Skip to main content

Advanced Search

Advanced Search

Current Filters

Filter your query

Publication Types



Issue Briefs


Are Medicaid Patients Seen in Office-Based Practices Getting High-Quality Primary Care?

  • Patients receive the same quality of care whether they visit a primary care practice that sees mostly Medicaid patients or a practice that sees mostly patients with other types of coverage, national data show

  • Office-based practices that primarily serve Medicaid patients are less likely to have an electronic health record, which could affect their ability to deliver high-quality care in the future

  • Patients receive the same quality of care whether they visit a primary care practice that sees mostly Medicaid patients or a practice that sees mostly patients with other types of coverage, national data show

  • Office-based practices that primarily serve Medicaid patients are less likely to have an electronic health record, which could affect their ability to deliver high-quality care in the future


  • Issue: About three-quarters of nonelderly adult Medicaid patients get their primary care from independent office-based practices, rather than hospital outpatient departments or health centers. But many independent physicians don’t accept Medicaid, in part because of its low payment rates. Medicaid-covered care is concentrated in a small share of independent, typically underresourced practices, and this potentially has implications for the quality of care delivered. The Affordable Care Act’s Medicaid eligibility expansion and physician fee bump, however, may have changed the distribution or care of patients across independent practices.
  • Goals: Examine the distribution of Medicaid patients across office-based primary care physicians and the impact of a practice’s Medicaid patient share on care quality.
  • Methods: Analysis of National Ambulatory Medical Care Survey findings for 2006–2016 and 2018–2019.
  • Key Findings and Conclusions: Medicaid visits have remained concentrated in a minority of practices since Medicaid expansion. From 2014 to 2019, about one-third of sampled office-based primary care physicians accounted for 90 percent of Medicaid office visits. Patients are as likely to receive guideline-recommended care at Medicaid-dominated practices as they are at practices with more non-Medicaid patients, but the former are much less likely than the latter to have adopted electronic health records. This suggests that low payment rates may affect practices’ ability to invest in technology that promotes high-quality care.


Independent office-based physician practices play an important role in providing primary care for Medicaid patients. In 2011, Medicaid patients between 19 and 64 years of age made 47 million visits to either an office-based or hospital outpatient primary care provider (PCP). Of these visits, about 75 percent took place at an independent office-based practice (Appendix 1). This percentage is somewhat greater than estimates from the late 1990s.1 Medicaid’s low payment rates for office-based visits have long led to concerns about the concentration of Medicaid patients in these practices, particularly regarding the quality of care they provide.2

In 2014, the Affordable Care Act (ACA) extended Medicaid eligibility to adults younger than age 65 with incomes up to 138 percent of the federal poverty level. As a result, some 11 million individuals who had previously been uninsured gained Medicaid coverage by the end of that year.3 But while coverage expanded, questions remained about whether newly insured Medicaid beneficiaries would be able to access the care they need. That’s because Medicaid physician payment rates have historically been well below those of Medicare or private insurance rates.4 This fee discrepancy has contributed to many physicians’ reluctance to accept new Medicaid patients, which has left them clustered in a subset of practices.5

Partially in response to the concern over access, the ACA included a Medicaid fee bump that raised payment rates for Medicaid primary care visits to Medicare levels in 2013 and 2014. In 17 states, these higher fees continued past 2014. However, research suggests that the fee bump had only modest effects on access to care and physician participation in Medicaid.6

Research conducted since the passage of the ACA has confirmed that newly insured Medicaid patients have indeed been able to access care.7 On the supply side, surveys have shown that although the proportion of primary care physicians willing to accept new Medicaid patients remained lower than the share of physicians willing to accept new Medicare or private patients, it increased from 67 percent in the 2011–2012 period to 75 percent in the 2014–2017 period.8 While these results are reassuring, it is not clear whether this increased willingness to accept new Medicaid patients has changed the distribution of these patients among physician practices.

