Ranta 2015.

Study characteristics
Methods	RCT Unit of randomisation: primary care practice/clinic
Participants	Place of recruitment: from local directories ‐ participants were selected if GP practices were involved Numbers randomised: total: 56 (I: 29; C: 27) % Completing final follow‐up: 100% Inclusion criteria: any TIA or stroke, never been exposed to this tool before, access to an organised TIA pathway consistent with the New Zealand TIA guideline Exclusion criteria: did not present to a participating primary or secondary health care providers during the study period or presented without neurologic/ophthalmologic symptoms Mean age years (SD): I: 69.8 (13.3) C: 72.3 (14.0) Gender (men): I: 67, C: 55 Ethnicity: I: European 156/172; C: European 101/119
Interventions	Intervention details (components, length, frequency): the tool is a Web‐based software program accessed via a GP computer desktop icon. Clicking the icon opens a single page of tick boxes asking for relevant aspects of the presenting illness. Depending on diagnosis and risk estimation, the tool recommends a guideline‐based management strategy. Location: primary care practice/clinic Mode of delivery: face‐to‐face Personnel responsible for delivery: primary care doctor Timing post‐stroke: after initial event Control: usual care
Outcomes	Stroke at 90 days, stroke and TIA at 90 days or vascular event/death
General Information	Funding: the New Zealand Health Research Council funded this trial Country of origin: New Zealand Publication language: English
Notes	Analysis method: generalised linear models Risk of bias: unclear
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	The clusters were general practices randomised one‐to‐one to intervention and control groups using a computer‐generated simple randomisation schedule
Allocation concealment (selection bias)	Low risk	Central allocation
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Low number of GP practices agreed to join in the study and none were excluded
Selective reporting (reporting bias)	Low risk	Outcomes were recorded electronically by individual GPs/from GP records
Other bias	Low risk	The study appears to be free of other sources of bias