DEA Registrant Enforcement: The Federal Database Behind Controlled Substance License Revocations

The Drug Enforcement Administration publishes every order to show cause, immediate suspension order, and final order revoking or denying a DEA registration — the controlled substance prescribing and dispensing licenses held by physicians, pharmacies, hospitals, and distributors — in the Federal Register and agency database.

This article covers the statutory basis for DEA registration under 21 USC 823, the enforcement authorities created by 21 USC 824, the three primary enforcement instruments (orders to show cause, immediate suspension orders, and final orders), the grounds for revocation and the role of state medical license proceedings, the opioid pill mill enforcement wave and landmark DEA actions against physicians, pharmacies, and distributors, the federal prescription drug monitoring program landscape, and how to access and analyze DEA enforcement documents from the Federal Register API using Python.

The DEA registrant system

The Controlled Substances Act, codified at 21 U.S.C. §823, requires every person or entity that manufactures, distributes, or dispenses controlled substances to obtain a registration from the Drug Enforcement Administration. The registration requirement applies across the entire supply chain: manufacturers who produce active pharmaceutical ingredients and finished dosage forms; wholesale distributors who move product from manufacturers to pharmacies and hospitals; pharmacies and hospital pharmacies that dispense to patients; and practitioners — physicians, dentists, veterinarians, and other authorized prescribers — who administer or prescribe controlled substances in the course of professional practice.

The scale of the registrant population is substantial. As of recent DEA figures, approximately 1.8 million practitioners hold active DEA registrations, representing the universe of physicians, physician assistants, nurse practitioners, dentists, veterinarians, and other mid-level practitioners authorized to prescribe Schedule II through V controlled substances. Approximately 80,000 pharmacies hold DEA registrations, including retail chain pharmacies, independent community pharmacies, hospital and institutional pharmacies, and mail-order pharmacies. At the manufacturing and distribution tier, approximately 1,800 entities hold manufacturer or distributor registrations. Each registration is specific to a physical address and a schedule of controlled substances; a practitioner who maintains offices at multiple locations must obtain a separate registration for each location.

The DEA registration number — informally called a DEA number — is the primary identifier for each registrant. The format is two letters followed by seven digits. The first letter encodes the registrant type: A, B, F, and G indicate practitioners; M indicates mid-level practitioners (nurse practitioners, physician assistants); P and R indicate manufacturers and distributors; and other prefixes apply to researchers, analytical labs, importers, and exporters. The second letter is the first letter of the registrant's last name or business name. The seven digits include an internal check digit that allows verification of number validity, which is why pharmacies and hospitals routinely verify DEA numbers before dispensing or processing controlled substance orders.

The application forms vary by registrant type. DEA Form 224 is the practitioner application, used by physicians, dentists, veterinarians, and other Schedule III-V prescribers; a separate Form 224a covers mid-level practitioners in states that authorize them to prescribe controlled substances. DEA Form 225 covers manufacturers, distributors, researchers, analytical labs, importers, exporters, and narcotic treatment programs. Form 226 is the distributor renewal form. Applications require disclosure of any prior criminal convictions relating to controlled substances, any prior DEA registration actions, and current state licensure status — because DEA registration requires an underlying state license to practice or operate.

Enforcement authorities under 21 USC 824

The legal basis for DEA enforcement against registrants is 21 U.S.C. §824, which authorizes the Attorney General (acting through DEA) to revoke, suspend, or deny a registration. Section 824 specifies both the grounds for action and the procedures DEA must follow, and it is the source of the three primary enforcement instruments: the order to show cause, the immediate suspension order, and the final order.

The order to show cause (OTSC) is the standard enforcement initiation. When DEA proposes to revoke or deny a registration, it serves an OTSC on the registrant setting out the factual and legal basis for the proposed action and notifying the registrant of the right to request a hearing before a DEA Administrative Law Judge (ALJ). The registrant has 30 days to request a hearing. If a hearing is requested, the ALJ conducts proceedings under 21 CFR Part 1316, takes evidence, hears testimony, and issues a recommended decision. The DEA Administrator then reviews the ALJ's recommended decision and issues a final order. Final orders are published in the Federal Register and constitute binding agency action subject to judicial review in the appropriate U.S. Court of Appeals.

