L’interface est localisée ; les présentations et documents des Skills sont traduits automatiquement par IA dans votre langue. Le code, les commandes, chemins et termes techniques restent en version originale, et le document de l’auteur fait foi.
Explorer les Skills
Développement · SKILL.md

langcare-opioid-risk

langcare-opioid-risk

Performs opioid risk assessment by calculating morphine milligram equivalents (MME), applying Opioid Risk Tool scoring, and evaluating CDC 2022 guideline thresholds. Flags concurrent benzodiazepines, missing naloxone, and high-MME prescriptions. Use when asked to assess opioid risk, calculate MME, check opioid safety, or review opioid prescriptions.

Independent GitHub snapshot · Automated classification · Original documentation governs

Documentation de la Skill

GitHub

name: langcare-opioid-risk description: > Performs opioid risk assessment by calculating morphine milligram equivalents (MME), applying Opioid Risk Tool scoring, and evaluating CDC 2022 guideline thresholds. Flags concurrent benzodiazepines, missing naloxone, and high-MME prescriptions. Use when asked to assess opioid risk, calculate MME, check opioid safety, or review opioid prescriptions.

Opioid Risk Assessment

When to Use This Skill

Use when a clinician needs to evaluate opioid prescribing safety including total daily MME calculation, concurrent CNS depressant screening, ORT scoring, and CDC guideline compliance.

Clinical Workflow

  1. Use fhir_read to retrieve Patient demographics (age, gender for ORT scoring)
  2. Use fhir_search to pull all active MedicationRequest resources; classify as opioids, benzodiazepines, gabapentinoids, muscle relaxants, Z-drugs, naloxone
  3. Calculate total daily MME for each opioid using dose x frequency x MME conversion factor (see references/mme-conversion-table.md)
  4. Evaluate against CDC 2022 thresholds: <20 lower risk, 20-49 moderate, 50-89 caution, >=90 high risk
  5. Use fhir_search to pull Condition resources for ORT scoring: substance use history, psychiatric diagnoses
  6. Apply Opioid Risk Tool scoring (gender-specific criteria)
  7. Use fhir_search to check for recent urine drug screen (LOINC 19295-5) and pain assessments (LOINC 72514-3)
  8. Assess naloxone co-prescription status; recommend if MME >= 50 or concurrent CNS depressants

FHIR Resources

  • MedicationRequest -- Active opioid and concurrent prescriptions with dosing details
  • MedicationDispense -- Fill history for pattern analysis
  • Patient -- Age and gender for ORT scoring
  • Condition -- Psychiatric diagnoses, substance use history for ORT
  • Observation -- Urine drug screen results, pain scores

FHIR Query Examples

Pull Active Medications

fhir_search(resourceType="MedicationRequest", queryParams="patient=[patient-id]&status=active&_count=100")

Pull Substance Use History

fhir_search(resourceType="Condition", queryParams="patient=[patient-id]&_count=100")

Pull Urine Drug Screen

fhir_search(resourceType="Observation", queryParams="patient=[patient-id]&code=19295-5,3426-4&_sort=-date&_count=5")

Clinical Guidelines

  • CDC 2022 Clinical Practice Guideline for Prescribing Opioids for Pain
  • FDA REMS for extended-release/long-acting opioids
  • PDMP (Prescription Drug Monitoring Program) check recommended before each opioid prescription

Interpretation Guide

  • Total daily MME thresholds: <50 standard monitoring, 50-89 prescribe naloxone and reassess, >=90 avoid or taper
  • Concurrent benzodiazepine + opioid = FDA Boxed Warning for respiratory depression
  • ORT scores: 0-3 low risk, 4-7 moderate risk, >=8 high risk (scores differ by gender)
  • UDS consistency: presence of prescribed opioid confirms adherence; absence suggests diversion; non-prescribed substances flag polysubstance risk
  • Include CAGE-AID screening questions for clinician to administer

Safety

  • Never fabricate clinical data -- only report what FHIR returns
  • Flag critical/abnormal values immediately
  • Scope all FHIR queries to the authenticated patient
  • Use standard terminology (LOINC, SNOMED CT, RxNorm, ICD-10)
  • Present data in clinician-friendly format with reference ranges
Skills enregistrées