Tamara Beetham, PhD, MPH

Ongoing projects

Findings are preliminary and subject to change.

  • Job market paper

    Primary Care Clinician Supply, Market Structure, and Access in the US

    With Jane M. Zhu (Oregon Health & Science University), Nicholas S.J. King (Brown University), Michael L. Barnett (Brown University)

    We used AI voice agents to call nearly 13,000 primary care clinics in 280 metropolitan areas and ask about the soonest appointment for a new Medicare patient. Wait times varied widely across markets and were better explained by market competition than by clinician supply.

    Abstract and figures → · Paper available upon request

    Mean wait time for a new Medicare appointment, by metropolitan area
    Mean wait time (days) 102030405060

  • Did Florida's Anti-Brokering Initiative Reduce the Out-of-State "Florida Shuffle"?

    With Brendan Saloner (Brown University)

    Florida's residential addiction treatment programs have long drawn patients from other states, at times through paid patient brokering, a pattern known as the "Florida Shuffle." Using commercial insurance claims data for patients with opioid use disorder, we evaluate whether Florida's crackdown on patient brokering reduced that inflow and changed the care patients received.

    Residential patients from out of state, Florida vs. the rest of the US
    FloridaRest of the US
    0% 20% 40% 60% 80% 100% 2017 law 2019 law 201620172018201920202021202220232024 From out of state (%)

    Share of commercially insured residential treatment patients with opioid use disorder whose home state is elsewhere, by quarter. Dashed lines mark Florida's 2017 and 2019 patient brokering laws. Preliminary.

  • Prescriber Variation in Urine Drug Testing During Buprenorphine Treatment

    With Brendan Saloner (Brown University)

    Urine drug testing is a routine part of buprenorphine treatment for opioid use disorder, but it does not follow clear protocols and has historically been prone to overuse. Using national multi-payer insurance claims data covering more than 1.2 million buprenorphine treatment episodes from 2016 to 2024, we evaluate how much of the variation in testing comes from prescribers.

    Change in a patient's testing after switching prescribers
    −3 −3 −2 −2 −1 −1 0 0 +1 +1 +2 +2 +3 +3 Full match Moves 83% of the way tothe new prescriber's testing frequency Change in prescriber's testing frequency (tests per episode) Change in patient's tests per episode

    Patients whose next 30-day buprenorphine episode was with a different prescriber, in 20 equal-size groups. Preliminary, 2016 to 2024.

  • Substance Use Treatment Facilities Not Admitting Patients Taking Medications for Opioid Use Disorder

    With Michael L. Barnett (Brown University), Brendan Saloner (Brown University)

    Using annual national data on substance use treatment facilities, we track facilities that report treating opioid use disorder but neither provide medications for opioid use disorder (MOUD) nor accept patients who take them, and which kinds of facilities are more likely to turn these patients away.

    Facilities not accepting patients taking MOUD, by setting
    ResidentialAll facilitiesOutpatient
    0% 10% 20% 30% 20152016201720182019202020212022 Facilities not accepting (%)

    Share of facilities that report treating opioid use disorder but neither provide MOUD nor accept patients taking it, 2015 to 2022.

  • Residential Addiction Treatment Access Equity: Effects of the Medicaid Section 1115 SUD IMD Waiver

    With Helen Newton (University of North Carolina), Chima Ndumele (Yale School of Public Health), David A. Fiellin (Yale School of Medicine), Susan Busch (Yale School of Public Health)

    Medicaid historically excluded most residential addiction treatment from federal funding. Section 1115 waivers allowed states to receive federal Medicaid funds for this care if participating facilities offered medications for opioid use disorder, which are highly effective at reducing overdose risk. Comparing states that adopted the waivers with those that did not, we assess whether improvements reached communities equitably. Medicaid participation grew broadly, but gains in medication treatment were concentrated in ZIP codes with fewer Black residents.

    Change in MOUD provision after the waiver, by ZIP code racial composition
    Fewer Black residentsMore Black residents
    -20 0 20 40 60 Waiver -4-3-2-101234 Years since waiver Change (percentage points)

    Event-study estimates for Medicaid-accepting residential facilities in 7 waiver states relative to 19 non-waiver states, 2012 to 2022, with 95% confidence intervals. ZIP codes with more Black residents are those in the top quartile of Black residents' share. The year before the waiver is the reference, and the first and last points include all earlier and later years.

  • Effects of Opioid Treatment Program Closures on Methadone Treatment Receipt for Medicaid Beneficiaries

    With Alina Denham (Stony Brook University), Mah-Afroze Chughtai (Brown University), Haiden Huskamp (Harvard Medical School), Ellen Meara (Harvard T.H. Chan School of Public Health), Marc LaRochelle (Boston University), Michael L. Barnett (Brown University)

    Methadone for opioid use disorder can only be dispensed through opioid treatment programs, so when a program closes, local access can change quickly. Using Medicaid claims data from 2016 to 2022 and an event-study design, we estimate how county-level methadone receipt changes in the months before and after programs close.

    Opioid treatment program closures
    Programs in area 1 4 10+ Closures in area 1 2 3–5 6+ State not included

    Each dot is an area about 70 miles across.

Other projects

  • Variation in Urine Drug Testing Among Patients Starting Buprenorphine Led by Brendan Saloner (Brown University), with Allison O'Rourke (Brown University), Alene Kennedy-Hendricks (Johns Hopkins University), Beth Ann Griffin (RAND Corporation), Adam J. Gordon (University of Utah), Minna Song (Brown University), Kristianny Ruelas-Vargas (Brown University), Michael Incze (University of Utah), Bradley D. Stein (RAND Corporation) Uses national insurance claims data to study how often patients starting buprenorphine receive urine drug testing.
  • Methadone treatment retention and urine drug testing in Medicaid Led by Alina Denham (Stony Brook University), with Michael L. Barnett (Brown University), Mah-Afroze Chughtai (Brown University) Uses Medicaid claims data to study urine drug testing and retention in treatment among patients receiving methadone.
  • Continuity of methadone treatment during hospital stays Led by Dustin Kee (Johns Hopkins University), with William Garneau (Johns Hopkins University), Rachel Haroz (Cooper University Health Care) A secret shopper study of communication between hospitals and opioid treatment programs when patients receiving methadone are hospitalized.
  • Access to inpatient psychiatric care Led by Alexander McClanahan (Yale School of Medicine), with Susan Busch (Yale School of Public Health) A secret shopper study of access to inpatient psychiatric care.