection and reporting of pertinent data are the following:
Provider compensation under managed care may diminish the inducement for timely reporting. Under FFS arrangements, participating physicians are indemnified for services performed. If procedures are not reported in an acceptable format, the physician is not paid. Under Medicaid managed care, DSS pays the MCOs on a capitated basis, i.e. a specified amount per patient per month, regardless of the care delivered. Certain MCOs reimburse providers per service; however, others parallel the state's payment method and compensate providers on a capitated basis. This latter category of providers, therefore, does not have the financial incentive inherent in a FFS system to report encounter data to the MCOs.
There appears to be some confusion regarding the precise codes to be used for recording procedures. Therefore, plans are unable to rely exclusively on data received from their providers to determine which screening procedures have been performed. In several categories, pertinent data can only be obtained from patient chart review, a time-consuming but essential process.
DSS and the MCOs are in the process of comprehensively reviewing the state's reporting requirements in light of one year's experience with Medicaid managed care. The current contract will require amendment in order to include the findings of this task force.
Data collection is further hampered by the federal requirement that enrollees be permitted to change plans every month. For example, an immunization series may require administration over a lengthy time-frame and eligibles may be enrolled in several different MCOs during that period. Consequently, providers will need to furnish immunization data on a single patient to several MCOs.
In order to alleviate this problem, DSS is working with the MCOs to expand the Hartford Immunization Registry -- a program originally designed to provide a centralized repository of immunization records for Hartford's Medicaid enrolled children. The intent is to have all MCOs maintain immunization data centrally. Implementation of this plan should facilitate the accurate reporting of immunization data on a statewide basis. Of course, many children have intermittent Medicaid eligibility and will often receive part of an immunization series outside of the Medicaid program. Consequently, chart reviews, including patient provided history, will remain an integral part of the reporting process.
One of the difficulties faced by both the MCOs and DSS in reporting and analyzing data is the absence of a standard set of performance measures for the Medicaid population.
In an effort to address this concern, the National Committee for Quality Assurance (NCQA) has adapted for Medicaid managed care a series of performance measures originally developed for use by employers, health plans, and other interested organizations. These performance measures are designed to assist health plans and employers evaluate and compare specific plan performance; the first Health Plan Employer Data and Information Set (commonly referred to as HEDIS) was completed in 1991.
HEDIS currently measures five major areas of health plan performance: (1) Quality of Care; (2) Member Access and Satisfaction; (3) Membership and Utilization; (4) Finance; and, (5) Health Plan Management and Activities.
Medicaid HEDIS, released in draft form in July 1995 and in final version in February of this year, contains certain measurements not found in the parent data set (currently HEDIS 2.5). Conversely, measurements contained in HEDIS 2.5 which are not considered relevant to the Medicaid population are deleted from Medicaid HEDIS.
The federal Health Care Finance Administration (HCFA) has endorsed use of Medicaid HEDIS by states in assessing performance under their respective managed care programs. Commissioner Thomas has similarly recommended modifying the reporting requirements originally formulated in the RFA in order to employ Medicaid HEDIS measurements wherever possible.
In advocating its use, DSS recognizes that in particular areas the state reporting requirements are preferable to the measurements delineated in Medicaid HEDIS. Significantly, Medicaid HEDIS does not include all the information Connecticut requires to measure EPSDT compliance. An explanation for this apparent anomaly is found in the pertinent HCFA regulations.
With certain exceptions (e.g., the Childhood Immunization Schedule) HCFA has provided only general guidelines for implementation of Medicaid's EPSDT program, leaving to each state the responsibility for developing the specific measurements by which it will gauge program performance. For example, in addressing the matter of a periodicity schedule -- the frequency and schedule at which screenings, vision, hearing, and dental services should occur -- HCFA simply requires that these interventions "... be provided at intervals which meet reasonable standards of medical practice. States must consult with recognized medical organizations involved in child health care in developing reasonable standards."(See Endnote 5).
As considerable latitude is extended to each state in fulfilling the EPSDT requirements, it is not surprising that Medicaid HEDIS fails to include all of the data elements by which Connecticut tracks performance. Consequently, at least for the present, there continues to be the need for separate reporting in this area.
Despite its current limitations, Medicaid HEDIS provides a nationally recognized standard for judging the performance of Connecticut's MCOs. For this reason, it can serve as the basis for comparison with Medicaid managed care programs in other states.
It is the review team's recommendation that DSS seek an active and direct role in any future revisions of HEDIS in an effort to have those measurements Connecticut has found valuable incorporated into this data set. Several states were represented on the Medicaid HEDIS Work Group. It is not unreasonable to assume that the current version of Medicaid HEDIS reflects in part the reporting concerns of those states. ++ See footnote.
