Survival Benefit and Cost Savings From Compliance With a Simplified 3-Hour Sepsis Bundle in a Series of Prospective, Multisite, Observational Cohorts

Daniel E. Leisman, BS; Martin E. Doerfler, MD; Mary Frances Ward, RN, MS, ANP; Kevin D. Masick, PhD; Benjamin J. Wie, BA; Jeanie L. Gribben, BS; Eric Hamilton, BA; Zachary Klein, MS; Andrea R. Bianculli, BS; Meredith B. Akerman, MS; John K. D'Angelo, MD; Jason A. D'Amore, MD

| Disclosures

Crit Care Med. 2017;45(3):395-406. 

 

Abstract and Introduction

Abstract

Objectives: To determine mortality and costs associated with adherence to an aggressive, 3-hour sepsis bundle versus noncompliance with greater than or equal to one bundle element for severe sepsis and septic shock patients.

Design: Prospective, multisite, observational study following three sequential, independent cohorts, from a single U.S. health system, through their hospitalization.

Setting: Cohort 1: five tertiary and six community hospitals. Cohort 2: single tertiary, academic medical center. Cohort 3: five tertiary and four community hospitals.

Patients: Consecutive sample of all severe sepsis and septic shock patients (defined: infection, ≥ 2 systemic inflammatory response syndrome, and hypoperfusive organ dysfunction) identified by a quality initiative. The exposure was full 3-hour bundle compliance. Bundle elements are as follows: 1) blood cultures before antibiotics; 2) parenteral antibiotics administered less than or equal to 180 minutes from greater than or equal to two systemic inflammatory response syndrome "and" lactate ordered, or less than or equal to 60 minutes from "time-zero," whichever occurs earlier; 3) lactate result available less than or equal to 90 minutes postorder; and 4) 30 mL/kg IV crystalloid bolus initiated less than or equal to 30 minutes from "time-zero." Main outcomes were in-hospital mortality (all cohorts) and total direct costs (cohorts 2 and 3).

Measurements and Main Results: Cohort 1: 5,819 total patients; 1,050 (18.0%) bundle compliant. Mortality: 604 (22.6%) versus 834 (26.5%); CI, 0.9–7.1%; adjusted odds ratio, 0.72; CI, 0.61–0.86; p value is less than 0.001. Cohort 2: 1,697 total patients; 739 (43.5%) bundle compliant. Mortality: 99 (13.4%) versus 171 (17.8%), CI, 1.0–7.9%; adjusted odds ratio, 0.60; CI, 0.44–0.80; p value is equal to 0.001. Mean costs: $14,845 versus $20,056; CI, –$4,798 to –5,624; adjusted β, –$2,851; CI, –$4,880 to –822; p value is equal to 0.006. Cohort 3: 7,239 total patients; 2,115 (29.2%) bundle compliant. Mortality: 383 (18.1%) versus 1,078 (21.0%); CI, 0.9–4.9%; adjusted odds ratio, 0.84; CI, 0.73–0.96; p value is equal to 0.013. Mean costs: $17,885 versus $22,108; CI, –$2,783 to –5,663; adjusted β, –$1,423; CI, –$2,574 to –272; p value is equal to 0.015.

Conclusions: In three independent cohorts, 3-hour bundle compliance was associated with improved survival and cost savings.

Introduction

Sepsis is a leading cause of death and healthcare spending globally.[1,2] U.S. prevalence alone conservatively exceeds 1,000,000 cases per year in patients greater than or equal to 65, accounting for approximately 350,000 deaths.[3] In 2011, sepsis accounted for $20 billion in payer costs.[4]

Despite substantial effort, advances in understanding and managing sepsis have been modest. For years, best-practice discourse focused on "early goal-directed therapy", based on a 6-hour septic shock bundle that espoused optimizing central venous pressure (CVP) and oxygenation with inotropes and blood transfusion, adjusted based on rigorous hemodynamic monitoring.[5] Three recent multisite randomized trials all failed to demonstrate mortality benefits from EGDT versus "usual care".[6–8] However, although standard-of-care has changed over ensuing years prohibiting direct comparisons, both study and control arm patients in all three recent trials, and in the trial by Rivers et al,[5] received early antibiotics and empiric IV fluid resuscitation with average times less than 3 hours. Several studies investigating bundle "compliance," of which early IV fluid and antibiotic administration is a component, report association between full compliance and better outcomes.[9–13] This raises the question as to whether the original benefit of EGDT lied not in invasive, resource-intensive, hemodynamic monitoring, but earlier sepsis identification and intervention. The impact of full compliance with a 3-hour bundle alone in a population where there was no requirement to apply a 6-hour bundle has not been investigated.

