Identifying individuals at high risk for dementia in primary care: development and validation of the DemRisk risk prediction model using routinely collected patient data

Health policy in the UK and globally regarding dementia, emphasises prevention andrisk reduction. These goals could be facilitated by automated assessment of dementiarisk in primary care using routinely collected patient data. However, existing applicabletools are weak at identifying patients at high risk for dementia. We set out to developimproved risk prediction models deployable in primary care.MethodsElectronic health records (EHRs) for patients aged 60-89 from 393 English generalpractices were extracted from the Clinical Practice Research Datalink (CPRD) GOLDdatabase. 235 and 158 practices respectively were randomly assigned to developmentand validation cohorts. Separate dementia risk models were developed for patientsaged 60-79 (development cohort n=616,366; validation cohort n=419,126) and  80-89(n=175,131 and n=118,717). The outcome was incident dementia within 5 years andmore than 60 evidence-based risk factors were evaluated. Risk models weredeveloped and validated using multivariable Cox regression