Supply Chain Delays in Antimicrobial Administration After the Initial Clinician Order and Mortality in Patients With Sepsis

Sepsis Prediction at Emergency Department Triage Using Natural Language Processing: Retrospective Cohort Study

BACKGROUND

Despite its high lethality, sepsis can be difficult to detect on initial presentation to the emergency department (ED). Machine learning-based tools may provide avenues for earlier detection and lifesaving intervention.

OBJECTIVE

The study aimed to predict sepsis at the time of ED triage using natural language processing of nursing triage notes and available clinical data.

METHODS

We constructed a retrospective cohort of all 1,234,434 consecutive ED encounters in 2015-2021 from 4 separate clinically heterogeneous academically affiliated EDs. After exclusion criteria were applied, the final cohort included 1,059,386 adult ED encounters. The primary outcome criteria for sepsis were presumed severe infection and acute organ dysfunction. After vectorization and dimensional reduction of triage notes and clinical data available at triage, a decision tree-based ensemble (time-of-triage) model was trained to predict sepsis using the training subset (n=950,921). A separate (comprehensive) model was trained using these data and laboratory data, as it became available at 1-hour intervals, after triage. Model performances were evaluated using the test (n=108,465) subset.

RESULTS

Sepsis occurred in 35,318 encounters (incidence 3.45%). For sepsis prediction at the time of patient triage, using the primary definition, the area under the receiver operating characteristic curve (AUC) and macro F1-score for sepsis were 0.94 and 0.61, respectively. Sensitivity, specificity, and false positive rate were 0.87, 0.85, and 0.15, respectively. The time-of-triage model accurately predicted sepsis in 76% (1635/2150) of sepsis cases where sepsis screening was not initiated at triage and 97.5% (1630/1671) of cases where sepsis screening was initiated at triage. Positive and negative predictive values were 0.18 and 0.99, respectively. For sepsis prediction using laboratory data available each hour after ED arrival, the AUC peaked to 0.97 at 12 hours. Similar results were obtained when stratifying by hospital and when Centers for Disease Control and Prevention hospital toolkit for adult sepsis surveillance criteria were used to define sepsis. Among septic cases, sepsis was predicted in 36.1% (1375/3814), 49.9% (1902/3814), and 68.3% (2604/3814) of encounters, respectively, at 3, 2, and 1 hours prior to the first intravenous antibiotic order or where antibiotics where not ordered within the first 12 hours.

CONCLUSIONS

Sepsis can accurately be predicted at ED presentation using nursing triage notes and clinical information available at the time of triage. This indicates that machine learning can facilitate timely and reliable alerting for intervention. Free-text data can improve the performance of predictive modeling at the time of triage and throughout the ED course.

Sepsis mortality and ICU length of stay after the implementation of an intensive care team in the emergency department

Emergency department patient boarding is associated with hospital mortality and increased hospital length of stay. The objective of the present study is to describe the impact of deploying an Intensive Care team in the ED and its association with sepsis mortality and ICU length of stay. Patients admitted to ICU through the ED with an ICD-10 CM diagnosis of sepsis were included. Preintervention and postintervention phases included 4 and 15 months, respectively. Sepsis time zero, SEP-1 compliance, and lag time from time zero to antibiotic administration were compared. Outcomes of interest were mortality and ICU LOS. 1021 septic patients were included. Sixty-six percent fulfilled compliance with 3 h SEP-1 bundle. Lag time from time zero to antibiotic administration was 75 min. Multivariate analysis showed no association between ICU team in the ED and hospital mortality (Log OR 0.94, CI 0.67-1.34; p = 0.73). The ICU team in the ED was associated with prolonged ICU LOS (Log OR 1.21, CI 1.13-1.30; p < 0.01). Septic shock and ED boarding time were associated with prolonged ICU LOS. Compliance with SEP-1 bundle was associated with its reduction. Implementation of an ICU team in the ED for the treatment of septic patients during high volume hospitalizations is not associated with a reduction of mortality or ICU LOS.

The association between mortality and door-to-antibiotic time: a systematic review and meta-analysis

PURPOSE

Previous studies evaluating the impact of antibiotic timing on mortality in sepsis have shown conflicting results. We performed a meta-analysis to evaluate the association between door-to-antibiotic time (each hour of delay) and mortality in sepsis.

