Clinical Commander

All dossiers
geriatrics.deprescribing.core.v1

Deprescribing in older adults

general_internal_medicinechronicgeriatricadultoutpatientinpatient
Start encounter →Print handoutPRODUCTION

Action arm of geriatrics.frailty-polypharmacy.core.v1 — operationalizes the Scott/Reeve 5-step deprescribing process with class-specific tapers; sibling owns frailty stratification + START proactive prescribing. No dedicated deprescribing manifest or problem-package folder on disk — manifest repointed to prisma/seed/manifests/symptom.falls.v1.ts (shared FRID/structured-medication-review spine). Dedicated manifest tracked in design brief Open gaps. Regimen axis is entirely non_pharm: every entry is a deprescribing ACTION, taper, or lower-risk substitute described as a process — NO rxcui anywhere by design (deprescribing removes rather than adds drugs). calc.acb_scale / calc.clinical_frailty_scale / calc.frail_scale and workup.stopp_start_v3 / workup.beers_screen / workup.cga reused from the geriatrics registry whitelist; calc.qsofa / calc.news2 optional for post-deprescribing destabilization detection. RxNav/DailyMed not applicable (no drugs prescribed); Bayesian LR layer deferred — see brief Open gaps.

Entry points (5)

  • problem_list
    >=5 chronic medications — polypharmacy / medication-optimization trigger (O'Mahony Age Ageing 2023)
    polypharmacy
  • medication
    Beers-listed / STOPP-flagged / high-anticholinergic medication on review (AGS Beers 2023)
    potentially_inappropriate_medication
  • demographic
    Limited life expectancy / advanced frailty / goals-of-care shift — time-to-benefit reassessment (Scott JAMA Intern Med 2015)
    limited_life_expectancy
  • history
    Suspected adverse drug event, prescribing cascade, or drug-related fall/delirium (Reeve Br J Clin Pharmacol 2014)
    adverse_drug_event
  • history
    Hospital admission / discharge / SNF transfer — transitions-of-care reconciliation prompt (NICE NG5 medicines optimisation)
    transition_of_care

Required inputs (12)

  • agerequired
    demographic • used at FRAME
    Deprescribing prioritization is age- and life-expectancy-anchored; Beers/STOPP age-stratified (AGS Beers 2023)
  • current_medsrequired
    medication • used at CONTEXT
    Full reconciled list incl OTC/supplements/PRN/topicals is the substrate for every deprescribing decision (Scott JAMA Intern Med 2015)
  • otc_and_supplementsrequired
    medication • used at CONTEXT
    OTC NSAIDs, antihistamines, PPIs, herbals/supplements are commonly omitted yet high-yield deprescribing targets (Reeve Br J Clin Pharmacol 2014)
  • indication_per_drugrequired
    history • used at CONTEXT
    Each drug must be mapped to an active indication; no-longer-indicated / indication-lapsed drugs are first-priority targets (Scott JAMA Intern Med 2015)
  • comorbidity_listrequired
    history • used at CONTEXT
    Active essential indications protect a drug from deprescribing; comorbidity burden shapes benefit-risk (O'Mahony Age Ageing 2023)
  • goals_of_carerequired
    history • used at DIFFERENTIAL
    Patient/caregiver goals + life-expectancy estimate drive time-to-benefit vs time-to-harm weighting and palliative deprescribing (Scott JAMA Intern Med 2015)
  • functional_statusrequired
    history • used at RISK_STRATIFICATION
    Frailty/ADL trajectory modifies benefit horizon (statin/antihyperglycemic/antihypertensive targets relax) (O'Mahony Age Ageing 2023)
  • withdrawal_risk_drugsrequired
    history • used at RED_FLAGS
    Flags drugs that must NOT be stopped abruptly (benzodiazepine, opioid, beta-blocker, clonidine, corticosteroid, SSRI/SNRI, gabapentinoid, levodopa, PPI rebound) (Reeve Br J Clin Pharmacol 2014)
  • cognitive_status
    history • used at INITIAL_WORKUP
    Anticholinergic/sedative burden vs cognition; affects self-management capacity and caregiver involvement (Boustani Aging Ment Health 2008)
  • orthostatic_bp
    vital • used at INITIAL_WORKUP
    Orthostasis / supine-standing delta is the key antihypertensive-deprescribing signal (O'Mahony Age Ageing 2023)
  • creatininerequired
    lab • used at INITIAL_WORKUP
    eGFR re-checks renally-cleared drug appropriateness and is the primary post-deprescribing safety lab (O'Mahony Age Ageing 2023)
  • hba1c
    lab • used at RISK_STRATIFICATION
    Relax glycemic target in frailty/limited life expectancy; over-tight HbA1c on sulfonylurea/insulin = hypoglycemia harm (AGS Beers 2023)

