Study	Reason for exclusion
Buffum 2004	A double blind, placebo controlled cross‐over trial, excluded as it examined acetaminophen (paracetamol) and not opioids
Husebo 2011	A cluster randomised controlled trial involving 352 participants, in 60 clusters in 18 nursing homes in five municipalities of western Norway. Participants in the intervention group received individual daily treatment of pain for eight weeks according to a step‐wise protocol, with paracetamol (acetaminophen), morphine, buprenorphine transdermal patch, or pregabalin. The control group received usual treatment and care. Excluded because not all participants received an opioid and there was no placebo arm
Kurrle 1995	Case series of 16 patients with severe dementia in a care home setting and treated with oral morphine, including six where morphine was given specifically for agitation, restlessness or disruptive behaviour such as calling out or shouting. Excluded because it was not a randomised controlled trial
Manfredi 2003	Study uses a novel trial design, a concealed 'before and after' design. Placebo was administered for 4 weeks and then a long‐acting opioid for 4 weeks. Patients and study nurses did not know whether placebo or opioid was administered. Although placebo was used, the trial was excluded as it was not a truly randomised trial
Passmore 2011	Case report of a single patient with dementia, incident pain with agitation treated with the short‐acting opioid sufentanil administered sublingually before personal care. Excluded because it was not a randomised controlled trial
Sloan 1989	Case report of two patients with severe dementia due to Huntington's disease treated with oral morphine. Excluded because it was not a randomised controlled trial