INFECTIOUS DISEASE/BRIEF RESEARCH REPORT

Antibiotic Delays and Feasibility of a 1-Hour-

From-Triage Antibiotic Requirement: Analysis of
an Emergency Department Sepsis Quality
Improvement Database
Michael R. Filbin, MD, MSc*; Jill E. Thorsen, BS; Tracey M. Zachary, RN, MSN-Ed; James C. Lynch, BS;
Minoru Matsushima, PhD; Justin B. Belsky, MD, MPH; Thomas Heldt, PhD; Andrew T. Reisner, MD
*Corresponding Author. E-mail: mfilbin@mgh.harvard.edu.

Study objective: We identify factors associated with delayed emergency department (ED) antibiotics and determine feasibility of a
1-hour-from-triage antibiotic requirement in sepsis.

Methods: We studied all ED adult septic patients in accordance with Centers for Medicare & Medicaid Services Severe Sepsis and
Septic Shock National Quality Measures in 2 consecutive 12-month intervals. During the second interval, a quality improvement
intervention was conducted: a sepsis screening protocol plus case-specific feedback to clinicians. Data were abstracted
retrospectively through electronic query and chart review. Primary outcomes were antibiotic delay greater than 3 hours from
documented onset of hypoperfusion (per Centers for Medicare & Medicaid Services Severe Sepsis and Septic Shock National
Quality Measures) and antibiotic delay greater than 1 hour from triage (per 2018 Surviving Sepsis Campaign recommendations).

Results: We identified 297 and 357 septic patients before and during the quality improvement intervention, respectively. Before
and during quality improvement intervention, antibiotic delay in accordance with Centers for Medicare & Medicaid Services
measures occurred in 30% and 21% of cases (–9% [95% confidence interval –16% to –2%]); and in accordance with 2018
Surviving Sepsis Campaign recommendations, 85% and 71% (–14% [95% confidence interval –20% to –8%]). Four factors were
independently associated with both definitions of antibiotic delay: vague (ie, nonexplicitly infectious) presenting symptoms, triage
location to nonacute areas, care before the quality improvement intervention, and lower Sequential [Sepsis-related] Organ Failure
Assessment scores. Most patients did not receive antibiotics within 1 hour of triage, with the exception of a small subset post–
quality improvement intervention who presented with explicit infectious symptoms and triage hypotension.

Conclusion: The quality improvement intervention significantly reduced antibiotic delays, yet most septic patients did not receive
antibiotics within 1 hour of triage. Compliance with the 2018 Surviving Sepsis Campaign would require a wholesale alteration in
the management of ED patients with either vague symptoms or absence of triage hypotension. [Ann Emerg Med. 2019;-:1-7.]

Please see page XX for the Editor’s Capsule Summary of this article.

0196-0644/$-see front matter

Copyright © 2019 by the American College of Emergency Physicians.
https://doi.org/10.1016/j.annemergmed.2019.07.017

INTRODUCTION publicly report the proportion of patients who receive a

Background sepsis treatment bundle that includes administration of
Early antibiotic administration for patients with sepsis broad-spectrum antibiotics within 3 hours from the time of
has been a focus of recent consensus recommendations and onset of new organ dysfunction in patients with systemic
national quality measures, driven by investigations inflammatory response syndrome and documented
demonstrating an association between increased mortality infection.5 Moreover, the 2018 update of the Surviving
and each additional hour of delay in antibiotic Sepsis Campaign recommendations advised even earlier
administration,1-3 although a meta-analysis4 showed no administration of antibiotics as a new standard of care for
benefit of antibiotic administration within 3 hours of triage sepsis patients; namely, that antibiotics should be
or 1 hour of shock onset. Despite uncertainty about the administered within 1 hour from triage.6 This
clinical value of timely antibiotics, the Centers for Medicare recommendation was controversial, and the Infectious
& Medicaid Services (CMS) currently requires hospitals to Diseases Society of America did not endorse the 2018

Volume -, no. - : - 2019 Annals of Emergency Medicine 1

Analysis of an Emergency Department Sepsis Quality Improvement Database Filbin et al

