Early recognition of sepsis through emergency medical services pre-hospital screening
Contents lists available at ScienceDirect
American Journal of Emergency Medicine
journal homepage:
Early recognition of sepsis through emergency medical services pre- hospital screening?
George Borrelli, MS3 a, Erica Koch, MS3 a, Ethan Sterk, DO b,
Shannon Lovett, MD b, Megan A. Rech, PharmD, MS b,c,?
a Loyola University Chicago Stritch School of Medicine, 2160 S 1st Ave, Maywood, IL 60153, United States
b Department of Emergency Medicine, Loyola University Medical Center, 2160 S 1st Ave, Maywood, IL 60153, United States
c Department of Pharmacy, Loyola University Medical Center, 2160 S 1st Ave, Maywood, IL 60153, United States
a r t i c l e i n f o
Article history:
Received 1 July 2018
Received in revised form 17 October 2018
Accepted 18 October 2018
Keywords:
Sepsis
Pre-hospital
Emergency medical services
a b s t r a c t
Background: The Surviving Sepsis Campaign implemented a 3-hour bundle including blood cultures, lactate, in- travenous fluids, and antibiotics to improve mortality in sepsis. Though difficult to achieve, bundle compliance is associated with decreased hospital mortality. We predict that the implementation of an Emergency Medical Services (EMS) sepsis screening tool will improve 3-hour bundle compliance.
Objectives: To determine if pre-hospital sepsis screening improves 3-hour bundle compliance.
Methods: Prospective implementation of an EMS sepsis screening tool (June 2016-November 2016) was com- pared to a historical control (August 2015-March 2016). The protocol was facilitated via communication be- tween nurses and EMS personnel. The primary outcome was 3-hour bundle compliance. Secondary outcomes included time to individual bundle components.
Results: Of 135 patients screened, 20 were positive and included in the study, and subsequently compared to 43 control patients. Baseline demographics were similar, except median Sequential Organ Failure Assessment score was higher for the pre-EMS tool group (5 [interquartile range (IQR) 2-8] vs. 2 [IQR 1-4], p b 0.01). Three-hour bundle compliance was significantly higher in the EMS tool group (80% vs. 44.2%, p b 0.01). The pre-EMS tool group had lower median time to lactate (15 [IQR 0-35] vs. 46 min [IQR 34-57], p b 0.001), 30 mL/kg IV fluids (6.5 [IQR 0-38] vs. 46 min [IQR 27.5-72], p b 0.001), and, although not significant, antibiotics (63.5 [IQR 44-92] vs. 72 min [IQR 59.5-112], p = 0.26).
Conclusion: Implementation of an EMS sepsis screening tool resulted in improved 3-hour bundle compliance compared to retrospective control.
(C) 2018
Introduction
Sepsis is a life-threatening response to infection that can lead to tis- sue ischemia, organ failure, and death [1]. Progression to the most se- vere form of sepsis, septic shock, can dramatically increase mortality [1]. Early intervention with antibiotics and fluid resuscitation is critical to patient outcomes. Septic shock mortality increases by 7.6% every hour the patient is left without appropriate antibiotics [2]. In 2012, The Surviving Sepsis Campaign (SSC) recommended that early inter- vention of sepsis patients is of vital importance for reducing mortality
? This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.The results in this paper have been presented in abstract form at the Society for Academic Emergency Medicine (SAEM) annual meeting in Indianapolis, Indiana, May 15-May 18, 2018.
* Corresponding author at: Emergency Medicine Clinical Pharmacist, Loyola University Medical Center, 2160 S 1st Ave, Maywood, IL 60153, United States.
E-mail addresses: [email protected] (G. Borrelli), [email protected] (E. Koch), [email protected] (E. Sterk), [email protected] (S. Lovett), [email protected] (M.A. Rech).
