Implementation and Evaluations of Sepsis Watch

Implementation and Evaluations of Previously Developed Novel Early Warning System to Detect and Treat Sepsis

The purpose of this study is to study the implementation and impact of an early warning system to detect and treat sepsis in the emergency room. We are observing the implementation of a Sepsis Machine Learning Model on all Adult patients. All data (observations field notes, interview recording & transcripts, and survey responses) will be stored on HIPAA-compliant Duke servers behind the Duke firewall, and requiring password-protected user authentication to access. The risk to patients is minimal. The two risks to interviewed clinical staff we have identified involve loss of work time and anonymity.

Study Overview

• Status: Completed
• Conditions: Sepsis; Septic Shock; Severe Sepsis
• Intervention / Treatment: Other: Sepsis Watch

Detailed Description

Sepsis represents a significant burden to the healthcare system. National predictions estimate 751,000 cases of severe sepsis per annum which will increase at a rate of 1.5%. Sepsis accounts for >$23 billion in aggregate hospital costs across all payers and represents nearly 4% of all hospital stays. Six percent of all deaths in the US can be attributed to sepsis. Protocol driven care bundles improve clinical outcomes but require early and accurate detection of sepsis. Unfortunately, identifying sepsis early remains elusive even for experienced clinicians leading to diagnostic uncertainty.

To improve diagnostic consensus, a task force in 2016 agreed upon a new sepsis definition. The task force also included a new risk stratification tool to improve early identification, the quick Sepsis-related Organ Failure Assessment (qSOFA) model, which was more accurate than the older Systemic Inflammatory Response Syndrome (SIRS) in predicting adverse clinical outcomes. However, due to the reliance of end organ dysfunction, the new definition has been criticized for its detection of sepsis late in the clinical course. Clinical decision support tools based on predictive analytics can provide actionable information and improve diagnostic accuracy particularly in sepsis.

Several early warning tools have been described in the published literature based upon predictive analytics and large datasets. One example is the National Early Warning Score (NEWS), which was developed to discriminate patients at risk of cardiac arrest, unplanned intensive care admission, or death. Scores such as NEWS are typically broad in scope and not designed to specifically target sepsis. They are also conceptually simple, as they use only a small number of variables and compare them to normal ranges to generate a composite score. In assigning independent scores to each variable and using only the most recent value, they both ignore complex relationships between the variables and their evolution in time.

In previous work, our group developed a framework to model multivariate time series using multitask Gaussian processes, accounting for the high uncertainty, frequent missing values, and irregular sampling rates typically associated with real clinical data can be read in our prior work. Our machine learning approach is superior to other sepsis detection models that use traditional analytics and machine learning techniques. A custom web application, Sepsis Watch, presents the risk score along with relevant patient information and prompts the user to further evaluate the patient and begin treatment, if appropriate. The Sepsis Watch system is now being implemented by clinical operations at Duke University Hospital.

Our study employs a sequential roll-out study design in the Emergency Department at Duke University Hospital. Our study will involve pods A, B, C, and the Resuscitation Bay. The operational project is not being implemented on the psychiatry wing, fast track, triage or any inpatient encounters. The operational project and thus our study period is based upon a two-phase roll out:

• 1st phase: The predictive model notifies the rapid response team through a dashboard. Nurse notifies team of the risk for sepsis and provides treatment recommendation to primary team and primary team will place orders. Rapid response team nurse documents assessment and actions taken in electronic health record.
• 2nd phase: Improvement and optimization of the workflow integrated in phase 1. One workflow improvement includes the development of an ordering protocol and process whereby the rapid response team can place orders for patients who are deemed appropriate for sepsis treatment. A second workflow improvement includes the development of a clinician feedback and auditing report that would be sent to front-line staff with sepsis bundle compliance performance measures.

In addition to observing patient outcome measures, we propose an additional mixed-methods study component to obtain richer information about the effects of the early warning system on clinicians' situational awareness, decision-making, and workflow. This part of our research will involve (1) gathering data from clinicians through a series of semi-structured interviews, surveys, and observations (2) analysis of this data and identification of relevant patterns and insights. Relevant clinicians include include rapid response team nurses, emergency department (ED) nurses, and ED physicians. These interviews will be conducted in three rounds over the implementation period: before the 1st arm, after the 1st arm, and after the 2nd arm. Electronic surveys will be administered at the end of the 1st arm and the 2nd arm to clinicians. The observations will take place during the 1st and 2nd arms.

The goal of the interviews, surveys, and observations will be to (1) evaluate the effect of the early warning system on the clinicians' situational awareness and decision-making, (2) understand how the early warning system fits into clinician workflow, and, (3) identify opportunities to improve the implementation of the early warning system for future scale-up.

We will be structuring interviews according to the situational awareness model which differentiates between 3 levels of situational awareness: 1) perception of relevant information, 2) comprehension of that information, and 3) anticipation of future events based on that information. Through the interviews, observations, and surveys, we also hope to learn more about clinicians' perceptions of and interactions with the early warning system, and its change on the existing Emergency Department workflow for sepsis diagnosis and management. Data analysis will be conducted with the help of trained qualitative researchers from Data & Society, a research institute in New York City that is focused on the social and cultural issues arising from data-centric technological development.

• Study Type: Interventional
• Enrollment (Actual): 32003
• Phase: Not Applicable

Contacts and Locations

Study Locations

• United States
  • North Carolina
    • Durham, North Carolina, United States, 27710
      • Duke University Hospital

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: 18 years and older (Adult, Older Adult)
• Accepts Healthy Volunteers: No
• Genders Eligible for Study: All

Description

Inclusion Criteria:

• Arrival to Duke University Hospital emergency department pods A, B, and C, or resuscitation bay

Exclusion Criteria:

• Under 18 years old at time of emergency department arrival

Study Plan

How is the study designed?

