Can even experienced orthopaedic surgeons predict who will benefit from surgery when patients present with degenerative meniscal tears? A survey of 194 orthopaedic surgeons who made 3880 predictions

Key points

1. 01194 orthopaedic surgeons made 3880 predictions about real patients from a randomised trial, blinded to which treatment each patient actually received.
2. 02Only 50.0% of predictions were correct, exactly the rate expected by chance.
3. 03Experienced knee surgeons were no better than other surgeons (50.4% vs 49.5%, p=0.29).
4. 04Surgeons were especially poor at spotting patients who did not improve, predicting non-responders correctly only 34% of the time.
5. 05Most surgeons (78.4%) already preferred exercise therapy, yet still recommended surgery most often for patients who in reality did not benefit from it.

How it was conducted

Design

Electronic survey of surgeons using 20 real patient profiles drawn from the ESCAPE randomised controlled trial

Participants

194 orthopaedic surgeons and residents in the Netherlands and Australia (out of 1111 invited, 17% response)

Patient profiles

Middle-aged patients (45 to 70 years) with symptomatic non-obstructive meniscal tears, 5 best and 5 worst outcomes from each treatment arm

Task

For each profile, choose APM or exercise therapy and rate expected change in knee function on a 5-point Likert scale, blinded to actual treatment

Primary outcome

Percentage of correct predictions of treatment outcome

Analysis

Chi-square tests comparing experienced versus other surgeons and responders versus non-responders, significance set at 0.05

What they found

• Overall, 50.0% (95% CI 39.6% to 60.4%) of predictions were correct, equal to the proportion expected by chance.
• Experienced knee surgeons did no better than other surgeons: 50.4% (95% CI 48.6% to 52.2%) vs 49.5% (95% CI 48.0% to 51.1%), p=0.58 (reported as p=0.29 in the abstract).
• Correct predictions were higher for treatment responders at 66.0% (95% CI 57.0% to 75.0%) than for non-responders at 34.0% (95% CI 21.3% to 46.6%), p<0.001.
• Across all profiles, 21.6% of surgeons chose APM and 78.4% chose exercise therapy as preferred treatment.
• Surgeons recommended meniscectomy most often (25.2%) for patients who were actually non-responders to surgery.
• In cases where surgery was preferred, mean expected change was 4.3 of 5 from meniscectomy versus 3.2 from exercise therapy (mean difference 1.1 points).
• Responders selected by treatment arm had mean IKDC improvement of 64.6 points (APM) and 54.0 points (exercise therapy); non-responders had mean IKDC change of -11.5 points (APM) and -13.1 points (exercise therapy).

Limitations

• The response rate was low at 17%, raising the potential for selection bias.
• A digital survey based on written profiles cannot fully capture real-world clinical assessment.
• Poor prediction may partly reflect unknown variables that determine outcome after meniscal treatment, not just surgeon judgement.
• Profiles were deliberately chosen as the clearest best and worst cases, so they may not represent typical patients seen in clinic.

Why it matters

For patients

If you are middle-aged with a degenerative meniscal tear, a surgeon's confidence that surgery will help you is not a reliable guide, and exercise therapy is a reasonable first choice.

For clinicians

Surgeon judgement about who will benefit from APM is no better than chance regardless of experience, so treatment decisions should lean on trial evidence favouring exercise therapy as first-line care.

For readers

Even expert clinical intuition can be no more accurate than a coin toss, underscoring the value of objective trial evidence over personal experience.