Diagnostic accuracy of a rapid checklist to identify delirium in older patients transported by EMS

摘要

Background. The presence of delirium in elderly patients is common and has been identified as an independent marker for increased mortality and hospital-acquired complications, yet it is poorly recognized by health care providers. Early recognition of delirium in the prehospital setting has the potential to improve outcomes, but is not feasible without valid assessment tools. Objective. To determine whether use of a rapid delirium checklist by prehospital providers is a valid way to identify cases of delirium compared with a criterion standard and whether the checklist is better at identifying delirium than the Glasgow Coma Score (GCS). Methods. We conducted a prospective study at two academic, tertiary-care emergency departments (EDs) where a convenience sample of matched dyads of emergency medical services providers and elderly patients (age ??65 years) were enrolled. Prehospital providers reported limited demographics and work history about themselves. They also reported vital signs and GCS for each patient and completed the checklist asking about presence of the four features of delirium. The patient then underwent a cognitive assessment using the Confusion Assessment Method for the Intensive Care Unit (CAM-ICU) by a trained investigator, which was used as the criterion standard. Criterion validity and concurrent validity of the delirium checklist and abnormal GCS were evaluated using sensitivity and specificity. Results. Two hundred fifty-nine matched dyads were studied. Delirium occurred in 24 (9%) of the elderly patients sampled. Prehospital providers' recognition of any delirium symptom resulted in a sensitivity of 0.63 (95% confidence interval [CI] 0.43-0.79) and a specificity of 0.74 (95% CI 0.73-0.84). Prehospital report of a GCS <15 has a sensitivity of 0.67 (95% CI 0.47-0.82) and a specificity of 0.85 (95% CI 0.80-0.89). Conclusions. A rapid delirium checklist can identify 63% of patients with delirium, but performed no better than the GCS. Future research should determine whether a rapid test of cognition improves early identification of elderly patients with delirium.