Machine-Learning Based Model May Identify Dementia in Primary Care

A study published in BJGP Open found a machine-learning based model using data routinely accumulated in primary care identified individuals with dementia in such settings, reports Healio . "Such a tool could be used to select high-risk persons who could be invited for targeted screening," note the researchers. The team used Read codes, a set of clinical terms employed in Britain to summarize data for general practice, to develop the model. The Read codes were chosen based on their significant association with individuals with dementia, and included codes for risk factors, symptoms, and behaviors that are collected in primary care. To test the model, researchers collected Read-encoded data from 26,483 persons living in England 65 and older. They determined the model achieved a sensitivity of 84.47 percent and a specificity of 86.67 percent for recognizing dementia. "With the expected growth in dementia prevalence, the number of specialist memory clinics may be insufficient to meet the expected demand for diagnosis," the team says. "Furthermore, although current 'gold standards' in dementia diagnosis may be effective, they involve the use of expensive neuroimaging (for example, positron emission tomography scans) and time-consuming neuropsychological assessments which is not ideal for routine screening of dementia." The model will be assessed with other datasets, and have its validation tested "more extensively" at general practitioner practices in the future, the team says.