Cummings 1988.

Methods	Study design: factorial RCT (2x2) Location: USA Setting: n/a (remote intervention) Recruitment: callers to hotline telephone number, advertised through direct mailings, posters and brochures, radio talk shows, radio ads, newspaper articles and ads, and television announcements. Advert read ‘If you want to quit smoking, we can help’
Participants	N = 1895 Specialist population?: no Participant characteristics: 1232/1895 (15%) female; average age: 42 y; average cig/day: 28; nicotine dependence (average years smoked): 24 y Preference for quitting abruptly versus gradually: not reported
Interventions	All participants received the following interventions in the form of self‐help booklets sent through the post Comparator 1: Control booklet (not eligible as not a quitting intervention, but can not be classed as no smoking cessation treatment): provided information on the health hazards of smoking and the nature of tobacco addiction, but did not give specific advice on how to stop smoking (15 pages) Modality of support: self‐help booklet Overall contact time: n/a Number of sessions: n/a Pharmacotherapy: none Quit date set?: unclear whether booklet indicated when to quit Comparator 2: Cold‐turkey: high structure: booklet (45 ‐ 58 pages) instructed participants to quit abruptly. Day‐by‐day format. Participants were instructed to read the booklet every day and to carry out the activities given for each day of the plan Modality of support: self‐help booklet Overall contact time: n/a Number of sessions: n/a Pharmacotherapy: none Quit date set?: unclear whether booklet indicated when to quit Comparator 3: Cold‐turkey: low structure: booklet (45 ‐ 58 pages) instructed participants to quit abruptly. The same advice was provided as in the high‐structure group, i.e. tips in the same sequence; however, this was not in a structured form, and participants were instructed to examine the menu of information and to select those exercises they felt would be helpful rather than working through them systematically Modality of support: self‐help booklet Overall contact time: n/a Number of sessions: n/a Pharmacotherapy: none Quit date set?: unclear whether booklet indicated when to quit Intervention 1: Gradual ‐ high structure: booklet (45 ‐ 58 pages) instructed smokers to gradually reduce the number of cigarettes smoked over a brief period before quitting altogether. Participants were given a number of suggestions of how to reduce, such as setting daily goals, switching brands, changing habits, and delaying the first cigarette of the day. Participants were instructed to read the booklet every day and to carry out the activities given for each day of the plan Modality of support: self‐help booklet Overall contact time: n/a Number of sessions: n/a Pharmacotherapy: none Quit date set?: unclear whether booklet indicated when to quit Intervention 2: Gradual ‐ low structure: booklets (45 ‐ 58 pages) instructed smokers to gradually reduce the number of cigarettes smoked over a brief period before quitting altogether. Participants were given a number of suggestions on how to reduce, such as setting daily goals, switching brands, changing habits, and delaying the first cigarette of the day. The same advice was given as tips in the same sequence as in the high‐structure booklet, but not in a structured form. Participants were instructed to examine the menu of information and to select those exercises they felt would be helpful rather than working through them systematically Modality of support: self‐help booklet Overall contact time: n/a Number of sessions: n/a Pharmacotherapy: none Quit date set?: unclear whether booklet indicated when to quit
Outcomes	Definition of abstinence: continuous Longest follow‐up: 6 m Biochemical validation: none
Funding source	The National Cancer Institute (CA36265)
Author conflicts of interest	Not reported
Notes	Relevant comparisons: (excluding control booklet study arm) 1) Reduction versus abrupt; 2) Reduction method versus reduction method (structured versus unstructured) Factorial RCT, but no interaction detected between factors
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "randomization was done from a pre‐randomized list so subjects were randomized as they called into the study and were defined as eligible" (email communication).
Allocation concealment (selection bias)	Low risk	Self‐help intervention, involving minimal contact with investigators/enrolling clinicians
Blinding of outcome assessment (detection bias) All outcomes	Low risk	No biochemical verification; however, there was no contact with researchers and the relevant study groups (excluding control booklet study arm) did not differ in intensity
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	19.1% of total randomised lost to follow‐up; reported not to vary by arm. However, 18 additional participants are missing from the report results table. These participants are included in the current analyses and treated as non‐abstainers, but their allocation to treatment arms is unknown