METHODS

We searched PubMed and Embase through 10 November 2022 to identity cohort studies that evaluated the adjusted association between door-to-antibiotic time (each hour of delay) and mortality in adult patients with sepsis. The primary outcome was mortality. Analysis was based on inverse-variance weighting using a fixed-effects model. The variances were derived from the logarithms of the reported confidence intervals (CIs) for associations. We estimated the odds ratio, 95% CI, and number needed to treat for the pooled data.

RESULTS

Fifteen cohort studies involving 106 845 patients were included in the meta-analysis. Door-to-antibiotic time (each hour of delay) was associated with increased risk of mortality (odds ratio: 1.07; 95% CI: 1.06-1.08; P < 0.0001; number needed to treat = 91), with high heterogeneity (I2 = 82.2%). The association was robust in sensitivity analyses and consistent in subgroup analyses. No publication bias was found.

CONCLUSION

In adult patients with sepsis, each hour of delay in antibiotic administration is associated with increased odds of mortality. Key messages What is already known on this topic Sepsis is a common and lethal syndrome that affects millions of people worldwide. The updated 2018 Surviving Sepsis Campaign guidelines recommended initiating empirical broad-spectrum antibiotic coverage within 1 hour of identification of sepsis and septic shock. Delay in antibiotic administration may increase the risk of mortality in patients with sepsis. What this study adds This meta-analysis evaluates and quantifies the association between door-to-antibiotic time (each hour of delay) and mortality in patients with sepsis. Each hour of delay in antibiotic administration is associated with increased odds of mortality in sepsis. The number needed to treat (NNT) with delayed antibiotic administration for one additional death was 91. How this study might affect research, practice, or policy: More efforts should be made to speed up the diagnosis of sepsis or sepsis shock.

Pharmacist Involvement in Sepsis Response and Time to Antibiotics: A Systematic Review

Introduction

Sepsis is a life-threatening medical emergency and a leading cause of morbidity and mortality worldwide. Reductions in time to antibiotics in patients presenting with sepsis or septic shock are associated with reduced mortality, and Surviving Sepsis Campaign guidelines recommend antibiotics within one hour of recognition. Pharmacists are well-equipped to help navigate the therapeutic and operational challenges associated with achieving this goal.

Objectives

To assess the association of pharmacist involvement in sepsis response with time to antibiotics in hospitalized patients with sepsis and septic shock.

Methods

A systematic review of the following databases was conducted: PubMed/MEDLINE, Embase, Cumulative Index to Nursing and Allied Health Literature (CINAHL), and Web of Science. Studies must have included a designated role of an individual pharmacist in the management of sepsis or septic shock and not be considered an operational change. The primary outcome of interest was time to antibiotic administration, with secondary outcomes including intensive care unit (ICU) and hospital length of stay as well as in-hospital mortality.

Results

We identified 10 studies including 1772 patients with sepsis or septic shock that evaluated a sepsis response in which a pharmacist was included. Studies included patients in the ICU, emergency department, and hospital ward setting. Seven studies demonstrated a significant reduction in time to antibiotics, with two other studies supporting this conclusion in extrapolation or sensitivity analysis. There was not a consistent reduction in ICU or hospital length of stay nor in-hospital mortality between those interventions involving a pharmacist compared with their defined control groups.

Conclusion

Pharmacist involvement in sepsis response, often as part of a multi-professional team-based approach to sepsis care, is associated with a reduced time to antibiotic administration for hospitalized patients with sepsis or septic shock.

Evaluation of Intravenous Push Piperacillin-Tazobactam on Time to Antibiotic Administration in Emergency Department Patients with Sepsis

