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. |
|
Effects of the Vermont Mental Health and Substance Abuse Parity LawHealth Plan Responses to the Vermont Parity LawThis chapter provides evidence on how two health plans responded to the Vermont parity law. This analysis is based on the experiences of Blue Cross Blue Shield of Vermont (BCBSVT) and Kaiser/Community Health Plan (Kaiser/CHP), which, together, accounted for 78 percent of the private insurance market in Vermont at the time parity was implemented in 1998. The first section describes the effects of parity on the terms and conditions of coverage for mental health and substance abuse (MH/SA) services. The second section presents empirical results of the effects of parity on access, use, and spending for MH/SA services. A. Effects on MH/SA Coverage Provisions1. Pre-Parity Coverage of MH/SA Services To understand the potential effects of parity on access, use, and spending, this study first examined the pre-parity MH/SA coverage limits and cost-sharing requirements for BCBSVT and Kaiser/CHP, based on the contracts with the highest enrollment in 1997.8 The two most prevalent plans offered by Kaiser/CHP in 1997 varied only in the level of cost sharing ($5 versus $10 per visit). BCBSVT offered a wide range of contracts that varied not only in coverage provisions, but also in the use of managed care for MH/SA services prior to parity:
BCBSVT also had an extensive system of riders that covered MH/SA benefits above and beyond the standard plan offerings for an additional premium. However, most people enrolled in the top plans of 1997 did not have a rider for MH/SA services. a. Variation in Covered Services As shown in Table III. 1, the types of MH/SA services covered by Kaiser/CHP and BCBSVT prior to parity were similar in many, but not all, respects. Kaiser/CHP covered inpatient psychiatric care in specialty and general hospitals, as well as outpatient therapy (including psychotherapy and medication management). It also covered inpatient and outpatient detoxification and outpatient substance-abuse counseling. Coverage for nonhospital residential care and intensive nonresidential care was approved on a case-by-case basis. BCBSVT covered a continuum of mental health services across all its plans: inpatient psychiatric care, nonhospital residential services, partial/day treatment,2 and outpatient therapy. Substance abuse coverage consisted of inpatient and outpatient detoxification, nonhospital residential services, intensive nonresidential services, and outpatient counseling. The FFS plans, however, covered treatment only for alcoholism; the PPO, POS, and HMO plans covered treatment for alcohol and other drugs. b. Variation in MH/SA Benefit Limits Prior to implementation of the Vermont parity law and the Federal Mental Health Parity Act, 3 indemnity plans typically set annual or lifetime benefit limits, while HMOs typically applied limits on the number of covered inpatient days or outpatient visits (Buck et al., 1999). The health plans in Vermont generally followed this national pattern (see Table III.1). Kaiser/CHP provided coverage for up to 30 days of inpatient treatment in psychiatric hospitals and up to 20 outpatient mental health visits per year. Similarly, the three BCBSVT plans that covered MH/SA services through managed care arrangements - VFP-individual, 4 VHP, and TVHP - set annual limits on inpatient days (30 to 45 days per year) and outpatient visits (20 to 30 visits per year). The BCBSVT indemnity plans (Basic, Comprehensive, and VFP-group products) typically had annual limits of $5,000 and lifetime limits of $10,000 for mental health services (inpatient and outpatient combined). The Basic plan capped allowable outpatient visits at 50 visits per year in addition to the dollar ceilings. Coverage of substance abuse services was subject to limits on inpatient days and outpatient hours (in compliance with the minimum benefit mandated by existing state law). Kaiser/CHP had a limit of 28 inpatient days per year and 56 inpatient days per lifetime. All BCBSVT plans similarly had a limit of 28 inpatient days per occurrence and 56 days per lifetime. The limit on outpatient hours of substance abuse services was the same for Kaiser/CHP and BCBSVT plans: 90 hours per year and 180 hours per lifetime. Kaiser/CHP officials indicated, however, that they had no system to manage SA benefits according to the number of hours and, instead, tracked the number of visits. Exclusions or adjustments to the MH/SA benefit limits were common, and affected what health plans counted toward the benefit limit prior to parity. For example:
