Delirium Screening in Aphasic Patients With the Intensive Care Delirium Screening Checklist (ICDSC): A Prospective Cohort Study

Background: Ten to thirty percent of stroke patients suffer from post-stroke delirium. This leads to a longer hospital stay and increased mortality. Therefore, early detection and treatment are needed. All established delirium screening tools require some degree of language function. We sought to investigate whether the Intensive Care Delirium Screening Checklist (ICDSC) is suitable for delirium screening in patients with post-stroke aphasia. Methods: A prospective cohort study was carried out in adult patients consecutively admitted to the Stroke Unit of University Hospital Tuebingen, between July 2017 and December 2018. The index test, ICDSC, was compared with the DSM-V diagnostic criteria as reference standard. Measures of diagnostic precision and the degree of agreement were obtained. Results: Three hundred and forty six patients were included in the analysis. Aphasia was present in 231 (66.8%) and absent in 115 (33.2%) patients. Delirium was present in 83 out of 231 (36%) patients with aphasia and 32 out of 115 (27.8%) patients without aphasia (p = 0.132). For patients without aphasia, sensitivity and specificity at the established cut-off value of ≥ 4 points were 100% and 78%, respectively. For patients with aphasia, the test demonstrated inferior performance, with a sensitivity and specificity of 98% and 55%, respectively. It was necessary to increase the cut-off value to ≥ 5 points. Through this, sensitivity was 90% (95% CI, 81.9–95.8%) and specificity was 75% (95% CI, 67.2–81.8%). The degree of agreement to the DSM-V criteria was “substantial” (Cohen's κ = 0.61). Conclusion: For the purpose of delirium screening in patients with aphasia, increasing the ICDSC cut-off value to ≥ 5 points enables effective screening. Further studies are necessary to characterize post-stroke delirium.