Purpose: The purpose of this study was to determine if intravenous push (IVP) administration of piperacillin-tazobactam reduced the time to antibiotic administration compared to intravenous piggyback (IVPB) in emergency department (ED) patients who present with sepsis. Methods: This was a retrospective cohort study of patients with sepsis who received piperacillin-tazobactam before and after implementation of an IVPB to IVP conversion protocol. Results: A total of 486 charts were reviewed and the final analysis included 127 patients in each group. The mean time to administration of piperacillin-tazobactam was 67 (± 48) minutes and 58 (± 36) minutes in the IVPB and IVP cohorts, respectively (P = NS). The time to administration of secondary antibiotics was reduced by 38 minutes in patients who received piperacillin-tazobactam by IVP (105 min ±69 vs 67 min ±37; P < .001). Nurse administration time was reduced by 11 min for piperacillin-tazobactam (54 min ±46 vs 43 min ±33; P = .034) and 40 min for secondary antibiotics (90 min ±67 vs 50 min ±32; P = < .001) in the IVP group. There was no difference in hypersensitivity reactions, hospital length of stay, or mortality. Conclusion: Conversion from piperacillin-tazobactam IVPB to IVP was associated with a reduction in time to piperacillin-tazobactam and secondary antibiotic administration in emergency department patients with sepsis. Further prospective research is needed to evaluate clinical outcomes associated with IVP administration.

Delays in Time-To-Antibiotics for Young Febrile Infants With Serious Bacterial Infections: A Prospective Single-Center Study

INTRODUCTION

Fear of missed serious bacterial infections (SBIs) results in many febrile young infants receiving antibiotics. We aimed to compare the time to antibiotics between infants with SBIs and those without.

MATERIALS AND METHODS

We recruited febrile infants ≤ 90 days old seen in the emergency department (ED) between December 2017 and April 2021. SBI was defined as (1) urinary tract infection, (2) bacteremia or (3) bacterial meningitis. We compared the total time (median with interquartile range, IQR) from ED arrival to infusion of antibiotics, divided into (i) time from triage to decision for antibiotics and (ii) time from decision for antibiotics to administration of antibiotics.

RESULTS

We analyzed 81 and 266 infants with and without SBIs. Median age of those with and without SBIs were 44 (IQR 19-72) and 29 (IQR 7-56) days, respectively (p = 0.002). All infants with SBIs and 168/266 (63.2%) infants without SBIs received antibiotics. Among 249 infants who received antibiotics, the median total time from ED arrival to infusion of antibiotics was 277.0 (IQR 236.0-385.0) mins for infants with SBIs and 304.5 (IQR 238.5-404.0) mins for those without (p = 0.561). The median time to decision for antibiotics was 156.0 (IQR 115.0-255.0) mins and 144.0 (IQR 105.5-211.0) mins, respectively (p = 0.175). Following decision for antibiotics, infants with SBIs received antibiotics much faster compared to those without [107.0 (IQR 83.0-168.0) vs. 141.0 (94.0-209.5) mins, p = 0.017].

CONCLUSION

There was no difference in total time taken to antibiotics between infants with SBIs and without SBIs. Both recognition and administration delays were observed. While all infants with SBIs were adequately treated, more than half of the infants without SBIs received unnecessary antibiotics. This highlights the challenge in managing young febrile infants at initial presentation, and demonstrates the need to examine various aspects of care to improve the overall timeliness to antibiotics.

Identifying High-Risk Subphenotypes and Associated Harms From Delayed Antibiotic Orders and Delivery

OBJECTIVES

Early antibiotic administration is a central component of sepsis guidelines, and delays may increase mortality. However, prior studies have examined the delay to first antibiotic administration as a single time period even though it contains two distinct processes: antibiotic ordering and antibiotic delivery, which can each be targeted for improvement through different interventions. The objective of this study was to characterize and compare patients who experienced order or delivery delays, investigate the association of each delay type with mortality, and identify novel patient subphenotypes with elevated risk of harm from delays.

DESIGN

Retrospective analysis of multicenter inpatient data.

SETTING

Two tertiary care medical centers (2008-2018, 2006-2017) and four community-based hospitals (2008-2017).

PATIENTS

All patients admitted through the emergency department who met clinical criteria for infection.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

Patient demographics, vitals, laboratory values, medication order and administration times, and in-hospital survival data were obtained from the electronic health record. Order and delivery delays were calculated for each admission. Adjusted logistic regression models were used to examine the relationship between each delay and in-hospital mortality. Causal forests, a machine learning method, was used to identify a high-risk subgroup. A total of 60,817 admissions were included, and delays occurred in 58% of patients. Each additional hour of order delay (odds ratio, 1.04; 95% CI, 1.03-1.05) and delivery delay (odds ratio, 1.05; 95% CI, 1.02-1.08) was associated with increased mortality. A patient subgroup identified by causal forests with higher comorbidity burden, greater organ dysfunction, and abnormal initial lactate measurements had a higher risk of death associated with delays (odds ratio, 1.07; 95% CI, 1.06-1.09 vs odds ratio, 1.02; 95% CI, 1.01-1.03).