[3,4]. Early intervention includes a bundle to be completed within 3 h of recognition of severe sepsis or septic shock. The bundle is comprised of measurement of serum lactate concentration, fluid resuscitation of
30 mL/kg intravenous (IV) fluids for hypotension or lactate
>=4 mmol/L, blood cultures prior to antimicrobials, and the administra- tion of broad spectrum antibiotics for suspected or documented infec- tion [4].
The evidence for the need of early identification and medical inter- vention for septic patients is clear; however, global compliance of the bundle guidelines is lacking. One study showed that overall compliance of the SSC 3-hour bundle is 19% [5]. In addition, patients who received care with full compliance of the 3-hour bundle had a 40% decrease in hospital mortality [5]. The 3-hour bundle guidelines significantly reduce sepsis mortality with appropriate compliance; however, further evi- dence is needed to determine how complete bundle compliance can be achieved.
One study found that half of septic patients that present to the Emer- gency Department (ED) arrive via Emergency Medical Services (EMS)
https://doi.org/10.1016/j.ajem.2018.10.036
0735-6757/(C) 2018
[6]. EMS provides great opportunity for early sepsis recognition and medical intervention. EMS early interventions have been shown to im- prove outcomes in time-sensitive cases of stroke, cardiac arrest, myo- cardial infarction, and trauma [7-9]. Evidence shows that vital sign based sepsis screening in the pre-hospital setting often leads to earlier initiation and completion of the resuscitation bundle protocol [10]. The impact of pre-hospital screening and initiation of treatment has been widely debated, with one study concluding that although the bun- dles were completed in less time, the overall mortality rate did not im- prove [11]. Another study showed that initiation of intravenous access and fluids in the field decreased the mortality rate [9]. There is further need to explore the pre-hospital setting to improve both the screening process of septic patients and 3-hour bundle compliance in order to im- prove sepsis patient outcomes.
Methods
This retrospective cohort study was conducted at a large, multi- specialty, quaternary-care academic center Emergency Department in the United States in partnership with all EMS providers that transported to the ED. This study was approved by the academic center’s Institu- tional Review Board. The Emergency Department has a yearly census of approximately 47,000 patients, with 21% of patients transferred via EMS. The study took place in a high acuity ED with an admission rate of 33%. During the 6-month period of the study, June 2016 to November 2016, there was a census of 22,476 patients with approximately 4800 patients arriving by ambulance. In 2017, the Emergency Department re- ceived 22,743 calls with 10,040 transported to the ED. An EMS pre- hospital sepsis screening tool was created and prospectively imple- mented (Fig. 1). The screening tool was compared to a retrospective co- hort historical control of EMS patients who presented to the Emergency Department with an initial presentation of severe sepsis or septic shock between August 2015 and March 2016. Severe sepsis was defined as documented two out of four Systemic Inflammatory Response Syn- drome (SIRS) criteria, a suspected infection, and evidence of organ dys- function such as an elevated lactate (N2 mmol/L), acute Mental status change, hypotension with a mean arterial pressure b65 mm Hg or sys- tolic pressure b90 mm Hg, or an oxygen saturation b90%. Septic shock was defined as severe sepsis with persistent hypotension or end organ dysfunction despite fluid resuscitation or vasopressors. The protocol was carried out with the communication between Emergency Commu- nications Registered Nurse (ECRN) and EMS personnel while the patient was en route to the ED. If the patient screened positive, EMS was
directed by the ECRN to administer a 500 mL 0.9% NaCl fluid bolus, un- less there was evidence of fluid overload. The ECRN then noted the time and notified the attending physician of a positive sepsis screen. SSC 3- hour bundle compliance and the time to SSC 3-hour bundle recommen- dations were recorded and compared between groups. The majority of providers in the study were ALS providers. Of all providers, only two were BLS providers. All providers participating in the study were edu- cated on sepsis and the sepsis protocol prior to initiation of the protocol. Both ALS and BLS providers were able to identify sepsis and utilize the screening tool. Only ALS providers were able to initiate the fluid bolus. The screening tool consisted of the following seven possible criteria: Re- spiratory Rate N20 breaths per minute, Heart Rate N90 beats per minute, Systolic Blood Pressure b90 mm Hg, Documented Fever or History of Temperature N38.3 ?C or b36 ?C (N100.9 ?F or b96.8 ?F), New Onset of Mental Status Change, Oxygen Saturation b90%, and Suspected Infec- tion. ECRN personnel directed the screening process by asking EMS re- sponders if any of the seven criteria were present in the patient. The ECRN completed a screening tool form which was filed in a collection folder located in the radio room. The form was labeled with a sticker that included the name and Electronic Medical Record number of the patient. All patients over 18 years of age who presented to the ED via EMS between June 2016 and November 2016 were screened with the tool. The presence of three or more of the criteria was considered a pos- itive screen and sepsis protocol was immediately initiated. Patients with the following conditions were excluded from both study groups: b18 years of age, trauma, cardiopulmonary arrest, pregnancy, ST- segment elevation myocardial infarction (STEMI), and stroke. All ECRN and EMS personnel received standardized education.
