GC079 (supp-1)stopp-start Criteria For Potentially Inappropriate Prescribing In Older 2023
The 2023 STOPP/START criteria are evidence-based screening tools designed to identify potentially inappropriate medications (STOPP) and potential prescribing omissions (START) in adults aged 65 years and older to optimize pharmacotherapy and reduce adverse drug events.
STOPP/START Criteria for Potentially Inappropriate Prescribing in Older People – Version 3 (2023)
Lecture Map
Older adults with multimorbidity and polypharmacy are at extremely high risk of adverse drug reactions (ADRs) and adverse drug events (ADEs). The STOPP/START framework is a validated, systems-based tool for medication review that tackles two sides of the same coin:
- STOPP (Screening Tool of Older Persons' Prescriptions) — identifies Potentially Inappropriate Medications (PIMs) that should be stopped or avoided.
- START (Screening Tool to Alert doctors to Right Treatment) — identifies Potential Prescribing Omissions (PPOs) where beneficial medications are inappropriately omitted.
The combination of deprescribing PIMs and introducing PPOs is called "represcribing" — a concept central to optimising pharmacotherapy in the elderly. [1]
- Understand why explicit prescribing criteria exist and the difference between STOPP/START and Beers criteria.
- Know the structure (physiological-systems-based) and scope of STOPP/START version 3.
- Be able to apply high-yield STOPP criteria (drugs to STOP) and START criteria (drugs to START) to clinical vignettes.
- Understand how STOPP/START fits into Comprehensive Geriatric Assessment (CGA) and medication review.
Core Concepts and Mechanisms
"Detection and avoidance of PIMs is a fundamental aspect of routine medication review and is considered beneficial in older people with multimorbidity and associated polypharmacy undergoing comprehensive geriatric assessment." [1]
| Factor | Explanation |
|---|---|
| Pharmacokinetic changes | ↓ renal clearance, ↓ hepatic metabolism (↓ first-pass), ↑ body fat (↑ Vd of lipophilic drugs), ↓ albumin (↑ free drug fraction) |
| Pharmacodynamic changes | ↑ sensitivity to CNS depressants, anticholinergics, cardiovascular drugs |
| Multimorbidity | Multiple diseases → multiple drugs → drug–drug and drug–disease interactions |
| Polypharmacy | ≥5 regular medications; risk of ADR rises exponentially with each additional drug |
| Frailty | Reduced physiological reserve means even "mild" ADRs (e.g., orthostatic hypotension) can cause catastrophic outcomes (falls → hip fracture → death) |
| Undertreatment | Equally dangerous — omitting anticoagulation in AF → preventable stroke; omitting osteoporosis Rx → preventable fracture |
| Type | Example | Strength | Weakness |
|---|---|---|---|
| Explicit (criterion-based) | STOPP/START, Beers | Reproducible, can be computerized, easy to apply | Does not consider individual patient context |
| Implicit (judgment-based) | Medication Appropriateness Index (MAI) | Patient-centred, nuanced | Time-consuming, depends on clinician expertise |
Key Definitions
- PIM (Potentially Inappropriate Medication) — a drug that poses more risk than benefit in a given clinical context. Detected by STOPP criteria.
- PPO (Potential Prescribing Omission) — failure to prescribe a beneficial medication despite a clear indication. Detected by START criteria.
- "Represcribing" — the combined process of deprescribing PIMs and initiating PPOs to optimise the medication list. [1]
STOPP/START v3 — Structure and Key Numbers
"STOPP/START version 3, with 190 criteria, is significantly larger than version 2 (114 criteria), reflecting the expansion of the pharmacopeia and clinical trials evidence base relevant to older people." [1]
| Version | Year | STOPP | START | Total |
|---|---|---|---|---|
| v1 | 2008 | 65 | 22 | 87 |
| v2 | 2015 | 80 | 34 | 114 |
| v3 | 2023 | 133 | 57 | 190 |
The 66.7% increase from v2 to v3 reflects new evidence on drugs like SGLT-2 inhibitors, DOACs, and newer antidiabetics. [1]
- 11 academic physicians from 8 European countries rated criteria on a 5-point Likert scale via SurveyMonkey®.
