SAMHSA Treatment Data: The Federal Database Behind Substance Abuse and Mental Health Program Statistics

The Substance Abuse and Mental Health Services Administration publishes the most comprehensive federal data on addiction treatment and mental health services in the United States — the National Survey on Drug Use and Health, the Treatment Episode Data Set covering 2 million annual admissions, and the National Mental Health Services Survey covering 12,000 treatment facilities.

What SAMHSA is

SAMHSA is a component agency of the Department of Health and Human Services created by Congress through the ADAMHA Reorganization Act of 1992, which dissolved the Alcohol, Drug Abuse, and Mental Health Administration and transferred its research functions to the National Institutes of Health while consolidating the service delivery and grant functions into the new agency. The reorganization reflected a deliberate policy choice: separating basic research — which went to NIAAA, NIDA, and NIMH within NIH — from the population-level service delivery mission that SAMHSA now carries.

The agency operates with roughly 500 staff and a budget that has grown from approximately $2 billion at its founding to more than $7 billion annually as of the mid-2020s, driven largely by congressional appropriations responding to the opioid epidemic. SAMHSA's core function is not providing treatment directly but funding it: the agency distributes the majority of its appropriation through formula grants to states, which then allocate funds to local providers.

SAMHSA is organized into three centers, each with a distinct mission:

• Center for Substance Abuse Treatment (CSAT). Responsible for expanding the availability of effective substance use disorder treatment services. CSAT administers the Substance Abuse Prevention and Treatment Block Grant (SABG), the primary federal funding stream for publicly funded SUD treatment, and manages the opioid treatment program (OTP) certification system.
• Center for Substance Abuse Prevention (CSAP). Funds community-based prevention programs, maintains the Strategic Prevention Framework, and administers the Drug-Free Communities support program in partnership with the Office of National Drug Control Policy.
• Center for Mental Health Services (CMHS). Administers the Mental Health Block Grant (MHBG), funds community mental health centers and first-episode psychosis programs, and oversees the 988 Suicide and Crisis Lifeline.

SAMHSA's data collection arm — the Center for Behavioral Health Statistics and Quality (CBHSQ) — sits outside the three centers but is the organizational unit responsible for the major public datasets covered in this article. CBHSQ publishes annual reports, maintains the SAMHSA data portal at data.samhsa.gov, and produces the National Survey on Drug Use and Health, which is the largest recurring federal survey of behavioral health in the country.

National Survey on Drug Use and Health (NSDUH)

The NSDUH is a household survey conducted annually since 1971 (under earlier names), with approximately 67,500 respondents ages 12 and older drawn from the civilian non-institutionalized population of the United States. The survey excludes people in correctional facilities, hospitals, nursing homes, homeless shelters, and military barracks — populations that tend to have higher rates of substance use disorder than the general household population — which means NSDUH estimates are best understood as a floor on the national burden of behavioral health conditions rather than a complete count.

The survey uses computer-assisted self-interviewing for sensitive questions about drug use and mental health, a methodology designed to reduce social desirability bias. Field interviews are conducted by RTI International under contract. The questionnaire covers substance use (alcohol, tobacco, marijuana, cocaine, heroin, methamphetamine, prescription drugs used non-medically, and a range of other drugs), mental health (major depressive episode, serious mental illness, any mental illness, suicidal ideation and behavior), and treatment utilization across both domains.

Key NSDUH estimates from the 2022 survey (released 2023) illustrate the scale of the behavioral health burden the federal system is trying to address:

• 48.7 million Americans ages 12 and older — 17.3 percent of the population — had a substance use disorder in the past year.
• 57.8 million adults had any mental illness in the past year; 14.1 million had serious mental illness.
• Only 1-in-8 people with a substance use disorder received any substance use treatment in the past year.
• Among people with serious mental illness, about half received mental health services.
• 21.5 million people ages 12 and older had both a substance use disorder and a mental illness in the past year — the co-occurring disorder population that strains both treatment systems simultaneously.

NSDUH produces national estimates and, since 2002, state-level estimates for all 50 states and the District of Columbia. State estimates use small area estimation methods because the per-state sample sizes are not sufficient for direct estimation; the statistical model borrows strength from neighboring states and demographic covariates to produce reliable state-level figures. These state estimates are the primary input to federal formula-grant allocations that distribute SABG and MHBG dollars.

