Interventions started during hospitalisation to help people to stop smoking

Key messages

- When people who smoke are admitted to a hospital, they can be helped to quit smoking if they receive stop-smoking counselling that begins in hospital and continues for at least a month after they return home, compared to no counselling.

- Medications, such as nicotine patches and varenicline, in combination with counselling also help people quit smoking post-discharge. These treatments work better than not starting counselling or medication during a hospitalisation.

- Evidence supports hospitals and hospital clinicians offering in-hospital and post-discharge cessation support to patients, and demonstrates that patients may benefit from beginning their quit-smoking journey prior to, or upon, hospital discharge, in order to stay quit post-discharge.

What did we want to find out?

We wanted to find out what interventions are helpful to support hospitalised people who smoke in quitting cigarette smoking. Our main goal was to find out which treatments can help hospitalised patients stop smoking for at least six months. This is important because smoking contributes to many health problems, including cancers, heart disease, and lung disease. People who smoke and are admitted to a hospital to treat a medical illness, especially an illness that is related to smoking, might be more receptive to advice to quit smoking. The smoke-free hospital environment may also help them to try out not smoking and to start treatment to remain smoke-free after leaving the hospital.

What did we do?

We searched for studies that looked at stop smoking interventions (medications versus no medications or dummy pill and/or counselling versus no counselling) that began during a medical hospitalisation. Smoking cessation medications generally work to reduce withdrawal symptoms and stave off cravings; NRT by providing low levels of nicotine without the poisonous chemicals, and drugs such varenicline and bupropion, which do not contain nicotine, by directly targeting the reward and pleasure/addictive centres in the brain. Providing these treatments before someone is discharged from the hospital allows them to get a headstart on quitting smoking, as their hospitalisation is smoke-free. We looked for randomised controlled trials or quasi-randomised controlled trials, in which the treatments people received were decided at random or semi-random. Randomised studies typically give the most reliable and robust evidence about the effects of a treatment.

What did we find?

We found 82 studies of smoking cessation interventions that began in the hospital with 42,273 participants. These studies compared counselling (versus no counselling) and/or medications such as nicotine replacement therapy, varenicline, and bupropion (versus no medication or placebo). The studies took place across 17 countries (Australia, Belgium, Brazil, Canada, China, Denmark, France, Ireland, Israel, Japan, the Netherlands, Norway, South Korea, Spain, Tunisia, United Kingdom, and the United States).

Main results

People are more likely to stop smoking for at least six months if they receive stop-smoking counselling that begins in hospital and continues for more than a month after discharge than if they receive no counselling (28 studies, 8234 people). Shorter and less intensive counselling was less effective. People are also more likely to stop smoking for at least six months if nicotine replacement therapy (NRT) is started in the hospital (8 studies, 3838 people), or if varenicline is started in the hospital (4 studies, 829 people), than if they do not receive medication or if they receive a placebo medication. There was minimal evidence for bupropion's effectiveness in helping individuals quit smoking at six months. People who receive counselling in the hospital are more likely to quit smoking after discharge when using a combination of counselling and medication than when receiving neither (7 studies, 5610 people). Finally, there was insufficient evidence of quitlines' effectiveness for providing support after hospital discharge.

What are the limitations of the evidence?

Our results are based on numerous studies. However, we need more studies to confirm some of our findings, including buproprion and varenicline and lower-intensity counselling. We also need more studies to test whether digital interventions work - these interventions are promising as they might have greater reach at a lower cost. Some of our studies also had issues with their design and conduct, which made us less certain about our findings. These results could change when more evidence becomes available.

How up to date is this evidence?

The evidence is up to date to September 2022.

Authors' conclusions: 

Offering hospitalised patients smoking cessation counselling beginning in hospital and continuing for over one month after discharge increases quit rates, compared to no hospital intervention. Counselling provided only in hospital, without post-discharge support, may have a modest impact on quit rates, but evidence is less certain. When all patients receive counselling in the hospital, high-certainty evidence indicates that providing both counselling and pharmacotherapy after discharge increases quit rates compared to no post-discharge intervention. Starting nicotine replacement or varenicline in hospitalised patients helps more patients to quit smoking than a placebo or no medication, though evidence for varenicline is only moderate-certainty due to imprecision. There is less evidence of benefit for bupropion in this setting. Some of our evidence was limited by imprecision (bupropion versus placebo and varenicline versus placebo), risk of bias, and inconsistency related to heterogeneity.

Future research is needed to identify effective strategies to implement, disseminate, and sustain interventions, and to ensure cessation counselling and pharmacotherapy initiated in the hospital is sustained after discharge.

Read the full abstract...

