How Useful a Measure Is the Provider Participation Rate for - - PowerPoint PPT Presentation

Michelle Kitchman Strollo, Dr.PH

NORC at the University of Chicago

How Useful a Measure Is the Provider Participation Rate for Consumer Health Plan Choice?

• Provider networks nationwide are becoming more narrow, as health plans

compete on premiums

• Five market study suggests that exchange plan networks have 1/3 fewer providers than

employer plans in the same markets. (Avalere)

• A study of hospital participation in 2015 exchange plans concluded that 55 percent of

such plans have either “ultra-narrow,” “narrow,” or “tiered” hospital participation. (McKinsey)

• A study of physician participation in 2015 exchange plans concluded that 41 percent of

qualified health plans have “small” or “x-small” networks. (University of Pennsylvania)

• Consumers have raised concerns about provider exclusions from networks,

which has spurred greater regulatory oversight of networks

Provider Networks Emerging as a Policy Issue

• Federal and state governments are pushing for greater transparency of provider

networks, including fining health plans for outdated or inaccurate network data

• California recently fined two plans ($600,000) for faulty data
• CMS fined national MA plan ($1M) for pharmacy directory errors
• Consumers separately have brought suits in response to inaccurate provider

directory information

• At least five lawsuits are pending against health plans for mis-representing provider

access – one recently settled for $15M

• Since 2016, Qualified Health Plans (QHP) operating on federally-facilitated

Marketplaces are required to post machine-readable provider network directory files (JSON Files)

• Machine-readable files are expected in Medicaid managed care and Medicare Advantage

in the next few years

2017: The Year of Provider Network Oversight?

• QHP Marketplace data has been available since 2016 in JSON machine-

readable format

• Machine-readable ≠ usable
• Machine-readable ≠ clean or complete
• NORC researchers downloaded, aggregated, cleaned, and linked these JSON

files with other QHP and provider files, providing a full picture of FFM provider networks in 2016 and 2017

Challenges of Using Provider Network Data

• For 2017, shoppers on healthcare.gov in four states (Maine, Ohio, Texas, and

Tennessee) can compare plans on provider network size in three specialties (adult primary care, pediatric primary care, and hospital facilities)

• Size is determined using a metric called the Provider Participation Rate (PPR), the proportion of all

providers in a given county in a given specialty that are participating in an issuer’s network

• PPR is a relative, not absolute benchmark. A network with a PPR:
• More than one standard deviation above the mean in the county = broad
• More than one standard deviation below the mean = basic
• Everything in between = standard
• The relative nature of the measure allows for even comparisons within counties

but uneven ones across counties

CMS Methodology for Determining Network Breadth

Average Provider Participation Rate Needed to Achieve a Broad or Basic Network in Adult Primary Care, by County in Selected States, 2017

• What constitutes a

Broad or Basic network differs across and within states.

• In some counties, it is

impossible to achieve a broad or basic network!

Average PPR by Population Density, 2017

27% 48% 80% 62% 91% 96%

Basic Standard Broad

Large Metro Rural Frontier

Note: Parameters for large metro and rural frontier defined by CMS and found at: https://www.cms.gov/Medicare/Medicare- Advantage/MedicareAdvantageApps/Downloads/CY2017_MA_HSD_Network_Criteria_Guidance.PDF

• The nature of the Provider Participation Rate means that what constitutes a

broad or basic network can vary widely across geographies.

• We also find that the distribution of network sizes on the Exchanges varies by issuer

group

• Narrow networks are not necessarily bad for consumers, but data need to be

accurate for consumers to make informed choices

• Narrow network plans on the Marketplaces are cheaper than broad-network plans

(McKinsey), and consumers prefer narrow networks with lower premiums to broader networks with higher premiums (Kaiser Health Tracking Poll)

• Challenge in coming years will be to improve upon metrics of network size (like

the PPR) to better reflect network quality

• Consider integrating measures of provider quality, time and distance standards, cost of

care, or performance on certain health conditions

Long-Term Considerations for PPR