Payers Now Publish Their Prior Authorization Numbers. They Still Do Not Count Your Staff's Time.

At 4:40 on a Thursday, the practice manager at a three-clinician group is on hold with a payer for the second time that day. The first call was to confirm that the plan requires authorization for psychological testing at all. This one is to find out why the portal has said "pending" for a week. The fax went out Monday. Somewhere in that plan's public report, this request will be counted once, as a decision made in a reasonable number of days.

That report is new. Since March 2026, plans covered by the CMS Interoperability and Prior Authorization rule have had to post, in public, how many prior authorization requests they received, how many they approved and denied, how many approvals came only after an appeal, and how long decisions took. That is real progress, and I want to give it credit before I say what it leaves out. What it leaves out is the practice manager. No report counts her afternoon.

Who this applies to

If you do routine outpatient psychotherapy, prior authorization may rarely touch you. If you do psychological testing, intensive outpatient or partial hospitalization work, TMS, extended or high-frequency sessions, or you bill a good number of Medicaid and Medicare Advantage plans, it touches you often, and the rules shift by state and by plan.

The reports cover Medicare Advantage plans, state Medicaid and CHIP programs, their managed care plans, and most marketplace plans sold on the federal exchange. They do not cover original Medicare or commercial employer plans, which is where many private-pay-leaning practices live. If you bill Medicaid or Medicare Advantage, your plan's report should be findable through ordinary website navigation, because CMS says a login-only portal does not count as public.

What the reports say, and what they cannot

Each covered plan posts the services that require authorization and a set of aggregate numbers: the share of standard requests approved and denied, the share approved after an appeal, the share approved after a permitted extension, and mean and median response times, with expedited requests reported separately.

Three 2025 reports show the range. PrimeTime Health Plan, a Medicare Advantage contract in Ohio, reported 971 standard requests, 84 percent approved, and four standard appeals. Aetna Better Health of Texas, a Medicaid and CHIP plan, reported 41,416 standard requests, 92.3 percent approved, and 448 standard appeals. Priority Partners, a Maryland Medicaid managed care plan, published the same families of metrics with a searchable service list. Do not read those as a ranking. Different programs, different populations, denominators that do not line up. What they prove is that the information exists and can be read plan by plan.

Here is what none of them can show. A payer's clock starts when the request arrives. The practice's clock started earlier, when someone checked whether authorization was needed, found the right submission route, and assembled the justification, and it keeps running after the decision, through information requests, status checks, resubmissions, and appeals. Two three-day turnarounds can contain ten active staff minutes in one case and an afternoon of fax, portal, and phone tree in the other. The report calls both of them three days.

What a federal audit found behind the numbers

In June 2026, the HHS Office of Inspector General published an audit of Community Behavioral Health, the Medicaid behavioral-health managed care organization for Philadelphia. OIG chose it because it had the most denied service requests among Pennsylvania's behavioral-health Medicaid plans, so this is a targeted case, not a national sample.

The plan received 96,130 behavioral-health prior authorization requests in 2023 and denied 1,102 of them. OIG pulled a simple random sample of 100 of those denials. None of the 100 met every applicable administrative and procedural requirement.

• 17 denial notices went to the wrong address because required address-change procedures were not followed.
• 1 notice was mailed 31 days after the request instead of within the 21-day limit, which under the rules should have made the request an automatic approval.
• All 100 left out the enrollee's right to free copies of the documents behind the denial.
• 14 did not clearly explain the decision.
• 34 lacked the documented consultation required for enrollees under 21.
• 58 had more than one problem.

From the sample, OIG concluded that none of the 1,102 denied requests complied with all applicable federal and state requirements. The plan agreed with two findings and disputed three. Pennsylvania accepted the recommendation aimed at the state and revised its 2026 contract language.

Two limits, stated plainly. The audit tested paperwork and procedure, not whether the denied care was clinically appropriate. And it describes one plan's 2023 denials, not Medicaid nationally. What it shows is exactly what an approval rate cannot: a request can be decided on time, counted correctly in a public report, and still reach a family at the wrong address in a notice that leaves out their rights.

A payer report, an audit, and a practice's own log measure three different things, and each is useful as long as it is read for what it measures. An OIG issue count is not a denial rate, payer turnaround is not staff time, and one practice's bottleneck is not a national pattern.

Why I built the second clock into the chart

I am not a therapist. I spent years running supply chain operations for large companies, where the first rule was that a process nobody measures is a process nobody can fix. When I sat down with therapists to build Therapy Companion, prior authorization was the process they described in the most detail and measured the least. Nobody could say where the hours went, only that they went.

So the tracker records the request side of the log inside the client record: submission route, status changes, information requests, appeal deadlines, and sessions used against each approved authorization. It lives where the protected data already lives, which is the part I would not compromise on. Names, member numbers, authorization numbers, service dates, diagnoses, and denial letters do not belong in a spreadsheet, a note-taking app, or a consumer AI tool, even for the sake of an easier analysis. Whatever leaves the protected system should be an aggregate, counts and averages by payer and by stage, never a row about one person.

You do not need my software to start. A log of predefined events per request, with the staff time attached to each step, will show you where the hours go within a month.

Five questions for a practice owner

1. Which of the services you bill actually require authorization, plan by plan? The new public service lists answer that without a phone call.
2. For each covered plan, what does its own report say about approval rate, approval after appeal, and median turnaround?
3. Where does your time go: before submission, during information requests, or after a denial?
4. How many status checks and channel changes does a typical request take, and who is doing them?
5. Can you answer all of that from aggregates, with every row about a specific client staying inside the protected system?

The new reports make the payer's side visible for the first time. The practice's side stays dark until the practice measures it, carefully, inside the walls that protect the people whose care is being authorized. The practice manager on hold at 4:40 deserves to be counted too.

Sources

1. CMS Interoperability and Prior Authorization Final Rule CMS-0057-F, fact sheet. https://www.cms.gov/newsroom/fact-sheets/cms-interoperability-prior-authorization-final-rule-cms-0057-f
2. CMS Prior Authorization Metrics Reporting Overview and Template. https://www.cms.gov/prior-authorization-metrics-reporting-overview-template
3. HHS OIG, Community Behavioral Health prior authorization audit, landing page. https://oig.hhs.gov/reports/all/2026/community-behavioral-health-did-not-comply-with-requirements-when-denying-prior-authorization-requests
4. HHS OIG Audit A-03-24-00204. https://oig.hhs.gov/documents/audit/11700/A-03-24-00204.pdf
5. PrimeTime Health Plan 2025 prior authorization metrics. https://www.primetimehealthplan.com/assets/PDFs/prior-authorization-metrics-template-for-payers_2025_Plan.pdf
6. Aetna Better Health of Texas 2025 prior authorization metrics. https://ch.aetnabetterhealth.com/content/dam/aetna/pdfs/2025_ABHTX_CMS_Web_Metrics-Provider.pdf
7. Priority Partners 2025 prior authorization metrics. https://www.ppmco.org/wp-content/uploads/2026/03/pp-pa-metrics_508.pdf

Public report figures (paragraphs on the three 2025 reports) come from sources 5, 6, and 7; the reporting scope and metric definitions from sources 1 and 2; the audit findings from sources 3 and 4. All public URLs were retrieved without login on September 2, 2026. No patient or client data was used. This article is not legal, billing, coding, or clinical advice.