The Extent and Growth of Prior Authorization in Medicare Advantage

ABSTRACT

Objectives: To assess trends in the use of prior authorization requirements among Medicare Advantage (MA) plans.

Study Design: Descriptive quantitative analysis.

Methods: Data were from the CMS MA benefit and enrollment files for 2009-2019, supplemented with area-level data on demographic and provider market characteristics. For each service category, we calculated the annual share of MA enrollees in plans requiring at least some prior authorization and plotted trends over time. We mapped the county-level share of MA enrollees exposed to prior authorization in 2009 vs 2019. We quantified the association between local share of MA enrollees exposed to prior authorization and characteristics of that county in the same year. Finally, we plotted the share of MA enrollees exposed to prior authorization requirements over time for the 6 largest MA carriers.

Results: From 2009 to 2019, the share of MA enrollees in plans requiring prior authorization for any service remained stable. By service category, the share of MA enrollees exposed to prior authorization ranged from 30.7% (physician specialist services) to 72.2% (durable medical equipment) in 2019, with most service categories requiring prior authorization more often over time. Several area-level demographic and provider market characteristics were associated with prior authorization requirements, but these associations weakened over time. The use of prior authorization varied widely across plans.

Conclusions: In 2019, roughly 3 in 4 MA enrollees were in a plan requiring prior authorization. Service-level, area-level, and carrier-level patterns suggest a wide range of approaches to prior authorization requirements.

Am J Manag Care. 2024;30(3):e85-e92. https://doi.org/10.37765/ajmc.2024.89519

_____

Takeaway Points

Prior authorization is a common utilization-management tool among Medicare Advantage (MA) plans. However, service-level, area-level, and carrier-level patterns suggest variation in how plans use prior authorization.

• Roughly 3 in 4 MA enrollees are in a plan requiring prior authorization for some services.
• Across all plans, the use of prior authorization requirements increased from 2009 to 2019 for the majority of service categories.
• The use of prior authorization varies geographically, with a growing number of counties exceeding 90% exposure to prior authorization from 2009 to 2019.
• Whereas some MA carriers impose prior authorization requirements across nearly all service categories, others use it more sparingly.
• Opportunities exist to improve and expand data collection on the use of prior authorization by MA carriers.

One of the most notable trends in the Medicare program is the increasing use of private managed care plans to deliver benefits to Medicare beneficiaries. The percentage of Medicare beneficiaries enrolled in a Medicare Advantage (MA) plan rose from 24% in 2009 to 50% in 2023.¹ MA plans rely on a wide range of managed care techniques to contain spending and to encourage high-value health care service use. These techniques include closed panels of providers that are covered by an MA plan, care management activities, and the use of prior authorization for medical services.

Prior authorization is the process through which a clinician must obtain approval from an insurer before a particular service will be covered by the insurer. Prior authorization practices may have benefits along with costs, which may not be borne equitably across patients and providers—and as such are controversial and polarizing.² Significant majorities of commercial health insurers that responded to a 2022 survey by AHIP reported that prior authorization had a positive impact on quality, affordability, and safety of health care.³ In contrast, physician groups have voiced significant concerns with prior authorization. For instance, among practicing physicians who responded to the 2022 American Medical Association prior authorization physician survey and reported doing any prior authorization requests in a typical week, significant concerns were noted about prior authorization leading to care delays, at least some serious adverse events for patients, and overall worse clinical outcomes.⁴ Furthermore, physician respondents to that survey frequently indicated that prior authorization requirements are administrative burdens that raise the costs of medical practice.

