Economic impacts of artificial intelligence (AI)-based risk analytics for clinical deterioration

Economic impacts of artificial intelligence (AI)-based risk analytics for clinical deterioration | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Short Report Economic impacts of artificial intelligence (AI)-based risk analytics for clinical deterioration Jessica Keim-Malpass, Marieke K. Jones, Sarah J. Ratcliffe, Matthew T Clark, and 5 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7917988/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 9 You are reading this latest preprint version Abstract The economic impacts of AI-based risk-analytic systems integrated in hospitals are not well understood. We assessed the hospital charges and costs of patients who were randomized to receive either a continuously displayed visual risk score or standard of care. There was evidence of differences in cost outcomes for the entire admission, favoring the standard of care, with differences ranging from 9.3% to 12.2% across study arms. Trial registration NCT04359641 (Registered 4/22/2020) Health sciences/Health care Physical sciences/Mathematics and computing Health sciences/Medical research Scientific community and society/Scientific community predictive monitoring economics economics of AI costs artificial intelligence clinical deterioration predictive analytics monitoring implementation of artificial intelligence Figures Figure 1 Introduction Late identification of clinical deterioration events is a major source of morbidity and mortality among patients hospitalized in the acute care setting and is among the ‘top three’ contributors of diagnostic error. 1 Attention has been placed on developing artificial intelligence (AI) for early prediction of clinical deterioration using continuous predictive analytics monitoring. 2 These algorithms use advanced computational approaches and multi-modal data (discrete data elements from the electronic medical record along with mathematically derived inputs from continuous electrocardiogram (ECG) monitoring) to allow early detection of illness, often before a patient exhibits clinical signs. 3 The resulting risk estimates are then displayed and updated every 15 minutes 4 For predictive analytics monitoring to successfully reduce diagnostic error for those at risk of clinical deterioration, clinicians must first respond to the score in the context of the patient’s complex care environment, care trajectory, and act in the realm of diagnostic uncertainty. 5 Frequent iterative clinical decision-making occurs, assessing the need for early action while balancing the potential for medical overuse with unnecessary diagnostics and treatment. The application of these predictive analytics monitoring systems has economic impacts that must be considered to determine optimal implementation. Clinical deterioration models have been assessed predominantly in retrospective analyses, with few implemented and even fewer studied in large RCTs. Further, the economic effects of these systems are not well understood. 6–8 In the parent RCT of this study, we investigated the overall impact of an AI-based system on the number of hours free from events of clinical deterioration. 4 Here, we aimed to assess the economic implications of this system. We assessed the hospital charges and costs of patients who were randomized to receive either a continuous visual risk score displayed versus standard of care in the acute care hospital setting. Methods We conducted a cluster-randomized controlled trial (RCT) of an AI-based risk display in 2021–2022 among patients admitted to acute care cardiology medical-surgical wards. The CoMET (Nihon Kohden Digital Health Solutions, Irvine, CA) system 9 was used as the intervention due to its multi-modal model development, which favors computational insights from continuous ECGs, the dynamic visual risk display that can be integrated into complex workflows 5 , and well-performing models. 10,11 The CoMET system updates every 15 minutes (see Appendix Fig. 1) with two axes: cardiorespiratory deterioration and cardiovascular deterioration, with sepsis often presenting on the diagonal between the two axes. Eleven clusters were identified among an 85-bed unit, and patients were randomized to the intervention group (CoMET display “on” plus standard monitoring practices) versus standard of care (standard monitoring practices only - no display). Full pre-registered study protocol details and additional randomization details can be found in (NCT04359641 Registered 4/22/2020). 4 The University of Virginia (UVA) IRB approved the study in accordance with the Declaration of Helsinki with a waiver of informed consent (IRB#22196). The randomization arms were analyzed as intention-to-treat. The trial concluded when meeting the pre-specified sample size and had an active Data Safety Monitoring Board. We collected both hospital charges and costs for each hospital admission associated with the study visit from the hospital data repository. In the hospital setting, "charge" refers to the listed price for a service, while “cost” represents the actual expense incurred by the hospital to provide that service; therefore, charges can be higher than the actual cost to the hospital. At-risk cohorts were defined as patients who experienced a rise of 2 or more on either axis of the CoMET display during their study visit. These scores are interpreted as an odds ratio, so a score of 2 indicates that you are at 2-fold risk of clinical deterioration relative to average, a score of 3 indicates a 3-fold risk, etc. Patient demographics in the analysis included race, ethnicity, age at visit start, and insurance type/payor (Medicare, Medicaid, self-pay/unknown, private insurance). We also describe the units patients were assigned to, and length of stay. We describe differences in patient characteristics across the trial arms. Generalized estimating equation models (GEE) were constructed using a log-normal distributional assumption to account for outliers associated with patients with high charges and costs incurred due to longer lengths of stay. Models assessed the relationship between the trial arms (AI-risk display intervention versus standard of care) and cost outcomes in both the full cohort and in those at-risk of clinical deterioration for patients who incurred a cost or charge above $ 0. The final model controlled for patient care unit, admitting diagnosis, and time period enrolled in the study to account for temporal differences in acuity. The {geepack} 12 package was used in R (R Foundation for Statistical Computing, Vienna, Austria) to calculate robust clustered standard errors that account for correlation between measurements from the same person using an exchangeable correlation structure. Data visualization techniques (density plots) were generated to display log-transformed total patient costs and charges across RCT groups. Results 10,138 admissions