Medicaid Managed Care: Denials of Prior Authorization for Services

Kudos to the OIG!  That would be the Office of the Inspector General of the Department of Health and Human Services.  Although OIG is better known for its fraud-fighting persona, it also has a broader mission of making government programs work better.  And with last month’s report, “High Rates of Prior Authorization Denials by Some Plans and Limited State Oversight Raise Concerns About Access to Care in Medicaid Managed Care,” OIG has nailed the problem. Along the way, it has also mapped out a path to reform.  The rest is up to CMS, state Medicaid agencies and health care advocates.

Before we get into the substance, let’s take a paragraph (or two) to appreciate just how groundbreaking this report is.  It took a lot of work.  OIG measured the performance of 115 different MCOs by rates of denial of prior authorization requests, and it surveyed 37 different state agencies about their use of prior authorization data.  The report clearly identifies the problems at both the MCO and state agency levels and makes thoughtful recommendations to CMS for policy changes to reduce high denial rates and improve access to care.  But most importantly, in presenting its findings, OIG identifies each of the MCOs, their parent companies, and each of the state Medicaid agencies.

From a transparency standpoint, this is a 180 degree turn from the report OIG issued just last year, “CMS Has Opportunities To Strengthen States’ Oversight of Medicaid Managed Care Plans’ Report of Medical Loss Ratios.” That report, which reviewed 495 MLR reports and surveyed 43 states, also took a lot of work.  It found that nearly half of the MLR reports were incomplete, even though state agencies indicated that they check the MLR reports for completeness, but crucially, OIG did not identify the MCOs or the states with incomplete reports.  In this report on high rates of prior authorization denials and lax state oversight, OIG names names.  In doing so, it creates opportunities for accountability that last year’s report did not. Low-performing MCOs—and the state Medicaid agencies that oversee them—are now known, and reform efforts can focus on them.  Here’s hoping this level of transparency is the new normal.

Now for the substance.  Prior authorization is widely used by insurers in employer sponsored plans as well public programs like Medicare, Medicaid, CHIP and the Marketplaces to manage the use of services by enrollees.  If a plan requires prior authorization for a particular item or service, the provider won’t be paid for furnishing that item or service without advance approval by the plan. Insurers argue that prior authorization requirements are part of good care management because they ensure that only care that is medically necessary is delivered.  The other view is that prior authorization is a tool that insurers use to delay or deny payment for services that patients need and should be covered, as exemplified by this KFF News exposé, one of several media reports cited by the OIG.

OIG collected data on prior authorization denials, appeals, and appeal outcomes from a total of 115 MCOs operated by seven parent companies in 37 states in 2019.  Each of the seven parent companies—Aetna Inc. (now CVSHealth), Americare Caritas, Anthem (now Elevance Health), CareSource, Centene, Molina Healthcare, and UnitedHealthcare—had at least one million Medicaid enrollees; combined, they had a total of 29.8 million enrollees in their Medicaid products that year.  OIG also surveyed state Medicaid agency officials in those 37 states in order to better understand state oversight of MCO prior authorization denials and appeals. OIG calculated the rates at which each MCO denied prior authorization requests, the rates at which enrollees or their providers appealed those denials, and the rates at which each MCO upheld those denials on appeal.

Among the many OIG findings:  the 115 MCOs reviewed denied one out of every eight requests for prior authorization of services, an average rate of 12.5 percent.  Denial rates varied from a low of 2 percent to a high of 41 percent.  Among the parent companies, one stood out: Molina Healthcare.  Its overall denial rate of 17.7 percent was the highest among all seven companies, and seven of its MCOs had denial rates greater than 25 percent.  (Perhaps this is what Molina, which derives over 80 percent of its revenues from Medicaid, is referring to when it tells investors its earnings reflect “strong operating performance and medical cost management.”).  In addition, OIG found that, of the 35 states (including DC) that responded to its survey, only 13 reported that they routinely reviewed a sample of MCO denials of prior authorization requests for appropriateness.  Several of these states told OIG they identified “inappropriate denials for medically necessary drug therapy, health screening services for children, and inpatient hospital care.”  Only 22 states reported that they used data on denials of prior authorization requests for oversight of the MCOs with which they contract.

There’s more, but you get the idea.  Hopefully, CMS will take note, although its responses to the OIG recommendations are not particularly reassuring.  Of the five OIG recommendations, CMS concurred with only one, indicating that it would issue guidance and provide technical assistance to states on using prior authorization data for oversight of MCOs.  To its credit, however, CMS has proposed to strengthen regulations against abusive prior authorization controls in Medicaid, Medicare, and the federally-run Marketplaces.  Hopefully, the final rule will reduce administrative burdens on providers and increase beneficiary access to medically necessary services.

Regardless of the outcome of the CMS rulemaking, here’s how the OIG report can help beneficiaries and their advocates at the state level now. The denial rates for each of the 115 MCOs are set forth in Appendix B.  AMGP Georgia Managed Care Company, a subsidiary of Anthem (now Elevance Health), which enrolled 382,000 Medicaid beneficiaries in 2019, had a denial rate of 33.7%, more than twice as high as the rates of the other three MCOs operating in Georgia that OIG reviewed.  Appendix D, which summarizes state agency responses to the OIG survey about their prior authorization oversight policies, shows that Georgia did not respond to the OIG survey.

This begs a number of questions: Why was AMGP’s denial rate so high?  What types of items and services did AMGP most frequently deny?  If they included EPSDT services for children, what was the rationale?  Which beneficiary populations, stratified by race and ethnicity, were impacted by the denials?  Did AMGP’s high denial rate continue in 2020 and 2021 and 2022?  What actions, if any, has Georgia’s Medicaid agency taken to address AMGP’s high denial rate(s)?  Does the agency review samples of MCO denials or monitor MCO denial data?  What evidence is there that beneficiaries enrolled in AMGP and other MCOs with high denial rates are getting access to the services they need?

AMGP and Georgia are just one example.  OIG found 12 MCOs with denial rates greater than 25 percent, or twice the average denial rate among the 115 MCOs.  And although only one other state (Louisiana) failed to respond to the OIG survey altogether, two states (Nevada and Virginia) reported that they do not regularly review denials for appropriateness or use denials data for oversight, and 13 other states (Arkansas, California, Colorado, Iowa, Michigan, Mississippi, Missouri, New Jersey, Nebraska, Tennessee, Utah and Wisconsin) and the District of Columbia reported that they do not use denials data for oversight of MCO performance.

As problematic as that might seem, there’s a more fundamental issue.  The OIG report points out that the 12.5 percent average denial rate for Medicaid was more than twice the average denial rate of 5.7 percent for Medicare in 2019.  In fact, only two of the 115 MCOs the OIG reviewed had prior authorization denial rates equal to or lower than 5.7 percent for Medicare:  AmeriHealth Caritas Pennsylvania Community Health Choices (1.9 percent) and AmeriHealth Caritas Keystone First Community HealthChoices (2.2 percent).  All seven parent companies also compete in the Medicare market.  Are the prior authorization denial rates in their Medicaid products significantly higher than those in their Medicaid products?  If so, what is that about?

The OIG report raises a lot of questions for MCOs, their parent companies, state Medicaid agencies, and CMS.  It’s time for some answers.

Andy Schneider is a Research Professor at the Georgetown University McCourt School of Public Policy.