Design: Multicentre, patient- and assessor-blinded randomised controlled trial
Population: 122 adults with a displaced patellar fracture
Primary outcome: KOOS sub-scales at 12 months
Effect: Symptoms 10.4 (95% CI 4.4 to 16.4), sport 14.1 (3.4 to 24.7), quality of life 10.9 (2.7 to 19.0) points in favour of plating

A patient- and assessor-blinded multicentre randomised trial in The Bone & Joint Journal (1 September 2026) assigned 122 adults with a displaced patellar fracture to locking plate fixation (63) or tension-band wiring (59). The primary outcome was knee function on the five KOOS sub-scales at 12 months. Superiority was defined in advance as a significant difference, larger than the minimal important difference, in at least three sub-scales. About nine in ten completed follow-up.

Plate fixation met that test. It improved symptoms by 10.4 points, sport and recreation by 14.1, and quality of life by 10.9. Removal of hardware, fixation failure and reoperation were all more common after wiring.

Tension-band wiring is cheap, familiar and still the default in many units, but its high rate of hardware irritation and removal is well known. This trial puts patient-reported function on the side of the plate, with a strict definition of superiority.

Follow-up was one year, so long-term arthritis and late hardware problems are not yet known. Locking plates cost more, which matters where patients pay for implants. Where they can be afforded, this supports plating as the preferred fixation for displaced patellar fractures.

Consider locking plate fixation as the first choice for displaced patellar fractures where implants are available.
Discuss implant cost against the lower reoperation and hardware removal rates with patients who pay.
Warn patients treated with tension-band wiring that hardware removal is common.
Record KOOS or a similar score at one year to audit your own results.

Why it matters

It challenges the default of tension-band wiring with blinded trial evidence on function patients notice.

The statistics, in plain English

The predefined rule required differences larger than the smallest change patients notice, in at least three of five sub-scales, which is a high bar. The confidence intervals are wide (for sport, 3.4 to 24.7 points), so the true benefit could be smaller than the average, but all three exclude no difference.