LO56: Rate of delirium recognition by nurses and physicians in a cohort of 1584 older emergency department patients: how many would have been sent home?

Abstract

Introduction: Unrecognized delirium in the ED remains common despite a 3 fold mortality increase for those discharged home. But previous studies have not assessed delirium recognition rate in a multicenter study nor assessed the management plans of ED staff when they fail to recognize delirium. Objectives: To document 1) the rate of delirium recognition by nurses and MDs in a national sample and 2) the intended management plans for patients with unrecognized delirium. Methods: This is a planned sub-study of a randomized clinical trial at 5 EDs in 4 provinces conducted in English and French. We included people ≥ 65 years old. We excluded those with an ED stay < 4 hours, critical illness, visual impairment or from a nursing home. Research assistants (RAs) assessed delirium using the validated Confusion Assessment Method. RAs then asked ED nurses and physicians if the patient had delirium according to their clinical assessment. RAs also asked how confident they were that the patient could be safely discharged home using a 10 point Likert scale. We report proportions and 95% confidence intervals. RAs notified all ED staff of unrecognized CAM + ve patients prior to actual discharge for safety reasons. Results: We recruited 1584 older people; 1496 (92.5%) had complete data. Mean age was 76.5; 49% were female. Nurses performed 1465 delirium assessments. There were 76 CAM + ve patients in our sample (5.2%, 95% CI 4.2 to 6.5%). Nurses recognized delirium in 34/76 (44.7%, 95% CI: 33.3 to 56.6%). MDs assessed 20 CAM + ve patients and recognized the delirium in 10/20 (50.0%, 95% CI: 27.2 to 72.8). Nurses felt that 11/42 patients with unrecognized delirium could be discharged (26.2% 95% CI: 13.9 to 42.0%). Their median confidence in the safety of their plan was 7.0/10. MDs felt that 2/10 patients with unrecognized delirium could be discharged (20.0% 95% CI: 2.5 to 55.0%). Their median confidence in the safety of their plan was 7.5/10. Conclusion: Despite the potential Hawthorne effect raising initial delirium recognition rates above clinicians’ usual practice outside of a study, delirium recognition by both nurses and MDs remains poor in a national sample of ED patients. We also showed that a significant number of these patients could have been discharged with unrecognized delirium. Further research to find novel ways to improve delirium recognition is needed.