The distribution of patients among practices is a concern because Medicaid’s lower payment rates disadvantage those practices that participate in the program. Lower payment rates might lead practices to reduce the resources they devote to individual visits (for example, by seeing patients for very brief visits or providing poor-quality care) or diminish their ability to invest in fixed assets such as electronic health records (EHRs).9 As most evidence suggests that physician practice patterns do not vary by individual patients’ insurance status,10 any effect of lower Medicaid payment rates on quality is likely to be a function of the proportion of a practice’s patients who are covered by Medicaid, rather than by the insurance status of individual patients.

For this brief, we analyzed 2006–2019 office-based visit data from the National Ambulatory Medical Care Survey to examine 1) recent changes in the distribution of Medicaid visits across primary care providers over time and 2) the relationship between the insurance composition of a PCP and measures of quality of care and technology adoption (for more details, see “How We Conducted This Study”). While previous studies have examined the relationship between hospital payer distribution and quality of care, there has been little prior analysis of this dimension of primary care.

Key Findings

Distribution of Medicaid Visits Across Physician Practices

We examined the payment-type composition of a physician practice at different percentile points along the Medicaid share distribution (Exhibit 1). For example, in a practice at the median of the Medicaid share distribution, 4 percent of patients in a practice panel are covered by Medicaid, 66 percent by private insurance, and 22 percent by Medicare. At the 99th percentile of the Medicaid share distribution, 81 percent of patients in a practice are covered by Medicaid, 13 percent by private insurance, and 6 percent by Medicare. Between the periods 2006–2013 and 2014–2019, the Medicaid share within practices increased, while the share of visits that were uninsured (self-pay or other) declined.


One-third (34%) of PCPs accounted for 90 percent of Medicaid visits (Exhibit 2). Nearly half (49%) of PCPs reported not seeing any Medicaid patients, but only 10 percent of physicians did not see any privately insured patients.


We also found that slightly more PCPs saw Medicaid patients in the 2014–2019 period (data not shown). While most of the additional visits made by Medicaid patients after Medicaid expansion were in practices that were already seeing many Medicaid patients, some visits occurred at practices where coverage was more mixed. These results are consistent with previous studies.11

Insurance Type and Quality of Care

To assess quality of care in Medicaid-dominated practices, we reviewed whether patients received medical treatments, counseling services, and preventive screenings according to written guidelines. On average, we found that there was no significant or systematic difference in estimates of mean, or average, primary care quality between a visit to a PCP who saw only Medicaid patients and one that saw no Medicaid patients, either before or after Medicaid expansion (Exhibit 3). This finding of comparable quality across practices, regardless of dominant payer type, is consistent with other recent evidence.12 We did observe statistically significant differences for a few specific medical treatment guidelines, but the findings were inconsistent across measures (data not shown).


In the post-ACA period, practices with a higher concentration of Medicaid patients were less likely to provide preventive screening services for female patients, such as breast exams, pelvic exams, and mammograms (data not shown). However, these differences disappeared when restricting the sample to Medicaid patients only. Insurance composition did not affect visit duration in our study, unlike in earlier research that showed Medicaid-dominated practices with significantly shorter visit duration.13

Insurance Type and EHR Adoption

We found that the Medicaid share of a practice’s patient panel was strongly negatively associated with the likelihood of EHR adoption. Between the 2006–2013 and 2014–2019 periods, the proportion of all practices that fully adopted an EHR system (all health records stored electronically) increased from 42 percent to 81 percent. The disparity in adoption by Medicaid share, however, widened over time. In the 2006–2013 period, practices that saw Medicaid patients only were 30 percentage points less likely to fully transition to an EHR system (p<0.001) compared with Medicaid-zero practices (Exhibit 4). By 2014–2019, this disparity had increased to 40 points (p<0.001). It is noticeable that nearly 80 percent of practices with a high concentration of Medicaid patients had begun adopting EHRs post-Medicaid expansion, but fewer than two-thirds of them had transitioned to a fully electronic system.



The distribution of Medicaid patients among independent office-based primary care practices remains highly concentrated: about one-third of PCPs account for 90 percent of Medicaid visits, consistent with other recent findings.14

While prior research has shown that more doctors were willing to accept Medicaid in the post-ACA period, we found that this pattern of concentration has declined only modestly. Research on the Medicaid fee bump has similarly shown limited improvements in physician availability.15 Beneficiaries of the Medicaid expansion did, however, increase their use of care. Taken together, these findings suggest that the increased use observed through the expansion occurred primarily among PCPs that had already focused on treating Medicaid patients.