- Acceptance rule: median Likert ≤ 2.0 AND ≥ 75% agreement.
- 4 Delphi rounds → final 190 criteria validated. [1]
| System | STOPP (% increase) | START (% increase) |
|---|---|---|
| Cardiovascular | 21 (+61.5%) | 11 (+37.5%) |
| Coagulation | 16 (+23.1%) | 2 (new section) |
| CNS | 25 (+78.6%) | 7 (+16.7%) |
| Renal | 10 (+66.7%) | 4 (new section) |
| GI | 8 (+100%) | 7 (+350%) |
| Respiratory | 4 (0%) | 3 (0%) |
| Musculoskeletal | 9 (0%) | 9 (+28.6%) |
| Urogenital | 8 (+400%) | 5 (+66.7%) |
| Endocrine | 10 (+66.7%) | 1 (0%) |
| Falls risk drugs | 12 (+300%) | N/A |
| Analgesics | 6 (+100%) | 3 (+50%) |
| Vaccines | N/A | 4 (+100%) |
STOPP Criteria v3 — Slide-by-Slide High-Yield Content
The full list is in Appendix 1 (supp-2). Below are the exam-highest-yield criteria organised by system. [2]
"A1. Any drug prescribed without a clinical indication." [2]
"A2. Any drug prescribed beyond the recommended duration, where treatment duration is well defined." [2]
"A3. Any duplicate drug class prescription for daily regular use (e.g., two concurrent NSAIDs, SSRIs, loop diuretics, ACE inhibitors, anticoagulants, antipsychotics, opioid analgesics)." [2]
Why this matters: These three criteria alone catch a large proportion of inappropriate prescribing. Before looking at any specific drug, always ask: (1) Is there an indication? (2) Is the duration appropriate? (3) Is there duplication?
| Criterion | What to STOP | Why |
|---|---|---|
| B1 | Digoxin for HF with normal systolic function | No clear evidence of benefit [2] |
| B2 | Verapamil or diltiazem with NYHA III/IV HF | May worsen HFrEF [2] |
| B3 | Beta-blocker + verapamil or diltiazem | Risk of heart block [2] |
| B4 | Rate-limiting drugs (BB, verapamil, diltiazem, digoxin) with bradycardia < 50, type II/complete heart block | Risk of complete heart block, asystole [2] |
| B5 | BB as monotherapy for uncomplicated HTN | No firm evidence of efficacy as first-line for HTN alone [2] |
| B6 | Amiodarone as first-line for SVT | Higher risk of major side-effects than BB/digoxin/verapamil/diltiazem [2] |
| B7 | Loop diuretic as first-line for HTN | Safer, more effective alternatives available (unless concurrent HF) [2] |
| B8 | Loop diuretic for dependent ankle oedema without HF/liver failure/nephrotic syndrome/renal failure | Leg elevation and/or compression hosiery usually more appropriate [2] |
Exam Trap: Loop Diuretics for Ankle Oedema
A very common scenario: an elderly patient with bilateral ankle swelling is given furosemide. If there is no evidence of HF, liver failure, nephrotic syndrome, or renal failure, this is inappropriate. The correct approach is leg elevation and compression stockings. This is a favourite exam question. [2]
Key themes: anticoagulation in AF, concurrent antiplatelets + anticoagulants, bleeding risk.
- Anticoagulants with concurrent significant bleeding risk (e.g., active GI bleeding, recent ICH) — STOPP.
- Aspirin + clopidogrel + anticoagulant (triple therapy) without clear indication — STOPP (massive bleeding risk).
