Carve‐out plan financial requirements associated with national behavioral health parity

Abstract

Objectives

To examine changes in carve‐out financial requirements (copayments, coinsurance, use of deductibles, and out‐of‐pocket maxima) following the Mental Health Parity and Addiction Equity Act (MHPAEA).

Data Source/Study Setting

Specialty mental health benefit design information for employer‐sponsored carve‐out plans from a national managed behavioral health organization's claims processing engine (2008‐2013).

Study Design

This pre‐post study reports linear and logistic regression as the main analysis.

Data Collection/Extraction Methods

NA.

Principal Findings

Copayments for in‐network emergency room (−$44.9, 95% CI: −78.3, −11.5; preparity mean: $56.2), outpatient services (eg, individual psychotherapy: −$7.4, 95% CI: −10.5, −4.2; preparity mean: $17.8), and out‐of‐network coinsurance for emergency room (−11 percentage points, 95% CI: −16.7, −5.4; preparity mean: 38.8 percent) and outpatient (eg, individual psychotherapy: −5.8 percentage points, 95% CI: −10.0, −1.6; preparity mean 41.0 percent) decreased. Probability of family OOP maxima use (29 percentage points, 95% CI: 19.3, 38.6; preparity mean: 36 percent) increased. In‐network outpatient coinsurance increased (eg, individual psychotherapy: 4.5 percentage points, 95% CI: 1.1, 7.9; preparity mean: 2.7 percent), as did probability of use of family deductibles (15 percentage points, 95% CI: 6.1, 23.3; preparity mean: 38 percent).

Conclusions

MHPAEA was associated with increased generosity in most financial requirements observed here. However, increased use of deductibles may have reduced generosity for some patients.

What This Study Adds.

• Changes in behavioral health utilization and patient expenditures, but not financial requirements, associated with the Mental Health Parity and Addiction Equity Act (MHPAEA) have been studied in carve‐out plans.

• What this study adds:

  1. This work examines changes in copayments, coinsurance, use of deductibles, and use of out‐of‐pocket maxima associated with MHPAEA implementation among carve‐out plans.

  2. Compared to preparity, postparity, specialty mental health copayments were more generous for INN emergency room and outpatient services, coinsurance was more generous for OON emergency room and outpatient services, and more plans implemented out‐of‐pocket maxima.

  3. However, compared to preparity, postparity, specialty mental health coinsurance for INN outpatient services was less generous, and more plans required a deductible for behavioral health and medical services combined.

1. INTRODUCTION

The landmark 2008 federal Mental Health Parity and Addiction Equity Act (MHPAEA) was intended to correct historical inequities for behavioral health (BH), including mental health (MH) and substance use disorder (SUD) conditions, compared to medical/surgical coverage. ¹ The law applied to all health insurance plans sponsored by large employers. According to statutory provisions, financial requirements (ie, copayments, coinsurance, and deductibles) and quantitative treatment limits (QTLs, eg, number of visits, days of coverage) were included in parity requirements. The Interim Final Rule (IFR) added the requirement of parity in nonquantitative treatment limits (NQTLs, eg, pre‐authorization and medical necessity review). ²

Notably, the law required parity for BH benefits within both carve‐in and carve‐out plans. Under the carve‐out model, BH benefits are typically covered and administered by a managed behavioral health organization (MBHO) whose staff specialize in managing BH care utilization. ³ In carve‐out plans, BH and medical/surgical insurance benefits are administered separately. In contrast, carve‐in plans administer these benefits within the same plan. Prior to the passage of MHPAEA, approximately 170 million individuals were receiving insurance for specialty BH services (ie, BH services provided by a specially trained provider, such as a psychologist or psychiatrist, not a primary care provider) through an MBHO carve‐out arrangement. ⁴

Research on changes to benefits following MHPAEA’s implementation among carve‐in plans may not generalize to carve‐out plans. Historically, there is some evidence that carve‐out plans had lower financial requirements than carve‐in plans. ⁵ To contain spending, they may have relied more heavily on managed care techniques (ie, prior authorization, concurrent review, and retrospective review) than on financial requirements. If this is true, we would expect MPHAEA to have a relatively smaller impact on financial requirements among carve‐out plans.

