Vandemheen 2009.

Study characteristics
Methods	Single blind RCT with 1‐year follow up. Compared an evidence‐based decision aid for patients with advanced CF considering referral for lung transplantation to usual care. Primary objectives: to assess whether the decision aid increased knowledge about patient options, improved realistic expectations, and decreased decisional conflict.
Participants	Population of interest: N = 155 (assessed for eligibility). Number randomised: 149. Number of participants received the intervention: 148 (intervention group n = 70; control group n = 79). Final sample size after 1‐year follow up: N = 125 (intervention group n = 60; control group n = 65). Recruited from nine Canadian and five Australian outpatient CF centres. Eligibility criteria: 1.  > 18 years of age; 2. FEV1 ≦ 40% predicted; 3. Able to read English or French. Exclusion criteria: 1. patients who previously received a lung transplant; 2. patients on lung transplantation waiting list.
Interventions	All study patients received an educational session on the process of referral and the risks and benefits of lung transplantation from their CF physician and/or CF team. All patients received a copy of a pamphlet entitled 'Cystic Fibrosis and Lung Transplantation' and were provided addresses to access two generic websites on lung transplantation. After randomisation: Intervention group: Received a sealed package containing a paper version of the decision aid and a website address with a user name and password, where they could access the decision aid online (http://decisionaid.ohri.ca/decaids.html). Control group: received a package contained blank pages and a letter explaining why blank pages were included.
Outcomes	The following primary outcomes were assessed at baseline and 3 weeks after randomisation: participants’ knowledge and realistic expectations with designed questionnaire (structured format as per the International Patient Decision Aids Standards Guidelines); decisional conflict scores (Decisional Conflict Scale, a validated 16‐item scale).
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	The authors reported that randomisation was performed via a "a randomization schedule prepared through a computer‐generated random listing of the two treatment allocations blocked in variable blocks of two or four and stratified by site and by infection status with Burkholderia cepacia" (Vandemheen 2009, p.762).
Allocation concealment (selection bias)	Low risk	The authors reported that a central allocation was conducted (Vandemheen 2009, p. 762).
Blinding (performance bias and detection bias) All outcomes	Low risk	The authors provided the following information regarding blinding: "Neither research staff nor medical staff as aware of the treatment assignment before and after randomization" (Vandemheen 2009, p.762).
Incomplete outcome data (attrition bias) All outcomes	Low risk	The number of drop‐outs for 3‐week follow up and 12‐month follow up were reported for each group.
Selective reporting (reporting bias)	Low risk	The published report includes all expected outcomes, including those that were pre‐specified as primary and secondary outcomes.