Improving the accuracy of brief cognitive assessments when used as part of the process for identifying dementia in general practice.

Abstract

Identifying dementia in general practice remains a considerable challenge, with mild to moderate stages of dementia potentially underdiagnosed in 30-50% of cases. The primary aim of this PhD thesis was to address the question “how can we improve the accuracy of brief cognitive assessments when used as part of the process for identifying dementia in general practice?”. This was carried out via a combination of secondary research through three evidence syntheses, and primary research via a survey of general practitioners with results triangulated with existing research and thesis findings. Through the conduct of a rapid review of clinical practice guidelines (CPGs),I found a lack of consistent recommendations for general practice regarding selection and application of brief cognitive assessment (BCA) tools There was also a paucity of guidance given within the identified CPGs on tailoring BCA choice and use for specific populations. The rapid review indicates that greater clarity and consistency is needed from CPGs relating specifically to the use of BCAs as part of the process for identifying dementia in general practice. The systematic review and overview identified an absence of existing evidence. Where evidence exists, BCAs performed inconsistently and were broadly inadequate as a tool for use in general practice dementia care. Other factors beyond diagnostic accuracy render established tests ill-suited for general practice such as administration time, cost and acceptability for clinicians and patients. A number of areas are identified both in cognitive testing and research methods where progress can be made relatively simply. This thesis demonstrates that many assumptions underlying current practice are without robust foundations, with severe implications for general practice and patient care at a time of scarce resource and growing demand. These assumptions need revising as a priority. What is needed is clear, specific, well-designed primary research to begin to unpick these complexities and realistically address the challenges presented by the identification of dementia within general practice and primary care. I provide explicit recommendations for the design and conduct of a primary comparative accuracy study alongside a trial of effectiveness and cost-effectiveness of using brief cognitive assessments as part of the process of identifying dementia in general practice in order to objectively and systematically assess the suitability of brief cognitive assessments as a tool for use in this population and setting.