were included in the pragmatic cluster-randomized controlled trial, with demographic characteristics detailed in Table 1 . GEE models revealed statistically significant differences in costs in the full cohort after controlling for unit, admitting diagnosis, and time period enrolled, favoring 9.3% lower model-predicted costs in the full cohort in the standard of care arm (p < 0.001). In the at-risk cohort, the 12.2% reduction in model-predicted costs in the standard of care arm reached borderline statistical significance (p = 0.06). Distributions of costs and charges were similar between groups (Fig. 1 and Appendix Fig. 2) and between payor types ( Appendix Fig. 3). In post hoc analysis among patients who had bed changes, we noted that there was a higher patient acuity among those transferred to an AI risk display bed (not shown). [cite preprint] Table 1 Demographic characteristics of randomized controlled trial Admission Characteristic Whole cohort Whole cohort At risk At risk Control Intervention Control Intervention # At Risk 1,426 (28%) 1,459 (28%) Sex Female 2,181 (42%) 2,105 (40%) 587 (41%) 602 (41%) Male 2,999 (58%) 3,136 (60%) 839 (59%) 857 (59%) Race White 3,910 (75%) 3,942 (75%) 1,041 (73%) 1,053 (72%) African American 1,013 (20%) 1,009 (19%) 322 (23%) 316 (22%) Asian 47 (0.9%) 58 (1.1%) 11 (0.8%) 18 (1.2%) Other 198 (3.8%) 215 (4.1%) 47 (3.3%) 69 (4.7%) Unknown 12 (0.2%) 18 (0.3%) 5 (0.4%) 3 (0.2%) Ethnicity Hispanic 162 (3.1%) 174 (3.3%) 44 (3.1%) 66 (4.5%) Non-Hispanic 5,001 (97%) 5,056 (96%) 1,378 (97%) 1,390 (95%) /Unknown 17 (0.3%) 12 (0.2%) 4 (0.3%) 3 (0.2%) Age at Visit Start 65 (54, 75) 65 (55, 74) 65 (55, 74) 65 (54, 74) Unit Medical-Cardiology 1,851 (36%) 1,888 (36%) 427 (30%) 424 (29%) Cardio-Thoracic Surgery 1,708 (33%) 1,752 (33%) 506 (35%) 561 (38%) Medical/Surgical 1,621 (31%) 1,602 (31%) 493 (35%) 474 (32%) Payor Selfpay or Unknown 1,246 (24%) 1,281 (24%) 475 (33%) 483 (33%) Private 895 (17%) 982 (19%) 178 (12%) 215 (15%) Medicaid 537 (10%) 515 (9.8%) 121 (8.5%) 121 (8.3%) Medicare 2,502 (48%) 2,464 (47%) 652 (46%) 640 (44%) Hospital Length of Stay (days) 2.6 (1.1, 4.9) 2.7 (1.1, 5.0) 4.5 (2.6, 7.9) 4.7 (2.8, 7.8) Admitted with Congestive Heart Failure 798 (15%) 799 (15%) 287 (20%) 281 (19%) Total Charge 65,118 (30,200, 164,306) 70,072 (32,498, 172,308) 107,701 (49,387, 235,388) 118,958 (55,289, 254,043) Unknown 60 59 22 28 Total Cost 17,448 (8,250, 38,137) 18,715 (8,767, 40,074) 30,201 (14,066, 66,195) 32,840 (15,816, 73,179) Unknown 60 59 22 28 n (%); Median (Q1, Q3) Discussion Studying the economic implications of AI-based risk scores and early warning systems remains a challenge. There is a real possibility that early warning systems drive proactive action, resulting in earlier and longer interventions with improved patient and clinician outcomes. Proactive intervention enables care teams to act sooner and sustain treatment longer, thereby improving outcomes for both patients and clinicians—for example, promptly recognizing escalating needs and transferring a patient safely to the ICU rather than waiting until the patient deteriorates and arrives critically unstable, requiring emergency resuscitation. Furthermore, costs and charges associated with the entire hospital stay may not be the most important economic indicator of effectiveness. Finally, these models fail to integrate the direct impact on the clinician in the context of improved clinical decision-making or the potential to assess alert fatigue. 13 Obtaining adequate cost and charge data can be challenging. We present our own findings as hypothesis-generating in nature. Instead of relying on overall costs of care, determining the impact of the analytics and display systems on proximal actions (e.g., obtaining blood cultures, initiating broad-spectrum antibiotics, and averting clinical deterioration events) may be more fruitful in determining the overall economic impact on the system of care. The extent of movement between beds was unexpected and likely undermined the random nature of assignment in this real-world pragmatic design. 14 This movement of sicker patients to intervention beds could have contributed to the increased costs in display beds. Developing nuanced simulation models can help health systems determine the cost impacts prior to implementation. These models can also evaluate the cost-effectiveness in the context of the system infrastructure required, as well as the clinician and system impacts related to use. Economic assessment is also critical following the implementation of AI-based analytics. Future work is needed to understand the economic impacts of diagnostic errors and how clinicians cognitively process the trade-offs between early action and medical overuse during clinical decision-making when AI systems are integrated into care. Declarations Data availability : Unidentified and aggregated data can be made of the PIs with reasonable request. COI: JKM has equity in ArteraAI whose products are not discussed in this unrelated work. MC works for Nihon Kohden Digital Health Solutions, which owns and licenses CoMET, described in this work. He did not participate in data analysis or interpretation. LPM and JRM own equity and are paid consultants for Nihon Kohden Digital Health Solutions, which owns and licenses CoMET, described in this work. They did not participate in data analysis or interpretation. Author contribution: JKM and JMB secured the funding and conceptualized the analysis SJR and MKJ conducted the analysis JKM wrote the initial draft of the manuscript JKM, SJR, MTC, KNK, KJM, LPM, JRM, JMB participated in the conduct of the randomized controlled trial JKM, MKJ, SJR, MTC, KNK, KJM, LPM, JRM, JMB reviewed and edited the final manuscript draft and approved it for submission References Newman-Toker DE, Schaffer AC, Yu-Moe CW, et al. Serious misdiagnosis-related harms in malpractice claims: The “Big Three” - vascular events, infections, and cancers. Diagnosis (Berl) . 2019;6(3):227–240. doi: 10.1515/dx-2019-0019 Keim-Malpass J, Moorman LP. Nursing and precision predictive analytics monitoring in the acute and intensive care setting: An emerging role for responding to COVID-19 and beyond. International Journal of Nursing Studies Advances . 2021;3:100019. doi: 10.1016/j.ijnsa.2021.100019 Moss TJ, Lake DE, Calland JF, et al. Signatures of subacute potentially catastrophic illness in the ICU: model development and validation. Crit Care Med . 2016;44(9):1639–1648. doi: 10.1097/CCM.0000000000001738 Keim-Malpass J, Ratcliffe SJ, Moorman LP, et al. Predictive Monitoring-Impact in Acute Care Cardiology Trial (PM-IMPACCT): Protocol for a Randomized Controlled Trial. JMIR Res Protoc . 2021;10(7):e29631. doi: 10.2196/29631 Keim-Malpass J, Kitzmiller R, Skeeles-Worley A, et al. Advancing continuous predictive analytics monitoring: Moving from implementation to clincial action in a learning health system. Crit Care Nurs Clin North Am . 