12-phase flow (12)

  1. 1FRAME
    Scope: older adult on polypharmacy where systematic medication optimization is indicated; deprescribing is iterative and patient-centred, NOT nihilistic withdrawal of all drugs (Scott JAMA Intern Med 2015)
    inputs: age
    advance: Patient confirmed as a structured-medication-review / deprescribing candidate (NICE NG5 medicines optimisation)
  2. 2ENTRY
    Capture trigger: polypharmacy threshold, PIM on review, life-expectancy/goals shift, adverse drug event/prescribing cascade, or care transition (Reeve Br J Clin Pharmacol 2014)
    inputs: age, current_meds
    advance: Deprescribing trigger documented (Scott JAMA Intern Med 2015)
  3. 3CONTEXT
    Step 1 — comprehensive medication reconciliation: every prescribed + OTC + supplement + PRN + topical + inhaled drug, dose, duration, prescriber, adherence, and the ACTIVE indication for each (best possible medication history) (Scott JAMA Intern Med 2015)
    inputs: current_meds, otc_and_supplements, indication_per_drug, comorbidity_list
    advance: Best-possible medication history complete; each drug mapped to an indication or flagged indication-lapsed (Scott JAMA Intern Med 2015)
  4. 4RED_FLAGS
    Identify high-withdrawal-risk drugs that must NOT be stopped abruptly (benzodiazepine, opioid, beta-blocker, clonidine, corticosteroid, SSRI/SNRI, gabapentinoid, levodopa, PPI rebound) and any drug treating an active essential indication that must NOT be deprescribed; screen for acute drug-related emergency (Reeve Br J Clin Pharmacol 2014)
    inputs: withdrawal_risk_drugs, current_meds
    actions: workup.delirium
    advance: Withdrawal-risk and protected-indication drugs flagged; no abrupt-stop hazard pending (Reeve Br J Clin Pharmacol 2014)
  5. 5INITIAL_WORKUP
    Step 2 baseline + safety labs: CMP/eGFR, CBC, electrolytes; CGA domains (cognition, function, falls, mood, nutrition); orthostatic BP; anticholinergic inventory (O'Mahony Age Ageing 2023)
    inputs: creatinine, cognitive_status, orthostatic_bp
    actions: workup.stopp_start_v3, workup.beers_screen, workup.cga, workup.falls, panel.cmp, panel.metabolic, panel.cbc
    advance: Baseline safety labs + CGA + explicit-tool screens complete (O'Mahony Age Ageing 2023)
  6. 6BRANCHING_WORKUP
    Step 2 — PIM identification with explicit tools: STOPP/START v3 violations + omissions, AGS Beers 2023 PIMs, drug-disease and drug-drug interactions, prescribing cascades, anticholinergic burden inventory (O'Mahony Age Ageing 2023)
    inputs: current_meds, comorbidity_list, creatinine
    actions: workup.stopp_start_v3, workup.beers_screen
    advance: PIM / no-longer-indicated / harm>benefit candidate list catalogued with explicit-tool citations (AGS Beers 2023)
  7. 7DIFFERENTIAL
    Step 3 — assess each candidate drug's ongoing benefit vs risk in THIS patient: active vs lapsed indication, time-to-benefit vs estimated life expectancy, time-to-harm, frailty, goals of care; classify each as continue / dose-reduce / substitute-lower-risk / taper-and-stop / stop-now (Scott JAMA Intern Med 2015)
    inputs: goals_of_care, indication_per_drug
    advance: Per-drug benefit-risk-in-context decision assigned with rationale documented (Scott JAMA Intern Med 2015)
