Studies that compared the diagnostic validity of the mini-mental state examination to a validated standard of either dementia or mild cognitive impairment were eligible for inclusion. The mini-mental state examination had to be the unmodified, original, English language version. Included studies also had to contain adequate data for inclusion. Validated standards for dementia included: Diagnostic and DSM (Statistical Manual of Mental Disorders, III, IIIR and IV), NINCDS-ADRDA (National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer's Disease and Related Disorders Association), AGECAT (Automated Geriatric Examination for Computer Assisted Taxonomy), CAMDEX (Cambridge Mental Disorders of the Elderly Examination) and expert diagnosis. For mild cognitive impairment, validated standards included Mayo clinic criteria, prospective validation and Clinical Dementia Rating of 0.5 without functional deficits.
Following a sample size calculation, studies that contained a sample size of less that 160 were excluded. Inclusion criteria were not defined for study design. In terms of the outcomes, studies were included only if sensitivity and specificity were either stated or calculable. The outcomes reported in the review were the sensitivity, specificity, positive and negative predictive values, predictive summary index and Youden score.
Studies from all settings were included, but separated into low prevalence (primary care, community studies and nursing homes) and high prevalence (memory clinics, mixed hospital) settings. Subject demographics were variable. Age range (where reported) varied between under 59 to 100 years. Some trials included subjects with psychiatric illness or predominantly involved ethnic minorities. Educational attainment was rarely reported. Diagnostic cut-off thresholds for the mini-mental state examination were variable but the majority of included studies used a threshold of 23v24 (the threshold recommended for people with at least eight years of education).
The author did not state how the papers were selected for the review, or how many reviewers performed the selection.