Discussion of memory during primary care visits of older adults with cognitive impairment and accompanying family

Objectives: Cognitive impairment is underdiagnosed in primary care. Understanding factors that precipitate memory-related discussion could inform strategies to improve diagnosis and counseling. We assessed whether: 1) having a cognitive impairment or dementia diagnosis, 2) ratings of cognition by clinicians, or 3) ratings of cognition by family companions were associated with memory-related discussion during primary care visits. Methods: We examined audio-recorded primary care visits of cognitively impaired patients aged 65 years and older, family companions (n = 93 dyads), and clinicians (n = 14). Cognitive impairment and dementia diagnoses were extracted from the electronic health record. Clinicians and family rated patient cognition on a 10-point scale in postvisit surveys. We measured memory-related discussion using a ratio of memory-related discussion episodes to total visit statements. Results: We observed more memory-related discussion during primary care visits of patients with a diagnosis of mild cognitive impairment (+7.8% episodes; P <.001) or dementia (+26.3% episodes; P <.001) than no diagnosis. Clinician and family ratings of cognition varied by diagnosis: among patients with no diagnosis, family rated worse impairment than clinicians (average: 2.4 versus 1.3; P =.004) while for patients with a dementia diagnosis, clinicians rated worse impairment than family (average: 7.1 versus 5.5; P =.006). Each unit increase in clinician-rated severity of cognitive impairment was associated with more memory-related discussion (+2.6% episodes; P <.001); this association was attenuated for family (+0.7% episodes; P =.095). Conclusions: Discussion of cognitive impairment appears largely driven by clinician ratings of cognition and presence of an established diagnosis. Findings suggest potential benefit of engaging family to improve cognitive impairment detection in primary care.