Access Explorer Methodology

← Back to Access Explorer

Overview

The Medi-Cal Provider Access Explorer identifies phantom network gaps by cross-referencing provider registration data with Medicaid billing activity. A phantom provider is one who appears in provider directories but does not actively see Medi-Cal patients.

Data Sources

National Plan and Provider Enumeration System (NPPES)

NPPES is maintained by CMS and contains National Provider Identifier (NPI) records for all healthcare providers in the United States. We extract California providers and classify them by specialty using their primary taxonomy code.

• Source: CMS NPPES NPI Registry (monthly updates)
• Coverage: All healthcare providers with active NPIs in California
• Geographic mapping: Practice location ZIP codes mapped to counties via HUD ZIP-County crosswalk

HHS Medicaid Provider Spending Data

The U.S. Department of Health and Human Services publishes detailed Medicaid provider spending data that identifies providers who actively billed Medicaid in each reporting period.

• Source: HHS Medicaid Provider Spending File
• Coverage: All fee-for-service Medicaid claims in California
• Limitation: Does not include managed care encounter data (see Limitations section)

Definitions

Term	Definition
Registered Provider	A provider with an active NPI in NPPES whose practice location is in the county and whose taxonomy code maps to the specialty category.
Active Provider	A registered provider who billed at least one Medicaid claim in the county during the reporting period (trailing 12 months).
Participation Rate	Active providers divided by registered providers, expressed as a percentage. Higher rates indicate better effective access.
Phantom Gap	The number of registered providers who are not active (Registered minus Active). These providers appear in directories but do not serve Medi-Cal patients.
Participation Index	Monthly participation rate indexed to January 2019 = 100. Values below 100 indicate declining participation relative to the pre-pandemic baseline.
Composite Cost Index	Weighted average of healthcare wages, facility rent, and purchased services costs, each normalized so the state median = 100. Based on Medicare GPCI Practice Expense sub-component weights (56/30/14).
Effective Reimbursement Index	10,000 / composite cost index. Shows the purchasing power of a flat Medi-Cal payment relative to the state average. Higher values indicate greater purchasing power.

Specialty Groupings

Providers are grouped into six categories based on their NPPES primary taxonomy code:

Category	Included Taxonomy Codes
Primary Care	Family Medicine, General Practice, Internal Medicine, Pediatrics, Geriatric Medicine
Behavioral Health	Psychiatry, Psychology, Clinical Social Work, Marriage & Family Therapy, Mental Health Counselor, Substance Abuse Counselor
Dental	General Dentistry, Pediatric Dentistry, Oral Surgery, Orthodontics, Endodontics, Periodontics
OB/GYN	Obstetrics & Gynecology, Maternal-Fetal Medicine, Reproductive Endocrinology, Certified Nurse Midwife
Other Surgical	General Surgery, Orthopedic Surgery, Ophthalmology, Otolaryngology, Urology, Cardiothoracic Surgery
Pharmacy & DME	Pharmacy, Durable Medical Equipment suppliers

Methodology Notes

Trend Index Construction

Monthly participation indices are constructed by calculating the participation rate for each county-specialty-month combination, then indexing to January 2019 = 100. This allows comparison of relative changes across specialties with different baseline participation levels.

State Medians

State median participation rates are calculated across all 58 counties (population-weighted) for each specialty. These serve as benchmarks for identifying counties with below-average access.

Change from 2019

The percentage-point change in participation rate comparing the most recent 12-month period to the January-December 2019 baseline. This captures the net impact of the pandemic, provider attrition, and policy changes on effective network adequacy.