Editor’s Capsule Summary the admission note; and development of persistent

hypotension (systolic blood pressure <90 mm Hg on at
What is already known on this topic least 2 measurements), or elevated lactate level greater than
Despite a lack of evidence, recent recommendations or equal to 4.0 mmol/L, or use of vasopressor medication in
suggest a 1-hour-from-triage requirement for the ED. We excluded patients receiving comfort measures
antibiotic administration in patients with sepsis. only, who were unlikely to receive aggressive management,
What question this study addressed and also those transferred from an outside facility and
already treated for infection before arrival.
The study addressed the feasibility of complying with
a 1-hour-from-triage requirement both before and
after a quality improvement intervention and factors Interventions
associated with a delay in antibiotic administration. Precisely halfway through the analysis interval, our ED
initiated a sepsis care quality improvement intervention. The
What this study adds to our knowledge
intervention involved a sepsis screening protocol plus timely,
In this single emergency department (ED) case-specific performance feedback e-mailed to physicians,
retrospective time series study of 654 septic patients, midlevel providers, and nurses involved in treating the septic
only a fraction of patients were treated with patient. The protocol was communicated to all clinical staff
antibiotics within 1 hour from triage, and quality through e-mail, flyers, and departmental presentations, with
improvement interventions did not significantly de minimus communication activities before the formal start
change this fraction. of the quality improvement intervention. The Shock
How this is relevant to clinical practice Precautions on Triage (SPoT) Sepsis rule was the basis for
In this ED, a 1-hour-from-triage requirement for sepsis screening. It is positive when a patient has mild vital
antibiotic administration in patients with sepsis is not sign abnormalities (pulse rate >systolic blood pressure or
obtainable in the majority of patients, even with systolic blood pressure <100 mm Hg) and a clinical concern
quality improvement interventions aimed at for infection (explicit symptoms of infection, or vague
achieving the goal. symptoms in a patient with major comorbidities, or any
patient in extremis). In a previous report (data from before
the current analysis interval), we found that the SPoT Sepsis
rule was significantly more sensitive than the quick
Surviving Sepsis Campaign in part because of their concern Sequential [Sepsis-related] Organ Failure Assessment
that the 1-hour guideline would encourage inappropriate (qSOFA) score for adult patients with sepsis in our ED, and
administration of broad-spectrum antibiotics.7 it had similar specificity.8
In the current analysis, we examined emergency
department (ED) patient data collected before and during a Data Collection and Processing
sepsis quality improvement initiative. Our goal was to Data were sourced from the hospital electronic medical
identify factors associated with antibiotic delay, validate the record. Laboratory results, vital signs, patient locations,
quality improvement initiative, and identify the patient demographics, and hospital outcome were downloaded
population whose antibiotic administration time would electronically. For each parameter in our research database,
need to be further shortened to comply with the 2018 at least 20 cases were randomly reviewed and compared
Surviving Sepsis Campaign recommendations. with the source (ie, the hospital’s electronic medical record)
to confirm perfect agreement, including relevant time
MATERIALS AND METHODS stamps, thus ensuring that we had accurately extracted and
Study Design, Setting, and Selection of Participants processed electronic data.
This was a retrospective cohort analysis approved by the Blinded to encounter date and outcome, 2 trained chart
institutional review board. We studied all adult ED patients reviewers independently reviewed clinical notes (triage
(18 years) from April 1, 2014, to March 31, 2016, who note, as well as nursing and providers’ initial assessments)
met criteria adapted from the current CMS Severe Sepsis and completed a standardized data entry form9 for
and Septic Shock (SEP-1) definition5 for septic shock: a presenting symptoms, documentation of difficult
final hospital International Classification of Diseases, 10th intravenous access or need for ultrasonographically guided
Revision diagnosis code for sepsis; either confirmed source access, medications, and other ED treatments. The
of infection or high suspicion for infection documented in reviewers also verified whether active infection was