Recently, a joint Society for Critical Care Medicine (SCCM) and European Society of Intensive Care Medicine (ESICM) Task Force recommended changes to definitions of septic disease and identification measures, focusing on sepsis-induced organ dysfunction and diagnostic criteria.[14–16] The literature and recommendations reflect ambiguity in both defining sepsis and executing evidence-based management. Despite this, consensus remains on the importance of early intervention with IV fluids and antibiotics.

Reliable timely parenteral antibiotics and fluid administration for severe sepsis and septic shock (SS/SS) is the accepted standard in randomized trials, but often not achieved in community settings. We began a quality improvement (QI) project based upon the hypothesis that reliably implementing an aggressive 3-hour sepsis treatment bundle for every patient, every time, in a large multihospital health system would improve outcomes. The adopted bundle emphasizes early recognition with a focus on "time-zero" entry points followed by aggressive, time-sensitive collection of blood cultures, ordering and return of lactate levels, and administration of fluids and antibiotics within 180 minutes. Importantly, this approach obligates only rapid delivery of simple interventions with a focus on preventing further organ injury, and has no reliance on physiologic endpoints. It leaves all management beyond 3 hours to physician discretion.

We hypothesize that time-dependent adherence to all elements of a 3-hour sepsis bundle, without reliance on specific physiologic goals, may be sufficient to improve in-hospital mortality while reducing spending for providers and payers alike in this high-incidence, high-acuity population. We conducted a series of sequential, prospective, observational cohort studies with the objective of determining mortality and costs associated with 3-hour bundle compliance.

 
1 of 4
 
 
Latest in Critical Care
Table 1.  Organ Dysfunction "Time-Zero" Criteria for 3-Hour Bundle Application and Study Inclusion
Organ Dysfunction Criteriaa Definition
Hyperlactemia Serum lactate ≥ 2.2 mmol/L
Hypotension Systolic blood pressure < 90 mm Hg or a mean arterial pressure < 65 mm Hg
Acute kidney injuryb Serum creatinine > 2.0 mg/dL in the absence of chronic kidney disease or 50% increase from known baseline
Thrombocytopenia Platelet count < 150,000 cells/μm3
Coagulopathyc International normalized ratio > 1.5, activated partial thromboplastin time > 30 s, or partial thromboplastin time > 60 s, not otherwise explained by medical history
Elevated bilirubin Serum bilirubin > 2.0 mg/dL in the absence of preexisting liver failure
Acute altered mental status New altered mentation unrelated to the patient's prior medical history
Compromised oxygenationd New increased O2 requirement to maintain arterial oxygen saturation > 90% or a Pao2/FIO2 ratio < 300
≥ 2 "Super-SIRS" criteria at triage Locally developed consensus criteria, where meeting ≥ 2 criteria at triage was a "time-zero" entry point for 3-hr bundle care. "Super-SIRS" criteria were as follows:
1) Heart rate ≥ 120
2) Respiratory rate ≥ 24
3) Systolic blood pressure < 90 mm Hg or mean arterial pressure < 65 mm Hg
4) Temperature ≥ 38.0°C (101.0°F) or ≤ 36.0°C (96.8°F)
5) Acutely altered mental status

SIRS = systemic inflammatory response syndrome.
aThese seven "time-zero" triggers were used as 3-hr bundle entry points as well as study inclusion criteria. The intention was for all patients with a suspected infection who met ≥ 2 systemic inflammatory response syndrome (SIRS) criteria and "any" one of these criteria to receive care fully adherent to all 3-hr bundle elements. All patients included in this study had a source of infection, met at least two SIRS criteria, and met at least one of the above organ dysfunction criteria.
bAdapted from Kidney Disease: Improving Global Outcomes criteria for defining acute kidney injury (18).
cAdapted from the 2001 International Sepsis Definitions Conference (Sepsis-2) report (19).
dAdapted from the 2001 International Sepsis Definitions Conference (Sepsis-2) report (19).