The immediate suspension order (ISO) is DEA's emergency enforcement tool. Under 21 U.S.C. §824(d), DEA may suspend a registration immediately and without prior hearing upon a finding that continued registration constitutes “an imminent danger to the public health or safety.” The ISO suspends the registrant's authority to handle controlled substances effective upon service. An OTSC must follow within 30 days of the ISO, initiating the ordinary revocation proceeding. The ISO standard — imminent danger — is higher than the standard for ordinary revocation, which is why ISOs are reserved for the most egregious situations: a physician writing thousands of opioid prescriptions outside any legitimate clinical relationship, a pharmacy dispensing controlled substances without valid prescriptions, or a distributor shipping to pharmacies it knows are engaged in diversion. ISOs receive immediate Federal Register publication and are among the most visible DEA enforcement actions.

The final order is the culmination of the revocation proceeding. It may revoke the registration entirely, suspend it for a defined period, restrict it (limiting the schedules or substances the registrant may handle), or deny a pending application. Final orders are published in the Federal Register with the full factual findings and legal analysis. They serve as precedent for subsequent DEA enforcement actions and are regularly cited in the DEA Administrator's decisions on contested revocation proceedings. Federal Register publication is mandatory; the record is permanent and publicly accessible.

Voluntary surrender is not a formal enforcement action but appears frequently in aggregate enforcement statistics. A registrant under investigation may surrender their DEA number rather than face a revocation proceeding; surrender avoids an adverse agency finding on the record but typically has the same practical effect as revocation. DEA accepts voluntary surrenders and counts them in its enforcement totals. Surrenders do not generate Federal Register publication in the same manner as formal enforcement actions, which creates a gap between the documented public record and the total number of practitioners who lose controlled substance authority in any given year.

Grounds for revocation

Section 824 enumerates five grounds for revocation. The first is material falsification of an application, which includes misrepresenting prior criminal history, failing to disclose prior DEA actions, or misrepresenting state licensure status. The second is conviction of a felony relating to controlled substances, which is an automatic disqualifier; a physician convicted of drug diversion loses DEA registration as a statutory matter, and DEA revocation follows as a ministerial act. The third and fourth grounds — suspension or revocation of state controlled substance authority and loss of state authority to handle controlled substances — reflect the dependency of DEA registration on state licensure. DEA registration requires a valid underlying state license; if a state medical board revokes a physician's license, DEA registration authority evaporates under the statute even without a separate federal proceeding.

The fifth ground is the “public interest” catch-all, which is the basis for the most legally complex and contested revocation proceedings. Under 21 U.S.C. §823(f), DEA considers five factors when evaluating whether registration is consistent with the public interest: the registrant's recommendation by state licensing boards; the registrant's experience in dispensing controlled substances; any conviction relating to controlled substances; compliance with applicable state and local laws; and such other conduct that may threaten public health or safety. “Prescribing outside the scope of professional practice” — the legal standard for what distinguishes a legitimate prescription from drug diversion — is the most frequently litigated public interest factor. DEA's position, affirmed in numerous final orders, is that a prescription issued without a legitimate medical purpose in the course of a bona fide physician-patient relationship is not a valid prescription and exposes the issuing practitioner to revocation regardless of whether criminal charges are filed.

