Article Text
PDF
Abstract
Introduction Artificial intelligence (AI) is moving into the hospital with prospects of improving patient safety and freeing up staff time. To inform an economic analysis of an AI-assisted continuous vital signs monitoring (VSM) system known as WARD (Wireless Assessment of Respiratory and circulatory Distress) we reviewed the literature and analysed data from the Danish national patient registry.
We derive key points for researchers undertaking analyses of AI-assisted VSM and we quantify the potential of AI in shortening hospital admissions and preventing readmissions.
The project is funded by Innovation Fund Denmark and is joint work by VIVE, DacHE, and the WARD project group.
Artificial intelligence (AI) is increasingly being applied to technologies within healthcare and one example is the integration of AI in continuous VSM. AI-assisted continuous VSM can potentially reduce adverse events in hospitalized patients, shorten the treatment process and allows for new workflows potentially freeing (human) resources.
However, the impact of AI-assisted continuous VSM on clinical and economic outcomes, including length of stay and impact on workload, is currently understudied but needed, to demonstrate if and how such technologies can improve general patient safety and current workflows.
Methods We conducted a systematic literature search to identify studies of AI-assisted continuous VSM that assessed the impact on economic outcomes, including nonmonetary outcomes such as resource use, impact on workload, length of stay and transfers to intensive care unit.
Using Danish national registries we used coarsened exact matching to analyse the difference in length of stay between patients exposed to an adverse event during admission and comparable patients unexposed to such events. This analysis can quantify the magnitude of adverse events that could potentially be reduced by AI-assisted continuous VSM. Relevant adverse events were defined according to a pre-specified manual currently used in Danish clinical studies of AI-assisted VSM.
Using the same registries, we examined the magnitude of short-term acute admission after discharge, hence demonstrating the potential of home monitoring of patients discharged from hospital. Relevant patients groups for the analysis were identified in collaborations with clinicians from the WARD Home project, an ongoing Danish research project on AI-assisted home monitoring.
Results A comprehensive systematic literature search emphasised an evidence gap as to the impact of AI-assisted continuous VSM on important outcomes, such as length of stay and resource use.
The matched analysis revealed a 2.4 days difference in length of stay between hospitalised patients exposed and unexposed to an adverse event, respectively. This gap could potentially be reduced by replacing current manual monitoring procedures with AI-assisted continuous monitoring.
AI-assisted continuous VSM may improve patient safety and free up resources by reducing the gap in length of stay between patients exposed and unexposed to an adverse event. Moreover, it may avoid short-term readmissions if the monitoring is performed in the patient’s own home in the first days after discharge. More research still needs to be done to demonstrate that AI-assisted continuous VSM can and will meet the high expectations put on it.
References
Abdel Latif Al-Shwaheen TI, Hau YW. A new model for tracking and detection of deterioration of vital signs based on artificial neural network. J Theor Appl Inf Technol. 2019;97(14):3809-3818.
Clifton L, Clifton DA, Pimentel MAF, Watkinson PJ, Tarassenko L. Predictive monitoring of mobile patients by combining clinical observations with data from wearable sensors. IEEE J Biomed Heal Informatics. 2014;18(3):722-730. doi:10.1109/JBHI.2013.2293059
Steinhubl SR, Feye D, Levine AC, Conkright C, Wegerich SW, Conkright G. Validation of a portable, deployable system for continuous vital sign monitoring using a multiparametric wearable sensor and personalised analytics in an Ebola treatment centre. BMJ Glob Heal. 2016;1(1). doi:10.1136/bmjgh-2016-000070
Loftus TJ, Tighe PJ, Filiberto AC, et al. Opportunities for machine learning to improve surgical ward safety. Am J Surg. 2020;220(4):905-913. doi:10.1016/j.amjsurg.2020.02.037
Vistisen ST, Johnson AEW, Scheeren TWL. Predicting vital sign deterioration with artificial intelligence or machine learning. J Clin Monit Comput. 2019;33(6):949-951. doi:10.1007/s10877-019-00343-7
Olsen RM, Aasvang EK, Meyhoff CS, Dissing Sorensen HB. Towards an automated multimodal clinical decision support system at the post anesthesia care unit. Comput Biol Med. 2018;101(July):15-21. doi:10.1016/j.compbiomed.2018.07.018
Elvekjaer M, Rasmussen SM, Grønbæk KK, et al. Clinical impact of vital sign abnormalities in patients admitted with acute exacerbation of chronic obstructive pulmonary disease: an observational study using continuous wireless monitoring. Intern Emerg Med. 2022;17(6):1689-1698. doi:10.1007/s11739-022-02988-w
Haahr-Raunkjaer C, Mølgaard J, Elvekjaer M, et al. Continuous monitoring of vital sign abnormalities; association to clinical complications in 500 postoperative patients. Acta Anaesthesiol Scand. 2022;66(5):552-562. doi:10.1111/aas.14048
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.