Deprescribing at the End of Life: Which Fourteen Pills Actually Stay
A man newly enrolled in hospice is still taking fourteen daily medications from a life expectancy that no longer applies to him, and the team has to decide, drug by drug, which ones are protecting him from something real and which are protecting him from nothing he'll live long enough to need protecting from.
H.K. lives alone above the shop he ran for thirty years selling and repairing bicycles, and his weekly pill organizer — fourteen compartments filled twice over, morning and evening — sits on the kitchen table between him and the hospice nurse doing his admission medication reconciliation. He enrolled in hospice eight days ago for advanced heart failure and progressive frailty, prognosis measured in weeks, and every one of those fourteen medications was prescribed by a version of his care team optimizing for a life expectancy that no longer describes him.
Deprescribing at this point in a life isn't a blanket instruction to stop everything; it's a drug-by-drug question of what each medication is still protecting him from, and whether he'll live long enough for that protection to matter. The clearest evidence on this exact question comes from Kutner and colleagues' 2015 randomized trial in JAMA Internal Medicine, which tested discontinuing statins in patients with a life expectancy under a year and found no difference in survival between the groups, alongside improved quality of life and lower cost in the group that stopped. Not every drug on his list resolves that cleanly. His metoprolol carries a real, distinct withdrawal risk — abrupt discontinuation in a patient with underlying cardiac disease can produce rebound tachycardia, hypertension, and in some cases precipitate ischemia, which argues for tapering rather than stopping outright even in a patient whose prognosis has changed entirely.
H.K. has his own opinion about the pill organizer, stated plainly before the nurse even opened her laptop: he wants the evening compartment lighter than the morning one, because evenings are when his hands shake worst and he's twice dropped a full week's dosing onto the shop floor below trying to manage it alone. That preference has quietly shaped today's visit as much as any pharmacology has — a drug that's otherwise a defensible keep may still move to once-daily morning dosing if an evening dose is what's actually landing on the floor rather than in him.
At the kitchen table, going through the pill organizer
Statin and alendronate both come off today, and I don't think that's controversial. The Kutner trial randomized patients with life expectancy under a year to stopping versus continuing statins and found no survival difference, with better quality of life and lower pill burden in the group that stopped. Alendronate's whole purpose is preventing a fracture years from now; it has real GI and esophageal irritation risk in the present, for a benefit he won't live to collect.
I'm with you on the statin and the bisphosphonate. I want to push back specifically on treating metoprolol the same way. His heart failure is stable on it, and abrupt beta-blocker discontinuation in a patient with underlying cardiac disease carries a real, distinct risk — rebound tachycardia, rebound hypertension, and in some patients precipitated ischemia. Stopping metoprolol the way we're stopping the statin isn't removing a future-benefit drug, it's removing an active stabilizer, and doing it abruptly could make his final weeks measurably worse rather than simpler.
That distinction is the right one to hold onto for every drug on this list, not just the beta-blocker: does stopping it remove a future benefit he won't live to see, or does it remove something actively stabilizing him right now. Apixaban falls closer to the metoprolol side of that line than the statin side — stopping anticoagulation in atrial fibrillation with his stroke-risk profile trades a bleeding-risk reduction for a real, near-term stroke risk, not a distant one. I'd taper the metoprolol over one to two weeks rather than stop it outright, and keep the apixaban conversation open rather than folding it into today's stop-list by default.
Agreed: statin and alendronate stopped today; metoprolol tapered over one to two weeks rather than stopped abruptly; the remaining ten medications reviewed individually over the following two visits rather than all at once, given how much a single reconciliation visit had already asked of H.K.
Not agreed: whether apixaban should ultimately stay or go. The Cardiologist wants to preserve stroke protection as long as his functional status allows safe administration; the Palliative Care Physician is concerned that a fall this frail is now more likely to cause a serious bleed than his atrial fibrillation is to cause a stroke in his remaining weeks. Left open for the next visit, with fall-risk documented as the specific factor that would tip the decision.