In the context of concerns about negative consequences of prior authorization for patient care,^5,6 a range of public policies have been proposed to regulate prior authorization practices. For example, state lawmakers have pursued legislation that requires prior authorization decisions within a certain amount of time⁷ or requires that insurers grant “gold cards” that exempt clinicians from prior authorization if they meet a threshold for prior authorization approvals.⁸ Federal policy makers have also considered prior authorization legislation. Specific to MA plans, during the 117th Congress (2021-2022), the US House of Representatives passed and the US Senate introduced the bipartisan Improving Seniors’ Timely Access to Care Act, which would require that MA plans adopt electronic prior authorization, deliver real-time decisions for certain services, and adhere to transparency and reporting standards around prior authorization practices.⁹

Despite the increasing policy attention to controversial prior authorization practices, there are limited systematic and nationally representative data on how prior authorization is used and how it has changed over time.^10-12 Much of the existing literature focuses on prior authorization for prescription drugs,¹³ although these practices are used much more widely by health insurers. In this article, we focus on prior authorization within MA plans. Some information on prior authorization practices is available from KFF reports, but those reports do not quantify trends over time and focus on very high-level measures of prior authorization use. In this descriptive article, we use the only multiyear source of data across all MA insurers to document how the use of prior authorization by MA plans has changed over time and across major service categories, how the use of prior authorization varies by geography, and variations in how large MA carriers impose prior authorization requirements. Our objective is to help inform policy discussions by establishing basic facts about how MA plans use prior authorization and how that has evolved over recent years. Our findings should be interpreted as an imprecise estimate on the use of prior authorization in MA for 2 reasons. First, our data source requires plans to merely report whether they impose prior authorization for any enrollees (rather than how prevalent its use is). Second, we document some inconsistencies in the data reported by plans to CMS and consumer-facing prior authorization information available online. For these reasons, we hope that our findings inform policy discussions not only about regulating the use of prior authorization in MA but also about improving data collection.

METHODS

Data

This analysis relied primarily on data from the CMS MA benefit and enrollment files for 2009 through 2019. The benefit files contain information submitted annually by MA plans to CMS as part of the annual bidding process.¹⁴ Relevant to this analysis, the benefit files include information on whether each plan imposes prior authorization requirements for particular service categories. Each binary variable must equal 1 if the plan imposes prior authorization requirements for in-network care, defined as “a process requiring the physician or other health care provider to obtain advance approval from the plan that payment will be made for a service or item furnished to an enrollee.”¹⁵ The enrollment files contain annual counts of Medicare beneficiaries enrolled in each MA plan–county combination. Similar to past analyses of these data,¹⁶ we excluded cost plans, Programs of All-inclusive Care for the Elderly plans, health care prepayment plans, and Medicare-Medicaid plans from our sample because their enrollment requirements and payment regulations may not be comparable to other MA plans (see eAppendix Table 1 [eAppendix available at ajmc.com] for sample sizes before and after exclusion). We also excluded employer plans, as the benefit data may not be complete in recent years. We used only data through 2019 because in 2020 there appeared to be a discrete—but significant—increase in the use of prior authorization by the largest MA insurer, UnitedHealthcare. Through discussions with UnitedHealthcare, we learned that this increase in reported prior authorization represented a change in how they reported to CMS’ Health Plan Management System, which was prompted by a review of CMS’ direction on how to report prior authorization rather than any change in their underlying processes or practices.

This analysis also included demographic and provider market characteristics at the area level. Demographic characteristics were obtained from the IPUMS National Historical Geographic Information System and included county-level share of the population identifying as Black, Indigenous, and People of Color (BIPOC) and the share of the population 65 years and older living below the federal poverty line. To classify the rurality of each county, we used the most recent version available at the time (2013) of the Rural-Urban Continuum Codes from the US Department of Agriculture Economic Research Service. Finally, we relied on American Hospital Association Annual Survey data to construct a Herfindahl-Hirschman Index measuring hospital market concentration at the core-based statistical area level (a geographic unit larger than county, which likely better approximates the size of a hospital market), with number of beds as the market share variable.

Analysis

For each service category, we calculated the annual share of MA enrollees in an MA plan requiring prior authorization and plotted trends over time. To formally test for changes over time, we regressed the share of MA enrollees exposed to prior authorization on year. We present results overall and by service category, focusing on the 15 categories that were present in every year of our study period (2009-2019). We do not present the share of enrollees exposed to prior authorization for supplemental benefits categories because these numbers would be more difficult to interpret and the denominator would change by service category.