These adjustments and exclusions resulted in variations within and across health plans in the "effective" limits that members faced prior to parity. c. Variation in Cost-Sharing Requirements In addition to setting dollar and service limits, the two plans used differential copayment and coinsurance amounts to control MH/SA utilization. Typically, the separate cost-sharing requirements applied to outpatient services; however, the BCBSVT HMO product (TVHP) had an inpatient copayment of $500 per mental health admission, while the VFP-group product had a 50 percent coinsurance on both inpatient and outpatient mental health services. More common among the managed care plans - such as Kaiser/CHP, VFP-individual, VHP, and TVHP - was the practice of a two-tiered copayment for outpatient visits. The copayment for the first five visits ranged from $0 to $10, while the remaining visits (up to the limit) were $25. The less managed plans of BCBSVT generally did not have a different cost-sharing structure for MH/SA services, relying instead on the same deductible and office coinsurance rate used for physical health services (usually 80 percent). The one exception was the VFP-group product, which had a 50 percent coinsurance rate for MH/SA services, compared to an 80 percent coinsurance rate for other services. 2. Changes Brought About by the Vermont Parity Law With the introduction of parity in 1998, Kaiser/CHP and BCBSVT eliminated differential benefit limits and cost-sharing requirements for MH/SA services. For Kaiser/CHP, the change was relatively straightforward, resulting in elimination of the 30-day limit on inpatient days, the 20-visit limit on outpatient services, and the two-tiered copayment structure for outpatient visits. All Kaiser/CHP contracts were brought into compliance with the parity provisions on January 1, 1998, regardless of the date of contract renewal. Kaiser executives indicated that, because few members reached the limit pre-parity, they did not make major changes in their approach to care management. Their philosophy - both pre-and post-parity - was that resources were limited and the health plan encouraged treatment planning to spread the benefit over a longer period of time (for example, through the use of intensive outpatient treatment as a substitute for inpatient treatment and group therapy rather than individual sessions). Following implementation of parity, Kaiser/CHP officials reported that they attempted to target inpatient services more efficiently, increasing the use of step-down and diversion programs to shorten the length of inpatient stays or to avoid hospitalization altogether. BCBSVT phased in the parity provisions upon contract renewal, beginning with contracts renewed on January 1, 1998. With the introduction of parity, BCBSVT streamlined the number of benefit packages and rider options for MH/SA services. The three basic types of post-parity benefit packages for MH/SA services included:
Most members enrolled in the BCBSVT indemnity products - Basic, Comprehensive, and VFP-group products - were shifted to a managed care carve-out for their MH/SA benefits, although their other benefits continued to be provided on an indemnity basis. As discussed in Chapter II, this initially caused disruption and confusion among providers and consumers because of a combination of such factors as limited communication about the change, tight provider networks, and aggressive management of the newly expanded benefit. B. Effects of Parity on Access, Use, and Spending The adoption of parity in Vermont provided a "natural experiment" in which to learn about the effects of benefit changes on MH/SA access, use, and spending under contrasting health plan experiences. Kaiser/CHP provides a measure of effects within an integrated managed care model before and after parity, whereas BCBSVT demonstrates effects in a plan that shifted a large share of members from indemnity coverage to managed care but retained some members in unmanaged care. The underlying framework for this analysis was a decomposition of per capita spending into its component parts: the proportion of enrollees receiving services (a measure of access to care), the number of services per user (a measure of intensity of care), and the spending per unit of service. This decomposition can be represented as follows:
This approach was used to quantify the extent to which parity affected access to care, intensity of care, and spending for MH/SA treatment. Refer to Appendix C for an overview of the data and methods used in this analysis. 2. Patterns of Access to and Use of Mental Health Services Before and After Parity a. Outpatient Treatment Access to outpatient MH services - measured by the number of MH users per 1,000 members per quarter - increased significantly for both Kaiser/CHP and BCBSVT enrollees post-parity. Kaiser/CHP experienced a 6.4 percent increase in the number of outpatient MH users per 1,000 members per quarter, while BCBSVT experienced a 7.9 percent increase (Table III.2). The likelihood of obtaining MH services increased by 18 to 24 percent as a result of parity.13 The intensity of outpatient MH treatment - that is, the number of MH services per user per quarter - varied between the two health plans. Among Kaiser/CHP members, the average number of visits per user per quarter increased slightly (from 3.26 to 3.48 visits). The combined effect of increased access to and intensity of outpatient MH treatment led to a 14 percent increase in the number of outpatient MH visits per 1,000 members per quarter. Relatively few Kaiser/CHP members received group therapy as part of their MH treatment before parity, and the percentage did not change significantly after parity. However, the average number of group therapy visits per user did increase, suggesting that Kaiser/CHP relied on group therapy to extend the number of visits per user post-parity but did not widen the use of group therapy to a larger share of the population in treatment. Among BCBSVT members, there was a 6 percent reduction in the average number of outpatient services per