Table 2.  Baseline and Presentation Characteristics
Baseline Characteristics Cohort 1 (Calendar Year 2012) Cohort 2 (Medicare Fiscal Year 2014) Cohort 3 (Medicare Fiscal Year 2015)
All Subjects Bundle Compliant Noncompliant All Subjects Bundle Compliant Noncompliant All Subjects Bundle Compliant Noncompliant
n 5,819 1,050 4,769 1,697 739 958 7,239 2,115 5,124
Tertiary-care center, n (%) 3,931 (67.6) 638 (60.1) 3,293 (69.1) 1,697 (100) 739 (100) 958 (100) 5,727 (79.1) 1,472 (69.6) 4,255 (74.3)
Age, yr (IQR) 76 (54–98) 77 (55–99) 75 (52–98) 75 (62–85) 75 (63–85) 75 (62–85) 74 (61–84) 75 (62–85) 73 (61–84)
Male sex, n (%) 33,032 (52.1) 527 (50.2) 2,505 (52.5) 909 (53.6) 418 (56.6) 491 (52.3) 3,687 (51.0) 1,124 (53.2) 2,563 (50.0)
Body mass index (SD) 27.0 (8.1) 26.7 (7.7) 27.2 (8.4) 27.4 (7.8) 27.0 (6.7) 27.6 (7.8)
Primary payer class, n (%)            
Medicare 1,226 (72.3) 526 (71.2) 700 (73.1) 4,519 (62.4) 1,339 (63.3) 3,180 (62.1)
Medicaid 141 (8.3) 60 (8.1) 81 (8.5) 466 (6.5) 125 (5.9) 341 (6.7)
Commercial I 182 (10.7) 92 (12.5) 90 (9.4) 1,712 (23.8) 490 (23.3) 1,222 (24.0)
Commercial II (high reimbursing) 128 (7.5) 52 (7.0) 76 (7.9) 581 (8.1) 172 (8.2) 409 (8.0)
Comorbidities at presentation, n (%)a
   Chronic renal failure 179 (10.6) 64 (8.7) 115 (12.0) 722 (10.0) 141 (6.7) 581 (11.3)
   Congestive heart failure 256 (15.1) 98 (13.3) 158 (16.5) 1,045 (14.4) 237 (11.2) 808 (15.8)
   Chronic obstructive pulmonary disease 245 (14.4) 101 (13.7) 144 (15.0) 502 (6.9) 135 (6.4) 367 (7.2)
   Diabetes 553 (32.6) 230 (31.1) 323 (33.7) 2,324 (32.1) 645 (30.5) 1,679 (32.8)
   Metastatic disease 97 (5.7) 51 (6.9) 46 (4.8) 1,944 (26.9) 487 (23.0) 1,457 (28.4)
   Chemotherapy within 30 d 170 (10.0) 82 (11.1) 88 (9.2)
   Immune modifying medications 445 (6.1) 138 (6.5) 307 (6.0)      
   HIV positive 12 (0.7) 7 (1.0) 5 (0.5) 75 (1.0) 20 (0.9) 55 (1.1)
   Leukemia, lymphoma, multiple myeloma 105 (6.2) 50 (6.8) 55 (5.7) 275 (3.8) 73 (3.5) 202 (3.9)
   Liver failure 117 (1.6) 37 (1.7) 80 (1.6)
   Organ transplant 74 (1.0) 16 (0.8) 58 (1.1)
Presentation acuity                  
   Initial lactate, mmol/L (SD) 3.6 (2.9) 3.6 (2.8) 3.6 (3.0) 2.8 (2.1) 2.9 (2.2) 2.7 (2.0) 3.2 (2.5) 3.3 (2.5) 3.2 (2.5)
   Hyperlactemia (≥ 2.2 mmol/L), n (%) 3,400 (64.1) 694 (68.4) 2,706 (63.0) 1,051 (61.9) 468 (63.3) 583 (60.8) 4,741 (65.5) 1,498 (70.8) 3,243 (63.3)
   Hyperlactemia (≥4.0 mmol/L), n (%) 1,560 (26.8) 332 (31.6) 1,228 (25.7) 263 (15.5) 122 (16.5) 141 (14.7) 4,169 (57.6) 1,324 (62.6) 2,845 (55.5)
   Systolic blood pressure < 90 or mean arterial pressure < 65 mm Hg 2,477 (42.6) 506 (48.2) 1,971 (41.3) 241 (14.2) 108 (14.6) 133 (13.9) 2,405 (33.2) 701 (33.1) 1,704 (33.3)
   Acute kidney injuryb 1,641 (28.2) 266 (25.3) 1,375 (28.8) 375 (22.1) 150 (20.3) 225 (23.5) 1,512 (20.9) 440 (20.8) 1,072 (20.9)
   Coagulopathyc 941 (16.2) 149 (14.2) 792 (16.6) 323 (19.0) 153 (20.7) 170 (17.8) 443 (6.1) 139 (6.6) 304 (5.9)
   Thrombocytopeniad 435 (7.5) 68 (6.5) 367 (7.7) 245 (14.4) 125 (16.9) 120 (12.6) 808 (11.2) 225 (10.6) 583 (11.4)
   Total bilirubin > 2.0 mg/dL 361 (6.2) 57 (5.4) 304 (6.4) 100 (5.9) 46 (6.2) 54 (5.6) 419 (5.8) 121 (5.7) 298 (5.8)
   Acutely altered mental status 933 (16.0) 154 (14.7) 779 (16.3) 150 (8.8) 47 (6.4) 103 (10.8) 1,721 (23.8) 449 (21.2) 1,272 (24.8)
   Hypoxemiae 1,030 (17.7) 136 (13.0) 894 (18.7) 46 (2.7) 5 (0.7) 43 (4.5) 1,536 (21.2) 403 (19.1) 1,133 (22.1)
   Super-systemic inflammatory response syndrome criteria at triage, n (%) 2,248 (38.6) 435 (41.4) 1,813 (38.0) 329 (19.4) 195 (26.4) 134 (14.0) 2,162 (30.3) 656 (31.1) 1,506 (29.9)
   Septic shock, n (%)e 1,199 (16.6) 368 (17.4) 831 (16.2)

