Clinical history for diagnosis of dementia in men: Caerphilly Prospective Study.

Abstract

Diagnosis of dementia often requires specialist referral and detailed, time-consuming assessments. To investigate the utility of simple clinical items that non-specialist clinicians could use, in addition to routine practice, to diagnose all-cause dementia syndrome. Cross-sectional diagnostic test accuracy study. Participants were identified from the electoral roll and general practice lists in Caerphilly and adjoining villages in South Wales, UK. Participants (1225 men aged 45-59 years) were screened for cognitive impairment using the Cambridge Cognitive Examination, CAMCOG, at phase 5 of the Caerphilly Prospective Study (CaPS). Index tests were a standardised clinical evaluation, neurological examination, and individual items on the Informant Questionnaire for Cognitive Disorders in the Elderly (IQCODE). Two-hundred and five men who screened positive (68%) and 45 (4.8%) who screened negative were seen, with 59 diagnosed with dementia. The model comprising problems with personal finance and planning had an area under the curve (AUC) of 0.92 (95% confidence interval [CI] = 0.86 to 0.97), positive likelihood ratio (LR+) of 23.7 (95% CI = 5.88 to 95.6), negative likelihood ratio (LR-) of 0.41 (95% CI = 0.27 to 0.62). The best single item for ruling out was no problems learning to use new gadgets (LR- of 0.22, 95% CI = 0.11 to 0.43). This study found that three simple questions have high utility for diagnosing dementia in men who are cognitively screened. If confirmed, this could lead to less burdensome assessment where clinical assessment suggests possible dementia.