Design: Qualitative evidence synthesis with reflexive thematic analysis
Population: 32 qualitative studies, 677 primary care practitioners internationally
Primary outcome: Practitioners' views and experiences of managing UTIs
Effect: Five themes; uncertainty concentrated in children, men, older adults and recurrent UTI

A qualitative evidence synthesis in the British Journal of General Practice (21 September) brought together 32 studies reporting the views of 677 primary care practitioners on managing urinary tract infections. Study quality was rated high.

Five themes emerged. Practitioners felt confident with the typical case — a young woman with classic symptoms. Confidence fell when 'patient X' arrived: children, men, older people and women with recurrent infection, where the diagnosis is less certain and the stakes of under- or over-treating are higher. Weighing antibiotic benefit against resistance was a constant tension, sharpened by time pressure, staffing shortages and patients expecting a prescription. The relationship with the patient was what often decided the outcome.

This is not evidence about which treatment works, but it names where decisions go wrong. The atypical groups are exactly where a structured approach — a urine culture before antibiotics, a short safety-net plan, a follow-up call — protects both patient and prescriber.

In children, men, older adults and recurrent UTI, send a urine culture before starting antibiotics.
Give a clear safety-net plan for fever, loin pain or no improvement within 48 hours.
Discuss expectations about antibiotics openly; a delayed prescription can suit mild cases in women.
Record the reasoning when you do not follow the usual pathway.

Why it matters

It shows the uncertainty lies with specific patient groups, which is where a structured approach helps most.

The statistics, in plain English

Qualitative syntheses describe experiences and reasons, not rates or effects. They are useful for understanding why practice varies, not for deciding which treatment works.