In 2020, 32.6% of the world's population used tobacco. Smoking contributes to many illnesses that require hospitalisation. A hospital admission may prompt a quit attempt. Initiating smoking cessation treatment, such as pharmacotherapy and/or counselling, in hospitals may be an effective preventive health strategy. Pharmacotherapies work to reduce withdrawal/craving and counselling provides behavioural skills for quitting smoking. This review updates the evidence on interventions for smoking cessation in hospitalised patients, to understand the most effective smoking cessation treatment methods for hospitalised smokers.


To assess the effects of any type of smoking cessation programme for patients admitted to an acute care hospital.

Search strategy: 

We used standard, extensive Cochrane search methods. The latest search date was 7 September 2022.

Selection criteria: 

We included randomised and quasi-randomised studies of behavioural, pharmacological or multicomponent interventions to help patients admitted to hospital quit. Interventions had to start in the hospital (including at discharge), and people had to have smoked within the last month. We excluded studies in psychiatric, substance and rehabilitation centres, as well as studies that did not measure abstinence at six months or longer.

Data collection and analysis: 

We used standard Cochrane methods. Our primary outcome was abstinence from smoking assessed at least six months after discharge or the start of the intervention. We used the most rigorous definition of abstinence, preferring biochemically-validated rates where reported. We used GRADE to assess the certainty of the evidence.

Main results: 

We included 82 studies (74 RCTs) that included 42,273 participants in the review (71 studies, 37,237 participants included in the meta-analyses); 36 studies are new to this update. We rated 10 studies as being at low risk of bias overall (low risk in all domains assessed), 48 at high risk of bias overall (high risk in at least one domain), and the remaining 24 at unclear risk.

Cessation counselling versus no counselling, grouped by intensity of intervention

Hospitalised patients who received smoking cessation counselling that began in the hospital and continued for more than a month after discharge had higher quit rates than patients who received no counselling in the hospital or following hospitalisation (risk ratio (RR) 1.36, 95% confidence interval (CI) 1.24 to 1.49; 28 studies, 8234 participants; high-certainty evidence). In absolute terms, this might account for an additional 76 quitters in every 1000 participants (95% CI 51 to 103). The evidence was uncertain (very low-certainty) about the effects of counselling interventions of less intensity or shorter duration (in-hospital only counselling ≤ 15 minutes: RR 1.52, 95% CI 0.80 to 2.89; 2 studies, 1417 participants; and in-hospital contact plus follow-up counselling support for ≤ 1 month: RR 1.04, 95% CI 0.90 to 1.20; 7 studies, 4627 participants) versus no counselling. There was moderate-certainty evidence, limited by imprecision, that smoking cessation counselling for at least 15 minutes in the hospital without post-discharge support led to higher quit rates than no counselling in the hospital (RR 1.27, 95% CI 1.02 to 1.58; 12 studies, 4432 participants).

Pharmacotherapy versus placebo or no pharmacotherapy

Nicotine replacement therapy helped more patients to quit than placebo or no pharmacotherapy (RR 1.33, 95% CI 1.05 to 1.67; 8 studies, 3838 participants; high-certainty evidence). In absolute terms, this might equate to an additional 62 quitters per 1000 participants (95% CI 9 to 126). There was moderate-certainty evidence, limited by imprecision (as CI encompassed the possibility of no difference), that varenicline helped more hospitalised patients to quit than placebo or no pharmacotherapy (RR 1.29, 95% CI 0.96 to 1.75; 4 studies, 829 participants). Evidence for bupropion was low-certainty; the point estimate indicated a modest benefit at best, but CIs were wide and incorporated clinically significant harm and clinically significant benefit (RR 1.11, 95% CI 0.86 to 1.43, 4 studies, 872 participants).

Hospital-only intervention versus intervention that continues after hospital discharge

Patients offered both smoking cessation counselling and pharmacotherapy after discharge had higher quit rates than patients offered counselling in hospital but not offered post-discharge support (RR 1.23, 95% CI 1.09 to 1.38; 7 studies, 5610 participants; high-certainty evidence). In absolute terms, this might equate to an additional 34 quitters per 1000 participants (95% CI 13 to 55). Post-discharge interventions offering real-time counselling without pharmacotherapy (RR 1.23, 95% CI 0.95 to 1.60, 8 studies, 2299 participants; low certainty-evidence) and those offering unscheduled counselling without pharmacotherapy (RR 0.97, 95% CI 0.83 to 1.14; 2 studies, 1598 participants; very low-certainty evidence) may have little to no effect on quit rates compared to control.

Telephone quitlines versus control

To provide post-discharge support, hospitals may refer patients to community-based telephone quitlines. Both comparisons relating to these interventions had wide CIs encompassing both possible harm and possible benefit, and were judged to be of very low certainty due to imprecision, inconsistency, and risk of bias (post-discharge telephone counselling versus quitline referral: RR 1.23, 95% CI 1.00 to 1.51; 3 studies, 3260 participants; quitline referral versus control: RR 1.17, 95% CI 0.70 to 1.96; 2 studies, 1870 participants).