Simple tests predicted outlet obstruction in men with urinary symptoms

Design

Prospective multicentre diagnostic accuracy study with independent validation

Population

601 men with LUTS in 67 general practices in England and Wales

Primary outcome

Prediction of urodynamic diagnoses against invasive urodynamics

Effect

Outlet obstruction c-index 0.82 in validation; detrusor under/overactivity 0.63 and 0.62

This prospective diagnostic accuracy study recruited 601 men with lower urinary tract symptoms (LUTS) from 67 general practices in England and Wales. Each had a symptom questionnaire, a bladder diary, examination, PSA, a urine flow test and a bladder scan for post-void residual, then invasive urodynamics as the reference standard. Models were built in 350 men and tested in a separate 251. Published in the British Journal of General Practice on 1 October 2026.

The model for bladder outlet obstruction discriminated well: c-index 0.80 in development and 0.82 in validation. Models for detrusor underactivity (0.63) and overactivity (0.62) performed only moderately.

Most men with LUTS are managed in primary care, often on symptoms alone. This shows that a handful of non-invasive tests, several of which are cheap, can identify the men most likely to have obstruction and to benefit from alpha-blockers or referral.

Uroflowmetry and bladder scanners are not available in every practice. The models are not yet packaged as a tool, so for now the message is about which tests carry the most information.

• Ask men with urinary symptoms to keep a three-day bladder diary before the next visit
• Use a validated symptom score such as the IPSS to grade severity
• Measure post-void residual with a bladder scanner where one is available
• Offer PSA after discussing its benefits and harms, as part of the assessment
• Refer when obstruction seems likely and symptoms persist despite treatment

Why it matters

Primary care can sort men with urinary symptoms into likely obstruction or not without invasive testing.

Don't overread it

The models predicted obstruction well but overactive and underactive bladder only moderately.

The statistics, in plain English

A c-index of 0.82 means that in about 82% of random pairs of men, one with and one without obstruction, the model ranked the man with obstruction higher. Holding up in a separate validation group is what makes the result credible.

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.