Key messages:

• Varenicline (a medicine that decreases nicotine cravings) when compared to placebo (dummy pill), likely helps people living with HIV who smoke tobacco to stop smoking for six months or more, and may not increase the chances of experiencing serious unwanted effects.
• We don't know if other methods that are used to help people stop using tobacco can help people living with HIV to quit for six months or more because we did not find enough information.
• Future studies should be bigger and provide information on serious unwanted events.

Why is tobacco use among people living with HIV a problem? Many people living with HIV worldwide use tobacco, that is, the tobacco plant leaf and its products, for example through smoking, chewing, sucking or sniffing. Tobacco use causes a range of health problems and many deaths, but people become addicted to nicotine in tobacco, and find it difficult to quit. Smoking rates are about four times higher in people living with HIV than in the general population. Unfortunately, even with access to effective HIV treatment, people living with HIV can lose about 12 years of their life because of smoking ‐ more than double the number of years that they are likely to lose because of HIV infection on its own.

How can people living with HIV stop using tobacco? Methods that are used to help people stop using tobacco include medicines such as nicotine replacement therapy (NRT), varenicline and cytisine (medicines that decrease nicotine cravings) and bupropion (an antidepressant). Other methods include behavioural therapies such as information on the risks of smoking (brief advice), or individual or group counselling (behavioural support). Some health services try changing the way they deliver care (system change). It is unclear whether these interventions can help people living with HIV quit tobacco use.

What did we want to find out? We wanted to find the best methods to help people living with HIV stop using tobacco, and see whether there were unwanted effects.

What did we do? We searched for studies that investigated quit methods aimed directly at adults with diagnosed HIV. Also, studies on system change, aimed at people living with HIV who were receiving care, or healthcare staff working in these facilities. Studies could compare quit methods with a placebo (dummy) or no treatment, or another method. We only included studies that examined quitting tobacco for six months or longer. Ideally, quitting had to be verified with a chemical test. We also wanted to find out if the quit methods caused serious unwanted effects.

We compared and summarised the results of the studies and rated our confidence in the evidence, based on factors such as study methods and sizes.

What did we find? We found 17 studies with 9959 people. Twelve studies were in the USA. Fifteen studies investigated quitting smoking, either with individuals or groups. Two studies evaluated changes in the way care was delivered to people living with HIV attending a health facility.

• Behavioural support or system‐change compared with no behavioural support or less intensive behavioural support
  There was no clear evidence that behavioural support was more effective for quitting tobacco than brief advice or no support (7 studies, 2314 people).
• Changing the quitline process from referring a patient to a quitline by fax to calling the quitline and transferring the call to the patient whilst they are still in hospital did not show clear evidence of better quit rates (1 study, 25 people).
• None of the studies in this comparison assessed serious unwanted effects.

Medications for quitting tobacco compared with placebo, no medication or another type of medication

• Varenicline likely helps people living with HIV to quit smoking compared to placebo (2 studies, 427 participants).
• Single studies looked at behavioural support plus NRT compared with brief advice, behavioural support plus NRT compared with behavioural support alone, varenicline compared with NRT, and cytisine compared with NRT. The evidence did not show that any of these methods helped people to stop using tobacco.
• There are probably no more unwanted effects with varenicline than placebo (2 studies, 427 participants). There may be no difference in the proportion of people who experience serious unwanted effects between behavioural support plus NRT and behavioural support only (1 study, 560 people). No studies assessed serious unwanted effects for behavioural support plus NRT compared with brief advice, varenicline compared with NRT, and cytisine compared with NRT.

What are the limitations of the evidence? Our overall confidence in the evidence is moderate to very low, mainly because there were not enough studies to be certain about the results, and sometimes we did not have all the information we needed to evaluate the quality of the studies.