GPs missing early dementia -- new study

Researchers from the University of Leicester in the UK and National Collaborating Centre for Mental Health, London, UK and the Department of General Practice, Dusseldorf, Germany examined 30 previous studies involving 15,277 people seen in primary care for cognitive disorders, including 7109 assessed for dementia.

Although GPs managed to identify eight out of ten people with moderate to severe dementia, most patients with early dementia were not recognized. Only 45% of people with early dementia and mild cognitive impairment were identified. Mild cognitive impairment is a condition that may precede dementia in some people.

Across the whole spectrum, GPs identified 3 out of 5 of people attending for broadly defined memory problems.

Dr Alex Mitchell, a consultant psychiatrist with the Leicestershire Partnership NHS Trust and a researcher at the University, said: "This study highlights for the first time that GPs trying to identify dementia actually make more false positive errors, with misidentifications outnumbering missed cases at least two to one."

"GPs working in busy settings struggle to identify early dementia and prodromal conditions based on their initial clinical judgement. This was particularly the case for patients living alone where no informant was available and when patients had relatively preserved daily function. Furthermore, GPs' attitudes towards dementia may play an important role in dementia recognition. A project within the German Competence Network Degenerative Dementias (CNDD) at the University of Dusseldorf is currently investigating this.

"Conversely patients with depression or hearing problems were more at risk of being misidentified with dementia. However, the main influence is severity. Patients with mild dementia may not volunteer troubling memory problems and GPs are often unsure about the value of screening tests. Given the problem of false positives and false negatives we found that the application of a simple cognitive screening test after a clinical diagnosis would help GPs to achieve about 90% accuracy. We report separately which screening test may be best in Am J Geriatr Psychiatry 2010;18:759."

Provided by University of Leicester