The primary outcome was 3-hour bundle compliance. Secondary
outcomes included time to completion of the following bundle compo- nents: measurement of serum lactate concentration, fluid resuscitation of 30 mL/kg IV fluids for hypotension or lactate >=4 mmol/L, and admin- istration of broad spectrum antibiotics for suspected or documented in- fection. Other secondary outcome measures included the following: blood cultures prior to antimicrobials, hospital mortality, 30-day mor- tality, hospital length of stay , Intensive Care Unit (ICU) LOS, vaso- pressors, and hydrocortisone.
Retrospective chart review was conducted on a historic cohort of pa- tients who presented to the Emergency Department via EMS with se- vere sepsis or septic shock between August 2015 and March 2016. time zero for the pre-EMS screening tool group was defined as the time when two out of four SIRS criteria, a suspected infection, and organ dysfunction were documented. Signs of organ dysfunction
Respiratory Rate > 20 breaths per minute Heart Rate > 90 beats per minute Systolic Blood Pressure < 90 mmHg
Documented Fever or History of Temperature >38.3?C or <36?C (>100.9?F or
<96.8?F)
New Onset of Mental Status Change
O2 Saturation less than 90% Suspected Infection
Stop if Sepsis criteria not met
Fig. 1. Emergency Medical Services (EMS) pre-hospital sepsis screening tool. All patients arriving to the Emergency Department via EMS were screened.
include an elevated lactate (N2 mmol/L), acute mental status change, hypotension with a mean arterial pressure b65 mm Hg or systolic pres- sure b90 mm Hg, or an oxygen saturation b90%. Time zero for the post- EMS screening tool group was defined as the time of the first vital sign upon presenting to the ED via EMS. Vital signs at time zero, creatinine, albumin, white blood cell count, suspected source of infection, blood culture results, presence of chronic comorbid conditions such as chronic obstructive pulmonary disease and chronic kidney disease, time to anti- biotics, amount and time to intravenous fluids, time to lactate and lac- tate level, hospital and ICU LOS, hospital and 30-day mortality, and use of vasopressors and hydrocortisone were recorded for both study groups. Median Sequential Organ Failure Assessment score was calculated for both study groups. Three-hour bundle compliance was assessed and recorded as a percentage. Prior to the implementation of the EMS pre-hospital sepsis screening tool, there was no sepsis screening program in place. Therefore, the study compared two groups. One pre-intervention historical control group with no existing sepsis screening program and a post-intervention group with the EMS pre- hospital sepsis screening tool implemented.
Descriptive statistics were used to characterize baseline demo- graphics. Chi-square test or Fisher’s exact test, as appropriate, were used to analyze categorical data. Comparisons of parametric continuous variables were conducted using a t-test. Non-parametric data was expressed as a median using the Mann-Whitney test. A P-value b0.05 was considered significant in the bivariate analysis. A multivariate logistic regression was used to account for confounding variables and to determine if the EMS pre-hospital sepsis screening tool was indepen- dently associated with improved 3-hour bundle compliance.
