Scheibein 2020.

Study characteristics
Methods	Design: Non‐randomized single‐arm Recruitment: From supported temporary accommodation (STA) service STA project workers and support staff identified potential study participants who smoked and wished to quit Setting: Dublin Simon Community, Ireland Study start date: Recruitment February 2019 (overall trial start date March 2018). Study end date: June 2019
Participants	Total N: 23 but only report baseline for the 9 that completed the study. % female 8.7% (2/23) at baseline, (22.2% 2/9) completed and reported; mean age 43.89 (SD 7.36); mean cpd 25.22 (SD 7.77); mean FTND 7.89 (SD 1.2); mean CO 21.89 (SD 14.4 corresp) E‐cigarettes use at baseline: no Motivated to quit: yes Inclusion criteria: > 5 CO ppm (carbon monoxide) Active smoking status Expressed intention to quit using ENDS‐device Exclusion criteria: Self‐reported pregnancy Exhibition of florid psychotic or substance use‐related symptoms which could have affected ability to consent
Interventions	EC: Refillable Electronic‐cigarette: Endura T22e Electronic Nicotine Delivery System and 2 10 ml bottles of fluid strengths (0, 6, 11, 18 and 20 mg/ml) and flavors ('Purple Berry', 'Ice Menthol', 'Regular Blend' and 'American Tobacco')
Outcomes	Baseline (‘week 1’), week 4, week 8, week 12: CO, adverse events Also number of cigarettes smoked; Fagerström Test Scores
Study funding	This study was completed as part of a Tobacco Harm Reduction Scholarship funded by Knowledge Action Change
Author declarations	FS was a recipient of a Tobacco Harm Reduction Scholarship provided by Knowledge Action Change. He is currently the recipient of an Enhanced Scholarship from the same organization. AM and KM acted as mentors for both the Tobacco Harm Reduction Scholarship and Enhanced Scholarship. AM is an associate of New Nicotine Alliance. KM is a recipient of a grant from the Foundation for a Smoke Free World. JW declares no interests. WR declares no interests
Notes	New for 2021 update
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	Only 1 arm
Allocation concealment (selection bias)	High risk	Not randomized
Blinding of participants and personnel (performance bias) All outcomes	High risk	Not randomized
Blinding of outcome assessment (detection bias) All outcomes	High risk	Self‐reported adverse events
Incomplete outcome data (attrition bias) All outcomes	High risk	9/23 completed. Reason was many people moved away so not linked to unacceptability of the study. Incomplete paperwork to enable to be followed
Selective reporting (reporting bias)	Unclear risk	Protocol published afterwards