Designing an intervention to improve cognitive evaluations in primary care.

Background: Early diagnosis is crucial to the optimal management of patients with cognitive impairment due to Alzheimer's disease (AD) or AD-related dementias. For some patients, early detection of cognitive impairment enables access to disease-modifying therapies. For all patients, it allows access to psychosocial supports. Patients typically first present their concerns about their cognition to a primary care provider, but in this setting, cognitive impairment is commonly underdiagnosed. There is also high variability in how cognitive evaluations are performed. We sought to understand barriers to and facilitators of cognitive evaluations in primary care, map barriers to implementation strategies, and gain consensus from stakeholders on possible strategies to improve dementia diagnosis in primary care.

Methods: Semi-structured interviews conducted with primary care providers (PCPs). We used the Consolidated Framework for Implementation Research to inform our question guide and analysis, and incorporated chart-stimulated recall - using actual patients who had cognitive complaints who had presented to these providers - to understand clinicians' medical decision-making processes. These data were used to map identified barriers and facilitators to targeted implementation strategies. Then, this candidate list of strategies was presented to an expert stakeholder panel including clinicians and clinical operations specialists. Through a modified Delphi process, the list was narrowed to select the most promising strategies to incorporate in an intervention to improve cognitive evaluations in primary care.

Results: Twenty PCPs were interviewed and mentioned barriers included lack of expertise to perform or interpret an assessment, time pressures, lack of incentives, competing priorities, lack of decision-making supports, and limited access to dementia specialists. Facilitators included the presence of an informant or caregiver and having additional staff to conduct cognitive testing. Implementation mapping resulted in a list of 15 candidate strategies. Using the modified Delphi process, these were narrowed to six.

Conclusions: We used a rigorous process to identify barriers to and facilitators of cognitive assessments in primary care, identify promising implementation strategies to address these barriers, and obtain the feedback of front-line users on these strategies. This holds substantial promise for improving cognitive assessments in primary care in future implementation trials.