From building parity compliance models for group health renewals, the pattern is consistent: plans that have never formally benchmarked MH/SUD provider reimbursement against medical/surgical rates discover gaps wide enough to move their allowed-cost assumptions by 3 to 8 percent on behavioral health lines. The September 2024 MHPAEA final rule (89 FR 77586) transforms that discovery from an optional exercise into a regulatory mandate. For plan years beginning on or after January 1, 2026, the Phase 2 provisions require health plans to collect outcomes data on claim denial rates, provider reimbursement, out-of-network utilization, and network adequacy for mental health and substance use disorder (MH/SUD) benefits, then compare those metrics against the corresponding medical/surgical (M/S) benchmarks within each benefit classification.

Where material disparities surface, plans must take “reasonable action” to address them: expanding networks, raising reimbursement, or loosening prior authorization requirements. Each of those remediation paths carries cost. Pricing actuaries building 2027 renewal assumptions need a framework for quantifying the incremental load, because this is no longer a theoretical compliance risk. Plans are collecting the data now, and the gaps are measurable.

5.4x
MH/SUD Prior Auth Rate vs. M/S (Kennedy Forum/Milliman)
85%
Typical MH/SUD Reimbursement as % of Medicare
25% vs 12%
MH/SUD vs M/S Out-of-Network Utilization

The NQTL Comparative Analysis Framework

The data evaluation provisions require actuaries to compare MH/SUD outcomes against M/S outcomes within each of six benefit classifications: inpatient in-network, inpatient out-of-network, outpatient in-network, outpatient out-of-network, emergency care, and prescription drugs. This is not a plan-level aggregate comparison. The classification-level granularity matters because parity violations often concentrate in one or two classifications while aggregate statistics mask the disparity.

Within each classification, the required metrics fall into four categories:

Metric Category MH/SUD Measure M/S Comparator Pricing Impact Channel
Claim Denial Rates % of MH/SUD claims denied by classification % of M/S claims denied in same classification Completion factor (more claims approved = higher allowed cost)
Provider Reimbursement Average allowed amount per unit, MH/SUD Average allowed amount per unit, M/S Network discount factor (higher reimbursement = less discount)
Out-of-Network Utilization % of MH/SUD services rendered OON % of M/S services rendered OON Network leakage cost (OON claims cost 40-80% more)
Prior Authorization Auth approval/denial rates, turnaround times Auth approval/denial rates for M/S Utilization trend (looser auth = higher expected utilization)

The pricing actuary’s job is to translate each metric gap into a cost adjustment. A plan that denies 18% of outpatient MH/SUD claims versus 9% of outpatient M/S claims has a 9-point denial rate disparity. If regulatory pressure or “reasonable action” requirements force that MH/SUD denial rate down to 12%, the completion factor on outpatient behavioral health claims increases by roughly 7%, flowing directly into the allowed-cost projection.

Provider Reimbursement Parity Creates a Fee-Schedule Floor

The reimbursement comparison is where pricing actuaries will see the most direct cost impact. The final rule requires plans to evaluate whether the methodology used to set MH/SUD provider reimbursement rates produces results that are comparable to M/S reimbursement rates within the same benefit classification. In practice, this means benchmarking both against a common reference point, typically Medicare fee schedules.

FAIR Health data and Milliman behavioral health benchmarking studies consistently show MH/SUD provider reimbursement running at 80 to 90% of Medicare rates, while M/S provider reimbursement for commercial plans typically runs at 110 to 130% of Medicare. The gap is structural: behavioral health providers historically accepted lower rates because demand exceeded network slots, giving plans little incentive to compete on reimbursement. The MHPAEA data evaluation requirement changes that dynamic by making the gap visible and actionable.

Plans do not need to achieve dollar-for-dollar parity between MH/SUD and M/S allowed amounts per unit; the standard is that the methodology for setting rates cannot be more restrictive for MH/SUD. But if the methodology produces allowed amounts at 85% of Medicare for behavioral health while the same methodology produces 115% of Medicare for medical/surgical, the plan must explain why the disparity does not reflect a more restrictive standard. In most cases, it cannot.

The pricing implication: if a plan must raise MH/SUD reimbursement from 85% of Medicare toward parity with M/S rates, the network discount factor on behavioral health claims changes. For a plan where behavioral health represents 12% of total allowed costs and the reimbursement gap is 25 to 30 percentage points (85% of Medicare vs. 110 to 115%), closing even half that gap increases the behavioral health allowed-cost base by approximately 15%, which translates to a 1.5 to 2.0% increase in total plan cost. That figure compounds with the utilization effects discussed below.