NSDUH microdata is available in two forms. The public-use file — which removes geographic identifiers below the regional level and top-codes some continuous variables — is available through the SAMHSA data portal at data.samhsa.gov and through ICPSR (Inter-university Consortium for Political and Social Research) at the University of Michigan. Restricted-use files with state identifiers are available to researchers through a formal data access agreement with SAMHSA's Office of Applied Studies.

Treatment Episode Data Set (TEDS)

TEDS is an administrative data system rather than a survey. It collects records from state substance abuse agencies on every admission to and discharge from publicly funded substance use disorder treatment facilities. Unlike NSDUH, which estimates population prevalence, TEDS documents actual treatment events — who entered the system, what they were treated for, and how the episode ended.

Annual TEDS coverage runs to approximately 2 million admissions, with consistent data available back to 1992. TEDS has two components: TEDS-A (Admissions Data Set) and TEDS-D (Discharge Data Set). Both are de-identified; records carry state of treatment as the finest geographic identifier in the public-use files. TEDS excludes facilities that receive no public funding, meaning the private-pay residential treatment segment — which serves a wealthier demographic — is largely absent from the data.

TEDS-A variables recorded at admission include:

• Primary, secondary, and tertiary substance. The presenting substance of concern, coded to a standardized scheme covering alcohol, opioids (with further distinction between heroin, non-prescription methadone, other opiates/synthetics, and prescription opioids), cocaine and crack, cannabis, methamphetamine and other stimulants, sedatives/hypnotics/tranquilizers, and other substances.
• Route of administration. Oral, smoking, inhalation, injection, or other. Route of administration is a key variable for tracking the shift from injected heroin to non-injected fentanyl in counterfeit pill form, which has changed harm reduction programming priorities.
• Treatment setting. Outpatient, intensive outpatient, ambulatory detox, non-hospital residential (short-term and long-term), hospital inpatient, and opioid treatment program. This variable distinguishes the intensity of the treatment episode.
• Source of referral. Individual/self, substance use disorder treatment provider, other health care provider, school (drug education or student assistance program), employer/EAP, criminal justice or DUI/DWI, and other community referral. Criminal justice is historically the largest single referral source nationally, accounting for 35 to 40 percent of admissions.
• Demographics. Age group (five-year bands), sex, race and ethnicity (OMB categories), education, employment status, and living arrangement at admission.
• Prior treatment history. Number of prior substance use disorder treatment episodes. First-time admissions represent roughly 40 percent of the universe; repeat admissions are common and do not indicate treatment failure so much as the chronic relapsing nature of substance use disorder as a medical condition.
• Expected payment source. Medicaid, Medicare, state-financed insurance, private insurance, self-pay, no charge, and other. Payment source is a proxy for insurance coverage and socioeconomic status.

TEDS-D adds, at time of discharge:

• Reason for discharge: completed treatment, dropped out, transferred to another program, incarcerated, death, or other.
• Length of stay in days.
• Medication-assisted treatment at discharge: whether the patient was discharged on methadone, buprenorphine, naltrexone, or no medication.

The TEDS time series is the primary federal dataset for tracking substance-specific treatment trends across the full arc of the opioid epidemic. The shift from prescription opioid admissions in the mid-2000s, to heroin admissions after 2010 as the prescription supply tightened, to synthetic opioid and polysubstance admissions after 2016 as illicit fentanyl displaced heroin — all of this trajectory is visible in TEDS. The dataset leads CDC overdose mortality by approximately two to three years, making treatment admissions data a leading indicator for drug supply shifts.

National Mental Health Services Survey (N-MHSS)

N-MHSS is an annual census of mental health treatment facilities in the United States, administered by SAMHSA each year since 2010 (prior to that year, a similar survey was conducted under the name SMHA Survey). The 2022 survey covered approximately 12,000 mental health treatment facilities, representing the near-universe of licensed specialty mental health treatment settings.