- Aspirin for primary CVD prevention — both STOPP/START v3 and Beers 2023 now flag this. [1][3]
| Criterion Theme | What to STOP | Why |
|---|---|---|
| Anticholinergics | In patients with delirium or dementia | Worsen cognitive impairment, cause confusion, urinary retention, constipation |
| Benzodiazepines | Long-term use (> 4 weeks); in those at risk of falls | Sedation → falls → fractures; cognitive impairment; dependence |
| Antipsychotics | As hypnotics; long-term without review; in Parkinsonism/DLB | Extrapyramidal effects, increased mortality in dementia, worsens Parkinsonism |
| TCAs | In patients with dementia, glaucoma, cardiac conduction abnormalities, prostatism, constipation | Highly anticholinergic; cardiac toxicity; orthostatic hypotension |
| SSRIs | If concurrent hyponatraemia (Na < 130) | SSRIs cause SIADH → worsen hyponatraemia |
| Neuroleptics in dementia | Long-term without clear indication | Increased stroke risk, mortality (NNH ~100 for death over 12 weeks) |
High Yield: Anticholinergic Burden
The concept of cumulative anticholinergic burden is critical. Even drugs not "obviously" anticholinergic (e.g., furosemide, digoxin, ranitidine, paroxetine) contribute. In older adults, total anticholinergic burden is associated with increased risk of falls, delirium, and dementia. Both STOPP/START v3 and Beers 2023 emphasise this. [1][3]
- NSAIDs with eGFR < 50 — risk of further renal deterioration.
- Metformin with eGFR < 30 — risk of lactic acidosis.
- Digoxin at full dose if eGFR < 30 — risk of toxicity due to reduced renal clearance.
- ACEI/ARB: monitor K+ and creatinine after initiation/dose change.
- PPIs beyond 8 weeks for uncomplicated peptic ulcer/erosive oesophagitis without ongoing indication — risk of C. difficile, hypomagnesaemia, osteoporotic fractures, renal disease. [2]
- Metoclopramide with Parkinsonism — D2 antagonist worsens Parkinsonism.
- Oral beta-blockers with history of asthma — risk of bronchospasm (note: cardioselective BB like bisoprolol may be acceptable in COPD but should be avoided in asthma).
- Benzodiazepines in respiratory failure — respiratory depression.
- NSAIDs long-term (> 3 months) for osteoarthritis — GI bleeding, renal, CV risk.
- Systemic corticosteroids (except short courses) for OA — side effects outweigh benefit.
- Colchicine long-term if eGFR < 10 — risk of toxicity.
- Anticholinergics (e.g., oxybutynin) in patients with dementia or chronic constipation — worsen cognition, worsen constipation.
- Alpha-blockers (e.g., tamsulosin) in those with postural hypotension — exacerbate falls risk.
- Sulfonylureas with long duration of action (e.g., glibenclamide/glyburide) — risk of prolonged hypoglycaemia, especially with renal impairment.
- Thiazolidinediones (pioglitazone) with HF — fluid retention.
- Insulin sliding scale as sole long-term regimen — not recommended.
This is a new major section in v3. [1]
Drugs that increase falls risk in older adults who have already fallen or are at high risk:
| Drug Class | Mechanism of Falls Risk |
|---|---|
| Benzodiazepines | Sedation, impaired balance, muscle relaxation |
| Neuroleptics/antipsychotics | Sedation, orthostatic hypotension, EPS |
| First-generation antihistamines | Anticholinergic + sedative |
| Vasodilators (alpha-blockers, nitrates) | Orthostatic hypotension |
| Opioids | Sedation, dizziness |
| Hypnotics (Z-drugs) | Sedation, impaired balance |
High Yield: STOPP Falls Criteria
In a patient who presents with recurrent falls, your medication review must specifically target FRIDs. The exam loves to ask: "Which medication should be stopped in this 82-year-old who has fallen 3 times in 6 months?" Look for benzodiazepines, anticholinergics, antipsychotics, opioids, and vasodilators. [1][2]
- Opioids without concurrent laxative — nearly universal constipation.
