Automating Delirium Severity in the ICU

Overview

The goal of this observational study is to develop a passive digital marker (PDM) for delirium severity and examine its performance in comparison to validated delirium severity tools in ICU patients \>50 years of age. The main questions it aims to answer are: * Is the trained convolutional neural network able to reliably measure delirium severity. * Is the Passive Digital Marker able to accurately measure delirium severity * Is the Passive Digital Marker acceptable and usable by frontline ICU nurse clinicians, patients, and their identified proxies (i.e., caregivers). Participants will: * Study participation involves a video camera recording you 24 hours per day while you are a patient in the Intensive Care Unit (ICU). * Study staff will visit you 4 times each day you are in the ICU. You will be asked questions each time they visit to train the digital marker and see differences between assessments and camera data.

Bedside cameras will stay up for at least 72 hours (about 3 days), it may be longer if delirium is being experienced and if the patient and or LAR (Legally Authorized Representative) agree to the cameras' continued recording. The study team will come and ask the patient questions 4 times daily. We will ask questions about their general, mental, and emotional health, their mood and memory. During these visits, the study team will check the alertness of the patient. Alertness will be checked by voice; if they do not respond, we will tap their shoulder gently. We will ask the patient to do activities like remembering words and saying them back, shaking hands, and drawing an object or clock. The study team will administer assessments 4 times daily. Participants will be asked questions about their general, mental, and emotional health, as well as mood and memory. While performing these sessions, the study team will observe the level of awareness. Awareness will be assessed by voice; if they do not respond, we will have to do a gentle tap on their shoulders. The patient will be asked to perform activities such as: remembering words and repeating them back, handshaking, and drawing an object or clock. In some cases, if the patient's level of awareness alertness is lowered, we will do observation only perform observation assessments. These assessments do not interfere with clinical care. Participation in this study is voluntary. The study team will continue to follow the patient until discharged from ICU. The bedside nurse, patient, or family can cover the cameras with a washcloth if needed for privacy. The faces of family and clinicians and the bathing suit areas of patients are blurred before analysis to protect privacy.