Facilities included in N-MHSS span the full continuum of mental health treatment settings:

• State psychiatric hospitals (long-term inpatient, the traditional backbone of public mental health systems)
• VA medical centers and outpatient clinics with mental health programs
• Community mental health centers (federally qualified mental health centers and their equivalents)
• Residential treatment facilities (RTFs) for adults and for children and adolescents
• Partial hospitalization and intensive outpatient programs
• Outpatient mental health clinics
• Multi-setting practices and hospital-based outpatient programs

For each facility N-MHSS collects: licensed bed count and current utilization (for inpatient and residential settings), diagnostic populations served (depression, bipolar disorder, schizophrenia, PTSD, personality disorders, co-occurring SUD, etc.), treatment modalities offered (individual therapy, group therapy, family therapy, psychotropic medication management, assertive community treatment, electroconvulsive therapy), special population programs (veterans, children and adolescents, older adults, LGBTQ+, criminal justice), and payment types accepted.

A parallel survey — the National Survey of Substance Abuse Treatment Services (N-SSATS) — covers the approximately 17,000 specialty substance use disorder treatment facilities in a structure mirroring N-MHSS. N-SSATS is the data source that powers the findtreatment.gov treatment locator: every facility in the public-facing search interface has a corresponding N-SSATS record. Both surveys are available through SAMHSA's Behavioral Health Services Information System (BHSIS) API at api.hhs.gov/bhsis.

SAMHSA and the opioid crisis

No federal agency has been more directly shaped by the opioid epidemic than SAMHSA. The agency's budget, regulatory posture, and public prominence all shifted dramatically starting in 2015 as overdose deaths accelerated and Congress began directing emergency appropriations through SAMHSA-administered grant programs.

Buprenorphine and the X-waiver history. The Drug Addiction Treatment Act of 2000 (DATA 2000) created the framework for office-based buprenorphine prescribing, allowing physicians to treat opioid use disorder with Schedule III-V controlled substances outside the OTP system — but required physicians to obtain a special DEA waiver (the “X-waiver”) and limited patient panel sizes. This restriction, which required separate DEA registration beyond the standard controlled substance registration, created a credentialing barrier that suppressed prescriber adoption for more than two decades. The Mainstreaming Addiction Treatment (MATE) Act, enacted as part of the Consolidated Appropriations Act of 2023, eliminated the X-waiver requirement entirely. As of 2023, any DEA-registered practitioner with authority to prescribe Schedule III controlled substances can prescribe buprenorphine for opioid use disorder without additional registration.

Opioid treatment programs and methadone regulations. SAMHSA regulates OTPs under 42 CFR Part 8, which establishes the federal certification requirements for methadone dispensing programs. OTPs must obtain SAMHSA certification, DEA Schedule II dispensing authorization, and state opioid treatment authority (OTA) approval. This multi-agency licensing process takes 12 to 18 months on average and requires significant capital investment. In 2021 SAMHSA issued guidance permitting OTPs to dispense methadone via mobile units, expanding geographic reach without requiring new fixed-site facility licensing. Pandemic-era regulatory flexibilities — expanded take-home methadone doses and audio-only telehealth for buprenorphine prescribing — were largely codified permanently in the 2024 SAMHSA rule.

State Opioid Response (SOR) grants. Congress directed roughly $1.5 billion annually through SAMHSA's State Opioid Response grant program starting with the 21st Century Cures Act (2016) and subsequent emergency supplemental appropriations. SOR grants flow to state substance abuse agencies, which distribute them to treatment providers for expanding medication-assisted treatment capacity, purchasing naloxone, and training first responders. Grant allocations are formula-driven, with overdose mortality rates from CDC data serving as a key input.

988 Suicide and Crisis Lifeline. SAMHSA administers the 988 Suicide and Crisis Lifeline, which launched on July 16, 2022, replacing the previous 10-digit National Suicide Prevention Lifeline number. The three-digit dialing code — analogous to 911 for mental health crises — was established by FCC order in 2020 following legislation directing SAMHSA to build out the network. In its first year of operation, the 988 system handled more than 5 million contacts (calls, texts, and chats). SAMHSA publishes monthly contact volume data, which shows steady growth in utilization and reveals geographic variation in response rates and average speed to answer across the network of approximately 200 local crisis centers.