2 Annals of Emergency Medicine Volume -, no. - : - 2019

Filbin et al Analysis of an Emergency Department Sepsis Quality Improvement Database

documented in the hospital admission note. For every adjudicated parameters (eg, difficult intravenous access and
subject, the codings of the 2 reviewers were compared, and individual symptoms that determined vague versus explicit
any disagreements were resolved by majority vote in a classification), Cohen’s k ranged from 0.71 to 0.90.
review session that included a third (physician) reviewer. Patient characteristics were generally similar between
Presenting symptoms were then labeled as explicit versus cohorts, including demographics, baseline comorbidities,
vague according to an objective categorization schema presenting symptoms, infection source, and illness severity
described by Filbin et al.10 Explicit symptoms included parameters (Table 1). More than half of the patients
fever, chills, or rigors; cough with productive sputum; screened positive for sepsis at triage by the SPoT Sepsis
dysuria; reported skin redness or concern for soft tissue rule, although a minority did not. Less than half of the
infection; referral for specific infection diagnosis; or patients had a positive triage qSOFA score (2) or triage
measured temperature greater than or equal to 100.4 F hypotension.
(38.0 C) at triage. Symptoms were defined as vague if they Rates of antibiotic delay were significantly reduced pre–
did not include any of the explicit symptoms. Cohen’s k versus post–quality improvement (Table 2): 30% versus
was computed for reviewer-coded parameters. 21% of cases (–9% [95% confidence interval {CI} –16% to
The primary outcome was proportion of patients with –2%]) according to CMS measures, and 85% versus 71%
delayed antibiotics. The CMS definition of delay was (–14% [95% CI –20% to –8%]) according to the 2018
greater than 3 hours until documented appropriate Surviving Sepsis Campaign recommendation. There was no
antibiotic administration from the first onset of difference in clinical outcomes between cohorts.
hypoperfusion (ie, systolic blood pressure <90 mm Hg or Vague presenting symptoms and triage location to
lactate level 2.0 mmol/L, occurring at triage or anytime nonacute areas were independently associated with
in the ED thereafter, which are the CMS-specified organ antibiotic delay for both CMS and 2018 Surviving Sepsis
dysfunction elements linked to perfusion5). The 2018 Campaign definitions of delay, as was receiving care before
Surviving Sepsis Campaign definition of delay was the quality improvement intervention and lower SOFA
documented appropriate antibiotics greater than 1 hour score (Table 3). For the 2018 Surviving Sepsis Campaign
after triage. Appropriate broad-spectrum antibiotics for definition, triage to nonacute areas had the largest odds
sepsis were explicitly determined in accordance with CMS ratio for delay. Of patients with triage hypotension, 95%
measures.5 Secondary outcomes were times to antibiotics were triaged to the acute area, indicating likely collinearity
and the rates of ICU admission and hospital mortality. between these parameters. Table 3 reports effect size and
95% CI for all covariates included in both models.
Primary Data Analysis
Descriptive statistics were calculated for patient cohorts
LIMITATIONS
before and during the quality improvement intervention.
First, our findings arose from a single center. Second, in
For comparing outcomes, we used the c2 test for
accordance with SEP-1, sepsis diagnosis was based in part
proportions and Wilcoxon rank sum test for continuous
on billing codes, which often depend on the subjective
variables. We also performed multivariable analysis to
diagnostic judgment of clinicians and billing staff, and may
identify independent factors associated with antibiotic
even be biased by financial incentives. Third, the pre- and
delay by both CMS and 2018 Surviving Sepsis Campaign
postintervention cohorts were enrolled at different times, so
definitions. We included 15 prespecified candidate
the observed reduction in antibiotic delays during the
parameters in 2 multivariable logistic regression models
quality improvement intervention may be due to other
(one for each specified outcome). Variables were selected to
factors or secular trends that we did not account for.
include those with univariate significance with either
Fourth, “appropriate” antibiotic was based on CMS
outcome, those with plausible theoretic association with
measures, whereas clinical effectiveness was not evaluated.
antibiotic delays, and variables standardly included in sepsis
Fifth, findings could be biased by systematic
models.
documentation errors or omissions (eg, inaccurate times for
antibiotic administration), and metrics of process care
RESULTS based solely on ED documentation are relatively crude.
A total of 654 patients were enrolled, 297 Also, we may have missed some information in the
preintervention and 357 during the quality improvement documentation, although having every subject chart
intervention. For independent adjudication of presence of independently reviewed by 2 reviewers provided
ED infection or not, Cohen’s k was 0.72. For other mitigation. Sixth, we used a basic multivariable analysis

Volume -, no. - : - 2019 Annals of Emergency Medicine 3

Analysis of an Emergency Department Sepsis Quality Improvement Database Filbin et al

Table 1. Patient characteristics.