  8. 8RISK_STRATIFICATION
    Step 4 — prioritize: highest harm / lowest benefit first; anticholinergic burden (ACB) and frailty severity rank the queue; relax HbA1c/BP targets in frailty/limited life expectancy; align order with patient/caregiver preferences (shared decision-making is central) (Boustani Aging Ment Health 2008)
    inputs: functional_status, hba1c
    actions: calc.acb_scale, calc.clinical_frailty_scale, calc.frail_scale
    advance: Ranked deprescribing list agreed with patient/caregiver; one-change-at-a-time order set (Scott JAMA Intern Med 2015)
  9. 9TREATMENT
    Step 5 — plan + implement class-specific TAPER with monitoring + contingency: PPI step-down/alternate-day/on-demand (rebound acid); benzodiazepine/Z-drug ~25%/2wk taper + CBT-I (withdrawal/falls); antipsychotic-in-dementia withdrawal when behaviourally stable (boxed mortality); anticholinergic substitution to lower burden; opioid slow taper; relax antihyperglycemic target (sulfonylurea/insulin hypoglycemia); individualize antihypertensive (orthostasis/falls vs CV benefit); stop statin in primary prevention with limited life expectancy; bisphosphonate drug holiday after 3-5 yr; antidepressant taper (discontinuation syndrome); stop unnecessary vitamins/supplements; stop DAPT beyond indicated duration — one change at a time, document rationale (Scott JAMA Intern Med 2015)
    inputs: current_meds, goals_of_care
    advance: Class-specific taper schedule + monitoring + re-prescribe contingency written and initiated (one change at a time) (Scott JAMA Intern Med 2015)
  10. 10DISPOSITION
    Outpatient structured medication review with pharmacist collaboration; inpatient = transitions-of-care reconciliation and a deprescribing plan handed to the receiving clinician; palliative-care referral when goals shift to comfort (NICE NG5 medicines optimisation)
    inputs: goals_of_care
    advance: Care setting + interprofessional/pharmacist + follow-up ownership confirmed; deprescribing plan communicated at the transition (NICE NG5 medicines optimisation)
  11. 11MONITORING
    Per-class withdrawal / rebound / symptom-recurrence surveillance after each single change: PPI rebound hyperacidity 2-4 wk, benzodiazepine/opioid withdrawal, beta-blocker rebound HTN/angina, antipsychotic BPSD recurrence, antidepressant discontinuation syndrome, glycemic/BP drift; re-prescribe if the original indication re-emerges (deprescribing is iterative) (Reeve Br J Clin Pharmacol 2014)
    inputs: current_meds, creatinine
    actions: calc.clinical_frailty_scale
    advance: Monitoring cadence set per stopped class; no uncontrolled withdrawal/rebound; re-prescribe trigger documented (Scott JAMA Intern Med 2015)
  12. 12FOLLOWUP
    Iterative re-review at every visit and care transition; update advance care planning / goals of care; palliative deprescribing at end of life; caregiver communication tools + updated medication list; schedule next structured medication review (NICE NG5 medicines optimisation)
    inputs: goals_of_care
    advance: Re-prescribe-or-continue decisions logged; next medication review scheduled; goals-of-care/ACP refreshed (NICE NG5 medicines optimisation)