- Long-acting opioids without prior short-acting trial.
- Opioids as first-line for mild-moderate pain — paracetamol and topical agents first.
| Removed Criterion | Reason |
|---|---|
| STOPP C2: Aspirin + previous PUD without PPI | Considered redundant (captured by other criteria) |
| STOPP H7: COX-2 selective NSAIDs + CVD | Redundant with newer CV criteria |
| START B1: Regular inhaled β2-agonist/antimuscarinic for mild-moderate asthma/COPD | Clinical practice has evolved; ICS-containing regimens now standard |
START Criteria v3 — High-Yield Potential Prescribing Omissions
START criteria are designed to detect potential prescribing omissions (PPOs) which represent another critically important aspect of inappropriate prescribing, i.e., undertreatment or failure to prescribe appropriate medications despite clear and valid indications. [1]
| What to START | When | Why |
|---|---|---|
| Anticoagulation (DOAC preferred) in chronic AF | CHA₂DS₂-VASc ≥ 2 (♂) or ≥ 3 (♀) | Prevents ~67% of ischaemic strokes; DOACs preferred over warfarin in elderly [1] |
| Statin in documented atherosclerotic CVD | Unless end-of-life or very limited life expectancy | Secondary prevention reduces CV events |
| ACEI/ARB in systolic HF (HFrEF) | All patients with HFrEF | Mortality benefit |
| Beta-blocker in stable HFrEF | Bisoprolol, carvedilol, metoprolol succinate, nebivolol | Mortality benefit (CIBIS-II, COPERNICUS, MERIT-HF) |
| SGLT-2 inhibitor in HFrEF | New in v3: dapagliflozin/empagliflozin | DAPA-HF, EMPEROR-Reduced — mortality and hospitalisation benefit |
| Appropriate antihypertensive | For BP consistently > target | Stroke and CV event prevention |
- VTE prophylaxis when indicated (e.g., immobilised hospitalised elderly).
- Levodopa or dopamine agonist in idiopathic PD with functional impairment
- Antidepressant for moderate-severe depression
- Acetylcholinesterase inhibitor for mild-moderate Alzheimer's disease (donepezil, rivastigmine, galantamine)
Note: Memantine for moderate-severe AD was REJECTED by the Delphi panel for inclusion in START v3. [1]
- ACEI/ARB in diabetic nephropathy with proteinuria
- SGLT-2 inhibitor in CKD with proteinuria (new in v3)
- PPI for severe GORD or peptic stricture
- Fibre/laxatives for chronic constipation
- Vitamin D + calcium in osteoporosis
- Bisphosphonate/denosumab in osteoporosis if indicated by fracture risk
- Urate-lowering therapy after recurrent gout attacks
- Annual influenza vaccine
- Pneumococcal vaccine
- Herpes zoster vaccine (new in v3)
- COVID-19 vaccine (new in v3, reflecting pandemic evidence)
| Feature | STOPP/START v3 (2023) | AGS Beers 2023 |
|---|---|---|
| Origin | European (Ireland-led, pan-European Delphi) | American (AGS) |
| Structure | Physiological systems-based, separate STOPP & START | Drug/organ-system tables; includes drug-disease, drug-drug tables |
| Unique strength | Includes PPOs (START) — detects undertreatment | Includes drugs to use with caution (Table 4) and drug-drug interactions |
| Falls risk | Dedicated FRID section (12 criteria) | Falls mentioned in drug-disease table |
| Anticholinergic focus | Dedicated section; cumulative burden emphasised | Table 7 lists drugs with strong anticholinergic properties |
| Total criteria | 190 | ~100+ across 5 tables |
| Validation | Delphi consensus by European physicians | Delphi consensus by US interprofessional panel |
| Endorsed by | NICE, BNF, BGS, RCGP, Dutch guidelines | AGS, widely used in US policy |
| Overlap | Significant overlap for PIMs (e.g., benzodiazepines, anticholinergics, NSAIDs) | — |
| Key exam point | START criteria are unique to STOPP/START — Beers does NOT systematically address prescribing omissions | Beers includes drugs to "use with caution" category |
