A critical aspect of hospital inpatient care is responsiveness. Especially in intensive care units, continuous close monitoring of patients as well as the urgency and accuracy in which drugs are administered are often key in determining a patient’s survival. The ideal solution encapsulates not only minimizing the time cost of doctors and nurses, who potentially have more urgent issues to attend to, but also replicating and potentially even improving their accuracy. To accomplish this, artificial intelligence can be synergistically paired with a medical device to automatically dispense of commonly administered drugs intravenously, assuring continuous stable vitals.
In recent years, a growing number of hospitals across the country have adopted the system of the Rothman Index, a measure of patient condition that utilizes the Electronic Medical Records. It focuses on 26 clinical measurements within the categories of nursing assessments, vital signs, laboratory results, and cardiac rhythms. Physicians are provided with a single value, created from summation of heterogeneous data, to determine the risk of any one hospitalized patient. Devices such as the ClearSight System produced by Edwards Lifesciences or the CGM (Continuous Glucose Monitoring) by Medtronic, have thrust us into an era of easy access to continuous patient monitoring. And even measurements that have traditionally required extra effort and time by healthcare providers, are rapidly becoming obsolete with technological progress. For example, recent developments allow for real time blood testing through quantitative phase imaging by utilizing label-free spectrally encoded flow cytometry.
We can thus take advantage of these measurements, specifically those in vital signs, cardiac rhythms, and laboratory results, to help further ICU care. Utilizing the vast troves of the Electronic Medical Records, the proposed system will monitor these many signs. When any one level is abnormal for the patient or has the potential to be dangerous, the system will first take into consideration the patient’s last recorded condition, lab results, family history, and other medications being taken, before searching through the records and finding all instances of similar presented conditions. By gathering all similar situations, then analyzing what drugs and what amounts were administered at those instances, the system will ultimately make the best possible calculation and decision. The machine, pre-stocked with many drugs commonly used in critical care, such as dopamine, propofol, heparin, fentanyl, or nitroglycerin, would then automatically measure this amount, and administer it to the patient through the IV. However, if the technology determines that the patient’s condition requires a more complicated next step or if it does not possess a certain confidence level in its decision, the physician on-call will immediately receive a notification about the issue.
Immediate intervention during any signs of early deterioration can help prevent extreme emergencies from occurring and minimize risk of death. Though there are currently bedside charts that assist healthcare professionals in the case of an emergency, this system will essentially eliminate the unnecessary extra time of the nurses and physicians having to be first notified and then making a correct decision in the heat of the moment.
Golan, L., Yeheskely-Hayon, D., Minai, L., Dann, E. J., & Yelin, D. (2012). Noninvasive imaging of flowing blood cells using label-free spectrally encoded flow cytometry. Biomedical Optics Express,3(6), 1455. doi:10.1364/boe.3.001455
Gommers, D., & Bakker, J. (2008). Medications for analgesia and sedation in the intensive care unit: an overview. Critical Care,12(Suppl 3). doi
Model Development & Scientific ValidationDevelopment of a Continuous Measure of Patient Acuity Across All Diseases, Conditions. (n.d.).
Popescu, G. (2013). Real time blood testing using quantitative phase imaging. Digital Holography and Three-Dimensional Imaging. doi:10.1364/dh.2013.dm4a.4
Rothman, M. J., Rothman, S. I., & Beals, J. (2013). Development and validation of a continuous measure of patient condition using the Electronic Medical Record. Journal of Biomedical Informatics,46(5), 837-848. doi:10.1016/j.jbi.2013.06.011
Salgado, D., Silva, E., & Vincent, J. (2013). Control of hypertension in the critically ill: a pathophysiological approach. Annals of Intensive Care,3(1), 17. doi:10.1186/2110-5820-3-17
DECISION SUPPORT & HOSPITAL MONITORING
Author: Ryan Hsieh
Status: Project Concept