user. Despite increases in initial access to outpatient services, there was no change in the overall number of services per 1,000 members, due to the reduction in intensity of treatment. The aggregate reduction in intensity of care was a function of the shift to managed care. As shown in Figure III.1, those shifting into managed care experienced a reduction in the average number of visits per user per quarter (all else being equal), while those remaining in an unmanaged product experienced a slight increase in the predicted number of visits per user. As a result, there was an estimated one-half visit differential during the quarter parity went into effect (3.4 visits managed versus 3.9 visits unmanaged). Thus, parity shifted the average level of use upward, while managed care exerted a downward pressure. These aggregate patterns of use were confirmed by examining distributions of the annual level of use. Among Kaiser/CHP members receiving any outpatient MH treatment, a higher proportion of users exceeded the pre-parity 20-visit limit in 1998 and 1999 (Table III.3). In contrast, BCBSVT members showed no increase in the proportion of outpatient MH users with more than 20 visits. Instead, a growing concentration of users was noted at the low end of the distribution (10 visits or less). A similar analysis was conducted on the subgroup of health plan members with a primary diagnosis of major depression, bipolar disorder, or schizophrenia to determine whether those with a serious mental condition may have been affected differently (data not shown). The results paralleled those in the general population of outpatient MH users. Among Kaiser/CHP members, the intensity of outpatient treatment increased, with a higher proportion exceeding the 20-visit pre-parity limit (11.9 percent in 1996 versus 16.4 percent in 1999). Among BCBSVT members, no significant change was observed in the level of outpatient use; for example, about one-fourth received 20 or more outpatient visits both before and after parity. b. Inpatient/Partial Treatment The two health plans exhibited opposite patterns of inpatient/partial treatment following implementation of parity. Fewer Kaiser/CHP members received inpatient MH treatment post-parity, as evidenced by a 38 percent reduction in the number of users per 1,000 members (Table III.2). The number of days per 1,000 members did not decline, however, as lengths of stay increased slightly (though not significantly). The distribution of annual levels of use shed further light on the complex patterns observed in the aggregate analysis (Table III.3). Kaiser/CHP experienced an increase in the proportion of inpatient users, with more than 30 inpatient/partial days following implementation of parity, as well as a growing concentration of inpatient users with 3 to 7 days per year. For BCBSVT enrollees, access to inpatient and partial MH treatment increased significantly following implementation of parity, despite the shift of the majority of BCBSVT members into managed care (Table III.2). The rate of inpatient users per 1,000 members per quarter rose steeply, leading to a 60 percent increase in the number of inpatient days per 1,000 members per quarter. This aggregate increase in inpatient days was due to increased access, rather than to increased intensity. No significant changes were found in the average number of days per user (Table III.2) or in the annual level of inpatient MH use for BCBSVT members (Table III.3). 3. Patterns of Access to and Use of Substance Abuse Treatment Before and After Parity a. Outpatient Treatment Access to outpatient SA treatment by Kaiser/CHP and BCBSVT members declined following implementation of parity (Table III.4). Among those in treatment, however, there was no significant change in the average number of outpatient SA visits per user per quarter. Nevertheless, BCBSVT experienced a 38 percent reduction in the total number of outpatient SA services per 1,000 members per quarter post-parity, given the substantial decline in the level of access. BCBSVT also relied increasingly on group therapy following parity. b. Inpatient/Partial Treatment Both health plans experienced large reductions in access to inpatient treatment following parity, coupled with increased access to partial treatment (although the latter change did not achieve statistical significance due to small sample sizes). The likelihood of obtaining inpatient/partial SA treatment dropped 51 percent for Kaiser/CHP members and 34 percent for BCBSVT members.14 The pattern of inpatient use differed across the two health plans. Kaiser/CHP members had shorter lengths of inpatient stays post-parity; BCBSVT members had longer stays and higher levels of partial treatment.15 The frequency distributions of annual levels of use confirmed these aggregate findings(Table III.5). Kaiser/CHP members demonstrated a noticeable shift in the distribution of inpatient/partial SA days over the 4-year period toward shorter stays, especially in the range of 3 to 7 days. Among BCBSVT members, treatment intensity increased, as 10 percent received more than 28 days of inpatient/partial SA treatment in 1999, compared to 2 percent pre-parity. These data suggest that BCBSVT (or its managed care carve-out) first raised the "threshold" for entering treatment and then provided more intensive treatment to fewer patients. 