Prescribing cascades and a chance for a bit of de-prescribing

Sylvia is a pretty fit 83-year-old who has a remote consultation with you. She is getting puffy feet and ankles, and she is fed up with it. She’s put up with it for 6 months without calling, as she knew you’d been busy with COVID pressures, but enough is enough. She cuts to the chase. ‘I just want some of those water tablets Dr Sweet used to give me for my annual trip to Spain when I got terrible swelling in my feet and couldn’t get my summer sandals on’. You look at the clock. It’s nearly 2 p.m., you haven’t finished your ‘morning’ surgery, no-one has been answering their (insert expletive) phone, and EMIS has crashed 3 times. The pull to the ‘re-start and prescribe’ button for the 40mg frusemide is stronger than a Death Star tractor beam. You take a deep breath and do a very quick medical sieve - heart failure? Nothing in the history to suggest that. Renal failure or nephrotic syndrome? Seems unlikely as she had a normal U&E a few months back. But do you remember to review her medication screen? Is it actually the 5mg amlodipine she is on for her hypertension?

A familiar story we all will have encountered over the years, and it brings up the interesting concept of prescribing cascades, which came back across my radar this month with a recent DTB article (Vol 59 Issue 2, Feb 2021). It reviewed a Canadian retrospective cohort study assessing whether older adults (mean age ~75) without heart failure or end-stage renal disease were more likely to be prescribed a loop diuretic after starting a calcium-channel blocker (CCB) compared to other antihypertensive drugs (they reviewed both renin-angiotensin drugs and ‘other’ antihypertensives as the two control groups). 80% of those who had a CCB were prescribed amlodipine which is in keeping with UK prescribing patterns. At 1-year loop diuretics were being prescribed at twice the rate in the CCB group compared to the two control groups. Whilst the indications for loop diuretic use could not be ascertained, it is likely that the higher rate of loop diuretic use was due to the common side effect of peripheral oedema from the CCB. However, the DTB points out that ‘The cause of CCB peripheral oedema is not thought to be related to fluid retention and has been attributed to precapillary arteriolar dilatation resulting in increased pressure and movement of fluid into the interstitial compartment’ and therefore that diuretic therapy is not recommended as it is unlikely to help.

This a good example of a prescribing cascade, leading to potentially problematic polypharmacy. Prescribing cascades were first described in the Lancet in 1995, and the concept was reviewed last year in the BMJ (BMJ 2020;368:m261). It describes the situation ‘when a drug is prescribed, an adverse drug event occurs that is misinterpreted as a new medical condition, and a subsequent drug is prescribed to treat this drug-induced adverse event’ (Lancet 2017). Since the original description in 1995, more than 20 prescribing cascades have been described, and the concept has been expanded to include situations where drug side effects trigger unnecessary investigations, or even intervention with medical devices, which may expose the patient to risk or harm. Common examples include

• Cholinesterase inhibitors causing urine incontinence leading to urinary anticholinergics.

• Thiazides causing gout leading to urate-lowering drugs (and NSAIDs).

• NSAIDs causing hypertension, leading to antihypertensives.

• Bisphosphonates causing oesophagitis leading to PPI therapy. 

But is this such a big deal? Well, yes. As discussed in the BMJ article evidence suggests that prescribing cascades can be associated with syncope, traumatic falls, invasive procedures (such as pacemaker insertion), drug toxicity, functional decline, and hospitalisation. This is not surprising as it tends to happen in elderly people with polypharmacy. In addition, prescribing cascades contribute to patients’ pill burden and increased drug spending. Not all prescribing cascades lead to problematic polypharmacy, and indeed in some situations are positively beneficial. We all know that co-prescribing PPIs in people on long-term NSAIDs can reduce GI bleeds (an example of an ‘anticipatory’ prescribing cascade where a risk is identified and a prescribing cascade is initiated to reduce risk). 

But we all know that problematic polypharmacy in our elderly population is an issue and a challenge to address. Time constraints, fragmented care, and trying to differentiate what is a symptom of disease, or what might be a drug side effect, are all significant barriers to safe de-prescribing. But an awareness of the concept of prescribing cascades can help. You discuss with Sylvia that her puffy ankles may simply be due to the amlodipine. She’s delighted to stop the amlodipine and review things with some home blood pressure readings in 4 weeks, to see if she does need an alternative antihypertensive or not. A highly satisfactory de-prescribing consultation, a prescribing cascade averted, you’re done in 5 minutes and lunchtime is somewhat closer….