Limitations

1. Managed Care Organization (MCO) data gaps: The HHS spending file covers fee-for-service Medicaid claims. Providers who see Medi-Cal patients exclusively through managed care plans may not appear as "active" in our data. As of 2024, approximately 85% of Medi-Cal enrollees are in managed care, meaning the true active provider count may be higher than reported.
2. Privacy thresholds: Provider counts below 11 in a county-specialty cell are suppressed to protect patient privacy. This primarily affects small rural counties.
3. Geographic assignment: Providers are assigned to counties based on their NPPES practice location ZIP code. Providers practicing in multiple counties are counted only in their primary practice location.
4. HCPCS-based identification: Specialty classification relies on self-reported taxonomy codes in NPPES, which may not always reflect actual practice patterns.
5. Billing threshold: A provider is classified as "active" based on any Medicaid billing in the trailing 12-month period. This is a low bar; some "active" providers may see very few Medi-Cal patients.

Affordability Context: Data Sources and Methodology

Research Question

Medi-Cal reimbursement rates are set statewide with no geographic adjustment. Medicare, by contrast, adjusts physician payments through the Geographic Practice Cost Index (GPCI), mandated by Congress in the Omnibus Budget Reconciliation Act of 1989. The Affordability Context module asks: how do local practice operating costs compare to the state average, and what does this imply for the purchasing power of flat Medi-Cal payments?

Affordability Data Sources

Source	Agency	Geography	Coverage	Role in Index
QCEW Annual Averages, NAICS 62	Bureau of Labor Statistics	County	2017–2022	Employee wage component (56%)
Fair Market Rents, 2-Bedroom	HUD	County	FY2018–FY2023	Facility rent component (30%)
QCEW Annual Averages, All Industries	Bureau of Labor Statistics	County	2017–2022	Purchased services proxy (14%)
Per Capita Personal Income (CAINC1)	Bureau of Economic Analysis	County	2017–2022	Supplementary (demand-side)
Medicare Geographic Variation	CMS	County	2017–2022	Comparison benchmark

Composite Cost Index Construction

The composite cost index is a weighted average of three components, each normalized so the state median equals 100:

Component	Weight	GPCI Analog	Data Source
Healthcare employee wages	56%	PE employee wage sub-index	BLS QCEW, NAICS 62
Facility rent	30%	PE office rent sub-index	HUD FMR, 2-bedroom
Purchased services	14%	PE purchased services sub-index	BLS QCEW, all-industry

Weighting Rationale

Weights are derived from the Medicare Practice Expense (PE) GPCI sub-component cost shares as reported in the 2006-based Medicare Economic Index, which remains current through the CY 2026 Physician Fee Schedule (90 FR 49266). Within the PE GPCI, the geographically-adjusted sub-components have the following cost shares: employee wages (19.15% of practice costs), office rent (10.22%), and purchased services (~5.07%). Equipment and supplies (12.81%) are set to 1.00 nationally because they are purchased in national markets. Normalizing the adjusted components to sum to 100% yields the weights used here: 55.6% → 56%, 29.7% → 30%, 14.7% → 14%.

Per Capita Income

Per capita personal income (BEA CAINC1) is reported as a supplementary variable but is not included in the composite cost index. Income is a demand-side measure (patient economic capacity), while the composite measures supply-side operating costs (what it costs to run a practice). Medicare's GPCI does not include patient income in any of its three components. Including income in the composite would conflate two distinct economic constructs.

Effective Reimbursement Index

The effective reimbursement index = 10,000 / composite_cost_index, showing how much purchasing power a flat Medi-Cal payment has in each county relative to the state average. A county with composite = 120 has an effective reimbursement index of 83.3, meaning a flat $100 payment buys only $83.30 worth of inputs compared to the state average.

Research by Alexander and Schnell (2024) found that closing the Medicaid-private payment gap would reduce more than two-thirds of access disparities for adults. Polsky et al. (2015) estimated a 1.25 percentage point increase in appointment availability per 10% increase in Medicaid reimbursement. These findings suggest that geographic variation in effective reimbursement may contribute to geographic variation in provider participation.