The treatment gap

The most consequential finding in NSDUH is the treatment gap: the difference between the estimated population with a substance use disorder and the population that actually receives treatment. As of the 2022 NSDUH, approximately 93 percent of the 48.7 million Americans with a substance use disorder in the past year did not receive treatment at a specialty facility. The gap is comparable, though less severe, for mental illness: roughly half of people with serious mental illness received no mental health treatment in the past year.

NSDUH surveys respondents who did not receive treatment about why. The most frequently reported barriers are:

• Cost and insurance coverage. The single most commonly cited barrier across income groups. Even with Medicaid expansion under the ACA, navigating benefit coverage and finding in-network providers creates access friction that stops treatment-seeking.
• Not feeling ready to stop using. Ambivalence about treatment is a clinical feature of addiction; motivational interviewing and low-threshold models (which allow treatment entry without requiring immediate abstinence) address this barrier but are not universally available.
• Stigma. Fear of what neighbors, employers, or family members would think if they sought treatment. Stigma affects help-seeking across all demographic groups but is particularly pronounced in rural communities where social networks are dense and anonymity limited.
• No program available or program inconvenient. Geographic barriers are significant outside metropolitan areas. Rural counties may lack any specialty SUD treatment facility, and even where facilities exist, transportation to daily or weekly appointments can be prohibitive.

SAMHSA's primary funding mechanisms for closing the treatment gap are the two block grants that flow to states as formula allocations:

Substance Abuse Prevention and Treatment Block Grant (SABG). Approximately $2 billion per year distributed to the 50 states, DC, and the territories. SABG funds the largest share of publicly funded substance use disorder treatment in the United States. States must spend a set-aside (historically 20 percent) on primary prevention programs. Allocation is formula-based, weighting population and state-level treatment need estimated from NSDUH. States have broad discretion in how they allocate SABG dollars within their systems, which creates substantial interstate variation in the mix of residential, outpatient, and medication-assisted treatment funded.

Mental Health Block Grant (MHBG). Approximately $1 billion per year distributed to states for community mental health services. MHBG funds community mental health centers, assertive community treatment teams, crisis stabilization programs, and supported housing and employment services for people with serious mental illness. The 21st Century Cures Act (2016) added a 10 percent set-aside within MHBG specifically for early serious mental illness programs, primarily first-episode psychosis programs modeled on the RAISE study.

Data access

SAMHSA's data infrastructure has multiple layers, each suited to different research uses:

data.samhsa.gov is the main public data portal. It hosts the NSDUH public-use files, TEDS annual data files, N-SSATS and N-MHSS public-use files, the DAWN emergency department drug mentions dataset, and a growing collection of state-level behavioral health indicator reports. File formats are generally SAS transport, SPSS, and CSV, with corresponding SAS and SPSS codebooks. Annual TEDS and N-SSATS files are released approximately 18 months after the survey year closes.

ICPSR (Inter-university Consortium for Political and Social Research, at the University of Michigan) archives the NSDUH public-use files with documentation and R/SAS/Stata analysis templates. ICPSR is the preferred access point for researchers who want version-controlled data files with full codebook documentation. NSDUH restricted-use files with state identifiers are available through a SAMHSA data use agreement for researchers with institutional review board approval.

BHSIS API (api.hhs.gov/bhsis) is the programmatic interface to the N-SSATS and N-MHSS facility databases. The API supports filtering by state, facility type, services offered, payment accepted, and special population programs. Authentication requires an API key issued by HHS; registration is straightforward and approval is typically within one business day. This is the same API that powers findtreatment.gov, so it returns the full facility records including name and address fields that the public-use files suppress.

findtreatment.gov Treatment Locator API. Accessible at https://findtreatment.gov/locator/api/facilities, this endpoint supports geographic search by zip code, city, or latitude/longitude with a distance radius parameter. Parameters include sType (SA for substance abuse, MH for mental health, or both), distance in miles, locationas a zip code or city/state string, and pageSize. The response includes facility name, address, phone, accepted payment types, services, and special programs. No API key is required for the locator endpoint, making it the most accessible starting point for facility-level data access.

