This Web site is a component of the SAMHSA Health Information Network
|
This Web site is a component of the SAMHSA Health Information Network |
 |
| | | | | | ||||||||||
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
This Web site is a component of the SAMHSA Health Information Network. |
|
Organization & FinancingSAMHSA Managed Care Initiative Training SessionsJanuary 1998 Dates: January 7, 1998-Parklawn Conference Center, Rockville, Maryland
Title: Behavioral Health Services for Youth in Managed Care Organizations Speaker: Sarah Scholle, Dr.P.H. Summary Dr. Scholle divided her presentation on the impact of managed care on behavioral services for children into several segments: (1) an overview of the diversity of behavioral health care arrangements in terms of their organization, financing, and coverage of mental health and substance abuse services, and the resulting challenges for researchers; (2) the presentation of preliminary data from the Allegheny County Children's Medicaid Evaluation (ACCME) Study, a Center for Mental Health Services (CMHS)-funded study, that compared managed care organizations (MCOs) with traditional fee-for-service (FFS), in terms of services used, enrollment, and service satisfaction, and (3) suggestions for future steps for policy and research. Child mental health services is a very complicated service sector. It involves a number of agencies, as well as the education, juvenile justice, child welfare, and mental health/retardation systems. Before trying to understand the impact MCOs have on the behavioral health care services that children receive, it is important to determine what existing systems have to offer. It is important to recognize that the roles undertaken by any service sector, and the benefits provided, can vary at each local, regional, and state level. Another dimension to consider is the complexity of the contracting issue in MCOs. For example, the payer may be the state Medicaid program, but dollars from the state mental health authority, child welfare, or the education system could be included. These payers contract with MCOs, which, in turn, may offer integrated or carve-out behavioral health care services. This produces a multiplicity of different relationships for both providers and patients. Additional issues that distinguish MCOs include ownership (providers or for-profit company); enrollment (mandatory or voluntary); funding streams; provider networks; type of reimbursement policies; and accountability standards. (Beginning in January, 1999, Allegheny County, through the State, will contract with behavioral MCOs for mental health and substance abuse services, and Health Maintenance Organizations (HMOs) for traditional medical services.) What this demonstrates is that comparing MCOs to FFS is not a simple dichotomous comparison. What we have is managed care that exists along a continuum, on a number of different dimensions. MCOs need to be profiled on a variety of dimensions; the way that various dimensions work together and how they affect the availability and utilization of services, as well as outcomes over a period of time, need to be examined. Given these parameters, Dr. Scholle summarized the challenges for research on MCOs as: (1) understanding the existing child service system; (2) defining managed care more specifically; and (3) understanding the contract provisions, interpretations, changes, and enforcement. Dr. Scholle next presented the preliminary results from the ACCME study. This study was begun more than two years ago, with the purpose of examining the enrollment, utilization, satisfaction, and outcomes of care for children with Serious Emotional Disturbance (SED) enrolled in the Medicaid program. The data sources for the 388-family study were: (1) administrative data from the state Medicaid program, including enrollment eligibility files and claims; (2) structured family interviews with the parents/guardians and with the child at baseline, and follow up interviews conducted over a 12-month period; and (3) an ethnographic substudy consisting of qualitative interviews with study families. The administrative data were used to identify children who received an intensive mental health service, and were still enrolled in Medicaid at the time of the baseline interview. In addition to services and satisfaction data, a variety of instruments were used to evaluate the child's symptoms and functional impairment. Dr. Scholle presented the following data concerning utilization in MCOs versus FFS:
This is preliminary data is based on parent reports. Validation currently is being sought using medical records from mental health providers and claims data records. "Use of services" is defined as at least one visit within the last three months of interview date. These data prompt questions concerning the reason(s) why there is lower specialty mental health care use in the MCOs. Potential hypotheses include the following:
Dr. Scholle drew on enrollment, disenrollment, and satisfaction data, obtained from the qualitative family interviews, to probe these questions. Previous literature on MCO enrollment suggests that there tends to be selection bias when there is voluntary enrollment in MCOs . The enrollment data were analyzed to determine which children joined MCOs and, among those who joined, who stayed in MCOs. Data from the study produced the following analysis: >
