Polypharmacy and Deprescribing — Written Clinical Reasoning

SAQ 1 — Polypharmacy, Falls and Adverse Drug Events in a Frail Older Woman (20 marks, 30 minutes)

Prompt: Outline your structured medication review of this patient, including the identification of potentially inappropriate medications using validated tools, the pharmacological reasoning for each change, the prescribing omissions you would address, the approach to deprescribing the highest-risk drugs, the discussion you would have with the patient and her family, and the discharge plan. Justify each decision with reference to evidence and guideline recommendations. [1]

Model Answer

Framing and problem list (3 marks): [1]

This is polypharmacy (14 regular medications) with multiple potentially inappropriate medications and several acute drug-related harms contributing to her fall, in a frail older woman (CFS 5) with multimorbidity and significant renal impairment (eGFR 29). The fall is multifactorial but the medications are the most modifiable contributors: she has orthostatic hypotension (104/58 lying, 88/52 standing), bradycardia (pulse 54), an over-anticoagulated INR (3.8) on dual antiplatelet-anticoagulant therapy, hypoglycaemic risk (glibenclamide with an HbA1c of 6.3 percent), and a high anticholinergic and sedative burden. The Kongkaew systematic review (PMID 18594048) established that adverse drug reactions account for approximately 10.7 percent of hospital admissions in older adults, and the Woolcott meta-analysis (PMID 19933962) identified benzodiazepines (OR 1.57), antidepressants (OR 1.68), and antipsychotics as the highest-risk drug classes for falls. This patient is on representatives of several of these classes. [1]

The structured approach is: build the verified medication list (already obtained), apply STOPP/START and Beers systematically, assess the renal function and dose-adjust, assess the anticholinergic burden, anchor each decision to her goals and prognosis, prioritise the changes, taper where needed, and communicate the plan. [1]

Applying STOPP/START and Beers — identifying the potentially inappropriate medications (6 marks): [1]

I will systematically identify the potentially inappropriate medications: [1]

• Glibenclamide (glyburide): This long-acting sulfonylurea is on the Beers list because of its prolonged hypoglycaemia risk, particularly in renal impairment (eGFR 29), where its active metabolites accumulate. Her HbA1c of 6.3 percent indicates over-treatment. STOP. Switch to gliclazide (shorter-acting, lower hypoglycaemia risk) at a low dose, or de-escalate to metformin alone with relaxed targets, or consider a DPP-4 inhibitor.
• Diazepam: A long-acting benzodiazepine, on the Beers list and flagged by STOPP, with a markedly prolonged half-life in older adults (increased fat distribution, reduced clearance) and a strong independent association with falls. She is taking it for insomnia. Plan a supervised taper (10 to 25 percent reduction every 1 to 2 weeks) combined with cognitive behavioural therapy for insomnia.
• Oxybutynin: A potent anticholinergic, the highest-burden agent on the Anticholinergic Burden Scale, on the Beers list, and directly opposed to her donepezil-equivalent cholinergic therapy (though she is not on donepezil, the principle of minimising anticholinergic burden in cognitive impairment applies). It contributes to her confusion, dry mouth, constipation, and falls risk. Stop. Switch to mirabegron (beta-3 agonist, no anticholinergic activity) if pharmacological treatment of her bladder is still needed, or manage with non-pharmacological measures (timed voiding).
• Amitriptyline: A tricyclic antidepressant with high anticholinergic burden, on the Beers list, contributing to her falls, orthostatic hypotension, and cognitive impairment. She is taking it for sleep and depression. Stop (or switch to an SSRI such as sertraline if an antidepressant is genuinely indicated for a current depressive episode, with caution for the hyponatraemia risk).
• Aspirin combined with warfarin: The combination of an anticoagulant and an antiplatelet is flagged by STOPP unless there is a clear indication (recent stent, mechanical valve). She has atrial fibrillation and ischaemic heart disease; without a recent coronary event or stent, the aspirin should be stopped (the risk of major bleeding, especially at an INR of 3.8, exceeds the marginal anti-ischaemic benefit). Her INR of 3.8 is above the therapeutic range and reflects her low weight, her age (increased warfarin sensitivity), and her reduced warfarin clearance — reduce the warfarin dose and monitor.
• Omeprazole: Started at some point (possibly for dyspepsia or gastrointestinal protection), now long-term without a documented specialist indication. Long-term PPI use is associated with hypomagnesaemia, vitamin B12 deficiency, Clostridioides difficile, and osteoporotic fracture (relevant given her osteoporosis). Reassess the indication and taper to the lowest effective dose or stop, managing rebound with on-demand therapy or an H2 antagonist.
• Spironolactone in renal impairment with an ACE inhibitor and a diuretic: This combination places her at high risk of hyperkalaemia and worsening renal function, particularly at an eGFR of 29. Review the indication (heart failure); consider holding or dose-reducing while in hospital, and monitor the potassium and renal function closely. [1]