Network Adequacy as an Explicit NQTL

The final rule elevates network composition to an explicit NQTL category for the first time. Plans must now measure and report out-of-network utilization rates for MH/SUD versus M/S, provider directory accuracy, and time-and-distance standards for behavioral health access. This codifies what health actuaries have long observed: MH/SUD out-of-network utilization rates persistently exceed M/S rates, often by a factor of two or more.

Kennedy Forum/Milliman research found that behavioral health inpatient services are utilized out-of-network at 5.2 times the rate of medical/surgical inpatient services. At the outpatient level, FAIR Health data shows MH/SUD out-of-network utilization running at approximately 25% of total behavioral health claims versus 10 to 12% for M/S outpatient services. The persistent gap reflects inadequate in-network behavioral health provider supply, which the rule now treats as a compliance problem rather than a market reality.

For pricing, network adequacy remediation costs flow through two channels. First, the direct cost of expanding in-network behavioral health provider panels: higher per-unit reimbursement to attract providers, credentialing and directory management costs, and potentially telehealth infrastructure investment. Second, the shift in claim mix from out-of-network (where balance billing and member cost-sharing reduce plan liability) to in-network (where the plan bears a higher percentage of allowed cost). A plan that successfully moves 10 percentage points of behavioral health utilization from OON to in-network will see the average plan-paid percentage on those claims increase, even if the per-unit reimbursement stays flat.

The Meaningful Benefits Standard and Benefit Design Constraints

Separate from the NQTL data evaluation requirements, the 2026-effective “meaningful benefits” standard constrains actuarial discretion in benefit design. Where a plan covers a mental health or substance use disorder condition, it must provide benefits for treatment of that condition in each classification where M/S benefits are provided, and those benefits must be “meaningful” as measured against “generally recognized independent professional medical or clinical standards.”

This standard effectively eliminates the practice of covering MH/SUD diagnoses at the outpatient level while excluding residential treatment or intensive outpatient programs (IOPs) from the inpatient classification. If medical/surgical benefits include inpatient rehabilitation or skilled nursing facility stays, excluding residential MH/SUD treatment becomes difficult to defend. The DOL has signaled that exclusions of applied behavior analysis (ABA) therapy for autism spectrum disorder, medication-assisted treatment (MAT) for opioid use disorder, and residential eating disorder treatment will receive heightened scrutiny.

For self-funded plans that previously excluded residential MH/SUD treatment, the cost impact of adding this benefit can be material. Residential SUD treatment episodes average $25,000 to $45,000 per admission, and residential mental health treatment for eating disorders or treatment-resistant depression can run $30,000 to $80,000. Even at low utilization rates of 1 to 3 admissions per 1,000 members per year, the per-member-per-month (PMPM) impact ranges from $2 to $10 depending on plan size and population demographics.

Worked Example: Quantifying the Parity Gap

Consider a 5,000-member self-funded employer plan where behavioral health represents 11% of total allowed costs ($6.2 million of $56.4 million annual claims). The plan’s NQTL data evaluation reveals three material disparities:

Reimbursement gap: MH/SUD provider reimbursement averages 85% of Medicare; M/S reimbursement averages 110% of Medicare. To close half the gap (moving MH/SUD to approximately 98% of Medicare), the actuary applies a 15.3% reimbursement increase to the behavioral health allowed-cost base: ($6.2M × 0.153) = $949,000 annual cost increase.

Denial rate differential: Outpatient MH/SUD denial rate is 17%; outpatient M/S is 8%. Reducing the MH/SUD denial rate to 12% increases the completion factor by approximately 6%: ($6.2M × 0.65 outpatient share × 0.06) = $242,000.

Induced demand from expanded access: As network adequacy improves and prior authorization loosens, utilization growth above medical trend compounds the cost. Using a conservative induced demand factor of 3 to 5% on the adjusted behavioral health base: ($6.2M × 1.153 × 1.06 × 0.04 midpoint) = $303,000.

Total incremental MHPAEA compliance cost: $949K + $242K + $303K = approximately $1.49 million, or $24.90 PMPM. Against a base plan cost of $940 PMPM, this represents a 2.6% plan-level cost increase attributable to parity compliance remediation.