IQR = interquartile range.
aComorbidities reflect status at time-zero, and would not reflect conditions that developed subsequently during hospital stay.
bAcute kidney injury defined as creatinine > 2.0 or 50% increase from a known baseline.
cCoagulopathy defined as an international normalized ratio > 1.5 or a partial thromboplastin time > 60 s.
dThrombocytopenia is defined as "platelets < 150,000 cells/μm3".
eHypoxemia defined as Pao2/Fio2 < 300 or an increased o2 requirement to maintain arterial oxygen saturation > 90%.
fSeptic shock defined as a lactate ≥ 4.0 mmol or hypotension refractory to IV fluid resuscitation.
Dashes signify that this field was not collected or not applicable for the indicated cohort.

Table 3.  Treatment and Postdischarge Characteristics
Baseline Characteristics Cohort 1 (Calendar Year 2012) Cohort 2 (Medicare Fiscal Year 2014) Cohort 3 (Medicare Fiscal Year 2015)
All Subjects Bundle Compliant Noncompliant All Subjects Bundle Compliant Noncompliant All Subjects Bundle Compliant Noncompliant
n 5,819 1,050 4,769 1,697 739 958 7,239 2,115 5,124
3-hr bundle interventionsa                  
   Fluid bolus initiated in <30 min, n (%)b 1,923 (55.6) 1,050 (100) 873 (36.2) 1,088 (64.1) 739 (100.0) 349 (36.4) 3,723 (51.4) 2,115 (100.0) 1,608 (31.4)
   Time to fluid bolus initiation, median (IQR)b 25 (–33 to 83) 4.5 (–32 to 41) 48 (–11 to 107) 0 (0–32) 0 (0–7) 34 (0–73) 12 (–38 to 61) 0 (–48 to 10) 42 (–20 to 120)
   Lactate order to result time < 90 min, n (%) 4,592 (87.5) 1,050 (100) 3,542 (84.4) 1,648 (97.1) 739 (100.0) 909 (94.9) 5,913 (81.7) 2,115 (100.0) 3,798 (74.1)
   Lactate order to result time, median (IQR) 40 (–12 to 92) 34 (–8 to 76) 44 (–20 to 108) 24 (14–38) 25 (15–38) 22 (14–37) 40 (24–64.5) 38 (24–54) 43 (24–75.5)
   Blood cultures drawn before antibiotics, n (%) 4,477 (93.4) 1,050 (100.0) 3,427 (91.5) 24 (14–38) 25 (15–38) 22 (14–.37) 5,119 (70.7) 2,115 (100.0) 3,004 (58.6)
   IV antibiotics in < 180 min, n (%) 4,294 (73.8) 1,050 (100.0) 3,244 (68.0) 1,456 (85.8) 739 (100.0) 717 (74.8) 5,480 (75.7) 2,115 (100.0) 3,004 (58.6)
   Time to antibiotic administration, median (IQR) 67 (–24 to 158) 38 (–19 to 95) 87 (–10 to 184) 46 (5–104) 32 (1–73) 66 (20–172) 57 (5–132) 29 (–4 to 66) 85 (20–208)
Critical care interventions                  
   Central line inserted, n (%) 1,758 (30.2) 277 (26.4) 1,481 (31.1) 30 (1.8) 10 (1.4) 20 (2.1) 1,208 (16.7) 307 (14.5) 901 (17.6)
   Central venous pressure monitoring, n (%) 920 (15.8) 133 (12.7) 787 (16.5) 11 (2.1) 0 (0) 3 (0.9) 362 (5.0) 67 (3.2) 295 (5.8)
   Scvo2 or Svo2 monitoring, n (%)c 292 (5.0) 48 (4.6) 244 (5.1) 3 (0.6) 1 (0.6) 10 (3.0) 30 (0.4) 8 (0.4) 22 (0.4)