Renal function and dose adjustment (3 marks): [1]

Her eGFR of 29 mL per minute per 1.73 metres squared, combined with her low weight (46 kg) and frailty, means her creatinine-based eGFR may still overestimate her true renal function because of her low muscle mass (Inker et al, NEJM 2021, PMID 34554658). I would request a cystatin C-based eGFR to refine the estimate, and I would dose-adjust the renally-cleared drugs: metformin (reduce or hold given the eGFR under 30 and the risk of lactic acidosis — consider switching to a DPP-4 inhibitor); the DOACs if she were on one (dose for renal function and weight); and review the bisoprolol dose given her bradycardia. The Cockcroft-Gault creatinine clearance would guide the formal drug-dosing reference. The principle: in a frail, low-muscle older patient, never trust the serum creatinine or the creatinine-based eGFR alone for a narrow-therapeutic-index renally-cleared drug. [1]

Prescribing omissions — applying START (2 marks): [1]

START also detects under-prescribing. I would assess:

• Whether she is on an SGLT2 inhibitor for her heart failure with reduced ejection fraction (guideline-directed therapy that may be indicated and may be missing).
• Whether her osteoporosis is adequately treated (she is on alendronate — I would assess the duration and whether a drug holiday or a reassessment of the fracture risk is due).
• Whether her vaccinations (influenza, pneumococcal, COVID-19, shingles, RSV) are up to date.
• Whether her heart failure therapy is optimised to guideline-directed therapy (beta-blocker at target, ACE inhibitor at target, mineralocorticoid antagonist) — though this must be balanced against her blood pressure and renal function. [1]

Deprescribing the highest-risk drugs — the plan (3 marks): [1]

I will prioritise by the balance of harm and the feasibility of the change: [1]

1. Immediate stops (in hospital): Stop the glibenclamide (prolonged hypoglycaemia risk in renal impairment), stop the aspirin (bleeding risk on dual therapy with an INR of 3.8), reduce the warfarin dose (over-anticoagulation), hold or reduce the spironolactone (hyperkalaemia risk), and reduce the frusemide and the antihypertensives to address the orthostatic hypotension and the bradycardia.
2. Short-term tapers: Plan a supervised diazepam taper (10 to 25 percent every 1 to 2 weeks) with CBT-I; plan a PPI step-down (daily to alternate day to on-demand).
3. Replacements: Switch the oxybutynin to mirabegron or non-pharmacological bladder management; switch the amitriptyline to sertraline if an antidepressant is indicated or stop it; switch the glibenclamide to gliclazide or de-escalate. [1]

Each change will be documented with the indication, the monitoring plan, and a review date. [1]

Discussion with the patient and family (2 marks): [1]

I would have a goals-of-care conversation with the patient and her family, anchored to her priorities (likely function, independence, and avoiding further falls and hospital admissions, rather than aggressive disease-modifying targets). I would explain that several of her medications are now causing more harm than benefit — that the sleeping tablet and the bladder tablet and the antidepressant are all contributing to her falls and confusion, that the blood-thinners together are too risky at her current levels, and that the diabetes tablet is lowering her blood sugar too much given how well-controlled her diabetes already is. I would frame the deprescribing as an active, positive intervention to improve her function and safety, not as 'giving up', and I would invite her preferences about which changes matter most to her. I would provide a written medication list and a deprescribing plan, and arrange the follow-up. [1]

