Lee 2013.

Methods	An RCT comparing a computerised errorless learning‐based memory training programme (CELP) vs a therapist‐led errorless programme (TELP) and a control condition in people with mild AD
Participants	19 Chinese persons who were 60 years of age and older, had early dementia (according to their scores on the Chinese Dementia Rating Scale), and were recruited from psychogeriatric day hospitals, an elderly daycare centre in Kwai Chung District, and Kwai Chung Hospital (outpatients)
Interventions	Participants in the 2 experimental conditions were trained on an errorless learning‐based memory training programme. Participants in the CELP condition (n = 6) were trained on a computerised version of this programme, and participants in the TELP condition (n = 6) completed a therapist‐led version of this programme These 2 interventions lasted for approximately 6 weeks; a total of twelve 30‐minute individual training sessions were delivered twice a week Occupational therapists who received specialist training delivered the interventions The control condition (n = 7) was a wait‐list condition
Outcomes	Primary outcomes were global cognition, verbal learning, and memory and prospective memory Secondary outcomes included mood (depression), self‐care, and instrumental activities of daily living Participants were evaluated before and after the intervention period, and at 3‐month follow‐up assessment
Country	China
Registration status	No information provided; presumed to be unregistered
Conflict of Interests	No
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Although study authors described an RCT, they provided no information on the method of randomisation
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	Unclear risk	Participants in all 3 conditions were engaged in some activity, and it is not clear whether participants were aware of the condition they were given or of the research hypothesis. Interventions were led by trained OTs, and the same feedback was provided to participants in both experimental groups
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Assessors who performed the evaluation were blinded
Incomplete outcome data (attrition bias) All outcomes	High risk	Five participants who were found eligible for the study dropped out due to 'deterioration in their medical condition', but study authors provided no details as to the stage at which this occurred, whether baseline characteristics were different, or to which conditions participants had been allocated. Non‐inclusion of their data is likely to have introduced bias due to incomplete reporting
Selective reporting (reporting bias)	Low risk	All tests mentioned in the "Methods" section were reported in the "Results" section
Other bias	Low risk	Study appears to be free of other sources of bias