PT - JOURNAL ARTICLE AU - Miriam J Johnson AU - David C. Currow TI - Opioids for breathlessness: a narrative review AID - 10.1136/bmjspcare-2020-002314 DP - 2020 Sep 01 TA - BMJ Supportive & Palliative Care PG - 287--295 VI - 10 IP - 3 4099 - http://spcare.bmj.com/content/10/3/287.short 4100 - http://spcare.bmj.com/content/10/3/287.full SO - BMJ Support Palliat Care2020 Sep 01; 10 AB - Chronic breathlessness is a disabling and distressing condition for which there is a growing evidence base for a range of interventions. Non-pharmacological interventions are the mainstay of management and should be optimised prior to use of opioid medication. Opioids are being implemented variably in practice for chronic breathlessness. This narrative review summarises the evidence defining current opioids for breathlessness best practice and identifies remaining research gaps. There is level 1a evidence to support the use of opioids for breathlessness. The best evidence is for 10–30 mg daily de novo low-dose oral sustained-release morphine in opioid-naïve patients. This should be considered the current standard of care following independent, regulatory scrutiny by one of the world’s therapeutics regulatory bodies. Optimal benefits are seen in steady state; however, there are few published data about longer term benefits or harms. Morphine-related adverse events are common but mostly mild and self-limiting on withdrawal of drug. Early and meticulous management of constipation, nausea and vomiting is needed particularly in the first week of administration. Serious adverse events are no more common than placebo in clinical studies. Observational studies in severe chronic lung disease do not show excess mortality or hospital admission in those taking opioids. We have no long-term data on immune or endocrine function. There are promising data regarding prophylaxis for exertion-related breathlessness, but given the risks associated with transmucosal fentanyl, caution is needed with regard to clinical use pending longer term, robust safety data.