Discharge and follow-up plan (1 mark): [1]

At discharge, I would conduct a formal medication reconciliation (the BPMH compared against the inpatient and discharge lists, every discrepancy resolved and documented), provide the patient and the GP with an accurate, reconciled medication list and the deprescribing plan, arrange a medication review with the GP and the pharmacist within 1 to 2 weeks, and schedule a clinic review at 4 to 6 weeks to assess the blood pressure, the renal function, the INR, the glycaemic control, the cognitive function, and the success of the benzodiazepine taper. I would arrange a comprehensive geriatric assessment and a falls assessment, review her osteoporosis management, and ensure her sensory aids (glasses, hearing aids) and her home environment are optimised. [1]

SAQ 2 — The Prescribing Cascade (10 marks, 15 minutes)

Prompt: A 79-year-old man is reviewed in the clinic for a routine check. He has type 2 diabetes, hypertension, and gastro-oesophageal reflux. Six months ago he was started on metoclopramide 10 mg three times daily for diabetic gastroparesis. Three months ago his general practitioner noted that he had developed a tremor and rigidity and referred him to a neurologist, who diagnosed Parkinson disease and started levodopa-carbidopa. He now has worsening tremor and rigidity despite the levodopa, and he has developed dizziness and has had one fall. Outline your assessment of this situation and your management plan. [1]

Model Answer

Recognise the prescribing cascade (3 marks): [1]

This is a prescribing cascade. The metoclopramide (a dopamine D2-receptor antagonist) has caused drug-induced parkinsonism — tremor, rigidity, and bradykinesia — which was misdiagnosed as idiopathic Parkinson disease and treated with levodopa. The levodopa cannot overcome the D2-receptor blockade by the metoclopramide, which is why his symptoms have worsened despite the levodopa, and the levodopa itself has contributed to his dizziness and fall (orthostatic hypotension is a common levodopa adverse effect). The cascade has added a second drug to treat the side effect of the first, and the second drug has caused its own harm. [1]

Assessment (3 marks): [1]

I would confirm the diagnosis clinically:

• The temporal relationship: the parkinsonian features began after the metoclopramide was started, which strongly supports a drug-induced cause.
• The levodopa non-response: idiopathic Parkinson disease typically shows a clear, early response to levodopa; a non-response or a worsening on levodopa is a red flag for drug-induced parkinsonism or an atypical parkinsonism.
• The examination: I would assess the tremor (a pill-rolling tremor suggests idiopathic Parkinson; a postural or action tremor may suggest drug-induced), the rigidity, the bradykinesia, the gait, and the postural blood pressure. I would exclude other causes of parkinsonism (vascular, normal-pressure hydrocephalus, other drug causes).
• I would review his full medication list for other drug-induced parkinsonism culprits (antipsychotics, metoclopramide, prochlorperazine, some calcium-channel blockers). [1]

Management plan (4 marks): [1]

1. Stop the metoclopramide — this is the cause of the parkinsonism and the first and most important step. Drug-induced parkinsonism is typically reversible over days to weeks after the causative drug is withdrawn.
2. Stop or taper the levodopa — it is treating a side effect, not a disease, and it is contributing to his dizziness and fall. I would coordinate this with the neurologist.
3. Manage the gastroparesis with an alternative that does not cross the blood-brain barrier — domperidone (a peripheral D2 antagonist, though it requires ECG monitoring for QT prolongation) or a 5-HT4 agonist (where available), and dietary measures (small, frequent, low-fat, low-fibre meals) and glycaemic optimisation.
4. Reassess the diagnosis at 4 to 8 weeks after the metoclopramide withdrawal — if the parkinsonian features have resolved, the diagnosis of drug-induced parkinsonism is confirmed and the levodopa can be stopped. If features persist, I would reassess for an alternative diagnosis with the neurologist.
5. Address the fall and the dizziness — review his antihypertensives (he has hypertension and may be over-treated), and arrange a falls assessment. [1]

The principle: whenever a new drug is started for a new symptom, ask whether the symptom could be an adverse effect of an existing drug, and address the first drug before adding the second. [1]