   Composite hemodynamic monitoring, n (%)d 1,104 (19.0) 164 (15.6) 940 (19.7) 13 (0.8) 1 (0.001) 12 (1.3) 392 (5.4) 75 (3.5) 317 (6.2)
Unadjusted primary outcomes                  
   In-hospital mortality, n (%) 1,437 (24.7) 224 (21.3) 1,213 (25.4) 270 (15.9) 99 (13.4) 171 (17.8) 1,461 (20.2) 383 (18.1) (CI, 1.3%) 1,078 (21.0) (CI, 1.1%)
   Total direct cost, mean (95% CI) $17,787 (SD, $18,205) $14,845 (CI, $1,608) $20,056 (CI, $2,021) $20,874 (SD, 20,882) $17,885 (CI, $1,135) $22,108 (CI, $888)
   Total direct cost, median (IQR) $9,654 ($5,439–19,087) $9,263 ($5,245–15,936) $10,250 ($5,560–20,621) $11,066 (5,307–23,274) $10,066 (4,912–20,027) $11,543 (5,468–24,487)
Unadjusted secondary outcomes                  
   ICU admission, n (%) 2,250 (38.7) 455 (43.3) 1,795 (37.6) 495 (29.2) 189 (25.6) 306 (31.9) 3,330 (46.0) 898 (42.5) (CI, 2.1%) 2,432 (47.5) (CI, 1.4%)
   ICU LOS (ICU admitted only), d (95% CI)e 6 (5.7–6.3) 5 (4.5–5.5) 6 (5.6–6.4)
   Vasopressors required, n (%) 1,370 (25.6) 240 (24.3) 1,130 (25.8) 68 (4.0) 27 (3.7) 41 (4.3) 1,528 (21.7) 378 (18.1) (CI, 1.7%) 1,150 (23.2) (CI, 1.2%)
   Mechanical ventilation Required, n (%) 2,097 (29.0) 502 (23.7) (CI, 1.8%) 1,595 (31.1) (CI, 1.3%)
   Hospital LOS, d (95% CI)e 68 (4.0) 27 (3.7) 41 (4.3) 9 (8.8–9.2) 8 (7.7–8.3) 9 (8.8–9.2)
   Adjusted net revenue, mean (95% CI) $25,233 (SD, $19,301) $23,120 (CI, $1,646) $26,864 (CI, $2,117) $31,270 (SD, 32,762) $28,367 (CI, $1,803) $32,469 (CI, $1,175)
   Adjusted net revenue, median (IQR) $16,662 ($11,011–26,585) $16,592 ($11,183–22,876) $16,730 ($10,901–28,625) $16,875 (14,147–35,234) $16,390 (13,804–30,649) $17,209 (14,333–37,622)
   Contribution margin, mean (95% CI) $7,141 (SD, $9,794) $7,991 (CI, $1,382) $6,487 (CI, $1,288) $6,613 (SD, 3,319) $7,760 (CI, $1,371) $6,139 (CI, $920)
   Contribution margin, median (IQR) $5,201 ($436–11,065) $5,878 ($859–11,723) $4,627 (–$387 to 10,143) $5,273 (–1,397 to 11,516) $5,228 (–646 to 11,736) $5,293 (–1,718 to 11,436)

IQR = interquartile range, LOS = length of stay.
aAll times are in minutes, and reflect the time elapsed from time-zero unless otherwise indicated.
bAll crystalloid was 0.9% normal saline solution administered at a volume ≥ 30 mL/kg.
cDefined as measurement of either central venous oxygenation (Scvo2) or mixed venous (i.e., pulmonary arterial) oxygenation (Svo2).
dComposite hemodynamic monitoring defined as measurement of either central venous pressure or Scvo2 or Svo2.
eLength of stay (LOS) values computed using Kaplan-Meier curve assessment with log-rank test, censored for mortality. We report median LOS in days with 95%
CIs of the medians.
Dashes signify that this field was not collected or not applicable for the indicated cohort.