On the other hand, aligning BH and medical benefits may have been a greater logistical challenge for carve‐out plans compared to carve‐in plans, since carve‐out vendors lack information about the medical benefits being administered by separate (and in some cases competing) vendors. In this case, carve‐out BH financial requirements may have been further from parity than carve‐in BH financial requirements, and MHPAEA may have had a relatively larger impact on carve‐out plans.

Literature on the law's impacts on carve‐out plans is limited. Two studies examined the effects of MHPAEA on BH financial requirements, but neither reported separately on carve‐outs. ⁶ , ⁷ Others reported changes associated with the law among carve‐out plans in QTLs and NQTLs, but not financial requirements. ⁸ , ⁹ One additional study examined the effects of the law on utilization and expenditure of BH services among carve‐out plan enrollees, ¹⁰ finding that some expenditures shifted from patients to plans postparity. The present study fits into the literature by investigating whether lower financial requirements postparity might explain this last finding. The detailed descriptive data it provides also address the dearth of benefit design information available on carve‐out plans.

This study used data from the BH division of Optum^® (hereafter called “Optum”), a subsidiary of UnitedHealth Group, to report levels and changes in copayments, coinsurance, deductibles, and out‐of‐pocket maxima for carve‐out plans. Notably, our benefit design variables were obtained from the actual Optum claims processing engine and therefore represent the “gold standard” in terms of what plans actually required. The analyses tested whether, compared to preparity, (a) copayments and coinsurance levels were lower, (b) use of deductibles was lower, and (c) use of out‐of‐pocket maxima was higher postparity, as plan‐years became parity‐compliant. To our knowledge, this study is the first to report changes in carve‐out plan financial requirements associated with MHPAEA.

2. METHODS

2.1. Study design and data

This study applied an analytic approach similar to prior work on carve‐in plans. ⁶ A “before vs. after” study design was used with 2008‐2013 data to estimate changes in financial requirements (copayments, coinsurance, out‐of‐pocket maxima, and deductibles) immediately after the MHPAEA statute went into effect (2010, the “transition” period) and the 3 years after MHPAEA’s IFR was in effect (2011‐2013, the “postparity” period, when legal compliance was required, including parity in NQTLs), relative to the 2 years prior to implementation (2008‐2009, the “preparity” period).

Data included (a) specialty behavioral health benefit design data for Optum's carve‐out plans; and (b) employer and plan characteristics from Optum's Book of Business. A unique strength of this study was our ability to use the actual benefit design data (ie, the information used to actually process claims), rather than values imputed from claims data.

2.2. Sample

We analyzed carve‐out plans sponsored by large employers subject to MHPAEA. Sample plans cover BH services, are not retiree, supplemental or indemnity plans, renew on a calendar‐year cycle, are not collectively bargained (to ensure uniform timing of MHPAEA compliance), and include at least one enrollee. The analysis sample contained 2257 plan‐years (the unit of analysis). Post‐MHPAEA, the number of carve‐out plans increased relative to carve‐ins. ¹⁰ This is because, pre‐MHPAEA, a carve‐out plan could use the same BH benefit design for multiple medical plans, but post‐MHPAEA, carve‐outs had to either choose a single benefit design that achieved parity with the most generous medical plan or else split into multiple plans with different BH benefits to conform to each medical plan offered by the employer.

2.3. Measures

Outcomes included copayment, deductible, and out‐of‐pocket (OOP) maximum dollar amounts and patient coinsurance rates for specialty mental health services. Copayment and coinsurance outcomes are reported separately by network (in‐network [INN] vs. out‐of‐network [OON]) and service type (see Table 1 for a list of services). Deductible and out‐of‐pocket maxima are reported separately by individual vs. family levels.

