Study designs and statistical analyses in randomized controlled trials of non-pharmacological preventive and therapeutic interventions for dementia continuum: A statistical review

Study designs and statistical analyses in randomized controlled trials of non-pharmacological preventive and therapeutic interventions for dementia continuum:A statistical review

Ryoichi Hanazawa¹⁾, Hiroyuki Sato¹⁾, Masanao Sasaki¹⁾, Keisuke Suzuki²⁾, Akihiro Hirakawa¹⁾

¹⁾Department of Clinical Biostatistics, Graduate School of Medical and Dental Sciences, Tokyo Medical and Dental University
²⁾Innovation Center for Translational Research, National Center for Geriatrics and Gerontology

Background: Various systematic reviews of non-pharmacological interventions involving the target population of normal cognition, mild cognitive impairment （MCI）, dementia, or caregivers of patients with dementia have attempted to synthesize the effects of interventions;however, the variations in study designs and the low quality of clinical trial methodology are barriers to deriving reliable conclusions.
Aim: We comprehensively investigated the study designs and statistical analyses used in randomized controlled trials （RCTs） of non-pharmacological preventive or therapeutic interventions for the four target populations, offering common practical recommendations for designing RCTs on non-pharmacological interventions.
Results: We reviewed 200 RCTs （31 on normal cognition, 19 on MCI, 126 on dementia, 22 on caregivers, and 2 on mixed populations）.　In each target population, >70% of the trials had a follow-up period ≥6 months.　About 50% of the studies on normal cognition or caregivers had a sample size of >100, while 50% of the studies on people with MCI or dementia had a sample size of ≥50.　Only 20-25% of studies in each target population performed a power analysis.
Conclusions: Based on the findings of this review, we offer suggestions for future consideration when planning and conducting RCTs on nonpharmacological interventions.
Keywords: Randomized controlled trial; dementia; non-pharmacological interventions; systematic review

Address correspondence to Dr. Akihiro Hirakawa, Department of Clinical Biostatistics, Graduate School of Medical and Dental Sciences, Tokyo Medical and Dental University （1-5-45 Yushima, Bunkyo-ku, Tokyo 113-8510, Japan）