Demographics Preintervention, n[297 Intervention, n[357
Age, median (IQR), y 65 (53–77) 66 (54–77)
Men, No. (%) 174 (58.6) 213 (59.7)
Race, No. (%)
Non-Hispanic white 236 (79.5) 280 (78.4)
Non-Hispanic black 20 (6.7) 16 (4.5)
Hispanic 10 (3.4) 9 (2.5)
Asian 12 (4.0) 20 (5.6)
Unknown 19 (6.4) 32 (9.0)
Charlson score, median (IQR) 3.0 (1.0–4.0) 2.0 (1.0–4.0)
Triage characteristics
Explicit symptoms of infection on presentation, No. (%) 189 (63.6) 220 (61.6)
Hypotension at triage (SBP <90 mm Hg), No. (%) 56 (18.9) 93 (26.1)
qSOFA score 2 at triage, No. (%) 84 (28) 94 (26)
Positive “SPoT Sepsis” screen result* at triage, No. (%) 183 (62) 207 (58)
Active ED patient census at triage, median (IQR) 73 (57–88) 85 (71–103)
Characteristics of infection
Documented infectious source,† No. (%)
Respiratory 73 (24.6) 96 (26.9)
Abdominal 70 (23.6) 76 (21.3)
Urinary 51 (17.2) 74 (20.7)
Wound/soft tissue/skin 19 (6.4) 21 (5.9)
Unclear 83 (27.9) 86 (24.1)
Other 15 (5.1) 16 (4.5)
Time to initial SBP <90 mm Hg, median (IQR), h 1.3 (0.2-3.6) 1.1 (0.1–4.3)
Initial serum lactate, median (IQR), mmol/L 4.0 (2.1–5.4) 3.5 (2.1–5.0)
High lactate level (>4 mmol/L), No. (%) 143 (48.1) 157 (44.0)
SOFA score, mean (SD) 7.2 (3.6) 6.8 (3.7)
ED management characteristics, No. (%)
Triage location, “acute care” area 217 (73.1) 275 (77)
Documented intravenous access difficulty 55 (18.5) 87 (24.4)
Sepsis flag set on electronic track board 151 (50.8) 288 (80.7)
Received vasopressors in the ED 145 (48.8) 172 (48.2)
Intubated in the ED 53 (17.8) 59 (16.5)

IQR, Interquartile range; SBP, systolic blood pressure.

*SPoT Sepsis screen is positive when a patient has mild vital sign abnormalities (pulse rate >SBP or SBP <100 mm Hg) together with a clinical concern for infection (explicit
symptoms of infection, or vague symptoms in a patient with major comorbidities, or any patient in extremis).7
†
Patients may have greater than one documented source of suspected infection.

methodology; alternative approaches are possible and could nonacute areas of the ED, pre–quality improvement
yield alternative findings. intervention period, and lower SOFA score. As for the
2018 Surviving Sepsis Campaign definition, we found that
DISCUSSION most septic patients had antibiotic delays greater than 1
We conducted a retrospective analysis to identify factors hour from triage, except for a small subgroup with frank
associated with antibiotic delay in ED patients with sepsis. hypotension at triage and explicit symptoms who were
For the CMS definition of delay, we found 4 independent treated after the start of the ED quality improvement
clinical predictors: vague presenting symptoms, triage to intervention.

4 Annals of Emergency Medicine Volume -, no. - : - 2019

Filbin et al Analysis of an Emergency Department Sepsis Quality Improvement Database

Table 2. Study outcomes.

