
A polypharmacy review is a structured pass through every medication an older adult takes, matching each one to a current indication, screening the list against the 2023 Beers Criteria and STOPP/START version 3, hunting for prescribing cascades and omissions, and then agreeing on one or two changes with a follow-up date. Done in a fixed order it fits into 10 minutes of a visit. On the NP exam the same order answers the item that asks which medication should be discontinued first.
A polypharmacy review is a fixed sequence, not a conversation that wanders through the medication list: reconcile what the patient actually takes, attach every drug to a current indication, screen the list against the 2023 Beers Criteria and STOPP/START version 3, look for the cascades and omissions that lists miss, and then agree on one or two changes with a date to check them. In a focused visit the whole pass takes about 10 minutes. On the FNP and AGNP exams the same sequence answers the recurring item "which medication should be discontinued first," because the item writer built the vignette by walking the same steps in reverse.
Older adults with multiple chronic conditions accumulate prescriptions from several clinicians, and each additional drug raises the probability of an interaction, an adverse event, a prescribing cascade, and nonadherence. The review also finds the opposite problem. START criteria exist because older patients are frequently undertreated for conditions where a drug has clear benefit (anticoagulation in atrial fibrillation, bone protection after a fragility fracture, an ACE inhibitor in HFrEF).
A workable definition for the visit: polypharmacy is present when the patient takes more drugs than are clinically indicated, whatever the count. The research cut-off of five or more is a trigger for review, not a diagnosis.
| Minute | Step | What you do | Output |
|---|---|---|---|
| 0 to 2 | Reconcile | Compare the chart list with what the patient brought (brown bag) or reads from bottles; add OTC products, supplements, eye drops, inhalers, patches, as-needed drugs, and anything another prescriber started | One accurate list |
| 2 to 4 | Indication match | For each drug write the condition it treats; mark any with no current indication, a resolved indication, or a duplicated class | Drugs with no indication; duplications |
| 4 to 6 | Screen | Run the list against Beers (avoid, drug-disease, interactions, renal) and STOPP; then START for omissions; add kidney function and anticholinergic burden | Flagged drugs and omissions |
| 6 to 8 | Hunt cascades and symptoms | Ask what each new drug was started for and whether that symptom began after another drug; link current complaints (falls, confusion, edema, constipation, dizziness) to candidates | Cascades; symptom-drug links |
| 8 to 10 | Decide and document | Prioritize one or two changes; agree with patient and caregiver; set the taper if needed, the monitoring, and the follow-up date | Plan, education, follow-up |
The chart list is wrong more often than it is right. Ask the patient to bring every container, including over-the-counter products, and read the labels with them. Ask specifically about sleep aids, allergy pills, pain relievers, antacids, laxatives, herbal products, and anything a spouse shares. Ask how each drug is actually taken, not how it was prescribed. Missing doses of an antihypertensive explains an "uncontrolled" blood pressure better than adding a fourth agent.
Write the indication beside each drug. Three findings follow. Drugs with no indication (a PPI started in hospital for stress-ulcer prophylaxis and never stopped). Drugs whose indication has resolved (an antidepressant from a bereavement 6 years ago, an allopurinol after a single gout flare). Duplications within a class (two anticholinergics, two sedatives, an SSRI plus a tricyclic). The Medication Appropriateness Index (Hanlon and colleagues, 1992) formalizes this with ten questions per drug: indication, effectiveness, dose, directions, practicality, drug-drug and drug-disease interactions, duplication, duration, and cost. You will not score it in a visit, but its first two questions (is there an indication, is the drug effective for it) carry most of the yield.
Two lists do the heavy lifting. The 2023 AGS Beers Criteria name the drugs to avoid in most older adults, the drug-disease pairs to avoid, interactions to avoid, and drugs to adjust by creatinine clearance. STOPP/START version 3 (2023) contains 133 STOPP criteria arranged by physiologic system and 57 START criteria that flag omissions. Add two numbers: the estimated creatinine clearance (Cockcroft-Gault, using actual body weight with judgment in obesity) and the anticholinergic burden, using the Anticholinergic Cognitive Burden scale (drugs scored 1 to 3; a cumulative score of 3 or more has been linked with cognitive decline in cohort studies).