Table 4.  Adjusted Outcomes From Multivariate Regression Analyses
Outcome Regression Type Model Fit Model Output Effect Size 95% CI p
Cohort 1 (2012)a            
   Primary outcome            
      Mortality Logistic X 2 = 2.5; p = 0.96 OR 0.74 0.62–0.89 0.001  
   Secondary outcomes            
      ICU admission Logistic X 2 = 37.2; p < 0.001 OR 1.24 1.07–1.43 0.004
      Vasopressors required Logistic X 2 = 25.0; p = 0.002 OR 0.83 0.69–0.94 0.043
Cohort 2 (MFY 2014)b,c            
   Primary outcomes            
      Mortality Logistic X 2 = 2.2; p = 0.98 OR 0.65 0.49–0.87 0.004
      Total direct cost Linear Adjusted r 2 = 0.45 β –$2,851 –$4,880 to –822 0.006
   Secondary outcomes            
      ICU admission Logistic X 2 = 14.7; p = 0.07 OR 0.68 0.52–0.87 0.002
      Hospital LOSd,e Cox N/A HR−1 0.88 0.79–0.98 0.022
      Adjusted net revenue Linear Adjusted r 2 = 0.30 β –$2,545 –$4,950 to –141 0.038
      Contribution margin Linear Adjusted r 2 = 0.18 β $1,418 –$303 to 3,139 0.106
Cohort 3 (MFY 2015)f,g            
   Primary outcomes            
      60-d in-hospital mortality Logistic X 2 = 3.4; p = 0.91 OR 0.84 0.73–0.97 0.019
      Total direct cost Linear Adjusted r 2 = 0.47 β –$1,571 –$2,746 to –397 0.009
   Secondary outcomes            
      ICU admission Logistic X 2 = 21.9; p = 0.005 OR 0.85 0.76–0.95 0.003
      Mechanical ventilation Logistic X 2 = 7.8; p = 0.45 OR 0.73 0.64–0.83 0.000
      Vasopressors required Logistic X 2 = 7.8; p = 0.45 OR 0.75 0.65–0.86 0.000
      Hospital LOSd,e Cox N/A HR−1 0.94 0.89–0.99 0.033
      Adjusted net revenue Linear Adjusted r 2 = 0.27 β –$2,315 –$4,183 to –448 0.015
      Contribution margin Linear Adjusted r 2 = 0.10 β $934 –$698 to 2,566 0.262  
Subanalysesh            
   Heart failure            
      60-d in-hospital mortality Logistic X 2 = 8.9; p = 0.35 OR 0.67 0.47–0.95 0.026
   Chronic renal failure            
      60-d in-hospital mortality Logistic X 2 = 11.4; p = 0.18 OR 1.27 0.83–1.94 0.269