TABLE 1.

Percent of carve‐out plan‐years ^a requiring specialty mental health benefit features; mean and range among subset of plans requiring them, pre‐ and postparity

		% of sample plan‐years a				Among plan‐years requiring benefit feature
		Pre b		Transition b	Post b	Pre b		Transition b		Post b
						Mean ($)	Range	Mean ($)	Range	Mean ($)	Range
Copayment for in‐network services
Inpatient hospitalization		27		37	22	231	64‐680	331	59‐878	316	100‐825
Inpatient emergency room		31		36	9	182	38‐384	332	59‐878	243	11‐330
Residential treatment		24		16	16	227	64‐680	319	59‐878	313	100‐825
Inpatient partial hospitalization		24		16	16	228	64‐680	319	59‐878	313	100‐825
Intensive outpatient		28		20	25	165	16‐545	234	21‐780	27	5‐412
Outpatient ECT		33		16	1	143	16‐545	283	52‐780	147	21‐258
Outpatient group psychotherapy		84		62	30	20	5‐44	22	5‐52	19	5‐46
Outpatient psychotherapy		84		62	30	21	5‐44	23	5‐52	19	5‐46
Outpatient medication management		84		60	30	21	5‐44	23	5‐52	19	5‐46
Outpatient diagnostic interview		84		62	30	21	5‐44	23	5‐52	19	5‐46
Copayment for out‐of‐network services c , d
Inpatient hospitalization	25			15	8	263	128‐768	399	117‐878	368	110‐825
Inpatient emergency room	25			14	4	192	64‐384	381	70‐878	205	100‐440
Residential treatment	23			11	7	232	128‐680	366	117‐878	360	110‐825
Inpatient partial hospitalization	24			12	7	249	128‐680	365	117‐878	360	110‐825
Intensive outpatient	24			12	0	203	106‐545	324	104‐780	258	258‐258
Coinsurance for in‐network services						Mean (%)	Range	Mean (%)	Range	Mean (%)	Range
Inpatient hospitalization	59			50	70	16	10‐30	17	10‐30	19	5‐40
Inpatient emergency room	56			50	67	16	10‐30	17	10‐30	19	5‐40
Residential treatment	58			49	70	16	10‐30	17	10‐30	19	5‐40
Inpatient partial hospitalization	58			50	70	16	10‐30	17	10‐30	19	5‐40
Intensive outpatient	54			49	69	16	10‐30	17	10‐30	19	10‐40
Outpatient ECT	53			49	69	16	10‐30	17	10‐30	19	10‐40
Outpatient group psychotherapy	13			37	69	21	10‐50	18	10‐50	19	10‐40
Outpatient individual psychotherapy	13			37	69	21	10‐50	18	10‐50	19	10‐40
Outpatient medication management	13			37	68	21	10‐50	18	10‐50	19	10‐40
Outpatient diagnostic interview	13			37	69	21	10‐50	18	10‐50	19	10‐40
Coinsurance for out‐of‐network services c , d
Inpatient hospitalization	99			96	97	41	15‐50	34	10‐50	36	10‐60
Inpatient emergency room	97			93	89	40	15‐50	33	10‐50	34	10‐60
Residential treatment	99			100	97	40	15‐50	33	10‐50	36	10‐60
Inpatient partial hospitalization	99			100	97	40	15‐50	33	10‐50	36	10‐60
Intensive outpatient	99			100	100	40	15‐50	33	10‐50	35	10‐60
Outpatient ECT	99			100	100	40	15‐50	33	10‐50	35	10‐60
Outpatient group psychotherapy	99			100	100	41	15‐50	34	20‐50	35	10‐60
Outpatient individual psychotherapy	99			100	100	41	15‐50	34	20‐50	35	10‐60
Outpatient medication management	99			100	100	41	15‐50	34	20‐50	35	10‐60
Outpatient diagnostic interview	99			100	100	41	15‐50	34	20‐50	35	10‐60
Deductible and out‐of‐pocket maximum c
Any individual deductible	51			76	89
Any family deductible	38			70	87
Any individual OOP maximum	40			86	98
Any family OOP maximum	36			84	97