Primary and Secondary Outcomes Preintervention, n[297 Intervention, n[357 D Absolute (95% CI)
Primary outcomes: pre- vs intervention cohorts (and subgroups)
Antibiotic delay >3 h from hypoperfusion,* % (n) 30 (90/297) 21 (76/357) –9 (–16 to –2)
Subgroup: explicit presenting symptoms and 17 (5/30) 14 (7/51) –2.9 (–19 to 13)
triage SBP <90 mm Hg, % (n)
Subgroup: all others, % (n) 32 (85/267) 23 (69/306) –9 (–17 to –2)
Antibiotic delay >1 h from triage, % (n) 85 (252/297) 71 (253/357) –14 (–20 to –8)
Subgroup: explicit presenting symptoms and 70 (21/30) 39 (20/51) –31 (–52 to –10)
triage SBP <90 mm Hg, % (n)
Subgroup: all others, % (n) 86 (231/267) 76 (233/306) –10 (–17 to –4)
Secondary outcomes: pre- vs intervention cohorts
Time to appropriate antibiotic from hypoperfusion, median (IQR), h 1.6 [0.5 to 3.0] 0.8 [0.2 to 1.9] –0.6 (–0.9 to –0.4)
Time to appropriate antibiotic from triage, median (IQR), h 2.5 [1.3 to 4.2] 1.7 [0.8 to 3.4] –0.6 (–0.9 to –0.3)
Appropriate antibiotics never received in ED, % (n) 8 (24/297) 7 (26/357) –1 (–5 to 3)
Admission to ICU within 48 h, % (n) 72 (213/297) 66 (234/357) –6 (–13 to 1)
Hospital mortality, % (n) 23 (67/297) 23 (82/357) 0 (–6 to 7)

95% CIs according to Wald asymptotic confidence limits for differences in proportions and according to Hodges-Lehmann estimates for differences in median times.
*Hypoperfusion defined as SBP less than 90 mm Hg or lactate level greater than or equal to 2 mmol/L.

Metrics of ED crowding were not predictive of A large fraction of septic patients presented with vague
antibiotic delay, nor was difficult intravenous access, symptoms (ie, nonspecific complaints such as lethargy,
suggesting that timely antibiotic administration depended somnolence, isolated hypotension, or near syncope), and
on diagnostic assessment, rather than strictly operational without fever or source-localizing symptoms. This finding
factors. that vague symptoms were common in patients with sepsis

Table 3. Multivariable logistic regression analysis for factors associated with antibiotic delays.
Hypoperfusion*D180 Minutes TriageD60 Minutes
(CMS Measure), (2018 SSC Recommendation),
Antibiotic Delay Definition: Characteristic Adjusted Odds Ratio (95% CI) Adjusted Odds Ratio (95% CI)
Pre–quality improvement intervention 1.7 (1.1–2.5) 2.7 (1.7–4.3)
Vague presenting symptoms 3.1 (2.1–4.7) 4.0 (2.5–6.4)
Triage to nonacute area 2.0 (1.3–3.1) 15.9 (5.6–44.7)
SOFA score 0.9 (0.9–1.0) 0.9 (0.9–1.0)
SBP 90 mm Hg at triage 0.9 (0.5–1.4) 2.5 (1.6–4.0)
Age 0.9 (0.6–1.3) 1.0 (0.7–1.6)
Male sex 0.9 (0.6–1.3) 0.7 (0.5–1.2)
Race (nonwhite) 0.9 (0.6–1.5) 1.0 (0.6–1.6)
Charlson score 1.0 (0.9–1.0) 1.0 (0.9–1.1)
Chronic liver disease 1.9 (0.9–3.7) 2.3 (0.9–6.0)
Immunosuppression 0.6 (0.4–1.0) 0.7 (0.4–1.2)
ED capacity at triage 1.1 (0.4–2.6) 1.8 (0.7–5.0)
IV access difficulty 1.4 (0.9–2.3) 1.0 (0.6–1.8)
Lactate level 4 mmol/L 1.1 (0.7–1.5) 0.9 (0.6–1.4)
Unclear infection source 1.5 (1.0–2.3) 0.9 (0.6–1.5)

SSC, Surviving Sepsis Campaign; IV, intravenous.

c Statistic for multivariable models is 0.71 and 0.80 for CMS and 2018 SSC, respectively.
*Hypoperfusion defined as SBP less than 90 mm Hg or lactate level greater than or equal to 2 mmol/L.