Opioid pill mills and DEA enforcement

The opioid crisis generated a wave of DEA enforcement against practitioners and pharmacies that the agency characterized as pill mills — operations that dispensed controlled substances, primarily opioids, with little or no legitimate clinical gatekeeping. The Florida pill mill phenomenon of the mid-2000s became the emblematic case study. Before Florida enacted its Prescription Drug Monitoring Program in 2011, the state had virtually no statewide mechanism for tracking opioid prescriptions across dispensing points. The result was an extraordinary concentration of high-volume opioid prescribing. At the peak, 98 of the top 100 oxycodone-prescribing physicians in the United States were located in Florida. Pain management clinics in Broward County and Palm Beach County were dispensing tens of thousands of opioid pills daily to patients who drove from Kentucky, West Virginia, Tennessee, and Ohio specifically to obtain prescriptions.

DEA's coordinated enforcement response was Operation Pill Nation, a multi-year joint task force with the Florida Department of Law Enforcement, local law enforcement, and the Florida Board of Medicine. Operation Pill Nation targeted the highest-volume pain clinics with a combination of ISO issuances, criminal referrals to U.S. Attorney's offices, and state medical board referrals. Between 2010 and 2012, the number of pain clinics in Florida dropped from approximately 900 to under 300. Oxycodone shipments to Florida pharmacies fell by more than 50 percent in the two years following implementation of the state PDMP and the Operation Pill Nation enforcement surge.

The Volkman case represents the most severe criminal outcome in the federal pill mill prosecution record. Paul Volkman, an Ohio physician, operated pain management clinics in southern Ohio between 2003 and 2006. DEA investigators documented that Volkman prescribed approximately 11,000 oxycodone pills to 100 patients over three years, with minimal or no physical examination, no drug testing, and prescriptions written in cash-only transactions. Four of his patients died from drug overdoses. Volkman was convicted in 2011 on four counts of drug distribution resulting in death under 21 U.S.C. §841(b)(1)(C), which carries a 20-year mandatory minimum per count. The sentencing judge imposed four consecutive life sentences, making Volkman's sentence the longest ever imposed for prescription drug diversion at the federal level. His DEA registration had been revoked by ISO years before the criminal conviction.

DEA enforcement extended to major pharmaceutical retailers. In 2013, Walgreens reached an $80 million civil penalty settlement with DEA — at the time the largest settlement with a retail pharmacy in DEA history — resolving allegations that six Walgreens distribution centers in Florida had failed to maintain effective controls against diversion and had shipped controlled substances in suspicious quantities to Florida pharmacies. DEA issued ISOs against two Walgreens distribution centers, suspending their DEA registrations and forcing temporary shutdown of controlled substance distribution from those facilities. The settlement required Walgreens to implement enhanced dispensing controls, restrict pharmacist discretion on certain opioid prescriptions, and submit to ongoing DEA monitoring.

At the distributor level, DEA enforcement actions against McKesson Corporation, Cardinal Health, and AmerisourceBergen established the regulatory standard for suspicious order monitoring failures. Cardinal Health's Lakeland, Florida, distribution center was subject to an ISO in 2012, suspending oxycodone distribution from that facility. McKesson paid $150 million in 2017 and agreed to suspensions at multiple distribution centers after DEA documented failures to report suspicious orders at 12 distribution centers across four states. These actions preceded and informed the larger opioid MDL litigation settlements detailed in the companion ARCOS article.

The Purdue Pharma enforcement trajectory illustrated the intersection of DEA registration authority and corporate criminal liability. Purdue held DEA manufacturing registrations for OxyContin and other controlled substances. As part of the 2020 Department of Justice settlement resolving criminal charges, Purdue agreed to plead guilty to three federal felonies including distribution of a controlled substance, pay $8.34 billion in combined criminal and civil penalties, and convert to a public benefit company with proceeds directed to opioid remediation. DEA's manufacturing registrations for Purdue and its successor entity were addressed in the settlement framework, with ongoing registration subject to enhanced compliance obligations.

Prescription Drug Monitoring Programs

Prescription Drug Monitoring Programs (PDMPs) are state-administered electronic databases that track controlled substance prescriptions dispensed within the state. All 50 states and the District of Columbia now operate PDMPs, though coverage, mandatory prescriber participation requirements, and real-time versus batch reporting vary by state. Most state PDMPs track Schedule II through V prescriptions dispensed at retail pharmacies; some also capture Schedule II through V dispensing by practitioners who maintain in-office dispensing registrations.

