Design: Registry cohort with cross-classified multilevel modelling
Population: 238,455 primary total hip arthroplasties in England, 2018–2022
Primary outcome: Variation in dual-mobility component use
Effect: Surgeon 32.4%, hospital 22.9%, patient factors about 10% of variation

This study linked 238,455 primary hip replacements from the National Joint Registry with hospital records for England between 2018 and 2022, and used multilevel modelling to split the variation in dual-mobility cup use between patients, surgeons and hospitals.

Dual-mobility cups were used in 7,032 hips, in patients who were older, frailer and more often treated for femoral neck fracture. But patient factors explained only about 10% of the variation. The surgeon explained 32% and the hospital 23%. The pattern held among high-volume surgeons.

That suggests whether a patient gets a dual-mobility cup depends more on who operates and where than on their dislocation risk. The authors argue for tools that identify high-risk patients and for trials in those groups. The finding is descriptive: it does not say which surgeons are right. It was published in September 2026.

Make dual-mobility decisions against explicit dislocation risk factors, not habit.
Consider agreeing unit criteria for dual-mobility use, for example in femoral neck fracture or neuromuscular disease.
Audit your own dual-mobility use against patient risk profiles.
Tell patients why a particular bearing is chosen for them.

Why it matters

Where you are operated on currently predicts your implant better than your risk does.

The statistics, in plain English

The percentages are shares of the explained variation in who received a dual-mobility cup. A marginal R² of 0.10 means patient factors alone explained about a tenth; adding surgeon and hospital raised it to about 60%. It describes practice patterns and says nothing about outcomes.