How physicians evaluate patients with dementia who present with shortness of breath. Issue 1 (20th September 2021)

• Record Type: Journal Article
• Authors: Ly, Dan P.; Shekelle, Paul G.
• Is Part Of: Journal of the American Geriatrics Society. Volume 70:Issue 1(2022)
• Page Start: 119
• Page End: 125
• Publication Date: 2021-09-20
• Subjects: dementia -- physician behavior -- physician decision‐making; Geriatrics -- Periodicals; 618.97
• DOI: 10.1111/jgs.17438
• Languages: English
• ISSNs: 0002-8614

Abstract

Abstract: Background: Whether the presence of dementia in patients makes it difficult for physicians to assess the risk such patients might have for serious conditions such as pulmonary embolism (PE) is unknown. Our objective was to examine the differential association of four clinical factors (deep venous thrombosis (DVT)/PE, malignancy, recent surgery, and tachycardia) with PE testing for patients with dementia compared to patients without dementia. Methods: We performed a cross‐sectional study of emergency department (ED) visits to 104 Veterans Affairs (VA) hospitals from 2011 to 2018 by patients aged 60 years and over presenting with shortness of breath (SOB). Our outcomes were PE testing (CT scan and/or D‐dimer) and subsequently diagnosed acute PE. Results: The sample included 593, 001 patient visits for SOB across 7124 ED physicians; 5.6% of the sample had dementia, and 10.6% received PE testing. Three of the four clinical factors examined had a lower association with PE testing for patients with dementia. For example, after taking into account that at baseline, physicians were 0.9 percentage points less likely to test patients with dementia than patients without dementia for PE, physicians were an additional 2.6 percentage points less likely to test patients with dementia who had tachycardia than patients without dementia who had tachycardia. We failed to find evidence that any clinical factor examined had a differentially lower association with a subsequently diagnosed acute PE for patients with dementia. Conclusions: Clinical factors known to be predictive of PE risk had a lower association with PE testing for patients with dementia compared to patients without dementia. These results may be consistent with physicians missing these clinical factors more often when evaluating patients with dementia, but also with physicians recognizing such factors but not using them in the decision‐making process. Further understanding how physicians evaluate patients with dementia presenting with common acute symptoms may help improve the care delivered to such patients.