4. Patterns of Mental Health and Substance Abuse Spending Before and After Parity The analyses of access and use present a complex picture of increased use of certain types of services and a decreased use of others. How did these changes in utilization patterns affect spending for MH/SA services? Spending is comprised of both health plan payments and out-of-pocket expenditures. The analysis shows how both of these spending components, as well as overall MH/SA spending, changed following implementation of parity. This section first presents data on patterns of BCBSVT spending for MH/SA services before and after parity and then imputes the effects of parity on Kaiser/CHP spending. The section concludes with a discussion of the effect of parity on cost sharing for those with serious mental conditions. a. Mental Health Spending Patterns On average, MH spending per BCBSVT member per quarter was not significantly different before and after parity (Table III.6). Moreover, MH spending as a percentage of total spending did not change following implementation of parity, averaging 2.31 percent during both periods. However, spending by type of service did change significantly over the study period. Despite an increase in outpatient utilization, spending on outpatient MH services per member per quarter declined 6.5 percent, and spending on outpatient MH services per user declined 13 percent - driven by a 10 percent reduction in average spending per outpatient visit. The unit cost reduction could be a function of a changing service mix, as well as of payment reductions negotiated by the carve-out plan. In contrast to declining outpatient costs, combined spending on inpatient and partial MH services doubled. Relatively few BCBSVT members incurred health plan payments of $5,000 or more pre-parity for MH services; and that pattern continued following implementation of parity (Table III.7). Over the 4-year period, the proportion of MH users with health plan payments over $1,000 fell from 26 percent to 20 percent. A more pronounced trend was a growing share of MH users spending between $101 and $1,000. This may include two groups of users: (1) those with chronic conditions who received shorter-term psychotherapy and crisis intervention post-parity; and (2) new users with less severe conditions who received a brief course of therapy. Both scenarios are consistent with the results of the descriptive analysis, suggesting that more BCBSVT members had access to MH treatment post-parity, but users received fewer services, on average. b. Substance Abuse Spending Patterns The pronounced reductions in SA utilization translated into substantial reductions in spending. Overall, average SA spending per BCBSVT member per quarter fell by 47 percent, with across-the-board reductions in spending for both inpatient and outpatient services (Table III.6). This resulted in a reduction in SA spending as a percentage of total spending from 0.37 to 0.24 percent. Per capita spending reductions were a function not only of lower rates of access but also of lower unit costs for treatment. Among the factors that might account for lower unit costs are differences in service mix, case mix, or lower reimbursements negotiated by the health plan. A more detailed analysis of the annual level of spending revealed little change in the distribution of health plan spending per user (Table III.7). c. Changes in Health Plan Payments In the aggregate, quarterly MH/SA spending declined by about 8 percent, while health plan payments for MH/SA services increased by 4 percent (Table III.8). Reductions in consumer out-of-pocket payments drove these increases in health plan payments. Prior to parity, health plan payments accounted for 70 percent of MH spending, while consumers paid for the remaining 30 percent. Following parity, the health plan share rose to 83 percent as consumer cost-sharing requirements were brought into compliance with the parity provisions. The health plan share of SA spending remained constant at 87 percent. Health plan payments for MH/SA services accounted for 2.47 percent of total health plan payments for all services post-parity, up from 2.30 percent pre-parity (Table III.8). This 0.17-percentage-point increase reflected a 0.26-point increase for MH services and a 0.09-point decrease for SA services. Overall, health plan payments for MH/SA services increased by 58 cents per member per quarter following the implementation of parity. In other words, the cost of full parity to BCB-SVT amounted to about $2.32 per member per year, or 19 cents per member per month. Multivariate analysis provided evidence of the joint effects of managed care and parity on the level of health plan payments per user (Figure III.2). Although implementation of managed care constrained both MH and SA spending, parity offset this effect for MH services but not for SA services. Thus, spending for MH services was highest in an unmanaged parity environment. In contrast, spending for SA services was higher preparity, and higher still before the transition to managed care. d. Changes in Kaiser/CHP Spending Estimates of changes in Kaiser/CHP spending were imputed by applying BCBSVT unit costs to Kaiser/CHP utilization patterns.16 Based on this approach, overall MH/SA spending per member per quarter was estimated to have decreased by nearly 18 percent. Furthermore, health plan spending (net of patient out-of-pocket expenses) was estimated to have decreased by about 9 percent following implementation of parity. 