These Medicaid-focused PCPs generally receive lower fees than PCPs that see mostly privately insured and Medicare patients. In our study, the effect of low fees was evident in PCPs’ ability to invest in electronic health records. Practices that treated a large share of Medicaid patients were much less likely to have EHR systems than their counterparts serving primarily other patients. These results may be particular to independent office-based practices and to Medicaid-dominated practices. Prior studies have found that federally qualified health centers (FQHCs), which predominantly treat Medicaid patients, had relatively high EHR adoption rates.16 A recent study that also included clinics and health centers found that practices with no Medicaid patients had fewer health information exchange capabilities than practices for which Medicaid constituted 10 percent or more of their revenue.17

Slow EHR adoption may have negative consequences for quality of care. While the literature on EHRs’ impact on the quality of primary care is sparse, some evidence suggests that these systems do improve quality, including in practices serving low-income populations.18 Moreover, because the beneficial effects of EHRs on quality may accumulate with experience and over time,19 the lack of technology may create disparities in the future. To help address such inequities, Medicaid offered incentives for EHR adoption through 2021 (our data ended during the period these incentives were still in effect). If the differences in adoption continue to exist, policymakers might consider adding new incentives to facilitate implementation of this technology in practices that serve mainly Medicaid patients.

Policymakers also might reexamine current penalties against practices that fail to adopt EHRs. Under Medicare’s physician payment policy (established by the Medicare Access and CHIP Reauthorization Act of 2015, or MACRA), some practices that delay their transition to EHRs may face Medicare reductions in payment and total incentives. Although there are no EHR-related penalties associated with Medicaid payments, practices that see predominantly Medicaid patients also see a significant number of Medicare patients. Thus, Medicare penalties could further disadvantage lower-paid providers. It will be important to assess whether penalties under MACRA’s Merit-Based Incentive Payment System differentially affect Medicaid-serving providers and consider adjustments to the penalty system, or offsetting incentives, to avoid these effects.


Data and Study Population

We analyzed data on all primary care visits to office-based practitioners from 2006 to 2019 in the National Ambulatory Medical Care Survey (NAMCS), an annual, nationally representative survey of randomly sampled office-based physician practices (excluding 2017 NAMCS, which is currently unavailable). Primary care physicians (PCPs) are defined following the definition of variable SPECCAT in the survey — that is, including physicians specialized in general/family practice, internal medicine, and obstetrics and gynecology. Insurance information is recorded by physicians based on the payment source of the visit. For visits covered by more than one type of payment, we assign a primary insurance type to each visit based on the following hierarchy: Medicaid, Medicare, private insurance, self-pay, and “other” (including workers’ compensation, no charge, other payment, unknown, and blank). For example, if a visit was paid with both Medicaid and Medicare, it was coded as a Medicaid visit. We did not include visits made to community health centers, as they were not sampled after 2015. More information, including data collection procedures, is available at the National Center for Health Statistics.20

We excluded visits made by patients younger than age 19 (to avoid including pediatricians) and collapsed visit-level data by physician and year. Then we restrict our sample to physicians with five or more adult visits in a year to examine the visit distribution across PCPs and the impact of a PCP practice’s Medicaid share on electronic health record (EHR) adoption. Our analytic sample includes 5,017 unique PCPs. In analyses of quality measures, we excluded visits made by patients younger than age 19 or older than age 64, for prenatal examination, or with a duration of zero or more than one hour, yielding a total of 74,450 visits.

Statistical Analysis

We first examined the insurance type composition within a practice at different percentile points of the practice’s Medicaid share distribution. The results of a sensitivity analysis, which prioritizes Medicare over Medicaid when coding a visit that was paid by both types of insurance, are consistent with our primary findings.

We then measured the extent of practice composition differences. To do this, we examined the allocation of patient visits across PCPs. We described the cumulative distribution of visits paid by private insurance, Medicaid, and Medicare across PCPs, ordered by the proportion of their visits covered by Medicaid.