^aThe unit of observation is the plan‐year, so one plan may count up to twice in the preperiod, once in the transition period, and three times in the postperiod. Analysis excludes plan‐years with one cost‐sharing level for some visits and another cost‐sharing level for other visits for a particular cost‐sharing feature, as well as plan‐years that do not cover a particular service and plan‐years with missing data for a particular cost‐sharing feature.

^bPreparity (2008‐2009; N = 374), transition (2010; N = 401), and postparity (2011‐2013; N = 1482).

^cAmong plan‐years with out‐of‐network benefits: preparity N = 316, transition N = 367, and postparity N = 1403.

^dFive out‐of‐network services very rarely had copayments: Outpatient ECT, Outpatient Group Psychotherapy, Outpatient Individual Psychotherapy, Outpatient Medication Management, and Outpatient Diagnostic Interview.

The variables of interest indicate whether the plan‐year observation is drawn from the transition period (2010) or the postparity period (2011‐2013) versus the preparity period (2008‐2009). ⁶ We focused on comparisons with the postparity period, when plans were legally required to comply with MHPAEA (including parity in NQTLs). Other covariates are employer group size, employer group industry, census region, whether the plan type is “more managed” (eg, health maintenance organization) versus “less managed” (eg, preferred provider organization), and whether plans were fully (insurer takes on financial risk) or self‐insured (ie, insurer does not take on financial risk). For in‐network outcomes, we also controlled for whether the plan included out‐of‐network benefits.

2.4. Statistical analyses

The main analyses use linear regression to study changes in continuous copayment and coinsurance values and logistic regression to study changes in use of deductibles and out‐of‐pocket maxima in postparity vs. preparity years. All models were adjusted for employer‐level clustering using generalized estimating equations cluster robust “sandwich” standard errors.

One sensitivity analysis uses a two‐part model. This technique is used to analyze data with large numbers of true zero values (rather than censored values). It examines whether overall changes in copayment and coinsurance values were due more to changes in the probability of use (using logistic regression) or, among plan‐years that used the feature, changes in the level of the feature (conditional gamma regression). ¹¹ , ¹²

A second sensitivity analysis repeats the main analysis but weights plan‐years by the number of enrollee‐month. Giving more weight to the plan‐years with more enrollee‐month observations highlights changes that affected the largest number of enrollees over time.

3. RESULTS

3.1. Sample statistics and unadjusted changes in financial requirements

The study sample represents employers (n = 40) and carve‐out plans (n = 1527) diverse in size, industry, and region (Appendix S1). Table 1 presents descriptive summaries of the average benefit feature values in 2008‐2009, 2010, and 2011‐2013.

3.2. Adjusted changes in copayments and coinsurance

Table 2 shows that, when significant, changes in the level of copayments for INN services indicated lower copayments postparity. These decreases were significant for inpatient emergency room (ER) (β = −$45; P = .01) and most outpatient services (β ranges from −$6 to −$10; P < .0001). Table 2 also shows lower levels of coinsurance for several OON services postparity. Coinsurance decreased for inpatient ER care (by −11 percentage points; P < .0001) and for all outpatient services (β ranges from −4 to −6 percentage points; P < .04). In contrast, postparity, coinsurance for INN outpatient services increased significantly (β ranges from 4 to 6 percentage points; P = .01).

TABLE 2.