Medicare Comparison

Medicare adjusts physician payments through three GPCIs—work, practice expense, and malpractice—mandated by OBRA 1989 (P.L. 101-239). The Geographic Adjustment Factor (GAF) can increase payments by 20% or more in high-cost areas. The Medicare Geographic Variation Public Use File reports actual versus standardized per capita spending by county, allowing direct computation of the adjustment effect. Medi-Cal applies no analogous geographic adjustment to its physician fee schedule.

Limitations of the Affordability Analysis

1. QCEW non-disclosure: BLS may suppress data for small counties with few healthcare employers. Suppressed counties are flagged but excluded from the composite calculation.
2. Rent proxy: HUD Fair Market Rents measure residential rent as a proxy for commercial office rent. CMS uses the same proxy approach for the PE office rent GPCI sub-index, but actual commercial rents may diverge in some markets.
3. Work component omitted: The composite cost index captures practice operating costs but not physician compensation, which the Medicare GPCI addresses through a separate work component. The work GPCI is statutorily limited to one-quarter of geographic variation.
4. Descriptive, not causal: The tool presents descriptive correlations between cost indices and participation rates, not causal estimates. The payment-participation relationship is established at the state level (Alexander & Schnell 2024; Polsky et al. 2015); the within-state application extends this logic to geographic variation against a uniform fee schedule.
5. Administrative barriers: Administrative barriers may affect participation as much as payment rates. MACPAC (2025) found physicians lose 17.6% of Medicaid visit contractual value to claims denials and resubmissions (vs. 4.7% for Medicare).

Policy Context

CMS finalized the Medicaid Access Rule (CMS-2439-F) with network adequacy provisions effective July 2025. The rule requires states to verify that provider directories reflect actual availability, making tools like this Access Explorer directly relevant to compliance monitoring.

References

1. Zhu, J. M., Charlesworth, C. J., Polsky, D., & McConnell, K. J. (2023). Phantom networks: Discrepancies between reported and realized mental health care access in Oregon Medicaid. Health Affairs, 42(1), 49-57.
2. U.S. Department of Health and Human Services, Office of Inspector General. (2025). Many providers in Medicaid managed care were not available to enrollees (OEI-02-23-00540).
3. Centers for Medicare & Medicaid Services. (2024). Medicaid and CHIP Managed Care Access, Finance, and Quality Final Rule (CMS-2439-F). Federal Register, 89(83).
4. Cantor, J. C., Thompson, F. J., & Farnham, J. (2013). State Medicaid fee-for-service reimbursement rates and access to care. Health Services Research, 48(2pt1), 689-710.
5. Decker, S. L. (2012). In 2011 nearly one-third of physicians said they would not accept new Medicaid patients, but is the problem growing? Health Affairs, 31(8), 1673-1680.
6. Alexander, D. & Schnell, M. (2024). The impacts of physician payments on patient access, use, and health. American Economic Journal: Applied Economics. DOI: 10.1257/app.20210227.
7. Polsky, D., Richards, M., Basseyn, S., et al. (2015). Appointment availability after increases in Medicaid payments for primary care. New England Journal of Medicine, 372(6), 537-545.
8. Centers for Medicare & Medicaid Services. (2025). Calendar Year (CY) 2026 Medicare Physician Fee Schedule Final Rule (CMS-1832-F). Federal Register, 90 FR 49266.
9. Bureau of Labor Statistics. Quarterly Census of Employment and Wages, Annual Averages, NAICS 62 Healthcare and Social Assistance.
10. Bureau of Economic Analysis. Personal Income by County, Metro, and Other Areas (CAINC1 Table).
11. Medicaid and CHIP Payment and Access Commission (MACPAC). (2025). Evaluating the effects of Medicaid payment changes on access to physician services. January 2025.
12. Institute of Medicine. (2012). Geographic Adjustment in Medicare Payment: Phase I — Improving Accuracy. Washington, DC: The National Academies Press.

Update Schedule

The Access Explorer data is updated quarterly as new HHS Medicaid Provider Spending data becomes available. NPPES data is refreshed monthly. The current dataset covers January 2018 through December 2024.

← Back to Access Explorer