Results
Sixty-three adults were included in this study (pre-EMS tool group: n = 43; post-EMS tool group: n = 20). The electronic medical records of 126 patients were reviewed retrospectively; 43 of the patients were in- cluded in the pre-EMS tool group. Of the 135 patients screened with the screening tool, 22 were positive and 2 of the patients who screened pos- itive were excluded due to age and study enrollment during a prior hos- pitalization (Fig. 2). Baseline characteristics and comorbidities were
similar between the study groups, with the exception of the SOFA score (Table 1). The median SOFA score was significantly higher for the pre-EMS tool group (5 [interquartile range (IQR) 2-8] vs. 2 [IQR 1-4], p b 0.01). The primary outcome of 3-hour bundle compliance was significantly higher in the post-EMS tool group compared to the pre-EMS tool group (80% vs. 44.2%, p b 0.01). The pre-EMS tool group had a lower median time to serum lactate concentration measurement (15 min [IQR 0-35] vs. 46 min [IQR 34-57], p b 0.001) and a lower me- dian time to 30 mL/kg IV fluids (6.5 min [IQR 0-38] vs. 46 min [IQR 27.5-72], p b 0.001). Median time to broad spectrum antibiotics was lower in the pre-EMS tool group (63.5 min [IQR 44-92] vs. 72 min [IQR 59.5-112], p = 0.26), however; not significant (Table 2).
Median hospital LOS was significantly shorter in the post-EMS tool group (8 days [IQR 5-12] vs. 5 days [IQR 3-6], p = 0.01). Median ICU LOS was also significantly shorter in the post-EMS tool group (3 days [IQR 0-6] vs. 0 days [IQR 0], p = 0.001). Hospital mortality was lower in the post-EMS tool group (11.6% vs. 0%, p = 0.14), however; not signif- icant. A significantly larger percentage of patients in the pre-EMS tool group required vasopressors (46.5% vs. 10.5%, p b 0.01) and hydrocorti- sone (39.5% vs. 10%, p = 0.02) (Table 2). Suspected sources of infection were similar between groups (p = 0.17), with genitourinary, pneumo- nia, and unknown the most common (Table 1).
Variables included in the multivariate logistic regression were pre- screening tool group vs. post-screening tool group, baseline SOFA score, atrial fibrillation, and heart failure. After accounting for con- founders, pre-screening tool group vs. post-screening tool group, base- line SOFA score, and atrial fibrillation were found to be independently predictive of 3-hour bundle compliance. Heart failure was found to be inversely predictive of 3-hour bundle compliance.
Discussion
We found that an EMS pre-hospital sepsis screening tool was associ- ated with an increased 3-hour bundle compliance. This provides evi- dence that the EMS pre-hospital sepsis screening tool was effective at promoting completion of the 3-hour bundle for sepsis patients. Despite improvement in bundle compliance, we found that time to the individ- ual bundle components of IV fluids, broad spectrum antibiotics, and
Fig. 2. Flowchart of included and excluded screens. EMS = Emergency Medical Services. Pre-group = Pre-screening tool group. Post-group = Post-screening tool group. Arriving to the Emergency Department via EMS is a requirement needed for patients to be included in the study for both the pre-screening tool group and the post-screening tool group.
