The Use of Patient Monitoring Systems to Improve Sepsis Recognition and Outcomes: A Systematic Review

Abstract

Introduction The aim of this systematic review was to determine the impact of automated patient monitoring systems (PMSs) on sepsis recognition and outcomes. Methods Systematic searches were conducted using CINAHL, MEDLINE, and Cochrane, for articles published from 2008 through 2018. English-language, peer-reviewed articles that reported the impact of PMS on sepsis care were included. For selected articles, the authors abstracted information, with the study designed to be compliant with Preferred Reporting Items for Systematic reviews and Meta-Analyses guidelines. Results Nineteen articles were identified for inclusion: 4 systematic reviews and 15 individual studies. Study design and quality varied, with some randomized controlled trials and quasiexperimental studies, as well as many observational studies. Study results for outcome measures (e.g., mortality, intensive care unit [ICU] length of stay, ICU transfer) were mixed, with more than half of the studies showing a significant improvement in at least one measure. Evidence for process measure (e.g., time to antibiotic administration, lactate measurement, etc.) improvement was of moderate strength across multiple types of hospital units, and evidence was most consistent outside the ICU. Conclusions Automated sepsis PMSs have the potential to improve sepsis recognition and outcomes, but current evidence is mixed on their effectiveness. More high-quality studies are needed to understand the effects of PMSs on important sepsis-related process and outcome measures in different hospital units.

Closed-Loop Medication Management with an Electronic Health Record System in U.S. and Finnish Hospitals

“…Interfaced when technology allows. Monitoring data included in dashboards; patient scoring tools, or machine learning used for early identification of diseases such as sepsis and acute kidney injury [43][44][45][46].…”

“…Monitoring data is frequently included in EHR order entry, dashboards, and patient scoring tools to provide just-in-time information and trending or scoring tools to identify patients in need of clinical intervention. Machine learning has also been integrated into EHR for the early identification of diseases such as sepsis and acute kidney injury [43][44][45][46].…”

“…Furthermore, artificial intelligence and machine learning are potential technologies to prevent adverse drug events and these have been already applied, e.g., for drug discovery and pharmacovigilance purposes [72,73]. Machine learning is integrated into some hospital EHRs in the U.S. for early identification of diseases such as sepsis and acute kidney injury [43][44][45][46]. Artificial intelligence is being investigated to improve patients' self-administration of medications [74].…”