"Medication review" = a structured evaluation of a patient's medicines with the aim of optimizing medicines use and improving health outcomes, entailing detecting drug-related problems and recommending interventions (PCNE, 2018). [1]
STOPP/START criteria should be applied as part of:
- Comprehensive Geriatric Assessment (CGA)
- Transitions of care (admission, discharge, care-home entry)
- Annual medication review in primary care for polypharmacy patients
The Role of Clinical Decision Support Systems (CDSS)
"Given the growing number of criteria from version 1 (87) to version 3 (190), deployment using appropriate electronic applications is highly desirable." [1]
Key findings from trials:
- SENATOR trial: CDSS-based STOPP/START intervention; only ~15% of recommendations were implemented by attending physicians. [1]
- OPERAM trial: ~62% implementation rate; better but still suboptimal. [1]
- Earlier single-centre trial (O'Connor et al., 2016): Face-to-face physician-to-physician contact → 81% STOPP and 87% START implementation → significant reduction in hospital-acquired ADRs. [1]
Take-home message: Electronic tools detect PIMs/PPOs, but a trained clinician must interpret and implement recommendations for real patient benefit.
Clinical Approach — Applying STOPP/START in Practice
- Include prescription, OTC, herbal, and PRN medications.
- Verify indication for each drug.
- Start with Section A (indication, duration, duplication).
- Go through each relevant physiological system.
- Pay special attention to FRIDs if falls history, and anticholinergics if cognitive impairment.
- Check for missing evidence-based medications.
- Common PPOs: anticoagulation in AF, statin in CVD, osteoporosis treatment, vaccines.
- Consider remaining life expectancy, goals of care, patient preferences.
- Use clinical judgment — STOPP/START criteria are potentially inappropriate, not definitely inappropriate.
- Deprescribe stepwise (especially CNS drugs: taper, don't abruptly stop).
- Initiate PPOs at low doses, titrate carefully.
- Schedule follow-up to assess response and adverse effects.
Exam Intelligence
| Question Type | Example Stem | Key Discriminator |
|---|---|---|
| Identify the PIM | "82-year-old on 8 medications falls repeatedly. Which should be stopped?" | Look for benzodiazepines, anticholinergics, opioids, antipsychotics |
| Identify the PPO | "75-year-old with AF and CHA₂DS₂-VASc 4, not on anticoagulant. What is missing?" | START anticoagulation (DOAC preferred) |
| Drug-disease interaction | "78-year-old with Parkinson's prescribed metoclopramide for nausea" | Metoclopramide is a D2 antagonist → worsens Parkinsonism (STOPP) |
| Deprescribing rationale | "Why should you stop glyburide in an 80-year-old with CKD?" | Long-acting sulfonylurea + reduced renal clearance → prolonged hypoglycaemia |
| STOPP vs Beers | "Which tool addresses both overtreatment AND undertreatment?" | STOPP/START (START = undertreatment) |
| Loop diuretic trap | "Is furosemide appropriate for bilateral ankle oedema?" | Only if HF/liver failure/nephrotic syndrome/CKD confirmed; otherwise compression + elevation |
| PPI duration | "Patient on PPI for 2 years for simple GORD. Is this appropriate?" | STOPP: PPI beyond 8 weeks without ongoing indication is potentially inappropriate |
Common Mistakes
- Assuming "potentially inappropriate" = "definitely contraindicated" — STOPP/START flags risks; clinical judgment is still needed.
- Forgetting START criteria — students focus on "what to stop" but miss "what to start." Omitting anticoagulation in AF is just as dangerous as prescribing the wrong drug.