2018;30(2). Hendrix N, Veenstra DL, Cheng M, Anderson NC, Verguet S. Assessing the economic value of clinical artificial intelligence: challenges and opportunities. Value Health . 2022;25(3):331–339. doi: 10.1016/j.jval.2021.08.015 Wolff J, Pauling J, Keck A, Baumbach J. The economic impact of artificial intelligence in health care: systematic review. J Med Internet Res . 2020;22(2):e16866. doi: 10.2196/16866 Vithlani J, Hawksworth C, Elvidge J, Ayiku L, Dawoud D. Economic evaluations of artificial intelligence-based healthcare interventions: a systematic literature review of best practices in their conduct and reporting. Front Pharmacol . 2023;14:1220950. doi: 10.3389/fphar.2023.1220950 Moss TJ, Clark MT, Calland JF, et al. Cardiorespiratory dynamics measured from continuous ECG monitoring improves detection of deterioration in acute care patients: A retrospective cohort study. PLoS ONE . 2017;12(8):e0181448. doi: 10.1371/journal.pone.0181448 Keim-Malpass J, Moorman LP, Moorman JR, et al. Prospective validation of clinical deterioration predictive models prior to intensive care unit transfer among patients admitted to acute care cardiology wards. Physiol Meas . 2024;45(6). doi: 10.1088/1361-6579/ad4e90 Ruminski CM, Clark MT, Lake DE, et al. Impact of predictive analytics based on continuous cardiorespiratory monitoring in a surgical and trauma intensive care unit. J Clin Monit Comput . 2019;33(4):703–711. doi: 10.1007/s10877-018-0194-4 Højsgaard S, Halekoh U, Yan J. The r packagegeepack for generalized estimating equations. J Stat Softw . 2006;15(2). doi: 10.18637/jss.v015.i02 Ruppel H, Dougherty M, Bonafide CP, Lasater KB. Alarm burden and the nursing care environment: a 213-hospital cross-sectional study. BMJ Open Qual . 2023;12(4). doi: 10.1136/bmjoq-2023-002342 Keim-Malpass J, Ratcliffe SJ, Clark MT, et al. A pragmatic randomized controlled trial of artificial intelligence (AI)-based predictive analytics monitoring for early detection of clinical deterioration. medRxiv . January 22, 2025. doi: 10.1101/2025.01.20.25320838 Additional Declarations Competing interest reported. JKM has equity in ArteraAI whose products are not discussed in this unrelated work. MC works for Nihon Kohden Digital Health Solutions, which owns and licenses CoMET, described in this work. He did not participate in data analysis or interpretation. LPM and JRM own equity and are paid consultants for Nihon Kohden Digital Health Solutions, which owns and licenses CoMET, described in this work. They did not participate in data analysis or interpretation. Supplementary Files APPENDIX.docx Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Revision requested 05 Dec, 2025 Reviews received at journal 30 Nov, 2025 Reviews received at journal 17 Nov, 2025 Reviewers agreed at journal 09 Nov, 2025 Reviewers agreed at journal 03 Nov, 2025 Reviewers invited by journal 03 Nov, 2025 Editor assigned by journal 28 Oct, 2025 Submission checks completed at journal 28 Oct, 2025 First submitted to journal 21 Oct, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7917988","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Short Report","associatedPublications":[],"authors":[{"id":542713649,"identity":"510f4cdf-05ae-4409-acb3-eaf43074a3bb","order_by":0,"name":"Jessica Keim-Malpass","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA2UlEQVRIiWNgGAWjYBACxnY2xgcMFQw8QGYDQwJI6AAhLc1szAYHzpCihYGZjU3iYBuyCCEtzM1sCdIf5x2WMe9f3PbgQQ2DHN+NBIIOO2BwcNthHpkbD9sNEo4xGEsS1sLekHBwWxqPBNB5EglsDIkbiNFy4OAcmJZ/DPVEaGE72HCwwYZHgr+xTSKxjSHBgAgtyQxnjgG1SDACtfRJGM488wC/FsP2NvMfFTUS9hL8x59J/vhmI893nIAthg0wlgRYpQR+5SAgD2fxHyCsehSMglEwCkYmAAAUmUXhPXZZNgAAAABJRU5ErkJggg==","orcid":"","institution":"University of Virginia","correspondingAuthor":true,"prefix":"","firstName":"Jessica","middleName":"","lastName":"Keim-Malpass","suffix":""},{"id":542713652,"identity":"425b9019-f4bb-4347-ac3d-96d6f50f6ce5","order_by":1,"name":"Marieke K. Jones","email":"","orcid":"","institution":"University of Virginia","correspondingAuthor":false,"prefix":"","firstName":"Marieke","middleName":"K.","lastName":"Jones","suffix":""},{"id":542713653,"identity":"6c1a914b-ef6e-46c4-9caa-38f00d471651","order_by":2,"name":"Sarah J. Ratcliffe","email":"","orcid":"","institution":"University of Virginia","correspondingAuthor":false,"prefix":"","firstName":"Sarah","middleName":"J.","lastName":"Ratcliffe","suffix":""},{"id":542713655,"identity":"1cfbdcdd-57a2-45e9-9d07-96df73bb8f82","order_by":3,"name":"Matthew T Clark","email":"","orcid":"","institution":"Nihon Kohden Digital Health Solutions","correspondingAuthor":false,"prefix":"","firstName":"Matthew","middleName":"T","lastName":"Clark","suffix":""},{"id":542713656,"identity":"0d97d0ac-cc37-4f99-98b4-f6caa15a3129","order_by":4,"name":"Katy N Krahn","email":"","orcid":"","institution":"University of Virginia","correspondingAuthor":false,"prefix":"","firstName":"Katy","middleName":"N","lastName":"Krahn","suffix":""},{"id":542713657,"identity":"5808d24e-6481-4a05-96bb-d99f64b9150c","order_by":5,"name":"K. Jane Muir","email":"","orcid":"","institution":"University of Pennsylvania","correspondingAuthor":false,"prefix":"","firstName":"K.","middleName":"Jane","lastName":"Muir","suffix":""},{"id":542713661,"identity":"a5eab034-d60e-4189-a9e4-7b5c25bfed0f","order_by":6,"name":"Liza P Moorman","email":"","orcid":"","institution":"Nihon Kohden Digital Health Solutions","correspondingAuthor":false,"prefix":"","firstName":"Liza","middleName":"P","lastName":"Moorman","suffix":""},{"id":542713662,"identity":"674563a4-881a-47e7-a773-d0d8aca38612","order_by":7,"name":"J. Randall Moorman","email":"","orcid":"","institution":"University of Virginia","correspondingAuthor":false,"prefix":"","firstName":"J.","middleName":"Randall","lastName":"Moorman","suffix":""},{"id":542713663,"identity":"1c265cf1-1ce6-4b0a-8499-bb3162f005cb","order_by":8,"name":"Jamieson M Bourque","email":"","orcid":"","institution":"University of Virginia","correspondingAuthor":false,"prefix":"","firstName":"Jamieson","middleName":"M","lastName":"Bourque","suffix":""}],"badges":[],"createdAt":"2025-10-22 08:13:36","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7917988/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7917988/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":96022450,"identity":"910a795a-927a-479e-ba39-98f2e069be84","added_by":"auto","created_at":"2025-11-16 16:28:59","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3758223,"visible":true,"origin":"","legend":"","description":"","filename":"EconRCTdraftsubmit.docx","url":"https://assets-eu.researchsquare.com/files/rs-7917988/v1/d8cf3374c010d251cc6fb9e6.docx"},{"id":96022445,"identity":"de772c3c-9f85-43eb-a39a-5d30223ed3f1","added_by":"auto","created_at":"2025-11-16 16:28:59","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":9263,"visible":true,"origin":"","legend":"","description":"","filename":"5e49e922d60d4fa79e7e363bb4ddffcf.json","url":"https://assets-eu.researchsquare.com/files/rs-7917988/v1/d79cc1b3365b72b35c126e52.json"},{"id":96022442,"identity":"7078ca22-963d-4985-8d2f-93453f366330","added_by":"auto","created_at":"2025-11-16 