HR = hazard ratio, LOS = length of stay, MFY = medicare fiscal year, N/A = not applicable, OR = odds ratio.
aAll outcomes from cohort 1 were tested in logistic regression models adjusted for the following variables: age, initial lactate, hypotension, two or more "Supersystemic inflammatory response syndrome (SIRS) criteria" at triage, acute kidney injury, compromised oxygenation, thrombocytopenia, coagulopathy, and altered mental status. The mortality model also adjusted for central line placement, and central venous pressure (CVP)/central venous oxygenation (Scvo2)/mixed venous oxygenation (Svo2) monitoring. Hosmer-Lemeshow tests assessed goodness-of-fit, where the null hypothesis that the model fit the data was accepted for p > 0.05.
bLogistic regression models from cohort 2 adjusted for age, congestive heart failure, chronic obstructive pulmonary disease, chronic renal failure, two or more "Super-SIRS criteria" at triage, initial lactate, hypotension, acute kidney injury, thrombocytopenia, altered mental status, and compromised oxygenation. Models did not adjust for central line placement or hemodynamic monitoring (CVP, Scvo2, or Svo2) because a prohibitively low number of subjects received these interventions. Model fit was assessed using Hosmer-Lemeshow test, where the null hypothesis that the model fit the data was accepted for p > 0.05.
cLinear regression models from cohort 2 adjusted for age, tertiary versus community site, payer class, congestive heart failure, chronic obstructive pulmonary disease, chronic renal failure, metastatic disease, nosocomial infection, initial lactate, hypotension, coagulopathy, altered mental status, surgical diagnosis-related group (DRG) product line, and case mix index.
dCox model from cohort 2 censored for mortality; adjusted for age, tertiary versus community site, payer class, congestive heart failure, chronic renal failure, nosocomial infection, hypotension, initial lactate, coagulopathy, altered mental status, surgical DRG product line, and case mix index.
eThe "event" in the Cox model was a live hospital discharge. The hazard ratio (HR) indicates rate of live discharge per unit time in the exposure group over the rate in the referent. The inverse HR (HR−1 ) is the ratio of time per unit live discharge between groups, that is, the relative length of stay.
fLogistic regression models adjusted for age, congestive heart failure, chronic obstructive pulmonary disease, chronic renal failure, nosocomial infection, two or more "Super-SIRS criteria" at triage, initial lactate, hypotension, acute kidney injury, thrombocytopenia, altered mental status, and compromised oxygenation. The mortality model also adjusted for central line placement, and CVP/Scvo2/Svo2 monitoring. Model fit was assessed using Hosmer-Lemeshow test, where the null hypothesis that the model fit the data was accepted for p > 0.05.
gLinear regression models adjusted for age, tertiary versus community site, payer class, congestive heart failure, chronic obstructive pulmonary disease, chronic renal failure, metastatic disease, nosocomial infection, initial lactate, hypotension, coagulopathy, altered mental status, surgical DRG product line, and case mix index. The total direct costs (TDC) model also adjusted for central line placement, and CVP/Scvo2/Svo2 monitoring.
hBoth subgroup analyses are drawn from cohort 3 and assess the primary mortality outcome. They utilize the same model variables employed in the primary analysis of the entire cohort.

References

  1. Liu V, Escobar GJ, Greene JD, et al: Hospital deaths in patients with sepsis from 2 independent cohorts. JAMA 2014; 312:90–92

  2. Fleischmann C, Scherag A, Adhikari NK, et al; International Forum of Acute Care Trialists: Assessment of global incidence and mortality of hospital-treated sepsis. Current estimates and limitations. Am J Respir Crit Care Med 2016; 193:259–272

  3. Iwashyna TJ, Cooke CR, Wunsch H, et al: Population burden of longterm survivorship after severe sepsis in older Americans. J Am Geriatr Soc 2012; 60:1070–1077

  4. Torio CMA, Andrews RMA: National Inpatient Hospital Costs: The Most Expensive Conditions by Payer, 2011. Statistical Brief No. 160. Rockville, MD, Agency for Healthcare Research and Quality, 2013

  5. Rivers E, Nguyen B, Havstad S, et al; Early Goal-Directed Therapy Collaborative Group: Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med 2001; 345:1368–1377

  6. Yealy DM, Kellum JA, Huang DT, et al; ProCESS Investigators: A randomized trial of protocol-based care for early septic shock. New Engl J Med 2014; 370:1683–1693

  7. Peake SL, Delaney A, Bailey M, et al; ARISE Investigators; ANZICS Clinical Trials Group: Goal-directed resuscitation for patients with early septic shock. N Engl J Med 2014; 371:1496–1506

  8. Mouncey PR, Osborn TM, Power GS, et al; ProMISe Trial Investigators: Trial of early, goal-directed resuscitation for septic shock. N Engl J Med 2015; 372:1301–1311

  9. Levy MM, Dellinger RP, Townsend SR, et al; Surviving Sepsis Campaign: The Surviving Sepsis Campaign: Results of an international guideline-based performance improvement program targeting severe sepsis. Crit Care Med 2010; 38:367–374

  10. Levy MM, Rhodes A, Phillips GS, et al: Surviving Sepsis Campaign: Association between performance metrics and outcomes in a 7.5-year study. Crit Care Med 2015; 43:3–12

  11. Gatewood MO, Wemple M, Greco S, et al: A quality improvement project to improve early sepsis care in the emergency department. BMJ Qual Saf 2015; 24:787–795

  12. Rhodes A, Phillips G, Beale R, et al: The Surviving Sepsis Campaign bundles and outcome: Results from the International Multicentre Prevalence Study on Sepsis (the IMPreSS study). Intensive Care Med 2015; 41:1620–1628

  13. Suarez D, Ferrer R, Artigas A, et al; Edusepsis Study Group: Cost-effectiveness of the Surviving Sepsis Campaign protocol for severe sepsis: A prospective nation-wide study in Spain. Intensive Care Med 2011; 37:444–452