- Not checking renal function before drug dose adjustment — eGFR < 30 changes everything (digoxin, metformin, DOACs, bisphosphonates, nitrofurantoin).
- Confusing Beers with STOPP/START — Beers is US-based and does NOT have a systematic "START" component.
- BB for HTN — STOPP says BB as monotherapy for uncomplicated HTN lacks evidence; but BB is indicated if there is a concomitant condition (HF, angina, rate control in AF).
- "According to STOPP criteria, [drug] should be deprescribed because [mechanism of harm in elderly]."
- "According to START criteria, [drug] should be initiated because [evidence of benefit / indication]."
- "This represents a PIM / PPO as defined by STOPP/START version 3."
Integration With Related Material
The parent lecture (GC 079) covers the broader principles of prescribing in older people, including pharmacokinetic/pharmacodynamic changes, polypharmacy, and medication review. STOPP/START v3 is the practical tool that operationalises these principles. [4]
Both Beers 2023 and STOPP/START v3 flag similar high-risk drugs:
- First-generation antihistamines (anticholinergic)
- Benzodiazepines
- Aspirin for primary CVD prevention
- NSAIDs in CKD/HF
- Warfarin when DOACs are preferred for non-valvular AF
Beers adds: drugs to use with caution (dabigatran, prasugrel/ticagrelor in > 75), and drug-drug interactions. [3][5]
Falls assessment includes medication review using Beers criteria or STOPP as part of the multidisciplinary multifactorial assessment and prevention approach. [6]
Principles: avoid nephrotoxic insult, correct dose for eGFR, beware of side effects in impaired renal function. STOPP renal criteria (Section E) directly reflect these principles. [7]
Active Recall - STOPP/START Criteria v3
High Yield Summary
- STOPP/START v3 (2023)
- is a Delphi-validated, physiological-systems-based tool with 190 criteria (133 STOPP, 57 START) for medication review in older adults (≥65) with multimorbidity and polypharmacy. STOPP identifies PIMs to deprescribe; START identifies PPOs to initiate — together they enable "represcribing." Key exam themes: (1) Always check indication, duration, and duplication first (Section A). (2) Falls Risk Increasing Drugs (FRIDs) are a major new section — target benzodiazepines, anticholinergics, antipsychotics, opioids. (3) STOPP/START is unique vs. Beers because it addresses both overprescribing AND underprescribing. (4) Electronic deployment is desirable but clinician interpretation is essential (SENATOR/OPERAM lessons). (5) STOPP/START cannot replace clinical judgment — criteria are "potentially" not "definitely" inappropriate.
[1] GC 079 (supp-1)STOPP-START criteria for potentially inappropriate prescribing in older 2023.pdf (O'Mahony et al., European Geriatric Medicine 2023; full text) [2] GC 079 (supp-2)STOPP-START-V3.pdf (Appendix 1: Full list of STOPP/START v3 criteria) [3] American Geriatrics Society 2023 updated AGS Beers Criteria for potentially inappropriate medication use in older adults.pdf (Tables 2–4) [4] GC 079. Prescribing in older people.pdf (Parent lecture) [5] GC 079 (supp-3)AGS Beers Criteria for potentially inappropriate med use_2023.pdf [6] Maksim Medicine Notes.pdf (Geriatrics section, p.114 — Falls assessment) [7] Block A - Drugs and the Kidney.pdf (Principles of prescribing in renal impairment)
GC079 Prescribing In Older People
Prescribing in older people involves the careful selection, dosing, and monitoring of medications to account for age-related pharmacokinetic and pharmacodynamic changes, polypharmacy, and increased vulnerability to adverse drug reactions.
GC079 (supp-2)stopp-start-v3
STOPP/START version 3 (GC079 supplement 2) is an updated evidence-based set of explicit criteria for potentially inappropriate prescribing (STOPP) and potential prescribing omissions (START) in older adults, designed to optimize medication appropriateness in people aged 65 and over.