Opioid Use Disorder (OUD) Topic Module¶

The oud module provides medication-assisted treatment (MAT) detection, care setting classification, and co-occurring condition identification for opioid use disorder research in Medicaid claims data.

Module Structure¶

Submodule	Purpose
`medication_and_behavioral_health`	MAT flags (procedure codes and NDC), behavioral health, concurrent prescriptions
`care_settings`	Care setting classification (FQHC, outpatient, physician office, etc.)
`cooccurring_conditions`	Mental health and substance use disorder co-occurrence

Medication-Assisted Treatment (MAT)¶

Each MAT medication has both a procedure-code flag (for IP/OT claims) and an NDC flag (for RX claims). All flag_proc_* functions accept a cms_format parameter (default "TAF").

Buprenorphine¶

from medicaid_utils.topics.oud import medication_and_behavioral_health as mat

# Procedure code flagging (J0571) on IP/OT claims
df_ip = mat.flag_proc_buprenorphine(ip.dct_files["base"], cms_format="TAF")

# NDC flagging on RX claims
df_rx = mat.flag_rx_buprenorphine(rx.dct_files["base"])

Buprenorphine/Naloxone¶

# Procedure codes (J0572–J0575)
df_ip = mat.flag_proc_buprenorphine_naloxone(ip.dct_files["base"], cms_format="TAF")

# NDC codes
df_rx = mat.flag_rx_buprenorphine_naloxone(rx.dct_files["base"])

Injectable Naltrexone¶

# Procedure code (J2315)
df_ip = mat.flag_proc_injectable_naltrexone(ip.dct_files["base"], cms_format="TAF")

# NDC codes (oral naltrexone)
df_rx = mat.flag_rx_oral_naltrexone(rx.dct_files["base"])

Methadone¶

# Procedure code (H0020)
df_ip = mat.flag_proc_methadone(ip.dct_files["base"], cms_format="TAF")

# NDC codes (splits ≤30 mg and >30 mg dosages)
df_rx = mat.flag_rx_methadone(rx.dct_files["base"])

Behavioral Health Services¶

flag_proc_behavioral_health_trtmt() flags claims with CPT/HCPCS behavioral health procedure codes:

df_bh = mat.flag_proc_behavioral_health_trtmt(ot.dct_files["base"], cms_format="TAF")

Concurrent Benzodiazepine and Opioid Prescriptions¶

flag_rx_benzos_opioids() identifies benzodiazepine and opioid prescriptions. Creates three columns: rx_benzodiazepines, rx_opioids, and rx_benzo_and_opioid (1 when both are present):

df_rx = mat.flag_rx_benzos_opioids(rx.dct_files["base"])

Care Setting Classification¶

flag_care_settings() creates claim-level flags for care settings using place of service and taxonomy codes:

from medicaid_utils.topics.oud import care_settings

df_ot = care_settings.flag_care_settings(ot.dct_files["base"])

Flags include:

FQHC — Federally Qualified Health Centers
Outpatient hospital — Hospital outpatient departments
Physician office — Office-based physician visits
Behavioral health centers — Specialized behavioral health facilities
Hospital — Inpatient hospital settings
Office-based — General office-based settings

Note

TAF claims use POS_CD (Place of Service Code) and BLG_PRVDR_TXNMY_CD (Billing Provider Taxonomy Code) for classification. MAX claims use TOS_CD (Type of Service Code).

Co-Occurring Conditions¶

Mental health¶

flag_cooccurring_mental_health_claims() flags claims with mental health diagnosis codes (based on MODRN 2021 definitions):

from medicaid_utils.topics.oud import cooccurring_conditions

df_mh = cooccurring_conditions.flag_cooccurring_mental_health_claims(
    ot.dct_files["base"], cms_format="TAF"
)

Substance use disorders¶

flag_cooccurring_sud_claims() flags claims with substance use disorder diagnosis codes, excluding OUD and tobacco:

df_sud = cooccurring_conditions.flag_cooccurring_sud_claims(
    ot.dct_files["base"], cms_format="TAF"
)

Example: Full OUD Analysis Pipeline¶

from medicaid_utils.preprocessing import taf_ip, taf_ot, taf_rx, taf_ps
from medicaid_utils.topics.oud import (
    medication_and_behavioral_health as mat,
    care_settings,
    cooccurring_conditions,
)

# Load claims
ip = taf_ip.TAFIP(year=2019, state="AL", data_root="/data/cms")
ot = taf_ot.TAFOT(year=2019, state="AL", data_root="/data/cms")
rx = taf_rx.TAFRX(year=2019, state="AL", data_root="/data/cms")

# Flag MAT in IP/OT claims
df_ip = mat.flag_proc_buprenorphine(ip.dct_files["base"])
df_ot = mat.flag_proc_methadone(ot.dct_files["base"])

# Flag MAT in RX claims
df_rx = mat.flag_rx_buprenorphine(rx.dct_files["base"])
df_rx = mat.flag_rx_methadone(df_rx)

# Classify care settings
df_ot = care_settings.flag_care_settings(df_ot)

# Flag co-occurring conditions
df_ot = cooccurring_conditions.flag_cooccurring_mental_health_claims(df_ot)
df_ot = cooccurring_conditions.flag_cooccurring_sud_claims(df_ot)