DRAGON is a 0–10 point pretreatment prognostic score for adults with acute ischaemic stroke receiving IV alteplase. It uses baseline non‑contrast CT and simple clinical variables to estimate 3‑month functional outcome on the modified Rankin Scale (mRS).

Evidence & context

Derived in 1,319 consecutive IV alteplase–treated ischaemic stroke patients at Helsinki University Central Hospital (≤4.5 hours from symptom onset), outcome was 3‑month mRS, with good outcome defined as 0–2 and miserable outcome as 5–6.
Externally validated in 330 patients treated at University Hospital Basel, with similar discrimination (area under the ROC curve about 0.80–0.84 for both good and miserable outcome).
A large multicentre registry from 12 stroke centres (n=4,519) confirmed good performance in routine practice and similar accuracy in anterior and posterior circulation strokes.

Risk gradient

Published cohorts show a steep relationship between DRAGON score and the probability of independence or severe disability/death at 3 months.

DRAGON score	Good outcome (mRS 0–2)	Miserable outcome (mRS 5–6)
0–1	≈96%	0%
2	≈88–93%	≈2%
3	≈74–78%	≈4–5%
8–10	0%	≈90–100%

These figures are cohort‑level probabilities and may not reflect outcomes in every service or era.

Further validation and limitations

External validations, including elderly Spanish cohorts and other national registries, generally support the risk gradient, several series report no independent survivors at 3 months with DRAGON scores ≥8.
The score was developed in IV alteplase–treated patients before widespread endovascular thrombectomy or tenecteplase use, so performance may differ in modern reperfusion pathways.
The tool predicts functional outcome and does not estimate symptomatic intracerebral haemorrhage risk, it should supplement, not replace, contemporary reperfusion guidelines.

Evidence is anchored to the DRAGON derivation study.

DRAGON Score for Post-TPA Stroke Outcome