American Psychiatric Association

Irvin l. ‘Sam’ Muszynski, JD
Senior Policy Advisor and Director, Parity Implementation and Enforcement

Federal Guidance on Network Adequacy and Reimbursement Rates is Necessary

Health plan participants lack of equitable access to in‐network specialty physician services for mental health and substance use disorder (MH/SUD) conditions, as compared to access for medical surgical conditions, is a significant parity problem. Network adequacy, or inadequacy, is a nonquantitative treatment limitation (NQTL) issue and subject to analysis per the established regulatory tests.

Numerous studies have documented that participation of psychiatrists in insurance networks and/or the availability of psychiatrists who are participating in‐network is very poor. This is especially evidenced by high out‐of‐network (OON) utilization rates as compared to those for medical‐surgical conditions, lack of timely appointment availability and sometimes grossly inaccurate health plan provider directories which is an undue burden for plan participants. A key contributing factor to physician non‐participation or availability are the payment rates plans reimburse. It is well documented that psychiatrists receive lower in‐network reimbursement (sometimes exceeding 60% less) than non‐psychiatrist physicians and mental health professionals for the same services.

Even though these issues are well documented and their status as NQTLs is clear, our experience is that there is a relative lack of compliance oversight by state regulators. We think that this is due in part to a lack of information or understanding as to how to review and evaluate these matters in the context of the NQTL regulatory tests. The Departments have previously stated that they may provide additional guidance if questions persist with respect to reimbursement rates.

We think that more definitive federal guidance on network adequacy and reimbursement rates is necessary and will greatly aid needed context for these issues per the NQTL rule. We also urge that due consideration be given to developing a Red Flag warning status to these issues accompanied by text which provides a framework for review and evaluation as part of the Agencies to be developed Action Plan as called for by the Cures Act. We have provided additional background discussion below regarding these issues and would welcome the opportunity to discuss these matters in more detail and provide additional documentation as to the extent of these problems.

Health Plan Provider Network Adequacy and MHPAEA

Health plan network adequacy (NA) is a NQTL¹

as defined by MHPAEA regulations. NA must be reviewed and evaluated by the regulatory tests for NQTLs established under MHPAEA, independent of any federal, state and/or health plan accreditation requirements. That is, health plan compliance with federal/state and/or accreditation network adequacy standards required for approval to operate is not dispositive of MHPAEA compliance. They are two separate inquiries.

This discussion will:

1. Provide information on the difference between federal, state and health plan accreditation standards and requirements for NA and parity as required for NQTLs under MHPAEA;
2. Provide a basis to understand NA as an NQTL and other factors that may contribute to an inadequate network that are NQTLs in and of themselves; and
3. Provide a framework for assessing MHPAEA compliance with the parity rules for an NQTL.

Federal and State NA Requirements, Health Plan Accreditation Standards and MHPAEA

NA refers to the ability of a health plan to provide enrollees with timely and geographically reasonable access to a sufficient number of in-network providers, including primary care and specialty physicians, as well as other health care providers and services included in the benefit contract. There are numerous measures a plan may utilize, or be required to measure, in order to comply with federal/state law or regulation and/or health plan accreditation standards (NCQA and/or URAC) to establish the adequacy of its provider network. These include: time and distance standards, provider to population ratios, among others.

Plan compliance with state/federal regulatory and/or accreditation NA standards can establish comparability as between MH/SUD and medical‐surgical (MS) NA respecting the “as written” or structured component of the NQTL test. That is, the networks are comparable on their face because they were developed and established by the same required standards and hence no more stringently applied to MH/SUD benefits. Of course, if this were not the case, the MHPAEA “as written” standard would not be met.

The more critical parity questions arise in the context of how the respective MH/SUD and MS networks actually perform or the “in operation” component of the NQTL test.

Parity Implications of Inadequate Networks

Whether a provider network fails as written or in operation, a plan’s failure to provide an adequate network has the following indicators that may portend parity non‐compliance:

1. Whether the ability of a plan participant to access on a timely basis in‐network providers for covered MH/SUD benefits is comparable to timely access for MS benefits;
2. Whether participants are more dependent on OON providers (or emergency room utilization) to access covered MH/SUD benefits as compared to MS benefits; and
3. The inability of a plan participant to access covered MH/SUD benefits at all, because a) an in- network provider is not available and the plan does not have an OON benefit or b) the plan has an OON benefit but patients that have them do not use them because they cannot afford to pay the difference between what the plan will pay and what they are charged. While plan data on OON utilization is generally available it does not account for the category of participants who need and seek services but do not get them.

