Moy 2015.

Methods	Location: United States and Puerto Rico Design: randomised controlled trial Unit of allocation: individual Start date: December 2011 End date: January 2013 Duration of intervention: 12 months Time points measured: baseline, 4 months, and 12 months (only baseline and 4 months reported in this article. Moy 2016 reports 12‐month data)
Participants	Population description: Participants were identified from the Veterans Association (VA) national database of patients with COPD who received care within the year before enrolment from a primary care provider, cardiologist, or pulmonologist within the VA healthcare system. All participants were over 40 years of age and had access to a computer with Internet access, used email regularly, were able to walk 1 block, and had received medical clearance from a doctor Setting: own home Method of recruitment: a random subset of 29,000 veterans (half urban and half rural) sent an email invitation Total number randomised: 238 participants were randomised in a 2:1 ratio to intervention (n = 154) or control (n = 84) groups Participants: 238, 223 were male with only 15 female participants. This is unusual but was not unexpected given that the target population consisted of veterans Age: average age, 67 years
Interventions	The intervention comprised online and wearable technology. Participants were instructed to wear a pedometer daily and to upload step‐count data regularly. Each participant had a weekly goal that was based on: • average of most recent 7‐day step counts + additional 600 steps; • previous goal + 600 steps; or • 10,000 steps/d Participants had Web access to step‐count feedback, allowing self‐monitoring; weekly goal setting, educational/motivational content and an online community forum. Valid wear days were those on which at least 100 steps and 8 hours of step counts were recorded
Outcomes	Self‐reported health‐related quality of life (SGRQ) and daily step counts (pedometer) were reported at 4 months and at 12 months. Hospitalisations and acute exacerbations were reported at 12 months only
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Interested participants completed an online questionnaire that assessed inclusion criteria, after which a computer algorithm determined eligibility and participants were randomized in a 2:1 ratio to intervention or control groups (p 129)
Allocation concealment (selection bias)	Low risk	Allocation was done by computer, and both groups received a pedometer; therefore, group allocation was unclear to all participants (p 135)
Blinding of participants and personnel (performance bias) All outcomes	High risk	Blinding was not possible; therefore, although limited as all study activity was online, participants may have been influenced by prior beliefs about whether or note the intervention is likely to work. Group allocation was revealed online (p 130). Also, participants were required to upload step‐count data weekly, which may have introduced bias
Blinding of outcome assessment (detection bias) All outcomes	High risk	No blinding of outcome assessment; therefore, participants may have given what was perceived as the 'right' response. However, this may have been limited by lack of direct contact with researchers and other participants, as questionnaires were completed by participants online (p 129)
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Reasons for missing outcome data not reported (p 130)
Selective reporting (reporting bias)	Low risk	Both groups reported but third time point measured at 12 months not reported; email correspondence with study authors indicates that paper reporting results at 12 months has been prepared for publication
Other bias	Low risk	Study appears to be free from other sources of bias