To visualize changes over time and understand geographic variation in prior authorization requirements, we plotted the county-level share of MA enrollees exposed to prior authorization in 2009 vs 2019. Because plotting this by service category would generate 30 plots, we focused on 4 sentinel service categories (see eAppendix for rationale): inpatient acute hospital services, psychiatric services, diagnostic procedures/labs/tests, and Part B drugs. We include a tabular presentation of this material for all service categories in the eAppendix.

We also tested for an association between the share of MA enrollees in a county covered by prior authorization for a particular service and characteristics of that county in the same year, repeating this analysis in 2009 and 2019. Finally, we plotted the share of MA enrollees exposed to prior authorization requirements over time by carrier for the 6 largest carriers: UnitedHealthcare, Humana, CVS (formerly Aetna), Kaiser Permanente, Anthem (formerly Wellpoint and now Elevance Health), and Cigna.

RESULTS

In 2019, 72.6% of MA enrollees were in a plan requiring prior authorization for at least 1 category of health care services compared with 71.3% in 2009 (Figure 1). Service categories most frequently requiring prior authorization included durable medical equipment (72.2% of MA enrollees in 2019; 62.1% in 2009), Medicare Part B drugs (72.0% in 2019; 60.9% in 2009), and skilled nursing facility care (71.9% in 2019; 60.7% in 2009). Conversely, nondiscretionary service categories least frequently requiring prior authorization included physician specialist services (30.7% of MA enrollees in 2019; 29.1% in 2009), dialysis services (45.0% in 2019; 36.7% in 2009), and diabetic supplies/services (60.3% in 2019; 42.3% in 2009). Service categories that experienced the greatest increase from 2009 to 2019 in the share of MA enrollees exposed to prior authorization requirements included diagnostic procedures, labs, and tests (63.7% of MA enrollees in 2019; 45.7% in 2009); diabetic supplies/services (60.3% in 2019; 42.3% in 2009); and psychiatric services (57.0% in 2019; 42.7% in 2009). eAppendix Table 2 formally tests for an annual time trend from 2009 to 2019, showing a statistically significant (P < .05) increase in the use of prior authorization for 12 of 15 services.

Figure 2 displays significant geographic variation in the share of MA enrollees exposed to prior authorization requirements. In 2009, of the nearly 3000 counties in the US and Puerto Rico with available information on MA prior authorization requirements, 234 had greater than 90% exposure to prior authorization requirements for inpatient services (ie, > 90% of MA enrollees were in a plan requiring prior authorization for inpatient care), whereas 1062 had less than 10% exposure. By 2019, 637 had greater than 90% exposure to prior authorization requirements for inpatient services, and 226 had less than 10% exposure. Psychiatric services, diagnostic procedures/labs/tests, and Part B drugs showed similar increases. eAppendix Table 3 presents similar information for all remaining service categories.

Several county-level characteristics were associated with the share of MA enrollees exposed to prior authorization requirements in that county, but the strength of these associations attenuated over time. The Table presents raw associations and associations standardized to the change in going from the 25th to 75th percentile county (ie, the interquartile difference) for each independent variable. In 2009, an interquartile increase in the percentage of a county’s residents identifying as BIPOC was associated with an increase of 5.59 to 10.82 percentage points in the share of MA enrollees whose plan required prior authorization, depending on the service category. The strength of this association shrank by 2019, with a similar interquartile increase associated with a change of –4.25 to 1.18 percentage points in the share of MA enrollees in that county exposed to prior authorization requirements. Depending on the year and the service category, the percentage of older adults in poverty was either positively or statistically insignificantly associated with the share of MA enrollees in that county who were exposed to prior authorization requirements. Rurality was strongly negatively associated with the county-level share of MA enrollees exposed to prior authorization requirements in 2009, with much weaker associations observed in 2019. Hospital market concentration (quantified at the core-based statistical area level rather than county to better approximate the large size of hospital markets) was negatively associated with prior authorization requirements in 2009, with an interquartile increase in county-level hospital Herfindahl-Hirschman Index associated with a change of –27.35 to –23.33 percentage points in the share of MA enrollees in that county exposed to prior authorization requirements. The strength of this association shrank by 2019, with a similar interquartile increase associated with a change of –5.51 to 2.64 percentage points in the share of MA enrollees in that county exposed to prior authorization requirements.

