Vasilopoulou 2017.

Study characteristics
Methods	RCT; follow‐up: 14 months, control group: usual care (all study arms have 2 months' pulmonary rehab)
Participants	Eligible: unknown Randomised: 150, I (A): 50, I (B): 50, C: 50 Completed: 147, I (A): 47, I (B): 50, C: 50 Mean age: I (A): 67 years, I (B): 67 years, C: 64 years Sex (% male): I (A): 94, I (B): 76, C: 74 Inclusion criteria: older than 40 years of age, diagnosis of COPD, FEV₁ to FVC < 0.7 with FEV₁ < 80% predicted, with optimal medical treatment according to Global Initiative for Chronic Obstructive Lung Disease (GOLD) without regular use of systemic corticosteroids, history of acute exacerbations of COPD 1 year before entry into the study Major exclusions: diagnosis of orthopaedic, neurological, and other conditions that significantly impair exercise tolerance; respiratory disorder other than COPD; cognitive impairment and/or difficulties in managing electronic devices
Interventions	Intervention consisted of 2 months' outpatient rehabilitation followed by a 12‐month maintenance rehabilitation programme that is home‐based (group A) or hospital‐based (group B) Home‐based maintenance telerehabilitation (group A) Intervention components ‐ Individualised action plan ‐ Individually tailored physical exercise sessions to remote monitoring, adaption of exercise load based on exercise vital sign data (144 sessions) ‐ Self‐measurement of exercise vital sign data (heart rate and oxygen saturation) along with ratings related to symptoms of dyspnoea and leg discomfort immediately after completion of home exercise programme ‐ Manual entry of data into tablet and transmission of self‐collected data to web‐based platform 3 times per week (exercise vital sign data) or 2 times per week (pedometer, spirometry, oximetry, and responses to questionnaires (HRQoL, CAT, HADS, mMRC)) ‐ Review of transmitted data on secure web‐based server platform regularly by different healthcare professionals (3 or 4 times per week) ‐ Self‐management; psychological support and dietary and self‐management advice via scheduled weekly contacts with a physiotherapist, an exercise scientist, a dietician, and a physician through telephone or video conference ‐ Access to a pulmonologist at a call centre 5 days per week, 10 hours per day Hospital‐based maintenance rehabilitation (group B) Intervention components ‐ Continuation of rehabilitation programme twice weekly for 12 months, including exercise training, physiotherapy, dietary and psychological advice Intervention duration: 14 months (2 months ' outpatient rehabilitation + 12 months ' home‐based or hospital‐based maintenance rehabilitation) Involved disciplines: physiotherapist, dietician, physician (as case manager for home‐based telerehabilitation)
Outcomes	Rate of moderate to severe acute exacerbation (GOLD) (primary outcome), hospitalisations due to acute exacerbation of COPD (primary outcome), ED visits (primary outcome), rate of severe exacerbations (hospitalisations), rate of ED visits due to acute exacerbation of COPD that did not require hospital admission, functional capacity (peak work rate, 6MWD), daily physical activity (activity monitoring via accelerometer), health‐related quality of life (SGRQ), respiratory symptoms (CAT, mMRC); compliance with intervention
Notes	Dominant component: (A) telemonitoring, (B) structural follow‐up
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: “patients were randomised into three groups using a set of computer‐generated random numbers”
Allocation concealment (selection bias)	Unclear risk	Comment: no details on allocation concealment provided
Blinding of participants and personnel (performance bias) All outcomes	High risk	Quote: “our study design was not blinded”
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Comment: investigator aware of allocation. However primary endpoint objective outcomes
Incomplete outcome data (attrition bias) All outcomes	Low risk	Comment: analysis performed per protocol. Dropouts (n = 3) in home‐based telerehabilitation group unlikely to affect outcome
Selective reporting (reporting bias)	Low risk	Comment: all outcomes have been reported
Other bias	Unclear risk	Comment: o utcome from baseline (Month 2, after outpatient PR) to 12 months. Usual care group also had no access to outpatient PR. Hence maintenance of benefits might be related to conduct of initial PR