Volume -, no. - : - 2019 Annals of Emergency Medicine 5

Analysis of an Emergency Department Sepsis Quality Improvement Database Filbin et al

is consistent with that in previous cohorts from our to identify most cases? We found that most septic patients
hospital,8 and it should not be assumed that vague lacked positive qSOFA scores at triage. The SPoT Sepsis
symptoms implied benign disease states; previous analysis screening rule is more sensitive,8 but even this rule failed to
suggests that mortality is notably high in patients with identify a sizable fraction of septic patients at triage.
vague symptoms even when timing of antibiotics is Second, what operational practices would enable patients
controlled for.10 with nonspecific triage presentation to receive antibiotics
Most septic patients lacked vital sign abnormalities at within 1 hour: rapid diagnostic testing by a sepsis team,
triage, having neither hypotension nor even a positive empirical antibiotics administered without diagnostic data,
qSOFA score (Table 1). The SPoT Sepsis screening rule or both? We note the Infectious Diseases Society of
was positive in a thin majority of cases. Absence of America concerns that the 2018 Surviving Sepsis Campaign
hypotension was not independently associated with delayed could encourage inappropriate administration of broad-
antibiotics according to CMS measures, but the majority of spectrum antibiotics.7 This leads to the third question:
patients who did receive antibiotics within 1 hour of triage what will be the effect on other ED patients if substantial
had frank hypotension at triage (Table 2). staffing resources are diverted to patients with stable triage
Triage to nonacute areas of the ED was independently vital signs or vague, nonspecific presenting symptoms?
associated with delayed antibiotics (Table 3). We speculate Given that adhering to the 2018 Surviving Sepsis
that septic patients who were not triaged to acute areas had Campaign would require a fundamental change in how our
more subtle presentations, and after they arrived in ED treats patients who present with vague symptoms or
nonacute areas, there was likely reduced staffing, reduced unremarkable vital signs, whereas the clinical effectiveness
monitoring, and possibly less alacrity in work flow. of requiring antibiotics within 1 hour of triage is wholly
The CMS definition of timely antibiotic administration uncertain, it is our opinion that currently the 2018
permits up to 3 hours after recognition of hypoperfusion, Surviving Sepsis Campaign recommendations about
which does allow for some degree of additional testing and antibiotic timing are unjustifiable.
assessment before antibiotics are due. Before the quality
improvement intervention, median time to antibiotics from Supervising editor: Alan E. Jones, MD. Specific detailed information
first time of hypoperfusion was 1.6 hours; after about possible conflict of interest for individual editors is available
at https://www.annemergmed.com/editors.
implementation of the quality improvement initiative, it
was reduced to 0.8 hours. The SPoT Sepsis rule, around Author affiliations: From the Department of Emergency Medicine,
which the quality improvement intervention was designed, Massachusetts General Hospital, Boston, MA (Filbin, Thorsen,
Zachary, Reisner); the Department of Electrical Engineering and
was intended to focus clinicians’ attention on mild
Computer Science (Lynch, Heldt) and Institute for Medical
hemodynamic abnormalities, major comorbidities, and Engineering and Science (Lynch, Matsushima, Heldt),
vague symptoms. The quality improvement intervention Massachusetts Institute of Technology, Cambridge, MA; Nihon
combined staff education (e-mail, flyers, and presentations) Kohden Innovation Center, Cambridge, MA (Matsushima); and the
and case-specific feedback to attending physicians, midlevel Department of Emergency Medicine, Yale–New Haven Hospital,
providers, and nurses alike. We speculate that the feedback New Haven, CT (Belsky).
intervention reduced CMS-defined antibiotic delays Author contributions: MRF, TMZ, and ATR conceived of the quality
through triage decisions, the timeliness with which improvement intervention. MRF, JET, TMZ, and JBB conducted the
providers evaluated potentially septic patients and ordered quality improvement intervention. MRF, TH, and ATR conceived of
the analytic methodology, conducted data analysis, and prepared
antibiotics, and the timeliness with which nurses the article. JCL, MM, and TH conducted data queries and led data
administered antibiotics. Our data support the feasibility of management and processing. MRF, JET, JBB, and ATR conducted
compliance with CMS measures. clinical adjudications. All authors reviewed the article. MF takes
Our findings (Table 2) suggest that administration of responsibility for the paper as a whole.
antibiotics to the majority of septic patients within 1 hour All authors attest to meeting the four ICMJE.org authorship criteria:
of triage would require a wholesale alteration of our current (1) Substantial contributions to the conception or design of the
practices for all patients arriving with stable vital signs, work; or the acquisition, analysis, or interpretation of data for the
vague nonspecific symptoms, or both. Such patients made work; AND (2) Drafting the work or revising it critically for important
intellectual content; AND (3) Final approval of the version to be
up the majority of our sepsis population.
published; AND (4) Agreement to be accountable for all aspects of
Contemplating how to possibly comply with antibiotics the work in ensuring that questions related to the accuracy or
within 1 hour of triage raises essential questions. First, how integrity of any part of the work are appropriately investigated and
does one screen for sepsis at triage with sensitivity sufficient resolved.