Baseline demographics. IQR = Interquartile Range. SOFA = Sequential Organ Failure As- sessment. Significant difference between groups: p b 0.05.
|
Characteristic |
Pre-screening tool (n = 43) |
Post-screening tool (n = 20) |
P-value |
|
Age, mean (IQR) |
63 (56-73) |
70.5 (45.5-76) |
0.86 |
|
Sex, male, n (%) |
24 (55.8) |
11 (55.8) |
0.95 |
|
Race, white, n (%) |
20 (46.5) |
7 (35) |
0.52 |
|
SOFA score, median (IQR) |
5 (2-8) |
2 (1-4) |
0.003 |
|
Comorbidities, n (%) Atrial fibrillation |
11 (26.2) |
2 (10) |
0.14 |
|
Coronary artery disease |
10 (23.3) |
2 (10) |
0.21 |
|
Cancer |
18 (41.9) |
7 (35) |
0.60 |
|
Stroke |
11 (25.6) |
3 (15) |
0.35 |
|
Chronic obstructive pulmonary |
5 (11.6) |
1 (5) |
0.40 |
|
disease Heart failure |
13 (30.2) |
3 (15) |
0.20 |
|
Hypertension |
27 (62.8) |
15 (75) |
0.34 |
|
Liver disease |
7 (16.3) |
1 (5) |
0.21 |
|
Chronic kidney disease |
13 (30.2) |
5 (25) |
0.67 |
|
Source of infection, n (%) |
0.17 |
have appeared less critical, but they still received more comprehensive care.
Patients in the post-EMS tool group also had improved outcomes in terms of hospital and ICU LOS, vasopressor and hydrocortisone use, and a non-significant trend towards improved mortality. This is likely due to the difference in baseline severity of illness indicated by the difference in SOFA scores rather than to the screening tool. We attempted to con- trol for this difference with a multivariate logistic regression analysis.
Previous studies have shown that sepsis Screening tools can be im- plemented in the pre-hospital setting using a variety of variables for criteria. A retrospective cohort study examining 555 EMS encounters created the Pre-hospital Severe Sepsis (PRESS) score to be used by EMS to screen for severe sepsis with a sensitivity of 86% and specificity of 47% [14]. Chief complaint, age, nursing home status, temperature, and blood pressure were the variables found to most predict severe sepsis [14]. To our knowledge, no further validation of the PRESS score has been published. This study also did not seek to determine the impact of the PRESS score on patient care and outcomes upon arrival to the ED [14]. Hunter et al. also piloted a pre-hospital sepsis screening tool
Genitourinary 8 (18.6) 8 (40)
|
Pneumonia |
11 (25.6) |
3 (15) |
using suspected infection, temperature, heart rate, respiratory rate, |
|
Abdominal |
4 (9.3) |
0 (0) |
and end tidal volume of CO2 (ETCO2) [15]. When sepsis was suspected, |
Bloodstream 3 (6.98) 1 (5)
Skin and soft tissue 3 (6.98) 1 (5)
Other 1 (2.3) 3 (15)
Unknown 13 (30.2) 4 (20)
serum lactate measurements were higher in the post-EMS tool group. This is likely due to higher disease severity in the pre-EMS tool group, as evident in the higher median SOFA score in the pre-EMS tool group. A prospective, multicenter study consisting of 1449 patients in 40 inten- sive care units in 16 countries found that increased SOFA scores were as- sociated with worsening organ failure and increased mortality [12]. Another prospective observational cohort study of 352 consecutive pa- tients admitted to the ICU calculated a SOFA score for each patient every 48 h while admitted to the ICU. Their results demonstrated that higher average SOFA scores throughout the ICU stay and highest SOFA scores overall corresponded to an increased likelihood of organ failure and death [13]. Therefore, patients with higher SOFA scores would naturally be triaged to receive faster care in the Emergency Department upon ar- rival. Since our post-EMS tool group had a lower median SOFA score, it would likely mean that those patients had a higher probability of appearing more stable upon arrival and a less likelihood of death over- all. This could explain why bundle completion took more time in the post-EMS tool group. However, the increase in bundle compliance in the post-EMS tool group, despite the lower SOFA scores, suggests that the screening tool performed its intended role. These patients may
Primary and secondary outcomes. IQR = Interquartile Range. IVF = Intravenous Fluids. LOS = Length of Stay. ICU = Intensive Care Unit. Significant difference between groups: p b 0.05.
