Grunau 2011.

Study characteristics
Methods	Single centre, open‐label randomised, cross,over trial, feasibility study.
Participants	Participants (n = 11) were adults aged > 18 (mean 73) years receiving warfarin therapy for > 3 months from a private family practice in British Columbia (Canada), and competent to use drug‐adjustment nomograms. Exclusions included: severe psychiatric disease; serious language barrier, poor physical acuity; primary care physician judgement of unsuitability.
Interventions	Self‐management vs usual care Participants were randomised to: a) self‐management (n = 6): written instructions on how to adjust dosage of anticoagulant, including a dose‐adjustment nomogram. Participants were asked to contact the study centre by phone/in person if they were having difficulty with the self‐management or if their INR value > 5. b) usual care (n = 5): physician anticoagulation management at the clinic. Measurement of INR among all participants was done by community laboratories and the results communicated by mail or in person. Participants were followed for 4.5 months, then allocation was reversed.
Outcomes	Primary outcome: proportion of INR values in therapeutic range. Secondary outcomes: number of days in therapeutic range. Feasibility end points included proportion of eligible participants consenting, preferred management strategy of participants at the end of the study, a treatment‐related satisfaction survey and additional office visits and phone calls pertaining to anticoagulation. Safety endpoints were bleeding or thromboembolic events.
Trial identification	Clinicaltrials.gov NCT00925028
Study duration	9 months
Oral anticoagulant used	Warfarin
Notes	Feasibility study: only 11 participants
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computerised random sequence generation.
Allocation concealment (selection bias)	Unclear risk	Not specified
Intention to treat analysis	Low risk	Outcomes were analysed according to ITT
Reporting of losses of follow‐up	Low risk	There were no losses to follow‐up.
Blinding	Low risk	Participants were not blinded; clinic staff were unaware of treatment allocation.