10 This reduction was driven entirely by the decline in use of SA treatment. e. Changes in MH/SA Spending for BCBSVT Members With Serious Mental Conditions A more in-depth analysis was conducted of changes in the level of health plan payments and cost sharing among BCBSVT members with serious mental conditions (major depression, bipolar disorder, or schizophrenia). This population has the most to gain from parity, both in terms of higher utilization and lower cost sharing. During the study period, the proportion of users with health plan payments of $5,000 or more increased from 3.9 percent in 1996 to 6.0 percent in 1999 (data not shown). At the same time, the proportion spending more than $1,000 out-of-pocket decreased from 5.8 to 2.7 percent, as health plans assumed a larger share of the costs post-parity. Median out-of-pocket payments for high users (those with total mental health charges more than $5,000 per year) declined from 9.0 to 4.4 percent of their total charges (Table III.9). Individuals with serious mental conditions who were relatively low users benefited substantially from the reduction of cost sharing (in relation to their total MH charges). For example, among those with total charges less than $500 per year, the median out-of-pocket payment as a percent of total charges declined from 50 percent to 19 percent, as the higher coinsurance rate for MH services was eliminated. Thus, the cost of initiating an episode of treatment was lower following implementation of parity. C. DiscussionThe two dominant insurers in Vermont at the time parity was enacted - BCBSVT and Kaiser/CHP - offered sharply contrasting parity-implementation experiences, but generally similar results. Across both plans, significant increases in access to MH services were observed following implementation of parity. Parity was associated with an increased likelihood of obtaining any MH treatment. Parity also had a positive effect on the average number of outpatient visits per user within the two health plans. However, these aggregate results do not mean that all health plan members experienced increases in outpatient MH access and utilization following implementation of parity. For those BCBSVT members who received their MH/SA benefits through the managed care carve-out, the effect of parity was offset by the use of managed care arrangements. Not only did the likelihood of obtaining outpatient treatment decline for those in the managed care carve-out, but also the average number of visits per user was lower. Results were mixed across the two health plans with regard to use of inpatient or partial MH services. Kaiser/CHP members had a significantly lower likelihood of obtaining inpatient or partial MH treatment following parity, suggesting that outpatient MH services may have substituted for inpatient treatment. In contrast, among BCBSVT members, access to inpatient or partial MH treatment increased following parity, coupled with increases in outpatient MH treatment noted above. There is considerable interest in how Vermont health plans responded to a fullparity law that includes SA treatment. Substantial reductions in access to substance abuse treatment were observed in both health plans (as measured by the number of users per 1,000 members), generally accompanied by large decreases in the number of services used per 1,000 members. BCBSVT members experienced an increase in the duration of inpatient and partial treatment; but, given the marked reduction in access to such treatment, this may have reflected the targeting of more intensive treatment to a higher-severity case mix. As a result of these changes in patterns of access and use, average SA spending per BCBSVT member per quarter was nearly halved after parity. This analysis revealed that overall spending for MH/SA services per BCBSVT member per quarter declined by 8 percent. However, due to declines in patient cost-sharing requirements, BCBSVT assumed an increasing share of total spending for MH/SA services. Thus, BCBSVT spending for MH/SA services rose by 4 percent. On the basis of this estimate, it is estimated that the cost of full parity in Vermont amounted to approximately $2.32 per member per year, or 19 cents per member per month. As a percent of total health spending (across all types of services), the share attributable to MH/SA services rose 0.17 percentage points, from 2.30 to 2.47 percent. Overall MH/SA spending per Kaiser/CHP member per quarter was estimated to have decreased by about 18 percent, while health plan spending decreased by about 9 percent following implementation of parity. This reduction was driven entirely by the decline in use of SA treatment by Kaiser/CHP members. The analysis of MH/SA spending and utilization during the 2 years after adoption of parity in Vermont suggests that the initial costs associated with movement to full parity were minimal. This is due, however, to large reductions in SA utilization, and only a minimal expansion of MH utilization above levels covered prior to parity. These findings reflect the effects of implementing parity for MH/SA services in a managed care context. |
|
| Home | Contact Us | About Us | Awards | Accessibility | Privacy and Disclaimer Statement | Site Map |