The comparison below shows how the two lists overlap and where each adds something.
| Area | 2023 Beers Criteria | STOPP/START version 3 (2023) | What the review gains |
|---|---|---|---|
| Organization | By drug class and by disease; US focus | By physiologic system; European origin | Beers for the exam's drug list; STOPP for a systematic sweep |
| Drugs to avoid | Anticholinergics, benzodiazepines, Z-drugs, antipsychotics in dementia, sulfonylureas, chronic NSAIDs, sliding-scale insulin, PPIs beyond 8 weeks | Similar list plus duration limits and duplication rules (two drugs of the same class, any drug beyond its recommended duration) | Catches duplicates and overlong courses |
| Drug-disease | Dedicated table (HF, syncope, delirium, dementia, falls, Parkinson, ulcer, CKD, incontinence, BPH) | Integrated into each system | Same decisions, two ways to find them |
| Omissions | Not addressed | START: anticoagulant in AF, statin with vascular disease, ACE inhibitor in HFrEF, bisphosphonate and vitamin D after fragility fracture or with osteoporosis, laxative with regular opioids, vaccines | Finds undertreatment |
| Kidney function | Renal table with CrCl thresholds | Criteria embedded per drug | Both prompt the Cockcroft-Gault step |
| End of life | Does not apply in hospice or end-of-life care | Includes criteria on burdensome therapy in limited life expectancy | Permission to stop preventive drugs |
A prescribing cascade starts when an adverse effect is read as a new disease and treated with another drug. The classic one is a dihydropyridine calcium channel blocker causing ankle edema, treated with a loop diuretic, which then causes urinary urgency, treated with an anticholinergic, which causes confusion. Savage and colleagues (2020) documented the calcium channel blocker to diuretic cascade at population scale in older adults with hypertension. Other common cascades:
The questions that expose a cascade are "when was this started" and "what was happening at the time." Then reverse the order: link the patient's current complaints to candidate drugs before accepting any as a new diagnosis. Falls point to sedatives, antihypertensives, anticholinergics, and hypoglycemics. Confusion points to anticholinergics, benzodiazepines, opioids, and H2 blockers. Edema points to calcium channel blockers, thiazolidinediones, and NSAIDs. Constipation points to anticholinergics, opioids, iron, and calcium. Orthostatic dizziness points to alpha blockers, diuretics, nitrates, and tricyclics.
Rarely can every flagged drug be changed in one visit. Prioritize in this order:
Change one or two drugs at a time so that a benefit or a withdrawal effect can be attributed. Decide whether the drug needs a taper: benzodiazepines, Z-drugs, antidepressants, beta blockers, clonidine, gabapentinoids, opioids, and PPIs taken long term all do; a statin, an aspirin for primary prevention, a vitamin, or a short-course PPI can stop at once. Write the indication for every drug that stays, including the Beers drugs you choose to continue, with the reason.
| Change | What to monitor | When |
|---|---|---|
| Benzodiazepine or Z-drug taper | Withdrawal symptoms (anxiety, insomnia, tremor, sweating), sleep diary, falls | Every 1 to 2 weeks during the taper |
| Antidepressant taper | Discontinuation symptoms, mood relapse, PHQ-9 or GDS | 2 weeks after each step; 4 to 8 weeks after stopping |
| PPI reduction or stop | Rebound heartburn, dyspepsia, alarm symptoms (dysphagia, weight loss, bleeding) | 4 and 12 weeks |
| Antihypertensive reduction | Blood pressure sitting and standing, home readings | 2 to 4 weeks |
| Sulfonylurea or insulin reduction | Hypoglycemia symptoms, fingerstick log, A1c | 2 weeks for symptoms; 3 months for A1c |
| Anticholinergic removal | Cognition (Mini-Cog or MoCA), urinary symptoms, bowel function | 4 to 6 weeks |
| Opioid taper | Pain, function, withdrawal, mood | Each dose step, typically every 2 to 4 weeks |
| Any new drug started | Target effect, adverse effects, labs the drug needs (sodium with SSRIs, potassium and creatinine with RAS inhibitors) | 2 to 4 weeks |
Refer to or consult a clinical pharmacist for a comprehensive medication review when the patient takes ten or more drugs, when kidney or liver function complicates dosing, when multiple prescribers are involved, or when a complex taper (multiple psychotropics, high-dose opioids) is needed. Refer to geriatrics when polypharmacy coexists with frailty, cognitive impairment, and recurrent falls, where goals of care should drive the plan. Refer to psychiatry or addiction medicine when a benzodiazepine or opioid taper is complicated by a use disorder or by severe anxiety relapse. Refer to the prescribing specialist before stopping a drug started for a condition you do not manage (antiplatelet after a recent stent, an antiepileptic, an immunosuppressant).