Another aspect supporting selection bias occurs through disenrollment. An analysis of the families enrolled in MCOs, followed over time, found that the disenrollent of children enrolled in MCOs, with general medical or pregnancy admissions, was about 20 percent after six months-at which point the disenrollment rate leveled off. For those children with a psychiatric admission, however, disenrollent continued-there was no leveling off of the rate. After two years, only approximately 35 percent of this population was still in the MCO. Based on this data, one of the hypotheses to emerge is that children with psychiatric admissions are encountering more barriers to services. Data from the ACCME study, where all the children had a psychiatric admission or other intensive mental health service, were analyzed to determine how these children compare in the MCOs versus FFS situations. The minority families were more likely than Caucasian families to join MCOs. Disabled children (for project purposes, "disabled" was defined as SSI recipients) were more likely to stay in FFS than MCOs; however, the functional impairment of the children in the MCO group and the FFS group was very similar. The interviewer-rated assessment of functional impairment was nearly identical between the two groups, as was the parent-rated impairment scale. Even though there was a greater proportion of children on disability in the FFS sample, no higher level of functional impairment was seen, as compared to the MCO group. The second hypothesis was to consider if MCOs present more barriers to using specialty mental health care services. At the baseline interviews, families were asked to rate their satisfaction with their health plan. The measure was based on the Group Health Association of America's Consumer Satisfaction Survey. The data were presented on four different subscales of satisfaction: (1) provider (quality of care, interpersonal skills, convenience of appointments); (2) plan (availability of providers, range of services covered); (3) doctor choice (number and location of doctors available); and (4) mental health care (quality, range, and availability of services). In addition, specific questions were asked concerning administrative barriers to services. The results were as follows:
Using these data to answer the question, "Do MCOs impose greater barriers to accessing specialty mental health services," the results demonstrate that in MCOs: (1) there is lower satisfaction among participants, and this difference is strongest among minorities, and (2) participants are more likely to cite administrative barriers. Dr. Scholle then discussed the results of the qualitative interviews conducted with a subset of the families in the study. These open-ended questions obtained information about obtaining services, choosing a health plan, and the positive and negative aspects of specific health plans. The family perceptions of MCOs were as follows:
In summary, these preliminary comparison data of MCOs versus FFS support the following findings:
Dr. Scholle then underscored the challenges of this specific research project, explaining it was a simple comparison of children in MCOs versus FFS, based on the parameters of (1) voluntary enrollment, and (2) for-profit MCOs that subcontract to for-profit behavioral health care companies. This is the current state of the managed care environment in the Pittsburgh area-which may change as enrollment in MCOs becomes mandatory in the near future. She then presented some policy and research issues that are raised by these findings, and suggested how these issues might be approached. Policy issues include: (1) selection bias, (2) benefit design, (3) problems with access/continuity, and (4) the attention that needs to be paid to vulnerable subgroups. Suggested solutions for preventing selection bias include:
Access to care, which is limited not only by case management, but also by the way the MCOs set up their networks, can be improved through the use of open access plans (i.e., use of any provider) and alternative qualifications for providers. MCOs often require only accredited providers in their networks, eliminating the mental health service agencies that use Bachelor's level providers. Alternative methods for determining qualifications for participation in MCO networks need to be developed to allow for the use of specific providers with a history of providing these services. Continuity of care can be maintained through (1) the use of grandfather clauses-ensuring that families can "grandfather in" providers that they have developed relationships with over time, (2) long-term contracts between the state and MCOs and between MCOs and their providers, and (3) multi-system coordination, where funding streams include not only Medicaid, but the education system as well. Maintaining continuity of care is a part of improving benefit design which also includes: (1) coordination among child-serving sectors (to determine who is responsible for what); (2) blended funding streams (to allow for broader, more comprehensive services); (3) clear definition of MCOs' responsibility (i.e., clear contract language); and (4) broader auxiliary coverage (e.g., vision, dental, pharmacy). Finally, fixed standards of accountability are essential for the purpose of determining if MCOs are meeting their commitments to their enrollees. This research has demonstrated that accountability is enhanced, not only by looking at overall service utilization or satisfaction with care, but by looking at these areas within specific subgroups that are vulnerable to limitations in services. In addition, is it important to use instruments that are sufficiently sensitive to identify problems that families with special needs are encountering, not just aggregate measures. Dr. Scholle concluded by itemizing additional research issues that emerge when conducting these types of studies, such as:
