Network adequacy

Read it this way All three measures land far from what a working directory should show: only 18% of calls reached an appointment, a third of listings were simply wrong, and a separate California study found 73% of calls failed to secure an appointment. These come from two different studies and geographies, so read them as three separate pieces of evidence for the same pattern rather than one consistent measurement. Use this chart to see whether the network problem is administrative, clinical, or parity-related, and why the recommendation focuses on audited access performance instead of nominal provider counts.

US Senate Committee on Finance Ghost Network Secret Shopper Study, and California directory study via KFF · 2023 · source

Read it this way In every setting the behavioral-health bar is several times taller than its medical or surgical comparator, most sharply in outpatient facility care (31.6% vs 5.5%). As the caveat notes, out-of-network use is a proxy for thin behavioral-health networks rather than a direct denial count, so the direction of the gap is more reliable than its exact size. Use this chart to see whether the network problem is administrative, clinical, or parity-related, and why the recommendation focuses on audited access performance instead of nominal provider counts.

Caveat Out-of-network utilization is the standard proxy in parity research because comparable claims-level denial data is not published. Figures are 2017.

Milliman, Addiction and Mental Health vs Physical Health: Widening Disparities in Network Use · 2019 · source

Read it this way In-person psychiatry waits more than three times as long as primary care (67 vs 20.6 days), and telepsychiatry cuts that gap roughly in half (43 days) without closing it. As the caveat notes, the figures come from two separate studies with different methods and geographies, so read the gap as directional rather than a precise multiple. Use this chart to see whether the network problem is administrative, clinical, or parity-related, and why the recommendation focuses on audited access performance instead of nominal provider counts.

Caveat The psychiatry and primary-care figures come from different studies with different methods and geographies, so treat the gap as directional.

General Hospital Psychiatry (2023), and AMN Healthcare 2022 Survey of Physician Appointment Wait Times · 2023 · source

Read it this way Behavioral health actually carries the strictest federal wait standard on paper, 10 business days, of the three provider types shown. This chart is only the rule, not whether plans actually meet it, so it can't be read alongside the ghost-network findings as proof the standard is being honored. Use this chart to see whether the network problem is administrative, clinical, or parity-related, and why the recommendation focuses on audited access performance instead of nominal provider counts.

CMS, Network Adequacy QHP Certification (2025 Payment Notice) · 2025 · source

Read it this way Tennessee allows five times as many enrollees per primary-care provider as Michigan, showing how much weaker a 'reasonable access' standard can be from state to state. Only three states are shown here, so this can't be read as a national ranking or a representative sample of state strength. Use this chart to see whether the network problem is administrative, clinical, or parity-related, and why the recommendation focuses on audited access performance instead of nominal provider counts.

KFF, Network Adequacy Standards and Enforcement · 2023 · source

Read it this way Violations cited hold roughly steady from FY2019 to FY2022 before nearly doubling in FY2023 (18 to 31), a sharp break from the earlier pattern. This counts violations DOL chose to cite, so the jump could reflect worsening plan behavior, intensifying enforcement, or both, and the chart alone can't separate the two. Use this chart to see whether the network problem is administrative, clinical, or parity-related, and why the recommendation focuses on audited access performance instead of nominal provider counts.

US DOL Employee Benefits Security Administration, MHPAEA Enforcement Fact Sheets FY2019 to FY2023 · 2023 · source

Read it this way Both reporting cycles found the large majority of comparative analyses insufficient on first submission, easing only from 100% to 76%, so most plans are still failing on their first try. As the caveat notes, the 2023 figure is a reconstructed ratio of 138 insufficiency letters to 182 plans reviewed, not a rate the report itself publishes. Use this chart to see whether the network problem is administrative, clinical, or parity-related, and why the recommendation focuses on audited access performance instead of nominal provider counts.

Caveat The 76% figure is the DOL ratio of 138 insufficiency letters to 182 plans reviewed. The 2023 report does not print a single top-line percentage.

DOL, HHS, and Treasury, MHPAEA Comparative Analysis Reports to Congress (2022 and 2023) · 2023 · source

Read it this way Only about 10 of 50 states set a numeric provider-to-enrollee standard at all, and fewer still, about 4, closely adopt the NAIC model act, so most states rely on vague reasonable-access language with no number attached. As the caveat notes, these are the only quantifiable counts available and are approximate, not an authoritative 50-state tally. Use this chart to see whether the network problem is administrative, clinical, or parity-related, and why the recommendation focuses on audited access performance instead of nominal provider counts.

Caveat These are the only quantifiable adoption counts in the source and are approximate. No authoritative 50-state tally of network-adequacy standard strength exists.

KFF Network Adequacy Standards and Enforcement, and NAIC Model Act #74 · 2023 · source