We calculated the insurance composition of the practice panel by computing the percentage of visits by each insurance group to each unique PCP. For example, at a PCP that contributed 10 observations — five visits paid by private insurance, three by Medicare, and two by Medicaid — visits paid by private insurance would yield a value of 5/10 (or 50%), visits paid by Medicare would yield a value of 3/10 (or 30%), and visits paid by Medicaid would yield a value of 2/10 (or 20%).

We then examined the correlation between the proportion of visits paid by Medicaid and the quality of care provided to patients by a PCP and the structure of care within the practice, including the adoption of electronic health record (EHR) systems. We measured primary care quality using the methods in Zhu and Glied21 (see Appendix 2).

We conducted regressions using visits-level data (N=74,450), clustering on the physician-year level, on each of the quality measurements and visit duration, and control for patient characteristics (age, sex, race, gender, major reason for visiting, and insurance status), visit characteristics (year), and practice characteristics (rural status, insurance composition, and physician employment status). We conducted regressions using physician-year-level data (N=5,017) to estimate the effect of Medicaid share on the EHR adoption rate and control for practice characteristics (rural status, office type, patient panel composition by insurance, race, age, and physician employment status). We report the estimated adjusted means of Medicaid-zero and Medicaid-only PCPs on aggregated measurements of medical management services, counseling services, preventive screening services, and EHR adoption rates. Coefficient estimates for specific treatments and care services are available upon request.

