Eyeballing frailty is good at ruling it out, poor at ruling it in

Design

Prospective diagnostic accuracy study, blinded paired assessment

Population

200 adults ≥70 years referred for coronary angiography

Primary outcome

Accuracy of physician eyeballing against the Fried test

Effect

Sensitivity 93.9%, specificity 80.8%, NPV 98.5%, PPV 49.2%, AUC 0.87

An Israeli prospective study assessed 200 adults aged 70 or over referred for coronary angiography. Two cardiologists, blinded to each other, rated frailty: one by the formal Fried test, the other by a quick visual impression scored on a Fried-like scale.

About one in six were frail by Fried criteria. The visual impression caught 94% of them and, when it judged a patient not frail, was right 98.5% of the time. But it labelled many robust patients as frail: only about half of those it flagged were frail on testing. Specificity was lower in women and in hospitalised patients. Frailty on the Fried test was associated with higher six-month mortality and readmission.

The practical message is a two-step approach. A clinician's quick impression can reasonably exclude frailty in a busy clinic. When the impression suggests frailty, a structured tool should confirm it before it shapes decisions about invasive treatment.

• Use your clinical impression to rule frailty out quickly in older patients before procedures.
• Confirm suspected frailty with a structured tool such as the Fried phenotype or Clinical Frailty Scale before it changes treatment.
• Take extra care in women and inpatients, where the visual impression over-called frailty more often.
• Record frailty status in procedure referrals; it predicted six-month mortality and readmission.

Why it matters

Labelling a robust older patient as frail can deny them an effective procedure.

Don't overread it

The study was in cardiology referrals at one centre; accuracy of eyeballing may differ in other settings and with other assessors.

The statistics, in plain English

Positive predictive value depends on how common frailty is. With about 16% frail, even a fairly specific test flags many people who are not, which is why only half of 'frail' impressions were confirmed.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.