[Cognitive assessment of elderly patients in the emergency department: a comparison between standard tools, medical...

We evaluated the accuracy of physician recognition of cognitive impairment in elderly patients in emergency departments (ED). In particular, we evaluated the accuracy of the subjective impression of the physician on patients' cognition (a comparison of the information obtained from the responsible physician with the S-IQCODE, a cognitive impairment screening test), and the accuracy of the medical records (a comparison of the information in the medical record with the S-IQCODE).

Cross-sectional study on 101 elderly patients selected at random from those attending ED, their ED physicians, and family member-carer. The study was conducted in the ED of four tertiary university teaching hospitals in a city, from July through November 2003. Cognitive data shown in the patient's medical records were compared against the S-IQCODE obtained from the family member-carer, using the kappa (kappa) concordance index. The physicians' impressions on the patients' cognitive status were also compared against the S-IQCODE results, using the kappa (kappa) concordance index. Each patient and their family member-carer were paired with a single physician. A logistic regression model was constructed to identify factors associated with the physicians' impressions of the patients' cognitive capacity.

The concordance between information on cognitive decline from medical records and the results of the S-IQCODE, was 0.47 (IC95%: 0.05-0.88). Concordance between the physicians' impression on the presence of cognitive impairment, and the S-IQCODE obtained from family member-carer was 0.26 (IC95% 0.06-0.45). The multivariate analysis demonstrated that the functional status of patients, as perceived by the physicians, were the variable that better explained the physicians' impressions of patient cognitive function.

The cognitive status of elderly patients is not properly assessed by emergency department physicians.