Paperonium

Anaesthesia · BMJ Glob Health · 2024

General doctors’ spinal injections: the paper’s range is 90 times too narrow, but they still pass

The paper’s summary90 times too narrowIts own counts2.75 fewer to 3.58 more, under 5The trial’s “no worse” limit5 more per 1000.36 to 0.43−40 failures per 100 patients456no differencelimitfewer failuresmore failures
The paper’ssummary90 timestoo narrowIts own counts2.75 fewerto 3.58 more,under 5The trial’s “noworse” limit5 more per 1000.430.36−505no differencefailuresper 100 patientslimitfewer failuresmore failures

Extra failures per 100 patients with general doctors

The paper says

Trained general doctors’ spinal injections (which numb the lower body for surgery) failed in 7 of 231 patients, 3.0 per 100; specialists’ in 5 of 191, 2.6: a gap of 0.4. The paper’s summary says the true gap lies somewhere between 0.36 and 0.43.

Its own data show

Those counts, from the paper’s own table, allow a true gap anywhere from 2.75 fewer to 3.58 more failures per 100 with general doctors: 90 times as wide. The summary’s range never reaches zero (no difference), so it says general doctors reliably fail more; the paper’s own test says the gap could easily be chance.

A range covers every true gap the data cannot rule out. The trial calls general doctors “no worse” if the range’s top stays under 5 more failures per 100. The real top, 3.58, passes with 1.4 to spare, not the 4.6 that 0.43 implies. The paper says this may support sharing the work with general doctors in rural Indian hospitals.

“The summary” means the paper’s abstract: the short summary printed at its top, which is the part most readers see.

Read the paper →Full technical write-up →

Task-sharing spinal anaesthesia care in three rural Indian hospitals: a non-inferiority randomised controlled clinical trial · BMJ Glob Health (2024) · doi:10.1136/bmjgh-2023-014170