Interstate PDMP data sharing occurs through the PMP InterConnect network, operated by the National Association of Boards of Pharmacy (NABP). PMP InterConnect enables a prescriber or pharmacist in one state to query the PDMP records of a patient in another participating state at the point of care, addressing the doctor-shopping problem that was central to the Florida pill mill phenomenon. As of 2024, the majority of states participate in PMP InterConnect with real-time query capability. Remaining gaps include states that have not enacted mandatory prescriber query laws (requiring prescribers to check the PDMP before prescribing Schedule II-III controlled substances) and states with technical systems that have not completed InterConnect integration.

PDMP data plays an important role in DEA investigations but is not directly accessible to DEA through a federal data-sharing agreement in the way that ARCOS transaction data is. DEA investigators obtain PDMP records through state-level cooperative agreements, administrative subpoenas, and grand jury subpoenas in the course of specific criminal investigations. The absence of a centralized federal PDMP database means that cross-state pattern analysis of prescription opioid dispensing — of the type that would have identified the Florida pill mills years earlier — requires stitching together state-level data sets rather than querying a single federal repository.

DEA's 2023 telemedicine rule addressed the expanded prescribing of Schedule II and III controlled substances via telemedicine that had been temporarily authorized during the COVID-19 public health emergency. The Ryan Haight Online Pharmacy Consumer Protection Act of 2008 generally prohibited prescribing Schedule II-III controlled substances via telemedicine without an in-person evaluation; the public health emergency waiver suspended this requirement for the duration of the declared emergency. DEA's 2023 proposed rule established a permanent telemedicine prescribing framework with differentiated requirements by substance schedule and prescriber specialty, attempting to preserve access to telemedicine-based care — particularly for buprenorphine in opioid use disorder treatment — while preventing the emergence of online pill mills that the Ryan Haight Act had been designed to suppress.

Accessing DEA enforcement data

The Federal Register is the primary public record of DEA enforcement actions. Every OTSC, ISO, and final order is published as a Notice in the Federal Register under the Drug Enforcement Administration agency heading. The Federal Register search interface at federalregister.gov allows keyword search filtered by agency and document type. Searching for “Drug Enforcement Administration” as agency, “Notice” as document type, with terms like “revocation” or “suspension” or “order to show cause” returns the relevant enforcement documents. Each document page includes the full text of the order, the factual findings, the registrant's DEA number and business address, and the effective date of the action.

The Federal Register API provides programmatic access to the same documents. The API endpoint is https://www.federalregister.gov/api/v1/documents, with parameters for agency, document type, and search terms. Responses are paginated JSON with fields including the document title, publication date, abstract, document number, and a URL to the full text. The API supports filtering by date range, allowing retrieval of all DEA enforcement notices published in a specified period. Rate limits are generous for research use, and the API does not require authentication. The full text of each document is available via the full_text_xml_url field, which links to a machine-readable XML version of the Federal Register document.

DEA's Diversion Control Division website at deadiversion.usdoj.gov provides a registration status lookup tool that allows anyone to verify whether a specific DEA number is currently active, expired, or revoked. The lookup is useful for due diligence — pharmacies use it to verify prescriber registrations before dispensing — but it is not a bulk download and does not expose enforcement history. Revoked registrations return a status indicator but do not link to the Federal Register enforcement record. For systematic analysis of enforcement patterns, the Federal Register API is the appropriate data source.

ARCOS transaction data — the controlled substance distribution database that tracks every manufacturer-to-distributor and distributor-to-pharmacy shipment — is a complementary but distinct data source. ARCOS documents the flow of controlled substances through the supply chain; the Federal Register enforcement record documents the DEA's regulatory response when that flow revealed diversion. Together, the two datasets allow reconstruction of the complete enforcement narrative: ARCOS shows that a particular distributor shipped 10 million oxycodone pills to a county of 25,000 people; the Federal Register shows the ISO that halted the distribution and the final order that revoked the distributor's registration. The ARCOS data is covered in the companion article on DEA ARCOS opioid distribution.