Python workflow: Treatment Locator API and NSDUH state estimates

The following snippet demonstrates two data access paths: the Treatment Locator API for facility discovery and the NSDUH published CSV tables for state-level prevalence estimates. The Locator API requires no authentication; the NSDUH tables are published as CSV files at fixed URLs on samhsa.gov.

import requests, pandas as pd

# SAMHSA Treatment Locator API — find treatment facilities
# API docs: https://findtreatment.gov/locator/apidoc
base = "https://findtreatment.gov/locator/api/facilities"
params = {
    "sType": "SA",        # substance abuse
    "distance": 50,
    "location": "10001",  # zip code (Manhattan)
    "pageSize": 20,
}
resp = requests.get(base, params=params, timeout=20)
data = resp.json()

print(f"Facilities near 10001: {data.get('totalCount', 0)} within 50 miles")
for f in data.get("rows", [])[:5]:
    name = f.get("name1", "")
    city = f.get("city", "")
    state = f.get("state", "")
    services = ", ".join(f.get("typeOfCare", [])[:3])
    print(f"  {name} — {city}, {state}")
    print(f"    Services: {services}")

# NSDUH state estimates (published table format)
# Download from SAMHSA CBHSQ reports — CSV format
nsduh_url = "https://www.samhsa.gov/data/sites/default/files/reports/rpt39456/NSDUHstateTab12-2022.csv"
try:
    df = pd.read_csv(nsduh_url)
    print("\nNSDUH state estimates loaded:", df.shape)
    print(df.head())
except Exception as e:
    print(f"Note: Download NSDUH tables from samhsa.gov/data: {e}")

The Treatment Locator API is the most practically useful SAMHSA data endpoint for geographic analysis because it returns facility-level records with full addresses — including the OTP facilities that require county-level analysis for the rural treatment gap research described above. The sType parameter accepts SA(substance abuse), MH (mental health), or SA MH for both. The pageSize maximum is 100 per request; paginate using the startparameter for large geographic areas.

NSDUH state estimates are published as Excel and CSV tables in SAMHSA's annual National Survey on Drug Use and Health: State Estimates reports. The URL pattern follows the CBHSQ report numbering scheme; the most reliable access method is downloading from data.samhsa.gov rather than constructing URLs to samhsa.gov/data directly, as the latter URLs change between report cycles.

Interpreting the treatment gap data

The 93 percent figure — the share of people with a substance use disorder who did not receive specialty treatment — requires careful interpretation. The NSDUH denominator is the full population meeting DSM-5 diagnostic criteria for a substance use disorder in the past year, which includes people with mild disorder (two or three criteria) who may be functioning adequately without formal treatment and who may not perceive themselves as needing it. The treatment gap at the severe end of the spectrum is smaller but still substantial.

NSDUH also asks about informal help-seeking: mutual aid programs like Alcoholics Anonymous and Narcotics Anonymous, pastoral counseling, and informal support networks. When informal help is counted alongside specialty treatment, the proportion of people with a SUD who received some form of help in the past year is somewhat higher — though still substantially below the treatment need estimates. Mutual aid participation is not captured in TEDS, so the formal treatment utilization data understates the total help-seeking activity in the population.

The treatment gap also varies dramatically by substance. Alcohol use disorder is the most prevalent SUD in NSDUH but has one of the lowest treatment utilization rates: fewer than 10 percent of people with alcohol use disorder receive treatment in a given year. Opioid use disorder has a higher treatment utilization rate than alcohol, partly because the consequences of untreated opioid addiction are more immediate and visible, but also because the 988 lifeline, SAMHSA treatment locators, and emergency department referral networks have become more active pathways into care.

Geographic concentration of treatment capacity

N-SSATS data consistently shows that substance use disorder treatment capacity — and particularly the OTP capacity required for methadone dispensing — is heavily concentrated in urban areas. The licensing burden for OTPs (requiring SAMHSA certification, DEA Schedule II authorization, and state OTA approval) creates a high barrier to entry that is more tractable for large urban providers than for small rural organizations. Community opposition to siting OTPs in residential neighborhoods further constrains geographic expansion.

SAMHSA's 2023 N-SSATS analysis identified approximately 1,200 of roughly 3,100 US counties with no OTP facility. Across those zero-OTP counties, a substantial fraction — concentrated in West Virginia, Kentucky, Tennessee, rural Ohio, and rural New Mexico — have above-median overdose mortality rates in CDC data, identifying the highest-priority geographic gaps for MAT expansion. The 2021 SAMHSA guidance permitting mobile OTP units began to address this gap by allowing methadone dispensing from vehicles that serve multiple rural locations from a single licensed hub facility.