Deprescribing fails when the patient hears "the clinician is giving up on me." Frame each change as a response to a specific risk for this person: "this sleep aid is linked to falls and memory problems, and you have had both this year." Give a written list of what stays, what stops, what tapers and on what schedule, what symptoms to report, and the follow-up date. Ask the caregiver to manage a pill organizer for a week and bring it back. Tell the patient that symptoms during a taper are expected and temporary, and that pausing at the current dose is acceptable while stopping and restarting the full dose is not the plan.
An 82-year-old man with hypertension, type 2 diabetes, benign prostatic hyperplasia, and chronic low back pain presents for an annual visit. He reports two falls in the past 3 months, both at night on the way to the bathroom, and his wife says he has seemed "foggy" for about 2 months. Medications: lisinopril 20 mg daily, amlodipine 10 mg daily, furosemide 20 mg daily (added 4 months ago for ankle swelling), glipizide 10 mg twice daily, tamsulosin 0.4 mg nightly, cyclobenzaprine 10 mg three times daily as needed (uses it nightly), and acetaminophen 1 g twice daily. Blood pressure is 118/68 mm Hg seated and 98/60 mm Hg standing with lightheadedness. Fingerstick glucose log shows several readings between 58 and 70 mg/dL before breakfast. A1c is 6.4 percent. Estimated creatinine clearance is 42 mL/min. Which of the following is the most appropriate first change?
Answer: B. The two active harms are falls with cognitive fog and documented hypoglycemia, and the drugs most plausibly responsible are cyclobenzaprine (a strongly anticholinergic, sedating Beers muscle relaxant taken nightly) and glipizide (a sulfonylurea causing hypoglycemia with an A1c already below the frail-adult target). Those are changed first, with follow-up at 2 weeks. Option A addresses a real cascade (amlodipine edema treated with furosemide) and the orthostatic drop, and it is a correct second step, but it does not address the drug most directly tied to the nighttime falls and confusion or the hypoglycemia, which is the more urgent harm. Option C removes the appropriate BPH drug and adds an anticholinergic that would worsen cognition and retention. Option D treats a probable drug-induced cognitive change as neurodegenerative before the reversible causes have been removed.
Common follow-up questions on Geriatrics & Prescribing (AGNP).
Five or more regular medications is the most common research definition, and ten or more is often called hyperpolypharmacy, but neither number is a clinical threshold. The question that matters is whether each drug still has an indication, an acceptable risk, and a patient who wants to take it.
Beers lists drugs and drug-disease pairs to avoid; STOPP does the same by organ system and adds criteria for duplication and duration, while START flags omissions such as a missing anticoagulant in atrial fibrillation or a missing bone-protective agent after a fragility fracture. Using both catches overuse and underuse.
Start with the drug causing a current symptom or the one with the highest harm for this patient (a benzodiazepine in a faller, an NSAID in CKD), then the drug with no indication, then preventive drugs whose time-to-benefit exceeds the patient's likely life expectancy. Change one or two at a time so that an effect can be attributed.
One to two weeks for drugs with withdrawal or rebound risk (benzodiazepines, antidepressants, beta blockers, PPIs), and four to twelve weeks for symptom recurrence after stopping a chronic-disease drug. Document the reason for the change and what the patient should report.
FNP and AGNP board questions written as primary-care cases, with guideline-based explanations, for the AANP and ANCC exams. Create a free account and pick your track.
Create Free Account →