Two of today's findings are negatives, and one of them is only interpretable because of how the control arm was built.
In the ablation trial, patients who received a sham improved by 15.7 points on a 100-point quality-of-life scale. Had the comparator been medical therapy, or nothing, that 15.7 points would have been counted as ablation's benefit and the trial would have read as a clear success. The sham did not create a placebo effect; it revealed one that was there all along and had been credited to the procedure.
The habit generalises beyond invasive trials. Whenever an outcome is subjective and both arms improve, the interesting number is the control arm's change, not the difference. A large control-arm improvement means regression to the mean, natural history and expectation are doing most of the work — and a trial without a credible control cannot tell you how much of the result is yours to claim.
- Read the control arm's change before the between-group difference
- Subjective outcomes plus an unblinded control means the effect is probably overstated
- A large sham response is information, not noise — it quantifies expectation
- Ask whether the comparator resembles what your patient would otherwise receive
- Useful in consent: say what the procedure adds, not what the patient will feel
The statistics, in plain English
A single-arm before-and-after comparison measures the treatment plus natural history plus regression to the mean plus expectation. A randomised comparator subtracts the first three; only a blinded sham subtracts the fourth. That is why a sham-controlled result can look disappointing while being the most trustworthy estimate available.
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 top clinical updates, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free