The buprenorphine picture improved substantially after the MATE Act eliminated the X-waiver requirement in 2023. However, DEA prescribing data shows that prescriber adoption has remained concentrated in urban and suburban practices; rural primary care adoption of buprenorphine prescribing has increased more slowly than anticipated. The combination of zero-OTP geography and below-median buprenorphine prescriber density continues to define the treatment desert counties visible in the N-SSATS data.

The 988 data infrastructure

SAMHSA publishes monthly contact volume data for the 988 Suicide and Crisis Lifeline at data.samhsa.gov. The published metrics include total monthly contacts by modality (call, text, chat), contacts handled by local crisis centers versus the national backup network, average speed to answer by center, and contact volume by state. The 988 data is unusual among SAMHSA datasets in that it is published with a short lag — typically 45 to 60 days after the reference month — compared to the 12 to 18 month lag for NSDUH and TEDS.

The 988 volume data enables analysis of crisis service utilization patterns that were previously invisible in federal data. The previous 10-digit National Suicide Prevention Lifeline was underutilized in part because the number was not widely known; the 988 transition produced a documented increase in contact volume that varied by state (reflecting differences in prior awareness and in the robustness of local crisis center networks). SAMHSA has signaled intent to develop a more comprehensive 988 data infrastructure, including outcome data on what happens to callers after crisis contact, which would make 988 analytics substantially more powerful as a policy tool.

Limitations and analytical caveats

NSDUH's exclusion of the institutionalized population — correctional facilities, hospitals, nursing homes — is the most consequential structural limitation. People in jails and prisons have substantially higher rates of substance use disorder than the general population; excluding them from the survey denominator means the treatment gap estimates are conservative. The Bureau of Justice Statistics' Survey of Inmates in State and Federal Correctional Facilities provides some coverage of this population but uses different methods and cannot be directly integrated with NSDUH.

TEDS' exclusion of privately funded treatment facilities creates a systematic bias toward the publicly funded system. The private-pay residential treatment segment — which serves wealthier patients and operates under different treatment philosophies than Medicaid-funded outpatient programs — is largely invisible in TEDS. Treatment outcomes, demographic profiles, and length-of-stay patterns in the publicly funded system may differ substantially from the system as a whole.

The 18-month publication lag for TEDS and N-SSATS means that the most recent data always reflects conditions from the prior year or before. During rapidly evolving episodes — the emergence of xylazine-adulterated fentanyl as a new clinical challenge for OTPs, the appearance of novel synthetic opioids in the nitazene class — TEDS will lag the clinical reality by several publication cycles. SAMHSA's Drug Abuse Warning Network (DAWN), which covers emergency department visits and has a shorter publication lag, is a better source for tracking novel substance emergence.

The N-MHSS and N-SSATS surveys undercount office-based providers. Primary care physicians, nurse practitioners, and other licensed practitioners who provide mental health or substance use disorder treatment as part of a general practice — without operating a dedicated specialty facility — are not systematically surveyed. This creates an undercount of treatment capacity that is most severe in urban areas, where integrated primary care and behavioral health is most common, meaning the geographic treatment gap may be modestly smaller than the facility-count data suggests.

For the CDC drug overdose mortality data that documents the public health crisis SAMHSA treatment programs are designed to address — including county-level death rates, ICD-10 coding, and the VSRR provisional data series: CDC Drug Overdose Mortality Data: The Federal Dataset Behind the Opioid Crisis →

For the CDC NNDSS notifiable disease surveillance system, which tracks infectious disease outcomes that intersect with substance use disorder — including HIV, hepatitis C, and wound botulism associated with injection drug use: CDC NNDSS: The Federal Notifiable Disease Surveillance System →

For IRS Criminal Investigation data on money laundering and drug trafficking prosecutions — the financial crime enforcement angle on the same illicit drug markets that generate SAMHSA treatment demand: IRS Criminal Investigation: The Federal Database Behind Tax Fraud and Financial Crime Prosecutions →