Examining prior authorization requirements by carrier and service type revealed different patterns (Figure 3). Across all 4 sentinel services, prior authorization requirements affected fewer than 20% of UnitedHealthcare (the largest MA insurer) enrollees in every year of our study period. Conversely, Humana, Anthem, CVS (Aetna), and Cigna increased prior authorization requirements during the early and middle years of the study period, such that nearly all MA enrollees were exposed to prior authorization requirements for all 4 sentinel services by 2019. Kaiser Permanente maintained inpatient acute hospital services and Part B drug prior authorization requirements for nearly all MA enrollees but decreased its prior authorization requirements for psychiatric services and diagnostic procedures/labs/tests. Carrier-specific analyses revealed considerable differences across and within plans over time. Although many differences likely reflect meaningful changes in benefit structure, they may also reflect changes in reporting accuracy. eAppendix Table 4 compares information available on prior authorization from the CMS MA benefit files with information available in consumer-facing documents published by carriers. Although the majority of carrier-service-year information matches across both sources of information, there are some examples of discordance.

DISCUSSION

As policy attention to prior authorization practices increases in MA, it is important to document trends and patterns in the use of prior authorization. We found that most MA enrollees were already in plans that used prior authorization in 2009, and the proportion with any prior authorization was essentially unchanged by 2019. However, there was considerable growth in the use of prior authorization across a breadth of service types along with variation in use of prior authorization across service types. For instance, in 2009, fewer than half of MA enrollees were exposed to prior authorization for diagnostic procedures, labs, and tests; psychiatric services; and diabetic supplies and services, but these 3 service categories saw the greatest growth in exposure to prior authorization by 2019. In contrast, the use of prior authorization for physician specialist services is notable both for its relatively low level of usage and for its lack of growth from 2009 to 2019.

Across the 4 sentinel services that we examined in detail, the association between prior authorization requirements and area-level characteristics varied over time. In 2009, exposure to prior authorization for all 4 measures was significantly higher in counties that had a greater share of BIPOC residents. By 2019, however, there was little association between a county’s proportion of BIPOC residents and exposure to prior authorization. In other recent survey research focusing on the population younger than 65 years, there was no significant correlation between racial and ethnic group and the likelihood of reporting any administrative health care or insurance tasks, which included prior authorization.¹⁷ Nevertheless, even if there is greater racial/ethnic equity in exposure to prior authorization in MA plans, it does not automatically indicate that prior authorization practices are implemented in equitable ways across racial and ethnic groups. The relationship between county-level poverty and exposure to prior authorization differed across the 4 sentinel services. In 2009, counties with greater poverty rates had higher levels of prior authorization for inpatient hospital services and for Part B drugs than areas with less poverty, whereas there were no statistically significant associations with prior authorization for psychiatric services and diagnostics. By 2019, the relationship between county-level poverty and prior authorization was insignificant for inpatient hospital services, diagnostics, and Part B drugs, whereas it became positive for psychiatric services but with a modest magnitude. However, all 4 sentinel services had much lower levels of prior authorization in 2009 in rural areas compared with urban areas and in more-concentrated hospital markets, with these associations becoming much weaker by 2019. In fact, by 2019, more-rural areas were more likely to have prior authorization for psychiatric services and for diagnostic procedures, labs, and tests. Similarly, the strong negative association between hospital market concentration and exposure to prior authorization in 2009 had attenuated (or even reversed, in the case of diagnostic procedures, labs, and tests) by 2019.

A notable result of our analysis is that levels and trends in the use of prior authorization vary substantially across MA insurers. Overall trends in prior authorization in MA plans may appear relatively smooth but reflect a series of discrete changes in prior authorization practices by specific insurers. As different insurers operate in different geographic regions, the differences in prior authorization use by insurers likely explain a meaningful share of the geographic variation in prior authorization use described in Figure 2. Another more concerning implication of the variation in prior authorization requirements across MA insurers (and within insurers, over time) is the accuracy and/or consistency of the data reported by insurers to CMS. The inconsistencies we note in data reported to CMS and information available online suggest room for improvement in the only known source of nationwide data on this policy-relevant topic.