Design Details

• Primary Purpose: Treatment
• Allocation: N/A
• Interventional Model: Single Group Assignment
• Masking: None (Open Label)

Number of Arms

1

Arms and Interventions

Participant Group / Arm

Intervention / Treatment

Experimental: Sepsis Watch on Duke University Hospital ED Adults; Patients older than 18 years old at time of presentation to Duke University Hospital emergency department.

Other: Sepsis Watch; The operational intervention comprises of a sepsis machine learning model, custom dashboard to present risk scores, and a rapid response team to monitor patients at-risk of sepsis and deliver sepsis treatment. Sepsis Watch was developed under operational management. The rapid response team will utilize information presented on the dashboard and follow a protocol that will enable them to support the primary teams of hospitalized patients.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Rate of Centers for Medicare and Medicaid Services (CMS) bundle completion for patients with sepsis	Proportion of patients with sepsis that complete Center for Medicare and Medicaid Services treatment bundle	Within 96 hours of emergency department arrival

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Mean time from ED arrival to sepsis for patients with sepsis	Mean time from to ED arrival to sepsis	Within 96 hours of emergency department arrival
Average number of patients who develop sepsis per day and month	Number of patients daily who meet sepsis phenotype	Within 96 hours of emergency department arrival
Average number of patients who develop sepsis and are not treated per day and month	Number of patients daily who meet sepsis phenotype who are not treated for sepsis	Within 96 hours of emergency department arrival
Mean ED length of stay for patients with sepsis	Emergency department length of stay for patients with sepsis	Within 96 hours of emergency department arrival
Mean Hospital length of stay for patients with sepsis	Hospital length of stay for patients with sepsis	Within 30 days of emergency department arrival
Mean Inpatient mortality for patients with sepsis	Inpatient mortality for patients with sepsis	Within 30 days of emergency department arrival
Mean ICU requirement rate for patients with sepsis	Intensive care unit requirement rate for patients with sepsis	Within 30 days of emergency department arrival
Mean time from sepsis onset to blood culture	Time of sepsis to blood culture order and collection for patients with sepsis	Within 96 hours of emergency department arrival
Mean time from sepsis onset to antibiotics	Time of sepsis to antibiotic order and administration for patients with sepsis	Within 96 hours of emergency department arrival
Mean time from sepsis onset to IV fluids	Time of sepsis to IV fluids order and administration for patients with sepsis	Within 96 hours of emergency department arrival
Mean time from sepsis onset to lactate	Time of sepsis to lactate collection for patients with sepsis	Within 96 hours of emergency department arrival
Mean time from sepsis onset to CMS bundle completion	Time of sepsis to CMS bundle completion for patients with sepsis	Within 96 hours of emergency department arrival
Rate of lactate complete for patients with sepsis	Proportion of lactate drawn within 3 hours and potentially re-drawn within 6 hours of sepsis for patients with sepsis	Within 96 hours of emergency department arrival
Number of sepsis diagnosis codes across Duke University Hospital patients per month	Number of billing diagnosis codes for sepsis	Within 30 days of emergency department arrival

Other Outcome Measures

Outcome Measure	Measure Description	Time Frame
Number of antibiotic orders in Duke University Hospital emergency department per month	Number of antibiotic orders in Duke University Hospital emergency department per month	Within 96 hours of emergency department arrival
Number of antibiotic days in Duke University Hospital emergency department per month	Number of antibiotic days	Within 96 hours of emergency department arrival
Number of blood culture orders in Duke University Hospital emergency department per month	Total blood culture orders	Within 96 hours of emergency department arrival
Number of lactate orders in Duke University Hospital emergency department per month	Total lactate orders	Within 96 hours of emergency department arrival
Number of IV fluid orders in Duke University Hospital emergency department per month	Total IV fluid orders	Within 96 hours of emergency department arrival
Number of vasopressor orders in Duke University Hospital emergency department per month	Total vasopressor orders	Within 96 hours of emergency department arrival
Number of vasopressor days in Duke University Hospital emergency department per month	Number of vasopressor days	Within 96 hours of emergency department arrival

Collaborators and Investigators

• Sponsor: Duke University
• Collaborators: Duke Clinical Research Institute; Data & Society Research Institute
• Investigators: Principal Investigator: Cara O'Brien, MD, Duke Health; Study Director: Mark Sendak, MD, Duke Institute for Health Innovation

Publications and helpful links

Helpful Links

• Learning to Detect Sepsis with a Multitask Gaussian Process Recurrent Neural Network Classifier
• An Improved Multi-Output Gaussian Process Recurrent Neural Network with Real-Time Validation for Early Sepsis Detection
• Society of Hospital Medicine Abstract
• American Thoracic Society Abstract

Study record dates

Study Major Dates

• Study Start (Actual): November 5, 2018
• Primary Completion (Actual): July 5, 2019
• Study Completion (Actual): July 5, 2019

Study Registration Dates

• First Submitted: August 30, 2018
• First Submitted That Met QC Criteria: August 30, 2018
• First Posted (Actual): August 31, 2018

Study Record Updates

• Last Update Posted (Actual): August 1, 2019
• Last Update Submitted That Met QC Criteria: July 31, 2019
• Last Verified: July 1, 2019

More Information

Terms related to this study

• Keywords: Sepsis; Implementation Science; Health Information Technology; Machine Learning
• Additional Relevant MeSH Terms: Pathologic Processes; Infections; Systemic Inflammatory Response Syndrome; Inflammation; Sepsis; Toxemia
• Other Study ID Numbers: Pro00093721

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: No

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No