The robotic knee trial records something rarely printed: recruitment fell as patients increasingly requested the robotic arm and surgeons lost clinical equipoise.
Equipoise — genuine uncertainty about which arm is better — is what makes randomisation ethical. When it goes, the trial cannot continue, and the results are interim by necessity rather than by design. That is why this trial reports two-year outcomes on the patients it managed to randomise before belief outran evidence.
The useful observation is about direction of travel. Here the technology's spread created the belief, and the belief then prevented the trial that would have tested it. The outcome measures that did move were early recovery and a sensitive patient-reported score; the established scales were null at two years. Had recruitment continued, we might know whether that pattern holds.
The habit: when a trial reports interim results because of lost equipoise, ask whether the belief that ended it came from the accumulating data or from outside the trial. If from outside — marketing, patient expectation, surgeon investment — the interim result is weaker evidence than its p values suggest, because the population that consented is no longer the population the question was about.
- Lost equipoise makes a trial's results interim rather than definitive
- Ask whether the belief came from the trial's data or from outside it
- Patients preferring one arm changes who consents to randomise
- Adoption preceding evidence is common with surgical technology
- Interim results from a truncated trial overstate precision
The statistics, in plain English
Trials stopped early for reasons unrelated to their own data suffer differently from those stopped for benefit. Here the problem is selection: as preference grew, the patients still willing to be randomised became less representative, so late enrolments differ systematically from early ones. No analysis recovers that, which is why the authors present the findings as interim.
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.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for orthopaedics, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free