The effect of technical filtering and clinical criteria on alert rates from continuous vital sign monitoring in the general ward

ABSTRACT

Objectives

Continuous vital sign monitoring at the general hospital ward has major potential advantages over intermittent monitoring but generates many alerts with risk of alert fatigue. We hypothesized that the number of alerts would decrease using different filters.

Methods

This study was an exploratory analysis of the alert reducing effect from adding two different filters to continuously collected vital sign data (peripheral oxygen saturation, blood pressure, heart rate, and respiratory rate) in patients admitted after major surgery or severe medical disease. Filtered data were compared to data without artifact removal. Filter one consists of artifact removal, filter two consists of artifact removal plus duration criteria adjusted for severity of vital sign deviation. Alert thresholds were based on the National Early Warning Score (NEWS) threshold.

Results

A population of 716 patients admitted for severe medical disease or major surgery with continuous wireless vital sign monitoring at the general ward with a mean monitoring time of 75.8 h, were included for the analysis. Without artifact removal, we found a median of 137 [IQR: 87–188] alerts per patient/day, artifact removal resulted in a median of 101 [IQR: 56–160] alerts per patient/day and with artifact removal combined with a duration-severity criterion, we found a median of 19 [IQR: 9–34] alerts per patient/day. Reduction of alerts was 86.4% (p < 0.001) for values without artifact removal (137 alerts) vs. the duration criteria and a reduction (19 alerts) of 81.5% (p < 0.001) for the criteria with artifact removal (101 alerts) vs. the duration criteria (19 alerts).

Conclusion

We conclude that a combination of artifact removal and duration-severity criteria approach substantially reduces alerts generated by continuous vital sign monitoring.

KEYWORDS:

• Continuous monitoring
• alert fatigue
• anesthesia
• vital signs
• artificial intelligence

Declaration of financial/other relationships

EK Aasvang and CS Meyhoff have founded a start-up company, WARD247 ApS, to pursue the WARD-projects regulatory and commercial activities. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Author contributions

Study conception and design: Karoline Kjærgaard, Eske Kvanner Aasvang, Christian Sylvest Meyhoff and Jesper Mølgaard. Data collection: Karoline Kjærgaard, Søren Møller Rasmussen and Jesper Mølgaard. Analysis and interpretation of results: Karoline Kjærgaard, Eske Kvanner Aasvang and Søren Møller Rasmussen. Draft manuscript preparation: Karoline Kjærgaard and Eske Kvanner Aasvang. All authors reviewed the results and approved the final version of the manuscript.

Additional information

Funding

This paper was funded by The Innovation Fund Denmark [8056-00055A], the Danish Cancer Society [R150-A9865-16-S48], Steno Diabetes Center Denmark [20200506] and institutional funding. WARD247 ApS has finalized terms for license agreement for any WARD-project software and patents, of which one has been filed. None of the above entities influences the study design, conduct, analysis, or reporting. All of the four studies were approved by the Danish Data Protection Agency, registered at http://ClinicalTrials.gov and approved or approval was waived by the Regional Ethics Committee [Surgery: H-17033535, COPD: H-18026653, CGM: H-20002220 and VASC: H-19086583].