Given that these factors do influence study design, she suggested:
Both MCOs and FFS are good systems; the challenge, however, is to blend the good of both to make a better system for families and youth. Question and Answer Session 17: Does your disenrollment data compare differentials by ethnic background? 1: We did and found that African Americans were slightly less likely to disenroll from MCOs than Caucasians. 17: Where is the "managed care" in an open-access plan? 1: MCOs limit access through prior authorization by their case managers, the setup of their provider network, and high utilization of "cost-effective" providers. In an open access plan, the MCO lose the ability to control the provider network, but can still manage the care through their case management. 17: In effect then, if open-access is more like FFS, only with more stringent utilization control, why not just recommend FFS? 1: The basic question is whether the MCO is limiting the network so severely so that it is affecting a certain subpopulation or region of base providers that these populations have been using for years. One of the things we are very interested in, with our findings of minority versus Caucasian differences, is whether it has to do with the kind of providers these families used before they enrolled in MCOs. If it is the restriction in the use of certain providers that is limiting the use of mental health services by certain populations we may want to think about how we can provide access. 17: Can you define "alternative qualifications?" 1: It means relaxing the accreditation requirements for providers to most suitably meet the needs of the patients. For example, agencies that provide therapeutic staff support or home-base services may have a team of providers-a psychiatrist or Doctorate-level provider who oversees the operation-but most of the one-on-one contact is between a Bachelor-level provider and the client. Bachelor-level providers have no standing in MCO networks, because the accreditation rules require that all of its providers must be a Masters-level therapist, with proper licensure in the state. So the Bachelor-level personnel, who may have a lot of experience, are excluded from the delivery of services. These people should somehow be included in the paradigm, as appropriate, to include a broader range of services. 17: Could there be different accreditation standards in rural areas than in urban areas, due to the difficulty of obtaining providers? 1: That has been debated. One of the usual provisions of the contract that the State develops with the MCO is a specific number of psychiatrists per population, psychologists per population, and the distance to care or services. In Pennsylvania we are very concerned about this-how to relax the accreditation standards to meet the needs of the population. On the other hand is it fair to relax the rules in some counties? Also, how can you force a MCO to create providers that don't exist. It is a problem for rural areas. MCOs, however, may in fact be positioned to offer more of a variety of services because they have a network. 17: I have a question about the service mix on one of the slides. I noticed that the rate of hospitalization was similar for both MCOs and FFS, although the ones with the psychiatric diagnosis would be considered the most impaired. Also, if MCOs are restricting access to care, one could argue that those who do not get early intervention care get sicker, therefore, you would expect the rate of hospitalization to be higher, but that was not the case. Can you explain? 1: The sample size is here is too small to draw an accurate conclusion. At three months, less than 10 percent of the children in MCOs had been in an inpatient or residential-type treatment facility. What will be interesting is if that same trend continues over time. We have mentioned that MCOs limit hospitalizations, and following that, there should be an increase in the other types of specialty services that are offered to children in MCOs. Over time, we may see more inpatient care, because these children are not receiving the types of services that prevent them from having to be admitted to an inpatient facility. We hope to address that as we see more children in the future. 17: Are you aware of any utilization rates/services satisfaction surveys comparing carve-out versus carve-in (integrated) models? 1: No, I have not yet seen a true integrated model. True integration would require the blending of case managers, data systems, funding streams, and medical information. 17: What is the age of the study population? 1: The children are ages 8 to 17, the mean age is about 12. 17: Some of your charts indicated that you controlled for some variables, such as pregnancy. Do you plan to include drug use in your analyses? 1: We have limited ability to extract that type of information from Medicaid claims as Medicaid generally doesn't pay for drug and alcohol services. We did capture that information-drug and alcohol use and the consequences-in our ACCME sample, but we have not analyzed that data at this time. 17: Are you planning to ask families in your 12-month follow up-those who have had FFS experience and then MCO experience-whether each experience was different, and how different? 1: We aren't asking that directly in our structured interviews, but perhaps we can capture some of that in our qualitative interviews. I'd be interested to ask that question.