Limitations

Our analyses have several limitations. First, we are able to assess only whether plans report the use of any prior authorization within service categories. We are unable to speak to the frequency with which prior authorization is used or for which specific services within a category among plans that report using it at all. Understanding variation in the frequency of use of prior authorization is a key priority for future research. Second, our analyses are entirely descriptive and do not represent causal relationships. Third, our analyses focus on the use of prior authorization of MA services that are covered by Part A and Part B, but our analyses do not include Part D prescription drug benefits, in which prior authorization is common. Fourth, as noted previously, our analyses only go through 2019 because of a significant change in 2020 in how information was reported from the largest MA insurer, so we cannot speak to the most recent data on prior authorization. We are also not making normative claims about whether levels and changes in prior authorization are, on net, good or bad. Despite well-known concerns about prior authorization practices, MA beneficiaries are choosing plans with these features and might be willing to trade off exposure to prior authorization for some combination of lower premiums and coverage of additional services.

Several facets of prior authorization in MA warrant additional research attention. As federal policy makers consider and implement regulations on the use of prior authorization, such as mandated maximum turnaround times and electronic prior authorization, it will be important to assess the effects on the overall use of prior authorization in MA and on access to care. Research is also needed to evaluate the extent to which the burdens of prior authorization are being distributed equitably across MA enrollees, as prior research on the population younger than 65 years suggests that administrative burdens in health care are inequitably distributed across racial and ethnic groups. Finally, a significant open question is to what extent MA plans use prior authorization in sufficiently well-targeted ways to optimize clinical value of services vs using prior authorization more indiscriminately in a way that introduces “sludge” into health care delivery and consequently reduces service use regardless of clinical value. These future research questions may be helpfully informed by better availability of more-detailed prior authorization data in coming years, in accordance with CMS’ proposed rules for MA insurers.¹⁸

CONCLUSIONS

From 2009 to 2019, the share of MA enrollees in a plan requiring prior authorization for any service remained stable. However, most service categories saw an increase in prior authorization over this time period. Several area-level demographic and provider market characteristics were associated with prior authorization requirements, but these associations weakened over time. Carrier-level analyses showed wide variation in the use of prior authorization across plans.

Author Affiliations: Division of Health Policy and Management, School of Public Health, University of Minnesota (HTN, JFM, EG), Minneapolis, MN.

Source of Funding: None.

Author Disclosures: The authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (HTN, EG); acquisition of data (HTN, JFM, EG); analysis and interpretation of data (HTN, JFM, EG); drafting of the manuscript (HTN, JFM, EG); critical revision of the manuscript for important intellectual content (HTN, JFM, EG); statistical analysis (HTN, JFM, EG); and supervision (HTN).

Address Correspondence to: Hannah T. Neprash, PhD, Division of Health Policy and Management, School of Public Health, University of Minnesota, 420 Delaware St SE, MMC 729, Minneapolis, MN 55455. Email: hneprash@umn.edu.

REFERENCES

1. Biniek JF, Freed M, Damico A, Neuman T. Half of all eligible Medicare beneficiaries are now enrolled in private Medicare Advantage plans. KFF. May 1, 2023. Accessed June 5, 2023. https://www.kff.org/policy-watch/half-of-all-eligible-medicare-beneficiaries-are-now-enrolled-in-private-medicare-advantage-plans/

2. Anderson KE, Darden M, Jain A. Improving prior authorization in Medicare Advantage. JAMA. 2022;328(15):1497-1498. doi:10.1001/jama.2022.17732

3. Prior authorization: selectively used & evidence-based: results of industry surveys. America’s Health Insurance Plans. November 2022. Accessed June 5, 2023. https://www.ahip.org/documents/Prior_Authorization_Survey_Infographic.pdf

4. 2022 AMA prior authorization (PA) physician survey. American Medical Association. 2023. Accessed June 5, 2023. https://www.ama-assn.org/system/files/prior-authorization-survey.pdf