CONCLUSIONS

Delays in antibiotic ordering and drug delivery are both associated with a similar increase in mortality. A distinct subgroup of high-risk patients exist who could be targeted for more timely therapy.

Management of sepsis and septic shock in the emergency department

Early management of sepsis and septic shock is crucial for patients' prognosis. As the Emergency Department (ED) is the place where the first medical contact for septic patients is likely to occur, emergency physicians play an essential role in the early phases of patient management, which consists of accurate initial diagnosis, resuscitation, and early antibiotic treatment. Since the issuing of the Surviving Sepsis Campaign guidelines in 2016, several studies have been published on different aspects of sepsis management, adding a substantial amount of new information on the pathophysiology and treatment of sepsis and septic shock. In light of this emerging evidence, the present narrative review provides a comprehensive account of the recent advances in septic patient management in the ED.

The Association Between Antibiotic Delay Intervals and Hospital Mortality Among Patients Treated in the Emergency Department for Suspected Sepsis

OBJECTIVES

Rapid delivery of antibiotics is a cornerstone of sepsis therapy, although time targets for specific components of antibiotic delivery are unknown. We quantified time intervals comprising the task of antibiotic delivery and evaluated the association between interval delays and hospital mortality among patients treated in the emergency department for suspected sepsis.

DESIGN

Retrospective cohort.

SETTING

Twelve hospitals in Southeastern United States from 2014 to 2017.

PATIENTS

Twenty-four thousand ninety-three encounters among 20,026 adults with suspected sepsis in 12 emergency departments.

MEASUREMENTS AND MAIN RESULTS

We divided antibiotic administration into two intervals: time from emergency department triage to antibiotic order (recognition delay) and time from antibiotic order to infusion (administration delay). We used generalized linear mixed models to evaluate associations between these intervals and hospital mortality. Median time from emergency department triage to antibiotic administration was 3.4 hours (interquartile range, 2.0-6.0 hr), separated into a median recognition delay (time from emergency department triage to antibiotic order) of 2.7 hours(interquartile range, 1.5-4.7 hr) and median administration delay (time from antibiotic order to infusion) of 0.6 hours (0.3-1.2 hr). Adjusting for other risk factors, both recognition delay and administration delay were associated with mortality, but pairwise comparison with a no-delay reference group was not significant for up to 6 hours of recognition delay or up to 1.5 hours of administration delay.

CONCLUSIONS

Both recognition delays and administration delays were associated with increased hospital mortality, but only for longer delays. These results suggest that both metrics may be important to measure and improve for patients with suspected sepsis but do not support targets less than 1 hour.

Infectious Diseases Society of America Position Paper: Recommended Revisions to the National Severe Sepsis and Septic Shock Early Management Bundle (SEP-1) Sepsis Quality Measure

The Centers for Medicare & Medicaid Services' Severe Sepsis and Septic Shock Early Management Bundle (SEP-1) measure has appropriately established sepsis as a national priority. However, the Infectious Diseases Society of America (IDSA and five additional endorsing societies) is concerned about SEP-1's potential to drive antibiotic overuse because it does not account for the high rate of sepsis overdiagnosis and encourages aggressive antibiotics for all patients with possible sepsis, regardless of the certainty of diagnosis or severity of illness. IDSA is also concerned that SEP-1's complex "time zero" definition is not evidence-based and is prone to inter-observer variation. In this position paper, IDSA outlines several recommendations aimed at reducing the risk of unintended consequences of SEP-1 while maintaining focus on its evidence-based elements. IDSA's core recommendation is to limit SEP-1 to septic shock, for which the evidence supporting the benefit of immediate antibiotics is greatest. Prompt empiric antibiotics are often appropriate for suspected sepsis without shock, but IDSA believes there is too much heterogeneity and difficulty defining this population, uncertainty about the presence of infection, and insufficient data on the necessity of immediate antibiotics to support a mandatory treatment standard for all patients in this category. IDSA believes guidance on managing possible sepsis without shock is more appropriate for guidelines that can delineate the strengths and limitations of supporting evidence and allow clinicians discretion in applying specific recommendations to individual patients. Removing sepsis without shock from SEP-1 will mitigate the risk of unnecessary antibiotic prescribing for noninfectious syndromes, simplify data abstraction, increase measure reliability, and focus attention on the population most likely to benefit from immediate empiric broad-spectrum antibiotics.