  14. Singer M, Deutschman CS, Seymour CW, et al: The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA 2016; 315:801–810

  15. Shankar-Hari M, Phillips GS, Levy ML, et al; Sepsis Definitions Task Force: Developing a new definition and assessing new clinical criteria for septic shock: For the Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA 2016;315:775–787

  16. Seymour CW, Liu VX, Iwashyna TJ, et al: Assessment of clinical criteria for sepsis: For the Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA 2016; 315:762–774

  17. Doerfler ME, D'Angelo J, Jacobsen D, et al: Methods for reducing sepsis mortality in emergency departments and inpatient units. Jt Comm J Qual Patient Saf 2015; 41:205–211

  18. Kellum JA, Lameire N; KDIGO AKI Guideline Work Group: Diagnosis, evaluation, and management of acute kidney injury: A KDIGO summary (Part 1). Crit Care 2013; 17:204

  19. Levy MM, Fink MP, Marshall JC, et al; SCCM/ESICM/ACCP/ATS/SIS: 2001 SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference. Crit Care Med 2003; 31:1250–1256

  20. Bone RC, Balk RA, Cerra FB, et al: Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee. American College of Chest Physicians/Society of Critical Care Medicine. Chest 1992; 101:1644–1655

  21. CMS SEP-1: NQF-endorsed voluntary consensus standards for hospital care. Available at: http://www.nhfca.org/psf/resources/Updates1/SEP-1%20Measure%20Information%20Form%20(MIF).pdf. Accessed November 27, 2016

  22. Gaieski DF, Edwards JM, Kallan MJ, et al: Benchmarking the incidence and mortality of severe sepsis in the United States. Crit Care Med 2013; 41:1167–1174

  23. Kumar A, Roberts D, Wood KE, et al: Duration of hypotension before initiation of effective antimicrobial therapy is the critical determinant of survival in human septic shock. Crit Care Med 2006; 34:1589–1596

  24. Ferrer R, Martin-Loeches I, Phillips G, et al: Empiric antibiotic treatment reduces mortality in severe sepsis and septic shock from the first hour: Results from a guideline-based performance improvement program. Crit Care Med 2014; 42:1749–1755

  25. Ferrer R, Artigas A, Suarez D, et al; Edusepsis Study Group: Effectiveness of treatments for severe sepsis: A prospective, multicenter, observational study. Am J Respir Crit Care Med 2009; 180:861–866

  26. Seymour CW, Cooke CR, Heckbert SR, et al: Prehospital intravenous access and fluid resuscitation in severe sepsis: An observational cohort study. Crit Care 2014; 18:533

  27. Neumar RW, Otto CW, Link MS, et al: Part 8: Adult advanced cardiovascular life support: 2010 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circulation 2010; 122:S729–S767

  28. Jauch EC, Cucchiara B, Adeoye O, et al: Part 11: Adult stroke: 2010 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circulation 2010;122:S818–S828

Authors and Disclosures

Daniel E. Leisman, BS1,2, Martin E. Doerfler, MD3, Mary Frances Ward, RN, MS, ANP4,5, Kevin D. Masick, PhD6, Benjamin J. Wie, BA1, Jeanie L. Gribben, BS1, Eric Hamilton, BA6, Zachary Klein, MS6, Andrea R. Bianculli, BS1, Meredith B. Akerman, MS5, John K. D'Angelo, MD1 and Jason A. D'Amore, MD1

1Department of Emergency Medicine, Hofstra-Northwell School of Medicine, Hempstead, NY.
2Icahn School of Medicine at Mount Sinai, New York, NY.
3Department of Medicine, Hofstra-Northwell School of Medicine, Hempstead, NY.
4Department of Neurosurgery, Hofstra-Northwell School of Medicine, Hempstead, NY.
5Feinstein Institute for Medical Research, Manhasset, NY.
6Krasnoff Quality Management Institute, Northwell Health System, New Hyde Park, NY.

Dr. Doerfler disclosed other support (ongoing employment) and received funding from Northwell Health (employer). Ms. Ward was a site subinvestigator for the National Institutes of Health funded Process Clinical Trial. Ms. Ward and Dr. D'Amore disclosed government work, and received support for article research from CMMI. Their institution received funding from CMMI subcontract grant with Dartmouth College. The remaining authors have disclosed that they do not have any potential conflicts of interest.

processing....