2026-09-21T19:00:52Zhttps://dspace.mit.edu/server/oai/request

oai:dspace.mit.edu:1721.1/1198442026-06-06T00:54:52Zcom_1721.1_7582com_1721.1_7581col_1721.1_131023

Thomas Heldt. Hensley, Sarah (Sarah L.) Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science. Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science 2019-01-03T20:55:53Z 2019-01-03T20:55:53Z 2018 2018 http://hdl.handle.net/1721.1/119844 1078699106 Thesis: M. Eng., Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, 2018. This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections. Cataloged from student-submitted PDF version of thesis. Includes bibliographical references (pages 71-73). Because the vast majority of monitoring alarms in the adult intensive care unit (ICU) do not require intervention, care providers are slow to respond to all alarms, endangering patients. We collect, characterize, and analyze alarms, alarm annotations provided by clinical staff while responding to alarms, and physiological data from a community hospital ICU. In order to suggest opportunities for suppressing irrelevant alarms, we examine monitoring device coverage across patients and analyze the alarms observed by device, priority, and type. On average, we observe 196.3 alarms per patient-day, for a total of 23,057 alarms. From these, the electrocardiogram and pulse plethysmogram produce 86.1% of all alarms. The lowest priority alarms represent 81.1% of all alarms, while the highest priority alarms compose just 5.5% of the total. While the rate of annotations is low, also just 5.5% of possible alarms, it is comparable to the rate of care provider interactions with alarms, as measured by alarm silencing, at 9.6%. Using these annotations, we find -- surprisingly -- that the annotated nuisance threshold-violation alarms tend to have higher excursions than actionable and advisory alarms, offering a statistic for separation. When focusing on threshold-crossing alarms, we find that 22.5% of Heart Rate Low alarms may actually indicate device error. Among ST segment alarms, 44.4% occur simultaneously with at least one other ST segment alarm, producing redundant alarms. Addressing these issues represent strategies for reducing excessive alarms in this community hospital cohort of ICU patients. by Sarah Hensley. M.Eng. 73 pages eng Massachusetts Institute of Technology MIT theses are protected by copyright. They may be viewed, downloaded, or printed from this source but further reproduction or distribution in any format is prohibited without written permission. http://dspace.mit.edu/handle/1721.1/7582 Electrical Engineering and Computer Science. Characterization of monitoring alarms in a community hospital intensive care unit Thesis application/pdf false Publication unknown unknown <Publication xmlns="https://www.openaire.eu/cerif-profile/1.1/" id="7dff6b67-3cb3-4bf7-8000-83c405bccd1b"> <Type xmlns="https://www.openaire.eu/cerif-profile/vocab/COAR_Publication_Types">http://purl.org/coar/resource_type/c_1843</Type> <Language>eng</Language> <Title>Characterization of monitoring alarms in a community hospital intensive care unit</Title> <PublishedIn> <Publication> </Publication> </PublishedIn> <PublicationDate>2018</PublicationDate> <Authors> <Author> <DisplayName>Hensley, Sarah (Sarah L.)</DisplayName> <Affiliation> <OrgUnit> </OrgUnit> </Affiliation> </Author> </Authors> <Editors> </Editors> <Publishers> <Publisher> <DisplayName>Massachusetts Institute of Technology</DisplayName> <OrgUnit /> </Publisher> </Publishers> <License>http://dspace.mit.edu/handle/1721.1/7582</License> <Keyword>Electrical Engineering and Computer Science.</Keyword> <Abstract>Because the vast majority of monitoring alarms in the adult intensive care unit (ICU) do not require intervention, care providers are slow to respond to all alarms, endangering patients. We collect, characterize, and analyze alarms, alarm annotations provided by clinical staff while responding to alarms, and physiological data from a community hospital ICU. In order to suggest opportunities for suppressing irrelevant alarms, we examine monitoring device coverage across patients and analyze the alarms observed by device, priority, and type. On average, we observe 196.3 alarms per patient-day, for a total of 23,057 alarms. From these, the electrocardiogram and pulse plethysmogram produce 86.1% of all alarms. The lowest priority alarms represent 81.1% of all alarms, while the highest priority alarms compose just 5.5% of the total. While the rate of annotations is low, also just 5.5% of possible alarms, it is comparable to the rate of care provider interactions with alarms, as measured by alarm silencing, at 9.6%. Using these annotations, we find -- surprisingly -- that the annotated nuisance threshold-violation alarms tend to have higher excursions than actionable and advisory alarms, offering a statistic for separation. When focusing on threshold-crossing alarms, we find that 22.5% of Heart Rate Low alarms may actually indicate device error. Among ST segment alarms, 44.4% occur simultaneously with at least one other ST segment alarm, producing redundant alarms. Addressing these issues represent strategies for reducing excessive alarms in this community hospital cohort of ICU patients.</Abstract> <Access xmlns="http://purl.org/coar/access_right" > </Access> </Publication>