Changes in mean copayment/coinsurance amounts and use of deductibles/out‐of‐pocket maximums among carve‐out plan‐years, pre vs postparity ^a

	Preparity mean	Postparity change in level, among all plans b , c
	$	$	P‐value	95% CI
Copayment for in‐network services d
Inpatient hospitalization	61.6	−17.8	.33	(−54.5, 19.0)
Inpatient Emergency Room	56.2	−44.9	.01	(−78.3, −11.5)
Residential treatment	55.0	10.4	.53	(−22.6, 43.3)
Inpatient Partial Hospitalization	55.6	9.6	.56	(−23.4, 42.6)
Intensive outpatient	46.4	−14.6	.31	(−43.3, 14.2)
Outpatient ECT	47.2	−9.6	.00	(−12.7, −6.5)
Outpatient Group Psychotherapy	16.9	−6.4	.00	(−9.6, −3.2)
Outpatient Individual Psychotherapy	17.8	−7.4	.00	(−10.5, −4.2)
Outpatient Medication Management	17.5	−7.1	.00	(−10.2, −3.9)
Outpatient Diagnostic Interview	17.8	−7.4	.00	(10.5, −4.2)
Copayment for out‐of‐network services e , f
Inpatient Hospitalization	64.5	−6.9	.74	(−48.1, 34.2)
Inpatient emergency room	47.0	−16.3	.16	(−39.4, 6.8)
Residential treatment	54.4	5.6	.78	(−34.1, 45.4)
Inpatient Partial Hospitalization	60.1	5.2	.79	(−33.8, 44.3)
Intensive Outpatient	49.0	−26.8	.13	(−16.8, 8.2)

	%	Percentage points	P‐value	95% CI
Patient coinsurance for in‐network services
Inpatient hospitalization	9.3	2.1	.11	(−0.5, 4.6)
Inpatient emergency room	8.8	2.2	.12	(−0.6, 5.0)
Residential treatment	9.2	2.3	.08	(−0.3, 4.9)
Inpatient partial hospitalization	9.1	2.3	.09	(−0.3, 5.0)
Intensive outpatient	8.5	1.3	.46	(−2.2, 4.8)
Outpatient ECT	8.3	6.3	.00	(2.6, 9.9)
Outpatient Group Psychotherapy	2.7	4.6	.01	(1.2, 7.9)
Outpatient Individual Psychotherapy	2.7	4.5	.01	(1.1, 7.9)
Outpatient Medication Management	2.7	4.4	.01	(1.0, 7.8)
Outpatient Diagnostic Interview	2.7	4.5	.01	(1.1, 7.9)
Patient coinsurance e for out‐of‐network services
Inpatient hospitalization	40.1	−3.6	.07	(−7.6, 0.3)
Inpatient emergency room	38.8	−11.0	.00	(−16.7, −5.4)
Residential treatment	39.8	−3.5	.08	(−7.4, 0.4)
Inpatient Partial Hospitalization	39.9	−3.6	.07	(−7.4, 0.3)
Intensive outpatient	40.1	−3.9	.04	(−7.7, −0.2)
Outpatient ECT	40.1	−5.8	.01	(−10.1, −1.6)
Outpatient Group Psychotherapy	41.0	−5.8	.01	(−10.0, −1.6)
Outpatient Individual Psychotherapy	41.0	−5.8	.01	(−10.0, −1.6)
Outpatient Medication Management	41.0	−5.8	.01	(−10.0, −1.6)
Outpatient diagnostic interview	41.0	−5.8	.01	(−10.0, −1.6)

	Preparity mean use (%)	Change in the probability of use, among all plans f		95% CI
Deductible and out‐of‐pocket maximum
Any individual deductible	51	12	.00	(4.1, 20.7)
Any family deductible	38	15	.00	(6.1, 23.2)
Any individual out‐of‐pocket maximum	40	29	.00	(18.3, 38.7)
Any family out‐of‐pocket maximum	36	29	.00	(19.3, 38.6)

^aPreparity (2008‐2009; N = 374) and postparity (2011‐2013; N = 1482).

