Home/community-based interventions to improve function in persons with mild cognitive impairment/early dementia.

The review included 18 studies, but they differed in design, intervention type, duration, dose, control group and outcome measure. Many studies did not measure the main function outcomes: seven multicomponent studies did not measure ADLs and five did not measure IADLs. Study quality was mostly IB, and participant blinding was a main source of high risk of bias. Many studies had small samples without a control group, reducing power. The included population excluded people with major neurological or psychiatric conditions, sensory deficits and cancer, so it may not represent older adults with co-occurring impairments. Race/ethnicity was measured in only four studies, and education was the only socioeconomic variable in most studies. Because interventions were bundled, the review cannot tell whether technology, caregiver involvement, dose, functional tasks or another component caused improvement.

No conflicts-of-interest statement is included in the supplied text. The metadata lists NIH extramural and non-U.S. government research support.

Do not read this review as showing that any specific home/community programme works. Only four studies reported significant ADL or IADL improvements, many studies did not measure those outcomes, and effective programmes bundled several components, so their separate contributions cannot be separated.