  1. Christopher B. Forrest and Ellen-Marie Whelan, “Primary Care Safety-Net Delivery Sites in the United States: A Comparison of Community Health Centers, Hospital Outpatient Departments, and Physicians’ Offices,” JAMA 284, no. 16 (Oct. 25, 2000): 2077–83.
  2. Janet B. Mitchell and Jerry Cromwell, “Large Medicaid Practices and Medicaid Mills,” JAMA 244, no. 21 (Nov. 28, 1980): 2433–37.
  3. Center for Medicare and Medicaid Services, Medicaid & CHIP: December 2014 Monthly Applications, Eligibility Determinations and Enrollment Report (CMS, Feb. 23, 2015).
  4. Stephen Norton and Stephen Zuckerman, “Trends in Medicaid Physician Fees, 1993–1998,” Health Affairs 19, no. 4 (July/Aug. 2000): 222–32; Stephen Zuckerman, Aimee F. Williams, and Karen E. Stockley, “Trends in Medicaid Physician Fees, 2003–2008,” Health Affairs 28, suppl. 1 (2009): w510–w519; and Stephen Zuckerman and Dana Goin, How Much Will Medicaid Physician Fees for Primary Care Rise in 2013? Evidence from a 2012 Survey of Medicaid Physician Fees (Henry J. Kaiser Family Foundation, Dec. 2012).
  5. Kayla Holgash and Martha Heberlein, “Physician Acceptance of New Medicaid Patients: New Findings,” presentation, Medicaid and CHIP Payment and Access Commission Public Meeting, Jan. 24, 2019; and Medicaid and CHIP Payment and Access Commission, Physician Acceptance of New Medicaid Patients: Findings from the National Electronic Health Records Survey (MACPAC, June 2021).
  6. Stephen Zuckerman, Laura Skopec, and Marni Epstein, Medicaid Physician Fees After the ACA Primary Care Fee Bump (Urban Institute, Mar. 2017).
  7. Madeline Guth, Rachel Garfield, and Robin Rudowitz, The Effects of Medicaid Expansion Under the ACA: Updated Findings from a Literature Review (Henry J. Kaiser Family Foundation, Mar. 2020).
  8. Sandra L. Decker, “Two-Thirds of Primary Care Physicians Accepted New Medicaid Patients in 2011–12: A Baseline to Measure Future Acceptance Rates,” Health Affairs 32, no. 7 (July 2013): 1183–87; and Robert Hest and Martha Heberlein, “Physician Acceptance of New Medicaid Patients: National and State-Level Findings from the National Electronic Health Records Survey,” presentation, 2021 AcademyHealth Annual Research Meeting: State Health Research and Policy Interest Group, June 2021.
  9. Clemens S. Kruse et al., “Barriers to Electronic Health Record Adoption: A Systematic Literature Review,” Journal of Medical Systems 40, no. 12 (Dec. 2016): 252; and Cindy Mann and Adam Striar, “How Differences in Medicaid, Medicare, and Commercial Health Insurance Payment Rates Impact Access, Health Equity, and Cost,” To the Point (blog), Commonwealth Fund, Aug. 17, 2022.
  10. Sherry Glied and Joshua Zivin, How Do Doctors Behave When Some (But Not All) of Their Patients Are in Managed Care? (National Bureau of Economic Research, Sept. 2000); and Bruce E. Landon, “Tipping the Scale. The Norms Hypothesis and Physician Behavior.,” New England Journal of Medicine 376, no. 9 (Mar. 2, 2017): 810–11.
  11. Liz Hamel et al., Experiences and Attitudes of Primary Care Providers Under the First Year of ACA Coverage Expansion (Commonwealth Fund, June 2015).
  12. Michael L. Barnett et al., “Low-Value Medical Services in the Safety-Net Population,” JAMA Internal Medicine 177, no. 6 (June 1, 2017): 829–37; and Hamel et al., Experiences and Attitudes, 2015.
  13. Glied and Zivin, How Do Doctors Behave?, 2000.
  14. Avital B. Ludomirsky et al., “In Medicaid Managed Care Networks, Care Is Highly Concentrated Among a Small Percentage of Physicians,” Health Affairs 41, no. 5 (May 2022): 760–68.
  15. Daniel Polsky et al., “Appointment Availability After Increases in Medicaid Payments for Primary Care,” New England Journal of Medicine 372, no. 6 (Feb. 5, 2015): 537–45; Sandra L. Decker, “No Association Found Between the Medicaid Primary Care Fee Bump and Physician-Reported Participation in Medicaid,” Health Affairs 37, no. 7 (July 2018): 1092–98; and Loren Saulsberry, Veri Seo, and Vicki Fung, “The Impact of Changes in Medicaid Provider Fees on Provider Participation and Enrollees’ Care: A Systematic Literature Review,” Journal of General Internal Medicine 34, no. 10 (Oct. 2019): 2200–9.
  16. Jamie Ryan et al., The Adoption and Use of Health Information Technology by Community Health Centers, 2009–2013 (Commonwealth Fund, May 2014); Kelly L. Myrick, Damon F. Ogburn, and Brian W. Ward, “Table 7. Percentage of Office-Based Physicians Using any Electronic Health Record (EHR)/Electronic Medical Record (EMR) System and Physicians That Have a certified EHR/EMR System, by Selected Characteristics: National Electronic Health Records Survey, 2017,” National Center for Health Statistics, July 2019.
  17. Steven B. Spivack et al., “Avoiding Medicaid: Characteristics of Primary Care Practices with No Medicaid Revenue,” Health Affairs 40, no. 1 (Jan. 2021): 98–104.
  18. Melinda B. Buntin et al., “The Benefits of Health Information Technology: A Review of the Recent Literature Shows Predominantly Positive Results,” Health Affairs 30, no. 3 (Mar. 2011): 464–71; and Robert Baillieu et al., “Impact of Health Information Technology Optimization on Clinical Quality Performance in Health Centers: A National Cross-Sectional Study,” PLoS One 15, no. 7 (July 15, 2020): 1–11.
  19. David Blumenthal, “The Electronic Health Record Problem,” To the Point (blog), Commonwealth Fund, Dec. 13, 2018.
  20. National Center for Health Statistics, “Ambulatory Health Care Data,” Centers for Disease Control and Prevention.
  21. Benjamin Zhu and Sherry A. Glied, More Is More: Expanding Access to Primary Care While Maintaining Quality (Commonwealth Fund, July 2021).

Publication Details



Sherry A. Glied, Dean, Robert F. Wagner Graduate School of Public Service, New York University

[email protected]


Dong Ding and Sherry A. Glied, Are Medicaid Patients Seen in Office-Based Practices Getting High-Quality Primary Care? (Commonwealth Fund, Jan. 2023).