Kao 2016.

Methods	A 2‐step cluster‐randomised trial to compare the long‐term effects of spaced retrieval (SR) training and SR training combined with Montessori activities (SR+M) and a control condition in persons with dementia suffering from hyperphagia
Participants	148 participants, recruited from 8 dementia special care units in China, who had received a diagnosis of dementia and hyperphagia‐related behaviours from a physician
Interventions	Participants in the SR condition (n = 48) and in the SR+M condition (n = 52) participated in thirty 40‐minute sessions over 6 weeks, which were held in dementia special care units. The SR sessions involved learning and reviewing a memory message. SR+M sessions involved the same activity, as well as practice‐based structured Montessori activities Interventions were delivered by 2 memory trainers who had received training in SR and Montessori activity methods Participants in the control condition (n = 48) participated in the usual activities of the institution, which did not involve any particular memory training activities
Outcomes	Outcomes included hyperphagic behaviours, measured by a 19‐item scale of hyperphagia in residents with dementia, and associated caregiver distress, measured by a scale of distress of the caregiver in response to hyperphagic behaviours Participants were assessed at baseline, immediately after the training period, and at months 1, 3, and 6 after completion of the training
Country	China
Registration status	Retrospectively registered
Conflict of Interests	No
Notes	We contacted study authors to ask for relevant scores (at all assessment occasions) and to find out more details about the scales used ("scale for hyperphagia in residents with dementia", and "scale of distress of the caregiver to hyperphagic behaviour"). No response was received
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Although the study was described as a randomised controlled trial, no information on the method of randomisation was provided
Allocation concealment (selection bias)	High risk	Study authors did not mention allocation concealment; for this reason, we assumed this was not done
Blinding of participants and personnel (performance bias) All outcomes	High risk	Study authors described the study as a single‐blind RCT; however, this refers to the outcome assessment
Blinding of outcome assessment (detection bias) All outcomes	Low risk	"The research assistants were blinded to the randomisation of subjects"
Incomplete outcome data (attrition bias) All outcomes	Low risk	No outcome data were missing. Study authors reported the results of those who completed the study and explained the reasons why some participants dropped out. They provided a flow diagram of the study that is adherent to CONSORT
Selective reporting (reporting bias)	High risk	Study authors did not present findings for some of the outcomes mentioned in the "Methods" section
Other bias	Low risk	Study appears to be free of other sources of bias