Question and Answer Session
17: Could you explain how these families initially were enrolled in either MCOs or FFS? 1: Currently, in western Pennsylvania, enrollment in MCOs for Medicaid recipients is voluntary (until January, 1999). Families are invited to join, but they have the option of choosing FFS programs as well. As a point of information, HMOs in Pennsylvania have contracts with the state to provide all Medical services, but the Medicaid HMOs contract with separate firms to managed behavioral care services. So it is either an internally separate organization or an external organization that is managing the mental health and substance abuse services. 17: When you asked the questions about satisfaction with provider, are the families thinking of their primary care provider or their mental health provider? 1: We asked them to think about their plan in general, so they are thinking of all the doctors that they use. 17: Do you have any way of knowing if the families were having any trouble obtaining other types of medical specialty care or were the problems more pronounced with mental health specialty care? 1: No. Our study only included those with defined mental health problems. 17: Many Medicaid MCOs require the enrollee to select a primary care physician. Was mental health ever chosen as the primary care provider? 1: No, mental health providers cannot be a primary care provider in western Pennsylvania, but this is probably true in most MCOs. When MCOs become mandatory next year, there will be even more of a divergence between general medical care and mental health service, as each will be managed by a different organization. Part of the requirements of the Request for Proposals (RFP) put out by the state does require responding MCOs to discuss how they will integrate services, but as of now there aren't very many good integrated models, and the integration is not tied to any financial restrictions. 17: There's no qualifications for primary care physicians dealing with children with mental health needs? 1: No. 17: What plans does Pennsylvania have for incorporating the new changes in the Balanced Budget Act, which now no longer requires mandatory enrollment of children with special needs or now forces the state to ask for a waiver, into their managed care of Medicaid populations?. 1: Since Pennsylvania's waiver authorization has been approved, managed care is a mandatory situation. It is, however, difficult to tell where they will be next year as the waiver was approved under a managed care carve-out and next year there will be a new company that is a provider-sponsored network-very different from what exists currently. We will have to look at what and where the incentives are placed to ensure quality care. 17: Is there anyway in the satisfaction survey to ask questions specifically about the mental health provider , that is, the family perceptions of his/her qualification and how they feel they are being treated? 1: The data I discussed was specific to the health plan satisfaction subscale. We also asked families to rate their satisfaction with the specific provider or agency that is giving the services to their children. We haven't analyzed those data yet, but I know that MCOs, typically, do conduct provider satisfaction surveys, and require that from their providers. They are monitoring the responses to develop their networks. That is part of a whole series of things that MCOs use to develop their networks, including the financial profiles of the providers. 17: One concern among families is how accurately can they respond if they feel obligated to the MCO providing the satisfaction survey. How "independent" can the answers be? 1: In Allegheny County, under the mandatory system, that is one thing that has been reserved as a county role. The country will have its own consumer satisfaction team to try to establish a level of independence. 17: Given the fact that you used Medicaid administrative data, how do the families feel about confidentiality? 1: There were only a few concerns. Our design was approved by the University of Pittsburgh Institute Review Board and we included family members on the oversight committee for our study. We felt this study was important, as it is the state's responsibility to track the services received by Medicaid families in MCOs. In addition, we used our own study's letterhead on all correspondence to emphasize our own independence. We also paid the families for their participation. 17: Did you make any effort to compare different MCOs-their benefits, structure, and consumer satisfaction? 1: No. Those MCOs that are active in our area are very similar in how they are developed and the benefits they offer. 17: Do you have any theories as to why minorities were the most dissatisfied group? 1: The answer to that question is something we will focus on in the remaining qualitative interviews. We know that, in general, minorities are less likely to use mental health services. There are a number of different hypothesis that we can generate which include: do minorities encounter greater administrative barriers or are the barriers more difficult to overcome? Is it a provider network issue-have the MCOs excluded the providers who traditionally have served these populations? What is their relationship with their providers? Why do they stay in MCOs, if they are so dissatisfied? Do they have lower expectations of MCOs than that of their Caucasian counterparts? These are some of the issues we want to explore in our qualitative data. 17: Did you get an opportunity to look at the benefits and licensure of carve-outs for behavioral services versus those that were part of the MCO? 1: We haven't addressed that at all in our study because there are no integrated plans in our region. |
|
|
| Home | Contact Us | About Us | Awards | Accessibility | Privacy and Disclaimer Statement | Site Map |
