Prescription Review practice question: a 71-year-old with six months of ankle swelling at a medication review

2 marks for the correct tick. Do this with a calendar rather than with a list of side effects. The symptom is bilateral pitting ankle oedema. It was absent at a review nine months ago and at every review before that, and it appeared six months ago. Now read the start dates: amlodipine seven years ago, atorvastatin seven years ago, colecalciferol five years ago, levothyroxine fifteen years ago, metformin ten years ago, naproxen and lansoprazole a little over nine months ago — and pregabalin three weeks before the ankles swelled. Only one thing on the page is new enough to be the cause.

That matters because the pharmacology points the other way. Amlodipine is the drug everybody associates with ankle oedema, and rightly: its Summary of Product Characteristics lists oedema as very common and ankle swelling as common, which is a stronger association than pregabalin's 'common'. But amlodipine oedema is dose-related and appears within weeks of starting or of an increase, and this dose has not changed in seven years, during which her ankles were fine and stayed fine for seven years. A very common adverse effect of a drug she has tolerated for seven years is less likely than a common adverse effect of a drug she started three weeks before the symptom. The strength of the association tells you what to think of first; the dates tell you what actually happened.

Rule out the things that are not drugs before you commit, because 'drug-induced' is a diagnosis of exclusion in a 71-year-old with swollen ankles: the jugular venous pressure is not raised, the heart sounds are normal, the chest is clear, the albumin is 41 g/L, the urine has no protein, the estimated glomerular filtration rate is 75 mL/min/1.73 m², the thyroid-stimulating hormone is normal and there is no hepatomegaly, no ascites and no varicose veins. Do not tick the naproxen: she has taken none for 5 weeks and her ankles were normal for nearly four months after it was started. The action is to review the pregabalin — reduce it or change the neuropathic agent, and reassess the ankles in a month.

2 marks for the correct tick. The furosemide was started six days after the swelling appeared and four weeks after the drug that caused it, and its indication is written on the list as 'ankle swelling' — a symptom, not a diagnosis. That is what a cascade prescription looks like: an adverse effect is read as a new problem, and a second medicine is added for it.

Two details tell you it is not working and was never going to. Her weight is 68 kg today and was 68 kg the day it was started, and the ankles are unchanged. Drug-induced peripheral oedema is a capillary and vasodilatory phenomenon at the ankles rather than an excess of total body water, so a loop diuretic has nothing to remove; it simply takes fluid from a normovolaemic woman. And it is not free. Her potassium is 3.6 mmol/L, at the bottom of the range, on a diuretic that lowers it; a loop diuretic in a 71-year-old also brings hypovolaemia, acute kidney injury, urinary urgency and night-time falls. Six months of that has bought nothing.

The row that catches people out is the lansoprazole. It was started because of another medicine — the indication says so — and a quick reader will tick it. It is a COMPANION prescription, not a cascade: it prevents an adverse effect that has not happened, on the instruction of a guideline, in a patient whose age puts her at risk, and it is licensed for precisely that ('Prophylaxis of NSAID-associated gastric ulcers and duodenal ulcers in patients at risk', 15 mg once daily). Look at the dates as well as the words: the lansoprazole and the naproxen were both started on the same day, a little over nine months ago, because they were written together. A cascade drug always comes later, because it is a response to something that happened after the first drug was started. Getting this the wrong way round has a cost in both directions — leaving an at-risk patient on a non-steroidal anti-inflammatory drug without gastroprotection is a defect, and stopping the gastroprotection because it looked like a cascade would create one.

The pregabalin cannot be ticked here, and it is worth knowing why properly rather than by reflex. It would be the right answer if something on this list had caused the neuropathy it treats, and there is a row with a real claim: metformin, taken for ten years, causes vitamin B12 malabsorption, and B12 deficiency causes a peripheral neuropathy. That would be a complete second cascade — metformin, deficiency, neuropathy, pregabalin — and it is the kind of thing this section exists to teach you to look for. Look, then: vitamin B12 421 ng/L (180–900), haemoglobin 131 g/L, mean cell volume 89 fL. All normal, so there is no deficiency and no cascade. What caused her neuropathy is her diabetes, and a disease is not an adverse effect. Do the two things in order — stop the furosemide, and review the drug that made the ankles swell — and check her ankles, her weight, her renal function and her potassium in a month.