Design: randomised, double-blind, double-dummy prospective trial, 1:1
Population: 1,194 adults aged 18-65 presenting to the emergency department with acute musculoskeletal trauma and moderate to severe pain
Primary outcome: reduction in numerical rating scale pain score at 30 minutes
Effect: −3.70 subcutaneous vs −4.42 intranasal; mean difference −0.72 (95% CI −0.95 to −0.48), below the prespecified 1.3-point threshold for clinical importance

A double-blind, double-dummy randomised trial enrolled 1,194 adults aged 18 to 65 presenting to the emergency department with acute musculoskeletal trauma and moderate to severe pain, assigning them to 20 mg ketamine subcutaneously or intranasally. Pain was measured on a numerical rating scale at 5, 10, 15, 30, 60, 90 and 120 minutes, with 30 minutes as the primary endpoint.

At 30 minutes the mean fall was 3.70 points with the subcutaneous route and 4.42 with the intranasal, a difference of 0.72 (95% CI −0.95 to −0.48) favouring intranasal. The authors then do the thing that makes this trial useful: they set the difference against the 1.3-point threshold for clinical importance and report that it — and every other timepoint — fell short of it. Secondary outcomes were the same apart from more minor adverse events in the subcutaneous group.

So the routes are interchangeable for this indication, with the needle-free one marginally ahead on both pain and minor side effects. That matters most where it removes a step: the patient with no intravenous access, the child or needle-phobic adult, the crowded department where a nurse and a sharps procedure are the rate-limiting resource, and the prehospital setting. Twenty milligrams intranasally is a dose a triage nurse can give under protocol.

Use the intranasal route where cannulation would delay analgesia or is unwanted
Twenty milligrams is the studied dose; do not extrapolate to other doses or to non-traumatic pain
Consider a nurse-initiated protocol at triage — the route makes that feasible
Watch for the usual ketamine effects regardless of route: emergence phenomena, nausea, transient hypertension
This population was 18 to 65 with musculoskeletal trauma; older patients and other pain types were not studied

Why it matters

It removes the needle and the cannula from the analgesia pathway without costing the patient anything measurable.

The statistics, in plain English

A difference of 0.72 points was statistically significant because the trial enrolled 1,194 patients, which makes small differences detectable. Pre-specifying 1.3 points as the threshold for clinical importance is what turns that into a usable answer: the trial found a real difference that is too small for any patient to notice. More trials should report both.