Frailty - Stopping the Drugs

It’s late in the day when the call comes in from the nursing home. 

“Mr Williams has a blood pressure of 172/98. We’ve checked it three times. He’s feeling a bit off. What should we do?”

You know Mr Williams well. He’s in his mid-80s, lived in the local care home for several years, cognitively good but physically he’s deteriorated in the past couple of years and now needs help with most of his basic needs. He had a coronary event more than a decade ago, has hypertension and type 2 diabetes, meaning he’s been on a raft of medication since.

To control his blood pressure to the NICE target in the over 80s of 150/90 he takes ramipril, amlodipine and indapamide. Mostly this is sufficient.

For cardiac secondary prevention he takes clopidogrel, and to hit the lipid target of an LDL <2mmopl/L we prescribe atorvastatin and ezetimibe.

National guidelines suggest he should have an ‘individualised glucose target’ for his type 2 diabetes, which, if we’re honest, meant we just went for the usual 58mmol/L (7.5%) as we are trying to make sure we hit our QOF this year. So, he’s on maximum dose metformin, dapagliflozin and semaglutide.

Less eye catching and pop-up triggering is his diagnosis of frailty. 

A recent Analysis article in the BJGP very reasonably explains why recognition of this diagnosis is crucial, and how it impacts every other aspect of management. It also highlights national guidance published by the British Geriatrics Society in November 2025, which recommends a radical approach to prescribing in people with moderate and severe frailty. One that I suspect will resonate with many clinicians in general practice.

The importance of frailty is often underestimated, unlike more prominent diagnoses. A diagnosis of moderate to severe frailty is accompanied by a 3 year mortality rate of 50%. 

Iatrogenic harm significantly contributes to this. A government report on overprescribing in 2021 highlighted that in the over 65 age group, up to 20% of hospital admissions are caused by adverse effects of medicines. People with frailty are the most susceptible to this.

But prescribers have a problem. The risk of iatrogenic harm is high in this group, but evidence-based guidelines recommend prescribing a wide range of treatments to supposedly improve outcomes. Something doesn’t add up.

The cause of this disconnect is that trials typically do not include patients with significant frailty. The BJGP points out that even trials specifically investigating treatments in older people fail to include those with the greater degrees of frailty, yet this rarely gets communicated to clinicians or patients, and guideline recommendations to prescribe pragmatically in certain groups rarely helps clinicians with decision-making, especially on a background of constant nationally mandated performance management.

How then, does this new BSG guidance fill this gap? 

It encourages us to explain that “the aim is that each patient is taking only medications that have real value for them and avoiding medication-related harm.” It states that “reduced life expectancy will attenuate prognostic benefits seen in less frail people” and that medicines used for prognostic reasons should be re-evaluated and deprescribing offered. It also points out that doing nothing and “not addressing problematic polypharmacy exposes patients to avoidable harm”.

The recommendations are too extensive to cover all in one blog (we will cover these more in the next Hot Topics GP Update course), but if we apply them to Mr Williams, we see huge changes to our approach to prescribing.

Firstly, we need to quantify his level of frailty. The most widely used tool is the Rockwood Clinical Frailty Scale, which categorises people from 1 – 9, ranging from 1: Very Fit, to 8: Very Severely Frail, where the person is completely dependent and approaching the end of life. (Category 9 represents Terminally Ill, where life expectancy is <6 months in a person not otherwise evidently frail). Mr Williams meets the severe frailty category. This correlates with the most aggressive recommendations on prescribing.

Second, we need to discuss with the patient if we are going to make any changes, explain the logic, and ensure they are happy. Many people are very pleased to reduce their medication burden.

Third, the reason the nursing home called – his raised blood pressure. 172/98. Already on three antihypertensives. In contrast to NICE guidance, the BGS suggests there should be NO BP target. The “harms likely outweigh the benefits. Deprescribing advised.” What about the risk of stroke and ACS? What if his decline, fatigue and unsteadiness are all iatrogenic? Due to his medication-induced hyponatraemia, orthostatic drops and the rest. 

Fourth, consider other issues. T2DM? The guideline recommends simplifying his prescription to simply avoid symptomatic hyper/hypoglycaemia. Medium dose metformin may well be adequate. Cholesterol? It suggests in severe frailty the harms are likely to exceed the benefits and advises deprescribing. Osteoporosis? Forget it (although denosumab may be more complex). Even heart failure with reduced ejection fraction it suggests continuing just loop diuretics to control fluid overload. 

The BJGP points out that there is not universal support amongst medical groups for these recommendations, but this is the sharp end of real-world medicine that you and I inhabit on a daily basis, and I suspect many of us will feel it is not just timely, but a righting of an imbalance we’ve felt now for years.