^bThe unit of observation is the plan‐year, so one plan may count up to twice in the preperiod and three times in the postperiod. Standard errors are adjusted for intraclass correlation at the employer group level. Analysis excludes plan‐years with one cost‐sharing level for some visits and another cost‐sharing level for other visits for a particular cost‐sharing feature, as well as plan‐years that do not cover a particular service and plan‐years with missing data for a particular cost‐sharing feature.

^cCoefficients are generated from an OLS regression. All regressions control for employer size, employer's region, employer's industry, more managed status, fully insured versus self‐insured, and, for in‐network benefits, whether the plan also covered OON services.

^dFour inpatient MH services rarely have an in‐network copayment (<1% of plan‐years): professional, emergency room professional, ECT facility, and ECT.

^eAmong plan‐years with out‐of‐network benefits: preparity N = 316 and postparity N = 1403.

^fIn carve‐out plans, most outpatient services rarely have an out‐of‐network copayment (<1% of plan‐years).

^gChange in probability determined using a logistic regression with the margins command. All regressions control for employer size, employer's region, employer's industry, more managed status, fully insured vs self‐insured, and, for in‐network benefits, whether the plan also covered OON services.

3.3. Adjusted changes in use of deductibles and out‐of‐pocket maxima

The bottom rows of Table 2 report the adjusted changes in the probabilities of using deductibles and out‐of‐pocket maxima. Postparity, the probability of using a deductible increased 12 percentage points for individual deductibles (P = .003) and 15 percentage points for family deductibles (P < .0001). Postparity, the probability of having an individual or family OOP maximum increased (29 percentage points, P < .0001). Increases in use of OOP maxima signify increased plan generosity (holding other features constant), since this benefit design feature caps the enrollee's out‐of‐pocket costs during a calendar year.

3.4. Sensitivity analyses

The two‐part model results lead to similar conclusions as the main analysis, but add more nuanced information (Appendix S2). Notably, the significant decreases in copayments and increases in coinsurance are primarily driven by decreases and increases in the use of any copayments and coinsurance, respectively. Additionally, the magnitude of the overall changes in the level among all plan‐years is slightly larger in the two‐part model, compared to the main results.

The weighted analysis suggests that changes observed in INN copayments and coinsurance in the main analysis were not uniform across plan‐years (Appendix S3). Among the INN services with significant copayment or coinsurance changes in the main model, only ER and outpatient ECT retain their significance (and magnitude and sign) after weighting. For the other INN services that were significant in the main model, the weighted estimates retain the same signs but are smaller in magnitude and lose significance. For OON services, no conclusions are changed based on the weighted regressions. The weighted results for deductible and OOP maxima were comparable to the main results.

4. DISCUSSION

Several studies document that MHPAEA had modest effects on financial requirements in carve‐in plans, but these results may not generalize to carve‐out plans. The current study describes changes in copayments, coinsurance, deductibles, and OOP maxima associated with MHPAEA among a sample of employer‐sponsored carve‐out plans provided by a large MBHO. MHPAEA was hypothesized to be associated with more generous benefits, correcting years of historical inequities between mental health and medical/surgical benefit generosity. Indeed, INN ER and outpatient copayments and OON ER and outpatient coinsurance values all decreased at the plan‐year level, although the changes were small. Additionally, more plan‐years used OOP maxima, (making those plans more generous), postparity. However, INN outpatient coinsurance increased, as did use of deductibles (although, as discussed below, this may not reduce generosity for all patients).

The primary study limitation is the lack of a control group, limiting our ability to distinguish parity effects from secular time trends. Candidates such as small employers (who were exempt from MHPAEA during the study period) and fully insured plans in states with prior parity laws (for which parity may have already been required for these cost‐sharing features prior to MHPAEA) were rejected for two reasons. First, they may be too dissimilar to provide valid comparisons, and introduced concern about selection bias due to poor matching. Second, they were too small in number to provide adequate power for hypothesis testing.