As noted previously, under FFS Medicaid providers are paid per procedure. In Connecticut, FFS Medicaid claims are reviewed and, where appropriate, paid by the state's agent, Electronic Data Systems Corporation (EDS).
In conjunction with implementation of Medicaid managed care, DSS and EDS are installing a new tracking system, Advanced Information Management (AIM). Under AIM, providers will continue to submit pertinent encounter data to the MCOs; the MCOs, in turn, will transmit the encounter data to EDS for storage on the AIM system. Information will be reviewed to insure that providers are submitting encounter data to the MCOs correctly and completely.
Unfortunately, implementation of AIM is behind schedule; only the pharmaceutical component is operational at the present time Moreover, AIM was designed to mirror the current EDS data system even though claims are no longer paid to the provider directly from the state. As a result, many of the data fields included in AIM are extraneous to a managed care program.
In particular, the billing data fields incorporated into AIM, essential in a FFS system, are irrelevant in a managed care program wherein the MCOs assume the financial risk for services in return for a capitated payment from the state. Consequently, AIM's design anticipates receipt of information which in many cases does not exist. It is unclear at this time how AIM is to be modified in order to accept the requisite encounter data from the MCOs, bypassing those fields for which data are unavailable.
Nevertheless, once completed the installation of AIM will afford DSS an excellent opportunity to analyze performance under managed care Again using EPSDT as an example, all the data elements required to track compliance will be included in the AIM system. AIM will allow DSS not only to replicate the reports currently required of each MCO but will also enable the agency to analyze aggregate data across plan lines. DSS will have the capability of assessing encounter data by any data field in the AIM system including, but not limited to, geographic region, provider, age, and procedure.
Consolidation of encounter data will also provide DSS the opportunity of issuing "report cards" for each MCO. These evaluations will both assist eligibles in plan selection and serve as an incentive to the plans themselves to continually improve health care delivery.
Clearly, additional resources will be required to maximize the full potential of AIM. However, this office believes that the benefits to be derived through centralized data analysis, in terms of program management, justify the additional costs.
Medicaid Managed Care Council
Through passage of Special Public Act No. 94-5, the legislature established the Medicaid Managed Care Council, charged with advising DSS on the planning and implementation of Medicaid managed care and with monitoring the program once established.
Council members represent a broad spectrum of stakeholders involved with delivery or oversight of health care in Connecticut (see Appendix "A"). The Council regularly receives reports from DSS and Benova Systems, the enrollment broker for the program. Additionally, Council members themselves may report on specific activities pertaining to Medicaid managed care in which they are individually involved. As warranted, other parties are asked to appear at Council meetings to address particular elements of program implementation.
The collective experience of Council members pertaining to health care generally and the Medicaid population specifically is considerable and the Council performs a valuable and much needed oversight function in relation to the managed care program.
Unfortunately, no one is assigned to the Council as staff on a full-time basis limiting its ability to effectively review the large volume of data brought to its attention by DSS, Benova Systems, the Children's Health Council, and the insurers themselves. Additionally, the Council is not reporting its activities and progress to the General Assembly quarterly, as required by statute.
It is the research team's recommendation that the Council hire a full-time staff person with the appropriate technical background who will be responsible for performing ongoing analyses of the managed care program and reporting the Council's findings.
In the coming months this office will look at additional elements of Medicaid managed care. In August, DSS, in conjunction with the Department of Children and Families, commenced enrollment of foster care and subsidized adoption children in the managed care program. This group's enrollment will focus attention on the state's ability through managed care to serve a population with significant psychological and substance abuse involvements. Further, the additional care required for these children provides an opportune time to evaluate the adequacy of the capitation rate structure.
DSS is currently negotiating with the Connecticut Peer Review Organization (CPRO) to perform an external quality review of the managed care program. CPRO's work has primarily involved the state's Medicare population with particular emphasis on quality of care issues in the hospital setting. The age of the AFDC population, and the medical issues associated with the young, will require significant adaptation by CPRO in order for it to effectively assess the quality of care being delivered under the managed care program. The scope and depth of CPRO's effort is a matter for future consideration by this office.
Finally, it is our intention to compare Connecticut's Medicaid managed care program with the experience of other states.
The Comptroller's office wishes to thank the staff of the Department of Social Services, and Judy Solomon of the Children's Health Council, for their generous and valuable assistance in the preparation of this report.
Back to Table of Contents
Back to Comptroller's Home Page