Python: analyzing DEA enforcement documents from the Federal Register API

The following script fetches DEA enforcement documents from the Federal Register API, filters for revocation and suspension notices, classifies each document by action type (OTSC, ISO, final order, voluntary surrender) and registrant type (physician, pharmacy, distributor, hospital), extracts the state from document text, computes enforcement action counts by year for 2018 through 2024, identifies states with the highest revocation counts, and flags documents naming major chain pharmacies or hospital systems. Requirements: requests andpandas.

import requests
import pandas as pd
from collections import defaultdict
import re

# ---------------------------------------------------------------------------
# DEA Registrant Enforcement: Federal Register API
# Fetches DEA revocation/suspension notices from the Federal Register,
# parses action type (OTSC/ISO/final order), extracts registrant type,
# computes actions by year (2018-2024), identifies states with highest
# revocation counts, and flags major chains/hospital systems.
# ---------------------------------------------------------------------------

FR_API = "https://www.federalregister.gov/api/v1/documents"

def fetch_dea_enforcement_notices(per_page=100, max_pages=20):
    """Paginate through Federal Register DEA enforcement notices."""
    records = []
    page = 1

    while page <= max_pages:
        params = {
            "conditions[agencies][]": "drug-enforcement-administration",
            "conditions[type][]": "Notice",
            "conditions[term]": "registration revocation suspension",
            "per_page": per_page,
            "page": page,
            "fields[]": [
                "title", "publication_date", "abstract",
                "document_number", "html_url", "full_text_xml_url",
            ],
        }
        resp = requests.get(FR_API, params=params, timeout=30)
        resp.raise_for_status()
        data = resp.json()
        results = data.get("results", [])
        if not results:
            break
        records.extend(results)
        total_pages = data.get("total_pages", 1)
        print(f"  Page {page}/{total_pages}: {len(results)} documents fetched")
        if page >= total_pages:
            break
        page += 1

    return records

# ---------------------------------------------------------------------------
# Action-type classifier: OTSC / ISO / Final Order / Voluntary Surrender
# ---------------------------------------------------------------------------
OTSC_PATTERN   = re.compile(r"order\s+to\s+show\s+cause", re.I)
ISO_PATTERN    = re.compile(r"immediate\s+suspension", re.I)
FINAL_PATTERN  = re.compile(r"final\s+order", re.I)
SURRENDER_PATTERN = re.compile(r"voluntary\s+surrender|surrender.*registration", re.I)

def classify_action(title, abstract=""):
    text = f"{title} {abstract}"
    if ISO_PATTERN.search(text):
        return "ISO"
    if FINAL_PATTERN.search(text):
        return "Final Order"
    if SURRENDER_PATTERN.search(text):
        return "Voluntary Surrender"
    if OTSC_PATTERN.search(text):
        return "OTSC"
    return "Other"

# ---------------------------------------------------------------------------
# Registrant-type classifier
# ---------------------------------------------------------------------------
PHARMACY_PATTERN     = re.compile(r"pharmacy|pharmacist|drug\s+store", re.I)
PHYSICIAN_PATTERN    = re.compile(r"physician|doctor|\bM\.?D\.?\b|practitioner|dentist|veterinarian", re.I)
DISTRIBUTOR_PATTERN  = re.compile(r"distributor|wholesale|distribution\s+center|McKesson|Cardinal\s+Health|AmerisourceBergen|Walgreen|CVS", re.I)
HOSPITAL_PATTERN     = re.compile(r"hospital|health\s+system|medical\s+center|clinic", re.I)

def classify_registrant(title, abstract=""):
    text = f"{title} {abstract}"
    if DISTRIBUTOR_PATTERN.search(text):
        return "Distributor/Chain"
    if PHARMACY_PATTERN.search(text):
        return "Pharmacy"
    if HOSPITAL_PATTERN.search(text):
        return "Hospital/Clinic"
    if PHYSICIAN_PATTERN.search(text):
        return "Physician/Practitioner"
    return "Other/Unknown"