16:28:59","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":62516,"visible":true,"origin":"","legend":"","description":"","filename":"5e49e922d60d4fa79e7e363bb4ddffcf1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7917988/v1/6ed997e1c7588135561c352f.xml"},{"id":96022448,"identity":"d984e4dc-dbcb-439f-b6c9-996f23995b7e","added_by":"auto","created_at":"2025-11-16 16:28:59","extension":"png","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":125335,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7917988/v1/54cf64b571f45541e31c7747.png"},{"id":96022444,"identity":"049071ac-3d1a-403e-8704-882bdef3d1d0","added_by":"auto","created_at":"2025-11-16 16:28:59","extension":"png","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":403978,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7917988/v1/bacf0a338da82ce5eee0e84b.png"},{"id":96246084,"identity":"ccd7823e-74d8-4811-9113-40aab65b1fc7","added_by":"auto","created_at":"2025-11-19 07:24:33","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":124276,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7917988/v1/388b073db86476324c7c53ce.png"},{"id":96022455,"identity":"0776ddcd-06b3-4564-9b85-03e2d7773ab4","added_by":"auto","created_at":"2025-11-16 16:28:59","extension":"png","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":185978,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7917988/v1/9cc0b302538a2067698172ac.png"},{"id":96245627,"identity":"b215c423-aab9-4b20-a2c0-42801162c08a","added_by":"auto","created_at":"2025-11-19 07:21:22","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":64252,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7917988/v1/f0c5b0006f508bcd61f8a1c3.png"},{"id":96022449,"identity":"a939a13a-c915-41ed-9c14-48c31a89a973","added_by":"auto","created_at":"2025-11-16 16:28:59","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":69353,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7917988/v1/aa249d0bb9c6095e4bba3353.png"},{"id":96022451,"identity":"765f8ec5-1d4b-432d-a0a4-66539f673152","added_by":"auto","created_at":"2025-11-16 16:28:59","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":61839,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7917988/v1/f7d08c3b49ac3a587a542f0c.png"},{"id":96246306,"identity":"9bb51b7d-8337-4ca3-b89c-53a98502a839","added_by":"auto","created_at":"2025-11-19 07:25:22","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":64901,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7917988/v1/19860f7c4317d57dab51d729.png"},{"id":96022454,"identity":"02c33985-e037-4f0e-b30d-b9d42fbea127","added_by":"auto","created_at":"2025-11-16 16:28:59","extension":"xml","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":59242,"visible":true,"origin":"","legend":"","description":"","filename":"5e49e922d60d4fa79e7e363bb4ddffcf1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7917988/v1/52009434bd2bdd8cb69bdcf8.xml"},{"id":96022456,"identity":"9bdbb528-2c3b-475a-b6fe-16fe34ad19b1","added_by":"auto","created_at":"2025-11-16 16:28:59","extension":"html","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":70632,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7917988/v1/5ef541c152cbdf62545b088e.html"},{"id":96246316,"identity":"c9fbfd9d-058c-41c7-be56-c47fb57b8ce7","added_by":"auto","created_at":"2025-11-19 07:25:24","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":352473,"visible":true,"origin":"","legend":"\u003cp\u003eTotal costs on the logarithmic scale among the full study cohort and the at-risk subgroup\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-7917988/v1/776fd8e60cedb7b17010c234.png"},{"id":96453003,"identity":"e375144e-a002-4b29-85d0-c17c6aeb2927","added_by":"auto","created_at":"2025-11-21 09:57:09","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":910297,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7917988/v1/4f3071fc-6de8-4c09-9c76-18b1e12c33f6.pdf"},{"id":96244975,"identity":"212c9e4b-da90-4bf8-aa7a-2b0e65978306","added_by":"auto","created_at":"2025-11-19 07:19:40","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":729276,"visible":true,"origin":"","legend":"","description":"","filename":"APPENDIX.docx","url":"https://assets-eu.researchsquare.com/files/rs-7917988/v1/e185c2748ca420ec9ac2fc85.docx"}],"financialInterests":"Competing interest reported. JKM has equity in ArteraAI whose products are not discussed in this unrelated work.\nMC works for Nihon Kohden Digital Health Solutions, which owns and licenses CoMET, described in this work. He did not participate in data analysis or interpretation.\nLPM and JRM own equity and are paid consultants for Nihon Kohden Digital Health Solutions, which owns and licenses CoMET, described in this work. They did not participate in data analysis or interpretation.","formattedTitle":"Economic impacts of artificial intelligence (AI)-based risk analytics for clinical deterioration","fulltext":[{"header":"Introduction","content":"\u003cp\u003eLate identification of clinical deterioration events is a major source of morbidity and mortality among patients hospitalized in the acute care setting and is among the \u0026lsquo;top three\u0026rsquo; contributors of diagnostic error. \u003csup\u003e1\u003c/sup\u003e Attention has been placed on developing artificial intelligence (AI) for early prediction of clinical deterioration using continuous predictive analytics monitoring.\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e These algorithms use advanced computational approaches and multi-modal data (discrete data elements from the electronic medical record along with mathematically derived inputs from continuous electrocardiogram (ECG) monitoring) to allow early detection of illness, often before a patient exhibits clinical signs.\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e The resulting risk estimates are then displayed and updated every 15 minutes \u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e\u003cp\u003eFor predictive analytics monitoring to successfully reduce diagnostic error for those at risk of clinical deterioration, clinicians must first respond to the score in the context of the patient\u0026rsquo;s complex care environment, care trajectory, and act in the realm of diagnostic uncertainty. \u003csup\u003e5\u003c/sup\u003e Frequent iterative clinical decision-making occurs, assessing the need for early action while balancing the potential for medical overuse with unnecessary diagnostics and treatment. The application of these predictive analytics monitoring systems has economic impacts that must be considered to determine optimal implementation.