For adults who smoke and are admitted to hospital for medical or surgical care (whether smoking‐related or not), high‐certainty evidence shows that high‐intensity smoking cessation counseling (involving in‐hospital contact plus follow‐up contacts lasting more than one month) increases quit rates at six months or longer compared with no counseling (209 vs 154 per 1000 people; all results on average). Similarly, nicotine replacement therapy (NRT) increases quit rates compared to placebo, routine care, or no NRT (279 vs 210 per 1000 people). Combining pharmacotherapy and behavioral support also leads to higher quit rates compared with no post‐discharge support (190 vs 155 per 1000 people). Moderate‐certainty evidence shows that counseling involving one or more in‐hospital contacts, lasting more than 15 minutes, with no follow‐up support likely increases quit rates compared with no counseling (99 vs 78 per 1000 people).

On the other hand, varenicline probably has little or no effect on quit rates compared with placebo or no varenicline (moderate‐certainty evidence). Bupropion may have little or no effect compared with placebo or no bupropion, and post‐discharge synchronous (real‐time) behavioral support may also have little or no effect compared with no support (low‐certainty evidence).

Evidence for the effect of other interventions is very uncertain, including asynchronous behavioral support (e.g. online or application‐based), quitline referral versus no referral, and health system telephone counseling versus quitline referral. Evidence for different intensities of counseling is also very uncertain, including the effects of a single brief in‐hospital counseling session lasting 15 minutes or less without follow‐up, or counseling involving in‐hospital contact with follow‐up lasting one month or less...