These patient access questions provide the context to define NA as an NQTL per the parity tests. That is, where:

1. Timely access to an in‐network provider for covered services for a MH/SUD condition is not comparable to that for MS services the parity NQTL rules are implicated;
2. Access to covered MH/SUD benefits is not comparable to that for other MS medical services, as evidenced by much greater MH/SUD utilization of OON providers and emergency rooms, the parity NQTL rules are implicated;
3. Participants are unable to access covered services at all because either: a) in‐network providers are not available and there is no OON benefit, or b) participants cannot find an in-network provider and while the plan has an OON benefit they cannot afford the cost, the parity rules are implicated; and
4. A plan has criteria governing the following, a) an OON benefit is available but the plan has criteria requiring authorization prior to use; or b) the plan has no OON benefit but permits access per established criteria to an OON provider where the service is not otherwise available from the plan’s provider network. The comparability and application of such criteria must pass muster per the NQTL rules, particularly in the operationalization of the criteria.

In any of the foregoing situations, the scope of the covered MH/SUD benefits is limited as per the NQTL definition and potentially noncompliant per the NQTL parity prohibition as follows:

1. When treatment must be secured from an OON provider a higher financial consequence is the result, ranging from higher cost sharing (if the plan has an OON benefit) to assumption of total cost where there is no plan OON benefit. This is de facto cost‐shifting to plan participants. In either case, the scope of the covered benefit has been truncated because lower cost sharing with a credentialed in‐network provider is not available.
2. Where no OON benefit is offered and in‐network care is not available and patients cannot afford to pay full fees for services, the patient does not receive care at all. If care is available on the MS side, the patient does not have parity in health care benefits. In a very real sense this situation can be characterized as a de facto but non‐transparent utilization control mechanism; i.e., the benefit is covered but it cannot be practically accessed.
3. There is also an issue concerning wait times for a consumer trying to access in‐network MH/SUD services as compared to MS. That is, the average time from the initial pursuit of care to the securing of the care regardless of the type (e.g., office visit or admission to a hospital). The average experience of a consumer trying to access MH/SUD services should be similar to that of a consumer seeking care for a MS condition. If it is not comparable, compliance with the parity test may be at issue.
4. The plan criteria governing OON utilization noted above are not parity compliant.

Interrelationship of Various NQTLs and Inadequate Networks

Ultimately a plan’s ability to assemble a proper network and fulfill the standards set for adequacy depends on robust provider participation via contract with the plan. Provider participation with a plan (whether to contract or not) is influenced by various factors, e.g., adequacy of the plan’s reimbursement schedule, ease of credentialing or the ability to be credentialed at all, the actual contract obligations such as continuity of care and conformity to often burdensome medical management protocols (especially for MH/SUD conditions), unsupportable or nontransparent plan medical necessity or clinical appropriateness criteria, among other factors. It is critical to understand that each of these factors, e.g., reimbursement rates, etc. in turn are NQTLs in their own right. These issues materially affect provider decisions to participate or not and/or the actual extent to which they accept and treat patients in plans in which they do participate. Exploration of these NQTLs may provide critical insight as to why observed deficiencies in in NA and reliance on OON utilization may be occurring.

A pictorial view is below:

Image

Framework for Assessing Parity Compliance

There are numerous measures a plan may utilize to gauge the adequacy of its network. As is the case with all NQTLs, evaluation for NA compliance requires looking beyond a plan’s assertion of comparability and theoretical results. That is, a plan may be able to demonstrate NA for MH/SUD with MS since the same required standards (e.g., geographical access) to establish adequacy are utilized. But the essential question is if the network is in fact functioning properly—if so, it will satisfy the “in operation” dimension of the test.

It is well established by various studies that provider directories are grossly inaccurate and/or providers listed accurately do not take patients from a particular plan they are listed with. Therefore, published provider directories which appear to be robust are no guarantee they are real in fact and in operation. Measures such as secret shopper surveys to determine appointment availability, use of out-network services and member wait times for distinct types of care are all indicators of how the network actually operates. This required assessment is actually consistent with what plans are already supposed to be doing themselves to maintain accreditation and provides some framework for assessing parity compliance. Gross discrepancies, that is mediocre performance of the MH/SUD network relative to the MS network (e.g., higher out-of-network utilization rates in various benefit classifications, longer wait times, etc.) is not likely MHPAEA compliant.

