Using Supervised Machine Learning to Classify Real Alerts and Artifact in Online Multisignal Vital Sign Monitoring Data*

Chen, Lujie; Dubrawski, Artur; Wang, Donghan; Fiterau, Madalina; Guillame-Bert, Mathieu; Bose, Eliezer; Kaynar, A. Murat; Wallace, David J.; Guttendorf, Jane; Clermont, Gilles; Pinsky, Michael R.; Hravnak, Marilyn

doi:10.1097/ccm.0000000000001660

Cited by 62 publications

(51 citation statements)

References 33 publications

(25 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Machine learning algorithms, for example, have been used with success to distinguish real events from artifact in online multisignal vital sign monitoring data streams. 12 They may also better predict clinical deterioration than current Early Warning Score systems. 9…”

Section: Ward Monitoring 30mentioning

confidence: 99%

Ward monitoring 3.0

Michard

Sessler

2018

British Journal of Anaesthesia

View full text Add to dashboard Cite

leadership: the effect of a superior's gender on respiratory therapists' ability to challenge leadership during a lifethreatening emergency. Br J Anaesth 2017; 119: 697e702 14. Adams TL. Gender and feminization in health care professions. Sociol Compass 2010; 4: 454e65 [Online] 15. Rajacich D, Kane D, Williston C, Cameron S. If they do call you a nurse, it is always a "male nurse": experiences of men in the nursing profession.

show abstract

Section: Ward Monitoring 30mentioning

confidence: 99%

Ward monitoring 3.0

Michard

Sessler

2018

British Journal of Anaesthesia

View full text Add to dashboard Cite

show abstract

“…when early mobilization is highly desirable), smart and robust software are needed to filter artifacts and prevent alarm fatigue. 68 Given the low clinician/patient ratio in surgical wards, various physiological signals and variables also need to be integrated (data fusion) into single warning scores or wellness indexes, [69][70][71] so that nurses can easily and accurately identify patients who are in the process of worsening condition (Fig. 5).…”

Section: Emerging Technologies For Pain Assessmentmentioning

confidence: 99%

Digital innovations and emerging technologies for enhanced recovery programmes

Michard

Gan

Kehlet

2017

British Journal of Anaesthesia

View full text Add to dashboard Cite

Enhanced recovery programmes (ERPs) are increasingly used to improve post-surgical recovery. However, compliance to various components of ERPs-a key determinant of success-remains sub-optimal. Emerging technologies have the potential to help patients and caregivers to improve compliance with ERPs.Preoperative physical condition, a major determinant of postoperative outcome, could be optimized with the use of text messages (SMS) or digital applications (Apps) designed to facilitate smoking cessation, modify physical activity, and better manage hypertension and diabetes. Several non-invasive haemodynamic monitoring techniques and decision support tools are now available to individualize perioperative fluid management, a key component of ERPs. Objective nociceptive assessment may help to rationalize the use of pain medications, including opioids. Wearable sensors designed to monitor cardio-respiratory function may help in the early detection of clinical deterioration during the postoperative recovery and to address 'failure to rescue'. Activity trackers may be useful to monitor early mobilization, another major element of ERPs. Finally, electronic checklists have been developed to ensure that none of the above-mentioned ERP elements is omitted during the surgical journey.By optimizing compliance to the multiple components of ERPs, digital innovations, non-invasive techniques and wearable sensors have the potential to magnify the clinical and economic benefits of ERPs. Among the growing number of technical innovations, studies are needed to clarify which tools and solutions have real clinical value and are cost-effective.

show abstract

“…[7]) and persisting for at least 80% of the time over a 3-minute window. Using the method described in our previous work (8,9), we identified a subset of clinically important CRI events to be used as the predictive endpoint in this study.…”

Section: Methodsmentioning

confidence: 99%

Dynamic and Personalized Risk Forecast in Step-Down Units. Implications for Monitoring Paradigms

et al. 2017

Self Cite

View full text Add to dashboard Cite

Rationale: Cardiorespiratory insufficiency (CRI) is a term applied to the manifestations of loss of normal cardiorespiratory reserve and portends a bad outcome. CRI occurs commonly in hospitalized patients, but its risk escalation patterns are unexplored.Objectives: To describe the dynamic and personal character of CRI risk evolution observed through continuous vital sign monitoring of individual step-down unit patients.Methods: Using a machine learning model, we estimated risk trends for CRI (defined as exceedance of vital sign stability thresholds) for each of 1,971 admissions (1,880 unique patients) to a 24-bed adult surgical trauma step-down unit at an urban teaching hospital in Pittsburgh, Pennsylvania using continuously recorded vital signs from standard bedside monitors. We compared and contrasted risk trends during initial 4-hour periods after step-down unit admission, and again during the 4 hours immediately before the CRI event, between cases (ever had a CRI) and control subjects (never had a CRI). We further explored heterogeneity of risk escalation patterns during the 4 hours before CRI among cases, comparing personalized to nonpersonalized risk.Measurements and Main Results: Estimated risk was significantly higher for cases (918) than control subjects (1,053; P < 0.001) during the initial 4-hour stable periods. Among cases, the aggregated nonpersonalized risk trend increased 2 hours before the CRI, whereas the personalized risk trend became significantly different from control subjects 90 minutes ahead. We further discovered several unique phenotypes of risk escalation patterns among cases for nonpersonalized (14.6% persistently high risk, 18.6% early onset, 66.8% late onset) and personalized risk (7.7% persistently high risk, 8.9% early onset, 83.4% late onset). Conclusions:Insights from this proof-of-concept analysis may guide design of dynamic and personalized monitoring systems that predict CRI, taking into account the triage and real-time monitoring utility of vital signs. These monitoring systems may prove useful in the dynamic allocation of technological and clinical personnel resources in acute care hospitals.

show abstract

Using Supervised Machine Learning to Classify Real Alerts and Artifact in Online Multisignal Vital Sign Monitoring Data*

Cited by 62 publications

References 33 publications

Ward monitoring 3.0

Ward monitoring 3.0

Digital innovations and emerging technologies for enhanced recovery programmes

Dynamic and Personalized Risk Forecast in Step-Down Units. Implications for Monitoring Paradigms

Contact Info

Product

Resources

About