# ---------------------------------------------------------------------------
# Major chain/hospital system flag
# ---------------------------------------------------------------------------
MAJOR_ENTITIES = re.compile(
    r"Walgreen|CVS|Rite\s+Aid|McKesson|Cardinal\s+Health|AmerisourceBergen|"
    r"Walmart|Kroger|Costco|HCA|Tenet|Ascension|Mayo|Cleveland\s+Clinic|"
    r"Johns\s+Hopkins|Kaiser|Express\s+Scripts|OptumRx",
    re.I,
)

def is_major_entity(title, abstract=""):
    return bool(MAJOR_ENTITIES.search(f"{title} {abstract}"))

# ---------------------------------------------------------------------------
# State extractor: looks for two-letter state abbreviations in document text
# ---------------------------------------------------------------------------
US_STATE_ABBREVS = {
    "AL","AK","AZ","AR","CA","CO","CT","DE","FL","GA","HI","ID","IL","IN",
    "IA","KS","KY","LA","ME","MD","MA","MI","MN","MS","MO","MT","NE","NV",
    "NH","NJ","NM","NY","NC","ND","OH","OK","OR","PA","RI","SC","SD","TN",
    "TX","UT","VT","VA","WA","WV","WI","WY","DC",
}
STATE_PATTERN = re.compile(r"\b([A-Z]{2})\b")

def extract_state(title, abstract=""):
    text = f"{title} {abstract}"
    for match in STATE_PATTERN.finditer(text):
        candidate = match.group(1)
        if candidate in US_STATE_ABBREVS:
            return candidate
    return "Unknown"

# ---------------------------------------------------------------------------
# Main
# ---------------------------------------------------------------------------
print("Fetching DEA enforcement notices from the Federal Register API...")
docs = fetch_dea_enforcement_notices(per_page=100, max_pages=15)
print(f"\nTotal documents retrieved: {len(docs)}")

# Build structured records
rows = []
for doc in docs:
    title    = doc.get("title", "")
    abstract = doc.get("abstract") or ""
    pub_date = doc.get("publication_date", "")
    year     = int(pub_date[:4]) if pub_date and len(pub_date) >= 4 else None

    rows.append({
        "title":            title,
        "document_number":  doc.get("document_number", ""),
        "publication_date": pub_date,
        "year":             year,
        "action_type":      classify_action(title, abstract),
        "registrant_type":  classify_registrant(title, abstract),
        "state":            extract_state(title, abstract),
        "major_entity":     is_major_entity(title, abstract),
        "url":              doc.get("html_url", ""),
    })

df = pd.DataFrame(rows)

# ---------------------------------------------------------------------------
# Filter to 2018-2024
# ---------------------------------------------------------------------------
df_period = df[df["year"].between(2018, 2024)].copy()
print(f"\nDocuments in 2018-2024 window: {len(df_period)}")

# ---------------------------------------------------------------------------
# Actions by year
# ---------------------------------------------------------------------------
by_year = (
    df_period.groupby(["year", "action_type"])
    .size()
    .unstack(fill_value=0)
    .sort_index()
)

print("\n=== DEA Enforcement Actions by Year (2018-2024) ===")
print(f"  {'Year':<6}", end="")
for col in by_year.columns:
    print(f"  {col:<22}", end="")
print()
print("  " + "-" * (6 + 24 * len(by_year.columns)))
for year, row in by_year.iterrows():
    print(f"  {int(year):<6}", end="")
    for col in by_year.columns:
        print(f"  {row[col]:<22}", end="")
    print()