5. Abelson R. Medicare Advantage plans often deny needed care, federal report finds. New York Times. April 28, 2022. Updated December 3, 2022. Accessed June 5, 2023. https://www.nytimes.com/2022/04/28/health/medicare-advantage-plans-report.html

6. Grimm CA. Some Medicare Advantage Organization Denials of Prior Authorization Requests Raise Concerns About Beneficiary Access to Medically Necessary Care.HHS Office of Inspector General report in brief OEI-09-18-00260; April 2022. Accessed June 5, 2023. https://oig.hhs.gov/oei/reports/OEI-09-18-00260.pdf

7. Becker C. Putting patients first: prescription drug policy in 2022. National Conference of State Legislatures. November 29, 2022. Accessed June 5, 2023. https://www.ncsl.org/state-legislatures-news/details/putting-patients-first-prescription-drug-policy-in-2022

8. Cass A. Texas physician ‘gold card’ rule takes effect Oct. 1. Becker’s Hospital Review. September 20, 2022. Accessed June 5, 2023. https://www.beckershospitalreview.com/finance/texas-physician-gold-card-rules-take-effect-oct-1.html

9. Improving Seniors’ Timely Access to Care Act, HR 3173, 117th Cong (2022). Accessed June 5, 2023.
https://www.congress.gov/bill/117th-congress/house-bill/3173

10. Schwartz AL, Brennan TA, Verbrugge DJ, Newhouse JP. Measuring the scope of prior authorization policies: applying private insurer rules to Medicare Part B. JAMA Health Forum. 2021;2(5):e210859. doi:10.1001/jamahealthforum.2021.0859

11. Anderson KE, Alexander GC, Ma C, Dy SM, Sen AP. Medicare Advantage coverage restrictions for the costliest physician-administered drugs. Am J Manag Care. 2022;28(7):e255-e262. doi:10.37765/ajmc.2022.89184

12. LaPensee KT. Analysis of a prescription drug prior authorization program in a Medicaid health maintenance organization. J Manag Care Pharm. 2003;9(1):36-44. doi:10.18553/jmcp.2003.9.1.36

13. Brot-Goldberg ZC, Burn S, Layton TJ, Vabson B. Rational medicine through bureaucracy: authorization restrictions in Medicare. National Bureau of Economic Research working paper 30878. January 2023. Accessed June 5, 2023. https://www.nber.org/system/files/working_papers/w30878/w30878.pdf

14. Benefits data. CMS. Updated September 6, 2023. Accessed June 5, 2023. https://www.cms.gov/data-research/statistics-trends-and-reports/medicare-advantagepart-d-contract-and-enrollment-data/benefits-data

15. Health Plan Management System (HPMS). CMS. Updated September 6, 2023. Accessed June 5, 2023. https://www.cms.gov/about-cms/information-systems/hpms

16. Freed M, Biniek JF, Damico A, Neuman T. Medicare Advantage in 2022: premiums, out-of-pocket limits, cost sharing, supplemental benefits, prior authorization, and star ratings.Kaiser Family Foundation. August 25, 2022. Accessed June 5, 2023. https://web.archive.org/web/20221230155208/https://www.kff.org/medicare/issue-brief/medicare-advantage-in-2022-premiums-out-of-pocket-limits-cost-sharing-supplemental-benefits-prior-authorization-and-star-ratings/

17. Kyle MA, Frakt AB. Patient administrative burden in the US health care system. Health Serv Res. 2021;56(5):755-765. doi:10.1111/1475-6773.13861

18. Medicare and Medicaid programs; Patient Protection and Affordable Care Act; advancing interoperability and improving prior authorization processes for Medicare Advantage organizations, Medicaid managed care plans, state Medicaid agencies, Children’s Health Insurance Program (CHIP) agencies and CHIP managed care entities, issuers of qualified health plans on the federally-facilitated exchanges, Merit-based Incentive Payment System (MIPS) eligible clinicians, and eligible hospitals and critical access hospitals in the Medicare Promoting Interoperability Program. Fed Regist. 2022;87(238):76238-76371.