\u003c/p\u003e\u003cp\u003eClinical deterioration models have been assessed predominantly in retrospective analyses, with few implemented and even fewer studied in large RCTs. Further, the economic effects of these systems are not well understood. \u003csup\u003e6\u0026ndash;8\u003c/sup\u003e In the parent RCT of this study, we investigated the overall impact of an AI-based system on the number of hours free from events of clinical deterioration. \u003csup\u003e4\u003c/sup\u003e Here, we aimed to assess the economic implications of this system. We assessed the hospital charges and costs of patients who were randomized to receive either a continuous visual risk score displayed versus standard of care in the acute care hospital setting.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e We conducted a cluster-randomized controlled trial (RCT) of an AI-based risk display in 2021\u0026ndash;2022 among patients admitted to acute care cardiology medical-surgical wards. The CoMET (Nihon Kohden Digital Health Solutions, Irvine, CA) system \u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e was used as the intervention due to its multi-modal model development, which favors computational insights from continuous ECGs, the dynamic visual risk display that can be integrated into complex workflows\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e, and well-performing models. \u003csup\u003e10,11\u003c/sup\u003e The CoMET system updates every 15 minutes (see \u003cspan refid=\"Sec5\" class=\"InternalRef\"\u003eAppendix\u003c/span\u003e Fig.\u0026nbsp;1) with two axes: cardiorespiratory deterioration and cardiovascular deterioration, with sepsis often presenting on the diagonal between the two axes. Eleven clusters were identified among an 85-bed unit, and patients were randomized to the intervention group (CoMET display \u0026ldquo;on\u0026rdquo; plus standard monitoring practices) versus standard of care (standard monitoring practices only - no display).\u003c/p\u003e\u003cp\u003eFull pre-registered study protocol details and additional randomization details can be found in (NCT04359641 Registered 4/22/2020).\u003csup\u003e4\u003c/sup\u003e The University of Virginia (UVA) IRB approved the study in accordance with the Declaration of Helsinki with a waiver of informed consent (IRB#22196). The randomization arms were analyzed as intention-to-treat. The trial concluded when meeting the pre-specified sample size and had an active Data Safety Monitoring Board.\u003c/p\u003e\u003cp\u003eWe collected both hospital charges and costs for each hospital admission associated with the study visit from the hospital data repository. In the hospital setting, \"charge\" refers to the listed price for a service, while \u0026ldquo;cost\u0026rdquo; represents the actual expense incurred by the hospital to provide that service; therefore, charges can be higher than the actual cost to the hospital. At-risk cohorts were defined as patients who experienced a rise of 2 or more on either axis of the CoMET display during their study visit. These scores are interpreted as an odds ratio, so a score of 2 indicates that you are at 2-fold risk of clinical deterioration relative to average, a score of 3 indicates a 3-fold risk, etc. Patient demographics in the analysis included race, ethnicity, age at visit start, and insurance type/payor (Medicare, Medicaid, self-pay/unknown, private insurance). We also describe the units patients were assigned to, and length of stay.\u003c/p\u003e\u003cp\u003eWe describe differences in patient characteristics across the trial arms. Generalized estimating equation models (GEE) were constructed using a log-normal distributional assumption to account for outliers associated with patients with high charges and costs incurred due to longer lengths of stay. Models assessed the relationship between the trial arms (AI-risk display intervention versus standard of care) and cost outcomes in both the full cohort and in those at-risk of clinical deterioration for patients who incurred a cost or charge above \u003cspan\u003e$\u003c/span\u003e0. The final model controlled for patient care unit, admitting diagnosis, and time period enrolled in the study to account for temporal differences in acuity. The {geepack} \u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e package was used in R (R Foundation for Statistical Computing, Vienna, Austria) to calculate robust clustered standard errors that account for correlation between measurements from the same person using an exchangeable correlation structure. Data visualization techniques (density plots) were generated to display log-transformed total patient costs and charges across RCT groups.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e10,138 admissions were included in the pragmatic cluster-randomized controlled trial, with demographic characteristics detailed in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. GEE models revealed statistically significant differences in costs in the full cohort after controlling for unit, admitting diagnosis, and time period enrolled, favoring 9.3% lower model-predicted costs in the full cohort in the standard of care arm (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). In the at-risk cohort, the 12.2% reduction in model-predicted costs in the standard of care arm reached borderline statistical significance (p\u0026thinsp;=\u0026thinsp;0.06). Distributions of costs and charges were similar between groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e and \u003cspan refid=\"Sec5\" class=\"InternalRef\"\u003eAppendix\u003c/span\u003e Fig.\u0026nbsp;2) and between payor types (\u003cspan refid=\"Sec5\" class=\"InternalRef\"\u003eAppendix\u003c/span\u003e Fig.\u0026nbsp;3).\u003c/p\u003e\u003cp\u003eIn post hoc analysis among patients who had bed changes, we noted that there was a higher patient acuity among those transferred to an AI risk display bed (not shown). [cite preprint]\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eDemographic characteristics of randomized controlled trial\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAdmission Characteristic\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eWhole cohort\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eWhole cohort\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eAt risk\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eAt risk\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eControl\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eIntervention\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eControl\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eIntervention\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e# At Risk\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1,426 (28%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e1,459 (28%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSex\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eFemale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2,181 (42%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2,105 (40%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e587 (41%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e602 (41%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2,999 (58%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e3,136 (60%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e839 (59%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e857 (59%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRace\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eWhite\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e3,910 (75%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e3,942 (75%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1,041 (73%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e1,053 (72%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAfrican American\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1,013 (20%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1,009 (19%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e322 (23%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e316 (22%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAsian\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e47 (0.9%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e58 (1.1%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e11 (0.8%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e18 (1.2%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eOther\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e198 (3.8%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e215 (4.1%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e47 (3.3%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e69 (4.7%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eUnknown\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e12 (0.2%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e18 (0.3%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e5 (0.4%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3 (0.2%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEthnicity\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHispanic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e162 (3.1%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e174 (3.3%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e44 (3.1%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e66 (4.5%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNon-Hispanic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5,001 (97%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e5,056 (96%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1,378 (97%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e1,390 (95%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e/Unknown\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e17 (0.3%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e12 (0.2%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4 (0.3%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3 (0.2%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge at Visit Start\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e65 (54, 75)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e65 (55, 74)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e65 (55, 74)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e65 (54, 74)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eUnit\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMedical-Cardiology\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1,851 (36%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1,888 (36%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e427 (30%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e424 (29%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCardio-Thoracic Surgery\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1,708 (33%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1,752 (33%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e506 (35%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e561 (38%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMedical/Surgical\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1,621 (31%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1,602 (31%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e493 (35%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e474 (32%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePayor\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSelfpay or Unknown\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1,246 (24%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1,281 (24%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e475 (33%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e483 (33%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePrivate\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e895 (17%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e982 (19%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e178 (12%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e215 (15%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMedicaid\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e537 (10%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e515 (9.8%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e121 (8.5%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e121 (8.3%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMedicare\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2,502 (48%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2,464 (47%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e652 (46%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e640 (44%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHospital Length of Stay (days)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2.6 (1.1, 4.9)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.7 (1.1, 5.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4.5 (2.6, 7.9)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e4.7 (2.8, 7.8)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAdmitted with Congestive Heart Failure\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e798 (15%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e799 (15%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e287 (20%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e281 (19%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTotal Charge\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e65,118 (30,200, 164,306)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e70,072 (32,498, 172,308)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e107,701 (49,387, 235,388)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e118,958 (55,289, 254,043)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eUnknown\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e60\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e59\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e22\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e28\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTotal Cost\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e17,448 (8,250, 38,137)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e18,715 (8,767, 40,074)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e30,201 (14,066, 66,195)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e32,840 (15,816, 73,179)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eUnknown\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e60\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e59\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e22\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e28\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e\u003cp\u003en (%); Median (Q1, Q3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eStudying the economic implications of AI-based risk scores and early warning systems remains a challenge. There is a real possibility that early warning systems drive proactive action, resulting in earlier and longer interventions with improved patient and clinician outcomes. Proactive intervention enables care teams to act sooner and sustain treatment longer, thereby improving outcomes for both patients and clinicians\u0026mdash;for example, promptly recognizing escalating needs and transferring a patient safely to the ICU rather than waiting until the patient deteriorates and arrives critically unstable, requiring emergency resuscitation. Furthermore, costs and charges associated with the entire hospital stay may not be the most important economic indicator of effectiveness. Finally, these models fail to integrate the direct impact on the clinician in the context of improved clinical decision-making or the potential to assess alert fatigue. \u003csup\u003e13\u003c/sup\u003e\u003c/p\u003e\u003cp\u003eObtaining adequate cost and charge data can be challenging. We present our own findings as hypothesis-generating in nature. Instead of relying on overall costs of care, determining the impact of the analytics and display systems on proximal actions (e.g., obtaining blood cultures, initiating broad-spectrum antibiotics, and averting clinical deterioration events) may be more fruitful in determining the overall economic impact on the system of care.\u003c/p\u003e\u003cp\u003eThe extent of movement between beds was unexpected and likely undermined the random nature of assignment in this real-world pragmatic design. \u003csup\u003e14\u003c/sup\u003e This movement of sicker patients to intervention beds could have contributed to the increased costs in display beds.\u003c/p\u003e\u003cp\u003eDeveloping nuanced simulation models can help health systems determine the cost impacts prior to implementation. These models can also evaluate the cost-effectiveness in the context of the system infrastructure required, as well as the clinician and system impacts related to use. Economic assessment is also critical following the implementation of AI-based analytics. Future work is needed to understand the economic impacts of diagnostic errors and how clinicians cognitively process the trade-offs between early action and medical overuse during clinical decision-making when AI systems are integrated into care.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e: Unidentified and aggregated data can be made of the PIs with reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCOI:\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eJKM has equity in ArteraAI whose products are not discussed in this unrelated work.\u003c/p\u003e\n\u003cp\u003eMC works for Nihon Kohden Digital Health Solutions, which owns and licenses CoMET, described in this work. \u0026nbsp;He did not participate in data analysis or interpretation.\u003c/p\u003e\n\u003cp\u003eLPM and JRM own equity and are paid consultants for Nihon Kohden Digital Health Solutions, which owns and licenses CoMET, described in this work. \u0026nbsp;They did not participate in data analysis or interpretation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contribution:\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eJKM and JMB secured the funding and conceptualized the analysis\u003c/p\u003e\n\u003cp\u003eSJR and MKJ conducted the analysis\u003c/p\u003e\n\u003cp\u003eJKM wrote the initial draft of the manuscript\u003c/p\u003e\n\u003cp\u003eJKM, SJR, MTC, KNK, KJM, LPM, JRM, JMB participated in the conduct of the randomized controlled trial\u003c/p\u003e\n\u003cp\u003eJKM, MKJ, SJR, MTC, KNK, KJM, LPM, JRM, JMB reviewed and edited the final manuscript draft and approved it for submission\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eNewman-Toker DE, Schaffer AC, Yu-Moe CW, et al. Serious misdiagnosis-related harms in malpractice claims: The \u0026ldquo;Big Three\u0026rdquo; - vascular events, infections, and cancers. \u003cem\u003eDiagnosis (Berl)\u003c/em\u003e. 2019;6(3):227\u0026ndash;240. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1515/dx-2019-0019\u003c/span\u003e\u003cspan address=\"10.1515/dx-2019-0019\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKeim-Malpass J, Moorman LP. Nursing and precision predictive analytics monitoring in the acute and intensive care setting: An emerging role for responding to COVID-19 and beyond. \u003cem\u003eInternational Journal of Nursing Studies Advances\u003c/em\u003e. 2021;3:100019. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.ijnsa.2021.100019\u003c/span\u003e\u003cspan address=\"10.1016/j.ijnsa.2021.100019\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMoss TJ, Lake DE, Calland JF, et al. Signatures of subacute potentially catastrophic illness in the ICU: model development and validation. \u003cem\u003eCrit Care Med\u003c/em\u003e. 2016;44(9):1639\u0026ndash;1648. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/CCM.0000000000001738\u003c/span\u003e\u003cspan address=\"10.1097/CCM.0000000000001738\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKeim-Malpass J, Ratcliffe SJ, Moorman LP, et al. Predictive Monitoring-Impact in Acute Care Cardiology Trial (PM-IMPACCT): Protocol for a Randomized Controlled Trial. \u003cem\u003eJMIR Res Protoc\u003c/em\u003e. 2021;10(7):e29631. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2196/29631\u003c/span\u003e\u003cspan address=\"10.2196/29631\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKeim-Malpass J, Kitzmiller R, Skeeles-Worley A, et al. Advancing continuous predictive analytics monitoring: Moving from implementation to clincial action in a learning health system. \u003cem\u003eCrit Care Nurs Clin North Am\u003c/em\u003e. 2018;30(2).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHendrix N, Veenstra DL, Cheng M, Anderson NC, Verguet S. Assessing the economic value of clinical artificial intelligence: challenges and opportunities. \u003cem\u003eValue Health\u003c/em\u003e. 2022;25(3):331\u0026ndash;339. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jval.2021.08.015\u003c/span\u003e\u003cspan address=\"10.1016/j.jval.2021.08.015\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWolff J, Pauling J, Keck A, Baumbach J. The economic impact of artificial intelligence in health care: systematic review. \u003cem\u003eJ Med Internet Res\u003c/em\u003e. 2020;22(2):e16866. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2196/16866\u003c/span\u003e\u003cspan address=\"10.2196/16866\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eVithlani J, Hawksworth C, Elvidge J, Ayiku L, Dawoud D. Economic evaluations of artificial intelligence-based healthcare interventions: a systematic literature review of best practices in their conduct and reporting. \u003cem\u003eFront Pharmacol\u003c/em\u003e. 2023;14:1220950. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fphar.2023.1220950\u003c/span\u003e\u003cspan address=\"10.3389/fphar.2023.1220950\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMoss TJ, Clark MT, Calland JF, et al. Cardiorespiratory dynamics measured from continuous ECG monitoring improves detection of deterioration in acute care patients: A retrospective cohort study. \u003cem\u003ePLoS ONE\u003c/em\u003e. 2017;12(8):e0181448. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1371/journal.pone.0181448\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0181448\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKeim-Malpass J, Moorman LP, Moorman JR, et al. Prospective validation of clinical deterioration predictive models prior to intensive care unit transfer among patients admitted to acute care cardiology wards. \u003cem\u003ePhysiol Meas\u003c/em\u003e. 2024;45(6). doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1088/1361-6579/ad4e90\u003c/span\u003e\u003cspan address=\"10.1088/1361-6579/ad4e90\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRuminski CM, Clark MT, Lake DE, et al. Impact of predictive analytics based on continuous cardiorespiratory monitoring in a surgical and trauma intensive care unit. \u003cem\u003eJ Clin Monit Comput\u003c/em\u003e. 2019;33(4):703\u0026ndash;711. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s10877-018-0194-4\u003c/span\u003e\u003cspan address=\"10.1007/s10877-018-0194-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eH\u0026oslash;jsgaard S, Halekoh U, Yan J. The\u003cem\u003er\u003c/em\u003e packagegeepack for generalized estimating equations. \u003cem\u003eJ Stat Softw\u003c/em\u003e. 2006;15(2). doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.18637/jss.v015.i02\u003c/span\u003e\u003cspan address=\"10.18637/jss.v015.i02\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRuppel H, Dougherty M, Bonafide CP, Lasater KB. Alarm burden and the nursing care environment: a 213-hospital cross-sectional study. \u003cem\u003eBMJ Open Qual\u003c/em\u003e. 2023;12(4). doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1136/bmjoq-2023-002342\u003c/span\u003e\u003cspan address=\"10.1136/bmjoq-2023-002342\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKeim-Malpass J, Ratcliffe SJ, Clark MT, et al. A pragmatic randomized controlled trial of artificial intelligence (AI)-based predictive analytics monitoring for early detection of clinical deterioration. \u003cem\u003emedRxiv\u003c/em\u003e. January 22, 2025. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1101/2025.01.20.25320838\u003c/span\u003e\u003cspan address=\"10.1101/2025.01.20.25320838\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"npj-digital-medicine","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"npjdigitalmed","sideBox":"Learn more about [npj Digital Medicine](http://www.nature.com/npjdigitalmed/)","snPcode":"41746","submissionUrl":"https://submission.springernature.com/new-submission/41746/3","title":"npj Digital Medicine","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"NPJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"predictive monitoring, economics, economics of AI, costs, artificial intelligence, clinical deterioration, predictive analytics monitoring, implementation of artificial intelligence","lastPublishedDoi":"10.21203/rs.3.rs-7917988/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7917988/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe economic impacts of AI-based risk-analytic systems integrated in hospitals are not well understood. We assessed the hospital charges and costs of patients who were randomized to receive either a continuously displayed visual risk score or standard of care. There was evidence of differences in cost outcomes for the entire admission, favoring the standard of care, with differences ranging from 9.3% to 12.2% across study arms.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTrial registration\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNCT04359641 (Registered 4/22/2020)\u003c/p\u003e","manuscriptTitle":"Economic impacts of artificial intelligence (AI)-based risk analytics for clinical deterioration","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-11-16 16:28:54","doi":"10.21203/rs.3.rs-7917988/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-12-05T18:45:31+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-12-01T04:53:52+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-11-17T13:23:04+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"243350407406057824265657916460906711492","date":"2025-11-09T13:17:10+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"14208043923648646232709056136409580318","date":"2025-11-04T01:46:28+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-11-03T20:51:02+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-10-28T14:15:57+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-10-28T04:46:27+00:00","index":"","fulltext":""},{"type":"submitted","content":"npj Digital Medicine","date":"2025-10-21T16:42:48+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"npj-digital-medicine","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"npjdigitalmed","sideBox":"Learn more about [npj Digital Medicine](http://www.nature.com/npjdigitalmed/)","snPcode":"41746","submissionUrl":"https://submission.springernature.com/new-submission/41746/3","title":"npj Digital Medicine","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"NPJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"705b9269-202d-4a51-a1df-c4e2828d6903","owner":[],"postedDate":"November 16th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[{"id":57733910,"name":"Health sciences/Health care"},{"id":57733911,"name":"Physical sciences/Mathematics and computing"},{"id":57733912,"name":"Health sciences/Medical research"},{"id":57733913,"name":"Scientific community and society/Scientific community"}],"tags":[],"updatedAt":"2025-12-05T18:53:25+00:00","versionOfRecord":[],"versionCreatedAt":"2025-11-16 16:28:54","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7917988","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7917988","identity":"rs-7917988","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}