DeepAISE - An interpretable and recurrent neural survival model for early prediction of sepsis

Sepsis, a dysregulated immune system response to infection, is among the leading causes of morbidity, mortality, and cost overruns in the Intensive Care Unit (ICU). Early prediction of sepsis can improve situational awareness among clinicians and facilitate timely, protective interventions. While the application of predictive analytics in ICU patients has shown early promising results, much of the work has been encumbered by high false-alarm rates and lack of trust by the end-users due to the 'black box' nature of these models. Here, we present DeepAISE (Deep Artificial Intelligence Sepsis Expert), a recurrent neural survival model for the early prediction of sepsis. DeepAISE automatically learns predictive features related to higher-order interactions and temporal patterns among clinical risk factors that maximize the data likelihood of observed time to septic events. A comparative study of four baseline models on data from hospitalized patients at three different healthcare systems indicates that DeepAISE produces the most accurate predictions (AUCs between 0.87 and 0.90) at the lowest false alarm rates (FARs between 0.20 and 0.25) while simultaneously producing interpretable representations of the clinical time series and risk factors.

Considerations for Empiric Antimicrobial Therapy in Sepsis and Septic Shock in an Era of Antimicrobial Resistance

Patients with sepsis present across a spectrum of infection sites and severity of illnesses requiring complex decision making at the bedside as to when prompt antibiotics are indicated and which regimen is warranted. Many hemodynamically stable patients with sepsis and low acuity of illness may benefit from further work up before initiating therapy, whereas patients with septic shock warrant emergent broad-spectrum antibiotics. The precise empiric regimen is determined by assessing patient and epidemiological risk factors, likely source of infection based on presenting signs and symptoms, and severity of illness. Hospitals should implement quality improvement measures to aid in the rapid and accurate diagnosis of septic patients and to ensure antibiotics are given to patients in an expedited fashion after antibiotic order.

Clinical significance of miR-19b-3p in patients with sepsis and its regulatory role in the LPS-induced inflammatory response

Background

MicroRNAs (miRNAs) play important roles in the development and progression of sepsis. This study investigated the clinical value of miR-19b-3p in sepsis patients, and explored its role in regulating inflammatory responses in HUVECs cells.

Methods

103 patients with sepsis and 98 healthy individuals were recruited. qRT-PCR was used for the measurement of miR-19b-3p level. Cell viability was evaluated using CCK-8. The protein levels of TNF-α and IL-6 were measured using ELISA. Receiver operating characteristic (ROC) curve and logistic regression analysis were constructed to evaluate the diagnostic and prognostic values of miR-19b-3p in sepsis patients.

Results

MiR-19b-3p level was significantly reduced in the serum from patients with sepsis compared with healthy controls (P < 0.001). Sepsis patients in the survival group had significantly high miR-19b-3p levels compared with the non-survival group (P < 0.001). MiR-19b-3p was of a good value in predicting sepsis risk, and was an independent prognostic factor for 28-day survival in sepsis patients (OR = 3.226, 95% CI 1.076–9.670, P = 0.037). MiR-19b-3p level was negatively associated with serum levels of IL-6 (r = − 0.852, P < 0.001) and TNF-α (r = − 0.761, P < 0.001). Overexpression of miR-19b-3p alleviated LPS-induced inflammatory response of HUVECs, which was reflected by the decrease of the levels of IL-6 and TNF-α induced by LPS treatment (P < 0.001).

Conclusion

MiR-19b-3p might be a potential biomarker for the early diagnosis and prognosis of sepsis patients. Overexpression of miR-19b-3p alleviated sepsis-induced inflammatory responses.