Additionally, the study sample's generalizability may be limited to plans from the source organization. However, the number of plans analyzed and their diverse size, region, and industry suggest some degree of generalizability. Additionally, enrollees in our sample plans had levels of outpatient mental health utilization comparable to those observed by the National Survey on Drug Use and Health among privately insured adults. ¹³ Further, the small effect sizes noted in this study are consistent with findings by a study with a nationally representative sample (using the Medical Expenditure Panel Survey) that concluded that MHPAEA did not affect behavioral health utilization, adding to the case for this study's generalizability. ¹⁴

Prior studies of changes in carve‐in plan financial requirements associated with MHPAEA discovered similar patterns among copayments and coinsurance use and levels to the ones reported here: Among all plans, small but significant decreases in INN copayments and OON coinsurance and small but significant increases in INN coinsurance were observed. ⁶ Slightly larger observed differences among carve‐out plan‐years might be attributable to the lack of a credible comparison group for the carve‐out plan‐years.

Interestingly, a previous study of MHPAEA's effects on behavioral health utilization and expenditures among enrollees in “carve‐out’ plans found no notable changes in utilization or total expenditures. ¹⁰ It did, however, find that on average, patient out‐of‐pocket expenditures decreased notably, while plan expenditures increased commensurately. This suggests that even as the benefits for some services became more generous, patients did not increase the volume of their use, but some may have experienced lower cost‐sharing, leading to shifting of costs onto plans. ¹⁰

Decreases in copayments and increases in coinsurance for INN outpatient services likely resulted in net decreases in patients' OOP costs. An ad hoc analysis used average total expenditure from the claims data to approximate OOP costs for the outpatient services that had significantly lower copayments but significantly higher coinsurance. For all of these services, absolute dollar reductions in OOP costs from lower copayments were larger than the absolute dollar increases in OOP costs from higher coinsurance. For example, for a patient needing an individual psychotherapy visit, the average copayment decrease of −$7.36 per visit more than offset the coinsurance increase of 4.59 percentage points per visit. The net change in OOP costs is small (−$4.22 per visit), but negative.

Changes in use of deductibles and OOP maxima should be interpreted in the context of how plans used these benefits preparity. Preparity, plans could have separate deductibles and OOP maxima for behavioral health and medical/surgical spending. Postparity, deductible and OOP maxima levels applied to combined behavioral health and medical/surgical service use. As a result, postparity, a new combined deductible would actually result in less generous plans for patients with behavioral health utilization only, but more generous plans for patients with sufficiently high levels of both medical and behavioral health utilization.

It is interesting to note that the changes observed at the plan‐year level may not reflect the experiences of enrollees in the largest plans (ie, most enrollees). The weighted sensitivity analysis was consistent with the main analysis for the OON, deductible, OOP maxima outcomes, and some, but not all, of the INN outcomes. For the outcomes for which the weighted analysis deviated from the main analysis, the changes in copayments and coinsurance observed in the main analysis may not have been experienced by large numbers of enrollees.

5. CONCLUSIONS

This study fills a gap in the literature on changes in financial requirements among carve‐outplans associated with the national behavioral health parity law, MPHAEA, using benefit data from the claims processing database from a large MBHO. MHPAEA was found to be associated with reduced OOP costs for ER and outpatient services, at least for patients without a deductible. Meanwhile, on average, use of deductibles increased, which may have increased OOP costs for some patients. Given carve‐out plans’ continued existence and use postparity, this study provides salient information about postparity cost‐sharing in this type of plan.

Supporting information

Author Matrix

Appendix S1

Friedman S, Xu H, Azocar F, Ettner SL. Carve‐out plan financial requirements associated with national behavioral health parity. Health Serv Res. 2020;55:924–931. 10.1111/1475-6773.13542