東華大學圖書館 |

語系: 繁體中文

說明(常見問題)

回圖書館首頁

手機版館藏查詢

登入

回首頁

切換: 標籤 | MARC模式 | ISBD

FindBook

Google Book

Amazon

博客來

Personalized Clinical Decision Support Using Machine Learning.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Personalized Clinical Decision Support Using Machine Learning./
作者:	Chen, Ji.
出版者:	Ann Arbor : ProQuest Dissertations & Theses, : 2022,
面頁冊數:	186 p.
附註:	Source: Dissertations Abstracts International, Volume: 83-10, Section: B.
Contained By:	Dissertations Abstracts International83-10B.
標題:	Pharmacology. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=28962643
ISBN:	9798426805064

Personalized Clinical Decision Support Using Machine Learning.
Chen, Ji.

Personalized Clinical Decision Support Using Machine Learning. - Ann Arbor : ProQuest Dissertations & Theses, 2022 - 186 p.

Source: Dissertations Abstracts International, Volume: 83-10, Section: B.

Thesis (Ph.D.)--New York University, 2022.

This item must not be sold to any third party vendors.

Although clinical decision support (CDS) alerts are effective reminders of best practices, their effectiveness is blunted by clinicians who fail to respond to an overabundance of inappropriate alerts. An electronic health record (EHR)-integrated machine learning (ML) algorithm is a potentially powerful tool to increase the signal-to-noise ratio of CDS alerts and positively impact the clinician's interaction with these alerts in general. The objective of this work was to develop and implement an ML-based signal-to-noise optimization system to increase the signal of alerts by decreasing the volume of low-value CDS alerts. We built and deployed the SmartCDS system, which builds personalized user activity profiles to suppress shingles vaccination alerts unlikely to yield a clinician's interaction. During the 6 weeks of pilot deployment, the SmartCDS system suppressed an average of 43.67% (15,425/35,315) potential shingles alerts (appointments) and maintained stable counts of weekly shingles vaccination orders (326.3 with system active vs 331.3 in the control group; P=.38) and weekly user-alert interactions (1118.3 with system active vs 1166.3 in the control group; P=.20). During 245 days in 2019 when the system was live, we suppressed 138,146 (43%) shingles alerts among 508 providers and 172,700 patients, and all key statistics remained stable while the system was turned on. Features derived from user profiling are the most impactful features, among which the short term alert accept rate of the shingles alert is the most valuable feature. We also found evidence that providers have a preference for off work hours in terms of accepting the shingles alert. Through over 100 user case reviews, we found little evidence that prospective intervention from the SmartCDS system changed user behavior patterns towards the shingles alert. The workflow from this case study was applied to other use cases through a generic model framework and proved decent generalizability. This work is the first alert suppression ML-based model deployed in practice and serves as foundational work in customization of alert display to maximize CDS effectiveness.

ISBN: 9798426805064Subjects--Topical Terms:

634543
Pharmacology.
Subjects--Index Terms:

Alert fatigue

Personalized Clinical Decision Support Using Machine Learning.
LDR:03384nmm a2200397 4500 001 2352178
005 20221118093834.5
008 241004s2022 ||||||||||||||||| ||eng d
020 $a 9798426805064
035 $a (MiAaPQ)AAI28962643
035 $a AAI28962643
040 $a MiAaPQ $c MiAaPQ
100 1 $a Chen, Ji. $3 1258307
245 1 0 $a Personalized Clinical Decision Support Using Machine Learning.
260 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2022
300 $a 186 p.
500 $a Source: Dissertations Abstracts International, Volume: 83-10, Section: B.
500 $a Advisor: Aphinyanaphongs, Yindalon;Mann, Devin.
502 $a Thesis (Ph.D.)--New York University, 2022.
506 $a This item must not be sold to any third party vendors.
520 $a Although clinical decision support (CDS) alerts are effective reminders of best practices, their effectiveness is blunted by clinicians who fail to respond to an overabundance of inappropriate alerts. An electronic health record (EHR)-integrated machine learning (ML) algorithm is a potentially powerful tool to increase the signal-to-noise ratio of CDS alerts and positively impact the clinician's interaction with these alerts in general. The objective of this work was to develop and implement an ML-based signal-to-noise optimization system to increase the signal of alerts by decreasing the volume of low-value CDS alerts. We built and deployed the SmartCDS system, which builds personalized user activity profiles to suppress shingles vaccination alerts unlikely to yield a clinician's interaction. During the 6 weeks of pilot deployment, the SmartCDS system suppressed an average of 43.67% (15,425/35,315) potential shingles alerts (appointments) and maintained stable counts of weekly shingles vaccination orders (326.3 with system active vs 331.3 in the control group; P=.38) and weekly user-alert interactions (1118.3 with system active vs 1166.3 in the control group; P=.20). During 245 days in 2019 when the system was live, we suppressed 138,146 (43%) shingles alerts among 508 providers and 172,700 patients, and all key statistics remained stable while the system was turned on. Features derived from user profiling are the most impactful features, among which the short term alert accept rate of the shingles alert is the most valuable feature. We also found evidence that providers have a preference for off work hours in terms of accepting the shingles alert. Through over 100 user case reviews, we found little evidence that prospective intervention from the SmartCDS system changed user behavior patterns towards the shingles alert. The workflow from this case study was applied to other use cases through a generic model framework and proved decent generalizability. This work is the first alert suppression ML-based model deployed in practice and serves as foundational work in customization of alert display to maximize CDS effectiveness.
590 $a School code: 0146.
650 4 $a Pharmacology. $3 634543
650 4 $a Artificial intelligence. $3 516317
650 4 $a Health care management. $3 2122906
650 4 $a Information science. $3 554358
650 4 $a Health sciences. $3 3168359
653 $a Alert fatigue
653 $a Clinical Decision Support
653 $a Machine Learning
653 $a Decision making
653 $a Electronic Health Record
690 $a 0419
690 $a 0723
690 $a 0800
690 $a 0769
690 $a 0566
710 2 $a New York University. $b Basic Medical Science. $3 1683852
773 0 $t Dissertations Abstracts International $g 83-10B.
790 $a 0146
791 $a Ph.D.
792 $a 2022
793 $a English
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=28962643