# ---------------------------------------------------------------------------
# States with highest revocation/suspension counts
# ---------------------------------------------------------------------------
state_counts = (
    df_period[df_period["action_type"].isin(["Final Order", "ISO", "OTSC"])]
    .groupby("state")
    .size()
    .sort_values(ascending=False)
    .head(15)
    .reset_index()
)
state_counts.columns = ["state", "enforcement_actions"]

print("\n=== Top 15 States by DEA Enforcement Actions (2018-2024) ===")
print(f"  {\'State\':<8}  {\'Actions\':<10}")
print("  " + "-" * 22)
for _, row in state_counts.iterrows():
    print(f"  {row[\'state\']:<8}  {row[\'enforcement_actions\']:<10}")

# ---------------------------------------------------------------------------
# Major chains / hospital systems flagged
# ---------------------------------------------------------------------------
major = df_period[df_period["major_entity"] == True][
    ["title", "year", "action_type", "url"]
].copy()

print(f"\n=== Major Chain / Hospital System Enforcement Actions ({len(major)} found) ===")
if major.empty:
    print("  None identified in this retrieval window.")
else:
    for _, row in major.iterrows():
        print(f"  [{row[\'year\']}] {row[\'action_type\']} | {row[\'title\'][:80]}")
        print(f"         {row[\'url\']}")

# ---------------------------------------------------------------------------
# Registrant-type breakdown
# ---------------------------------------------------------------------------
reg_type_counts = (
    df_period.groupby("registrant_type")
    .size()
    .sort_values(ascending=False)
    .reset_index()
)
reg_type_counts.columns = ["registrant_type", "count"]

print("\n=== Enforcement Actions by Registrant Type (2018-2024) ===")
for _, row in reg_type_counts.iterrows():
    print(f"  {row[\'registrant_type\']:<30}  {row[\'count\']}")

The Federal Register API returns documents in reverse chronological order by default. The per_page parameter accepts values up to 1,000, which reduces the number of API calls needed for bulk retrieval. For comprehensive enforcement history going back to 2000, increase max_pagesaccordingly. The action-type and registrant-type classifiers above use regular-expression pattern matching against document titles and abstracts; full-text parsing of the XML body would yield higher classification accuracy, particularly for distinguishing OTSCs from final orders in edge cases where both terms appear in the abstract. The state extractor is a heuristic; for systematic geographic analysis, cross-referencing the DEA number against the DEA's registrant address is more reliable than extracting state abbreviations from free text.

Limitations and research notes

The Federal Register enforcement record has several systematic gaps that affect quantitative analysis. Voluntary surrenders, as noted above, do not generate Federal Register publication in the same format as formal enforcement actions; they appear only if DEA publishes a notice confirming the surrender, which is not always the case. The surrender of a DEA number under investigation is a common outcome — practitioners facing potential revocation frequently surrender to avoid the formal record — but the aggregate count of surrenders is not publicly reported in a structured format. DEA's annual Diversion Control statistics include surrender totals, but these are aggregate figures without registrant-level detail.

State medical board proceedings that result in license revocation or suspension also trigger DEA registration loss under 21 U.S.C. §824(a)(3), but the Federal Register record of these actions is uneven. DEA sometimes publishes a notice confirming that it is revoking a registration based on a state board action; in other cases, the state board action effectively terminates DEA registration as a matter of law without a separate federal proceeding and Federal Register publication. The result is that the Federal Register enforcement record understates the total number of practitioners who lose controlled substance authority in any given year.

The DEA registration number format enables linkage of enforcement records to other federal datasets. The ARCOS transaction data, the DEA Diversion Control registration lookup, the CMS Medicare Part D prescriber data, and the CMS Open Payments database all use the DEA number as a practitioner identifier. Cross-referencing a revoked DEA number against Medicare Part D prescribing data can establish the volume of controlled substance prescriptions a revoked practitioner wrote before losing registration authority — a useful metric for characterizing the severity of the underlying conduct. CMS Part D prescriber data is covered in the related articles on Medicare Part D drug spending.