A trial randomised 1,028 emergency patients with acute non-traumatic headache to intravenous magnesium sulfate added to paracetamol, or paracetamol with placebo. Treatment success occurred in 78.9% of the magnesium group and 65.1% of placebo, a difference of 13.8 percentage points with an interval from 8 to 19.
That headline sits awkwardly with the pain scores. Every timed difference on the numerical rating scale fell below 1.3 points — the threshold generally accepted as the smallest change a patient can actually perceive. So the binary success measure moved substantially while the continuous measure it derives from did not move meaningfully.
The secondary outcomes are where the case for magnesium actually rests. Rescue analgesia was needed in 7.1% against 15.3%, a difference of 8.2 percentage points — that is a halving, and it is a real operational and clinical outcome rather than a scale artefact. Satisfaction was higher, 91.7% against 85.1%. Adverse events were more frequent, 15.4% against 11.1%, described as mild.
So: a reasonable adjunct where rescue opioids are what you are trying to avoid, and not something to describe to a patient as better pain relief.
- Success rose 13.8 points while every pain-score difference stayed under the threshold
- Rescue analgesia halved, from 15.3% to 7.1% — the most defensible reason to use it
- Mild adverse events rose from 11.1% to 15.4%
- Non-traumatic headache, undifferentiated — not a migraine-specific trial
- Do not promise a patient better pain relief; promise less need for a second drug
The statistics, in plain English
A dichotomous outcome like treatment success can shift sharply while the underlying continuous scale barely moves, because a success threshold converts small changes near the cut-point into category changes. That is why both were reported, and why the authors flagged the discrepancy. The rescue analgesia difference is the more trustworthy result: it is a discrete event, not a threshold applied to a subjective scale.
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 emergency & critical care, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free