Objective: Though many cognitive screening tests are validated for dementia surveys, we wanted to see if the diagnostic utility of one such test (AMT) can be enhanced by prefacing it with a single question about PF.
Design: A two-phased community survey of randomly selected elderly (³50y) Chinese residents in the Ang Mo Kio housing estate, (n=4707) was conducted from April to June 2001.
Materials and Methods: Of these, 2566 completed Phase 1 which included doing the AMT and asking for subject’s PF. Those who either had PF or abnormal (abn) AMT(based on previously validated, education-adjusted cut-off scores) underwent and completed Phase 2 evaluation for dementia and depression (n=121). An additional 35 randomly selected normal individuals were also studied for dementia and depression. Prevalence of dementia, PF and abn AMT were computed. Dementia status of subjects was analyzed by logistic regression with PF, abn AMT, depression and age as covariates. Sensitivity and specificity of PF for dementia were determined. The prevalence of dementia patients in the 4 possible diagnostic combinations of the PF question and AMT performance was subsequently studied.
Results: Overall prevalence of PF, abn AMT and dementia were 2.4%, 2.2%, 0.9% respectively. Age-adjusted prevalence of the same variables in the 50-74y and (>75y) age groups were 1.6%(7.9%), 1.2%(8.8%), 0.4%(4.2%) respectively (all p<0.01). Age and depression, however were not statistically associated with dementia status in the LR model which only showed PF (OR 38.0, p<0.01) and abn AMT (OR 20.5, p<0.01) to be significantly associated with dementia. The sensitivity of PF for dementia was 95.7% with specificity of 45.1%. Among our 55-74y old subjects with dementia, 0% were in the (no PF/pass AMT), 0% in the (no PF/abn AMT), 25% in the (PF/pass AMT) and 75% in the (PF/abn AMT) groups. The corresponding percentages for the >75y old patients were 0%, 6.7%, 6.7% and 86.7% respectively. These percentages showed the dementia prevalence for the 50-74 and (>75y) age groups to be 0%(0%) in the (no PF/pass AMT), 0%(0.28%) in the (no PF/abn AMT), 0.1%(0.28%) in the (PF/pass AMT) and 0.3% (3.6%) in the (PF/abn AMT) groups.
Conclusion: Our study indicates that it is diagnostically useful and efficient in primary care settings to ask for PF before administering the AMT. If PF is absent, there is epidemiological justification for not doing the AMT (total dementia prevalence in this category = 0.28%) and when PF is present, further evaluation of dementia is better initiated when AMT is abnormal. While age was marginally insignificant for dementia in our study, these conclusions are especially salient for the 50-74y age group.
Back to Epidemiological aspects of psychogeriatric practice
Back to Oral and Poster Sessions
Back to The IPA European Regional Meeting (1-4 April 2003) of IPA