There are metrics available to measure a plan’s actual network performance and serve as indicators as to whether a deeper and more comprehensive review is in order. High levels of discrepancy in performance between MH/SUD and MS network performance should be red flags that suggest more in‐depth regulatory review, especially as to the above noted NQTLs, which may be causative factors as to actual provider participation or nonparticipation in a plan’s network and the plan’s network performance. Plan criteria governing access to OON benefits also require examination where applicable.

Reimbursement Rates are a Central issue

The preamble to MHPAEA Parity Rule states: “Plans and issuers may consider a wide array of factors in determining provider reimbursement rates for both MS services and MH/SUD services, such as service type; geographic market; demand for services; supply of providers; provider practice size; Medicare reimbursement rates; and training, experience and licensure of providers. The NQTL provisions require that these or other factors be applied comparably to and no more stringently than those applied with respect to MS services.”

Our experience is that these factors along with others are in fact those considered by plans when establishing payment rates. What is lacking however is a principled analytic framework that enables appropriate review and evaluation respecting the precepts of the NQTL regulatory metrics of comparability and application stringency both as written and in operation. How do plans in fact establish comparability of the factors in setting reimbursement schedules between MH/SUD and medical surgical services. Moreover, how are they applied to comport with the applied no more stringently prong of the NQTL test? It is not adequate for plans to assert that they utilize the same factors when establishing fee schedules as proof positive of the matter.

It is essential to recognize that plan payment methodologies are not straight forward quantitative formulas applied in a simple addition and subtraction manner. Many of the variables plans aver they utilize are qualitative or judgmental and their application is often subjective or discretionary. Language from the Agencies in the preamble to the interim final rule is important in this regard.

The phrase, ‘‘applied no more stringently’’ was included to ensure that any processes, strategies, evidentiary standards, or other factors that are comparable on their face are applied in the same manner to MS benefits and to MH/SUD benefits. Thus, for example, assume a claims administrator has discretion to approve benefits for treatment based on medical necessity. If that discretion is routinely used to approve MS benefits while denying MH/SUD benefits and recognized clinically appropriate standards of care do not permit such a difference, the processes used in applying the medical necessity standard are considered to be applied more stringently to MH/SUD benefits. The use of discretion in this manner violates the parity requirements for NQTLs. Several types of illnesses or injuries may require different review, as well as different care. The acute versus chronic nature of a condition, the complexity of it or the treatment involved, and other factors may affect the review. FR, Vol. 75, No.21, p.5416

The facts are that there are invariably significant differentials between reimbursement for psychiatrists and other physicians for the same services. An unpublished study performed for SAMHSA - Differential Reimbursement of Providers of Psychiatric Services - offered the following conclusions from its large claims based study; “In 2014, both psychiatrist and non-psychiatrist visits for treatment of patients with a primary psychiatric diagnosis used the same two CPT established patient office visit E/M codes for more than forty per cent of their patients with behavioral health diagnoses. Yet, the in‐network reimbursement for psychiatrists was lower compared to other types of physicians for these services. This finding might explain why psychiatrists are less likely to participate in insurance networks... we (also) demonstrate that consumers who are covered by private insurance go out‐of‐network more than twice as often for psychiatrists than for other non‐medical doctor mental health professionals for the same services to treat behavioral health conditions.” There is a plethora of other unpublished data which correlates with these findings and in many cases, indicates the differentials are greater than indicated here.

It may well be that plan application of discretionary factors in setting rates passes muster under the NQTL parity tests. We acknowledge that there is an allowance for variance in results. However, result variance is only valid where in fact the methodologies utilized by plans can justify the result per the regulatory tests. We do not think plan methodologies have been adequately vetted in this regard and the fact that psychiatric physicians are consistently paid less for identical medical work suggest that this is in order and if the discretionary variables in a plan’s payment methodology are being routinely used in a discriminatory manner which results in lower rates. This lack of oversight and review by authorities of jurisdiction is in part due to a lack of informed guidance on the issue. As noted above, the Departments noted that additional guidance may be forthcoming if questions persist. We would note that at the July 27, 2017 Listening Session the request for more guidance was also made by groups representing health plans. We concur and would welcome the opportunity to present additional information on a methodology to review the rate question and offer the views of the actuarial consultant - Milliman - we work with.

Back to complete list of written comments