Using Telehealth to Improve the Accuracy of Delirium Screening by Bedside Critical Care Nurses

Liron Sinvani, MD; Craig Hertz, DO; Saurabh Chandra, MD; Anum Ilyas, BS; Suzanne Ardito, MA; Negin Hajizadeh, MD


Am J Crit Care. 2022;31(1):73-76. 

In This Article


Our health system, composed of 23 diverse hospitals in the greater New York metropolitan area, has tele-ICU capabilities for 350 ICU beds in 22 ICUs across 14 hospitals (4 tertiary and 10 community hospitals). The health system uses telehealth software (eCare platform, Philips) for integrated audio/video and algorithmic data acquisition.

This pilot feasibility study (institutional review board number HSRD 20-0075) included 13 bedside nurses from 4 ICUs who had been selected by nurse managers to become CAM-ICU champions. Before tele-delirium training, standard CAM-ICU training for nurses consisted of a mandatory 15-minute online module that provided an overview of delirium and CAM-ICU features. Direct observation and CAM-ICU champions were not used.

During a predetermined time, a remotely located geriatrician-hospitalist with clinical expertise in delirium across the hospital setting directly observed each bedside ICU nurse administering the CAM-ICU to a patient. These observations were done in real time via the tele-ICU platform. Before the session, the bedside nurse obtained permission from the patient and/or the patient's health care proxy. The nurse introduced the geriatrician-hospitalist to the patient and proceeded to conduct the CAM-ICU assessment from start to finish, verbalizing the response to each component (Richmond Agitation-Sedation Scale [RASS], features 1–4 of the CAM-ICU, and overall CAM-ICU). The geriatrician-hospitalist and a delirium-trained research assistant observed and independently scored the assessment by using a validated spot check form[7] to evaluate differences between the bedside nurse and an expert spot checker. Disagreements between the geriatrician-hospitalist and the research assistant were discussed and reconciled.

Following observation, the geriatrician-hospitalist provided training on CAM-ICU performance (accuracy of the RASS and of each individual CAM-ICU feature) and briefly discussed delirium prevention and management strategies. The second CAM-ICU nurse assessment occurred 4 to 7 days after the initial tele-delirium training session, depending on the bedside nurse's schedule. A paired-sample t test was used to analyze differences between pretraining and posttraining assessments.

After completion of the second observation session, each nurse was given an anonymous 2-question survey on the benefits of training and likelihood to change practice. Survey responses used a 5-point Likert scale (strongly agree, agree, neither agree nor disagree, disagree, strongly disagree). The nurse manager then asked the nurses to provide feedback regarding the tele-delirium training.

We hypothesized that (1) at baseline, CAM-ICU assessments by bedside ICU nurses would be inaccurate; (2) it would be feasible to observe and provide training on bedside nurses' CAM-ICU performance by using the tele-ICU platform; and (3) brief observation and training sessions for ICU nurses with patients in real time would improve the accuracy of CAM-ICU use.