6 Annals of Emergency Medicine Volume -, no. - : - 2019

Filbin et al Analysis of an Emergency Department Sepsis Quality Improvement Database

Funding and support: By Annals policy, all authors are required to shock from the first hour: results from a guideline-based
disclose any and all commercial, financial, and other relationships performance improvement program. Crit Care Med.
in any way related to the subject of this article as per ICMJE conflict 2014;42:1749-1755.
of interest guidelines (see www.icmje.org). Funding for this study 4. Sterling SA, Miller WR, Pryor J, et al. The impact of timing of
antibiotics on outcomes in severe sepsis and septic shock: a
was sponsored by the institutional award MIT-MGH Strategic
systematic review and meta-analysis. Crit Care Med.
Partnership, grants from the CRICO Risk Management Foundation; 2015;43:1907-1915.
Nihon Kohden Corporation; and the National Foundation of 5. Centers for Medicare & Medicaid Services (CMS). Sepsis Bundle
Emergency Medicine. Drs. Filbin, Heldt, and Reisner report Project (SEP), national hospital inpatient quality measures. Available
receiving grant support from Nihon Kohden Corporation and have at: https://www.qualitynet.org. Accessed February 11, 2018.
a patent application related to electronic decision support for 6. Levy MM, Evans LE, Rhodes A. The Surviving Sepsis
sepsis management. Dr. Matsushima is an employee of the Nihon Campaign bundle: 2018 update. Intensive Care Med.
Kohden Corporation. 2018;44:925-928.
7. IDSA Sepsis Task Force. Infectious Diseases Society of America
Publication dates: Received for publication March 8, 2019. (IDSA) position statement: why IDSA did not endorse the
Revision received June 19, 2019. Accepted for publication July 9, Surviving Sepsis Campaign guidelines. Clin Infect Dis.
2019. 2018;66:1631-1635.
8. Filbin MR, Thorsen JE, Lynch J, et al. Challenges and opportunities for
emergency department sepsis screening at triage. Sci Rep.
REFERENCES 2018;8:11059.
1. Seymour CW, Gesten F, Prescott HC, et al. Time to treatment and 9. Harris PA, Taylor R, Thielke R, et al. Research Electronic Data Capture
mortality during mandated emergency care for sepsis. N Engl J Med. (REDCap)—a metadata-driven methodology and workflow process for
2017;376:2235-2244. providing translational research informatics support. J Biomed Inform.
2. Liu VX, Fielding-Singh V, Greene JD, et al. The timing of early antibiotics 2009;42:377-381.
and hospital mortality in sepsis. Am J Respir Crit Care Med. 10. Filbin MR, Lynch J, Gillingham TD, et al. Presenting symptoms
2017;196:856-863. independently predict mortality in septic shock: importance of a
3. Ferrer R, Martin-Loeches I, Phillips G, et al. Empiric antibiotic previously unmeasured confounder. Crit Care Med.
treatment reduces mortality in severe sepsis and septic 2018;46:1592-1599.

Volume -, no. - : - 2019 Annals of Emergency Medicine 7