This estimate uses plan-specific experience for the base and FAIR Health/Optum behavioral health benchmarks for the induced demand assumption. Where plan data has fewer than 1,000 behavioral health claimants, credibility-weighted blending with external benchmarks is essential. Plans with credibility below 50% on behavioral health experience should weight external sources at 60% or more in trend selection, per ASOP No. 25 principles.

Self-Funded Plans and Stop-Loss Exposure

Self-funded plans face a distinct set of MHPAEA pricing pressures that fully insured plans partially absorb through carrier risk pooling. The final rule eliminates much of the historical flexibility self-funded plans exercised in MH/SUD benefit design. Plans can no longer apply “biased” evidentiary standards, meaning that claim review criteria for MH/SUD cannot rely on historical plan data from periods when the plan was not MHPAEA-compliant. This prohibition forces actuaries to use external benchmarks rather than plan-specific experience for MH/SUD cost projections in many cases.

The stop-loss implications are material. Residential MH/SUD treatment episodes routinely exceed $50,000 and can reach $150,000 to $250,000 for extended stays. For plans with specific deductibles at $150,000 or $200,000, the meaningful benefits standard’s expansion of covered residential treatment creates a new class of shock claims that was previously excluded by benefit design. Stop-loss carriers pricing 2027 renewals should expect increased specific claim frequency from residential MH/SUD admissions, particularly for substance use disorder treatment where episodes of 60 to 90 days are clinically standard.

The aggregate stop-loss corridor also tightens. Plans that see a 2 to 3% total cost increase from MHPAEA compliance remediation are closer to their aggregate attachment point, increasing the probability of aggregate corridor penetration. Pricing actuaries setting aggregate factors should model the MHPAEA compliance cost as a separate trend layer rather than absorbing it into the medical trend assumption, because the compliance cost is a step function (it hits once as plans remediate) rather than a gradual trend.

The Feedback Loop: Why Costs Compound Beyond Fee-Schedule Adjustments

Pricing actuaries modeling MHPAEA compliance costs must account for a compounding feedback loop. Higher provider reimbursement attracts more behavioral health providers to in-network panels, improving network adequacy. Better network adequacy reduces barriers to access, increasing utilization. Increased utilization generates more claims volume, raising plan costs beyond the initial fee-schedule adjustment. Simultaneously, looser prior authorization standards (driven by denial rate parity requirements) further accelerate utilization growth.

This feedback mechanism means a simple multiplicative adjustment to the allowed-cost base understates the true cost. The reimbursement increase, completion factor adjustment, and induced demand factor interact rather than summing independently. In the worked example above, the $1.49 million estimate represents first-year remediation cost. By renewal year two, the induced demand component may double as the full effect of expanded access materializes in claims experience.

Health actuaries pricing 2027 renewals should build a two-year cost trajectory: the initial remediation year (2026 plan year) and the first full experience year (2027), with the induced demand assumption loaded more heavily in year two. Plans that remediate early in the 2026 plan year will see more utilization growth in their first full experience period than plans that remediate late, creating a timing dependency in the trend assumption that varies by client.

Why This Matters for 2027 Renewal Pricing

The MHPAEA data evaluation requirements create a pricing challenge distinct from standard medical trend analysis. This is not a utilization trend or a unit cost trend in the traditional sense; it is a structural compliance cost that varies by plan based on the severity of existing parity gaps. Plans that have historically maintained robust behavioral health networks and competitive reimbursement will see minimal impact. Plans that relied on restrictive prior authorization, narrow behavioral health networks, and below-market reimbursement face material cost adjustments.

For pricing actuaries, the practical workflow involves four steps. First, obtain the plan’s NQTL comparative analysis data (denial rates, reimbursement benchmarks, OON utilization rates) as a standard part of the renewal data request. Second, identify material disparities by classification using the MH/SUD-versus-M/S comparison framework. Third, model the remediation cost for each disparity through the appropriate pricing channel (completion factor, network discount, utilization trend). Fourth, layer the compliance cost adjustment on top of standard medical and pharmacy trend assumptions, treating it as a one-time step for the remediation year with a trailing induced demand component in subsequent years.

Plans that ignore the compliance cost in pricing will underprice. Plans that overreact by assuming immediate full parity across all metrics will overprice. The actuarial value is in quantifying the specific gaps for each plan and building credibility-weighted cost projections that reflect both plan-specific data and external behavioral health benchmarks.

Further Reading

Sources