|
Primary outcome |
Pre-screening tool |
Post-screening tool |
P-value |
|
(n = 43) |
(n = 20) |
||
|
3-Hour bundle compliance, n (%) Secondary outcomes |
19 (44.2) |
16 (80) |
b0.01 |
|
Time to Lactate, median (IQR) |
15 (0-35) |
46 (34-57) |
b0.001 |
|
30 mL/kg IVF, n (%) |
19 (46.3) |
11 (73.3) |
0.07 |
|
Time to IVF, median (IQR) |
6.5 (0-38) |
46 (27.5-72) |
b0.001 |
|
Blood cultures prior to antibiotics, |
42 (97.7) |
20 (100) |
0.49 |
|
n (%) time to antibiotics, median (IQR) |
63.5 (44-92) |
72 (59.5-112) |
0.26 |
|
8 (5-12) |
5 (3-6) |
0.01 |
|
|
ICU LOS, median (IQR) |
3 (0-6) |
0 (0) |
0.001 |
|
Hospital Mortality, n (%) |
5 (11.6) |
0 (0) |
0.14 |
|
Vasopressors, n (%) |
20 (46.5) |
2 (10.5) |
b0.01 |
|
Hydrocortisone, n (%) |
17 (39.5) |
2 (10) |
0.02 |
a sepsis alert was initiated to the incoming hospital. They found that sepsis alert patients were more likely to be diagnosed with sepsis or se- vere sepsis upon arrival [15]. They also determined that ETCO2, which could be measured in the back of an ambulance, was a better predictor of severe sepsis and mortality than other variables such as vital signs [15]. This study demonstrated the feasibility of implementing a pre- hospital screening tool, but it did not examine the impacts of these tools on treatment quality measures. A prospective observational study of 160 patients transported by EMS showed that 33 patients who had a documented suspicion of sepsis received treatment and an- tibiotics faster than those 127 who had no documented suspicion [16]. The study did not analyze the effects of earlier treatment on ICU stay and overall survival. This study also relied on EMS documentation in- stead of using a standardized screening tool. Similar to our study, Guerra et al. used EMS education and application of a vital sign based severe sepsis screening tool to investigate the effect of pre-hospital screening on the detection of severe sepsis in 112 patients by EMS providers over 1 year. A Sepsis Alert Protocol was then initiated on all patients EMS deemed to have severe sepsis [17]. Retrospective analysis showed that the 32 of 67 severe sepsis patients correctly identified by EMS pro- viders had a faster time to antibiotics, more fluids at 2 and 6 h, and a shorter hospital LOS compared to patients who were not identified by EMS [17].
Our use of education and a screening tool aimed to remove any var- iation in EMS provider knowledge and judgement. Therefore, our proto- col is based on variables that are included in standardized measures such as the SIRS criteria and the qSOFA score. The protocol also has the ability to easily and quickly evaluate patients in the pre-hospital set- ting. Since no pre-hospital sepsis screening tools have been validated by multiple robust studies, we created our own screening tool to standard- ize the EMS assessment of patients with objective criteria. Therefore, our screening tool still requires external validation. Our primary aim was to investigate the effects of a pre-hospital sepsis screening tool on bundle compliance and time to treatment in the ED since studies are lacking. These previous studies have demonstrated the ability to imple- ment a reliable screening tool, while our study aimed to measure the ef- fects of these tools on care quality and patient survival. We were able to determine that a combined effort of the ED and EMS to implement and act on the screening tool led to increased SSC 3-hour bundle compliance.
Limitations
Data collection heavily depended on the ability of ECRN personnel to successfully communicate with EMS via radio and fill out the sepsis
screening tool forms. As a result, many potential patients were missed due to missing screening tools, especially on weekends and overnights when nursing staff and radio room coverage was limited. Inconsistent form filling, incorrect Electronic Medical Record number labeling of screening forms, and physical misplacement of the sepsis screening forms all led to data loss. This process led to a lower than anticipated sample size. Data collection has continued and will be analyzed at a later date to generate more robust results. Documentation was provided by different providers in each case, and there is no way to ensure accu- racy and consistency after the fact. The objective nature of our screening tool lessened the likelihood of Inaccurate data, but we still relied on ED personnel to accurately determine a time zero for notifying the emer- gency physician of possible sepsis. The screening tool created for the study still requires external validation. Due to the exact number not being saved in hospital records, the number of patients between June 2016 and November 2016 who arrived by ambulance, as noted in the methods section, is an estimation. Lastly, the number of patients transported by ALS versus BLS providers could not be obtained due to information not being saved in hospital records.
Conclusions
There is great potential in the pre-hospital setting for earlier recogni- tion of sepsis in patients arriving to the ED via EMS. Early recognition of sepsis will allow medical personnel to initiate sepsis protocol and de- crease an important variable in sepsis mortality, time. We demonstrated that the use of a pre-hospital sepsis screening tool was associated with significant improvement in 3-hour bundle compliance, likely due to im- proved recognition of sepsis. Further studies should be conducted to in- crease sample size and power. More data is needed to determine the impact that pre-hospital sepsis screening has on outcomes such as mor- tality and progression to septic shock.
The authors would like to thank the participating EMS providers and ED faculty.
References
- Lever A, Mackenzie I. Sepsis: definition, epidemiology, and diagnosis. BMJ 2007;335:
Kumar A, Roberts D, Wood KE, Light B, Parrillo JE, Sharma S, et al. Duration of hypo- tension before initiation of effective antimicrobial therapy is the critical determinant of survival in human septic shock. Crit Care Med 2006;34:1589-96.
- Levinson A, Casserly B, Levy M. Reducing mortality in severe sepsis and septic shock. Semin Respir Crit Care Med 2011;32:195-205.
- Dellinger R, Levy M, Rhodes A, Annane D, Gerlach H, Opal SM, et al. Surviving sepsis campaign: international guidelines for management of severe sepsis and septic shock: 2012. Crit Care Med 2013;41:580-637.
- Rhodes A, Phillips G, Beale R, Cecconi M, Chiche JD, De Backer D, 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: 1620-8.
- Groenewoudt M, Roest A, Leijten F, Stassen P. Septic patients arriving with emer- gency medical services: a seriously ill population. Eur J Emerg Med 2014;21:330-5.
- Iwami T, Nichol G, Hiraide A, Hayashi Y. Continuous improvements in “chain of sur- vival” increased survival after out-of-hospital cardiac arrests: a large-scale population-based study. Circulation 2009;119:728-34.
- Stiell I, Nesbitt L, Pickett W, Munkley D. The OPALS major trauma study: impact of advanced life-support on survival and morbidity. CMAJ 2008;178:1141-52.
- Minnerup J, Wersching H, Unrath M, Berger K. Effects of emergency medical service transport on acute stroke care. Eur J Neurol 2014;21:1344-7.
- Lane D, Ichelson R, Drennan I, Scales D. prehospital management and identification of sepsis by emergency medical services: a systematic review. Emerg Med J 2016; 33:408-13.
- Smyth M, Brace-McDonnell S, Perkins G. Identification of adults with sepsis in the prehospital environment: a systematic review. BMJ Open 2016;6:1-10.
- Vincent JL, de Mendonca A, Cantraine F, Moreno R, Takala J, Suter PM, et al. Use of the SOFA score to assess the incidence of organ dysfunction/failure in intensive care units: results of a multicenter, prospective study. Working group on “sepsis-re- lated problems” of the european society of intensive care medicine. Crit Care Med 1998;26:1793-800.
- Ferreira F, Bota D, Bross A, Melot C, Vincent J. Serial evaluation of the SOFA score to predict outcome in critically ill patients. JAMA 2001;286:1754-8.
- Polito C, Isakov A, Yancey A, Wilson D. Prehospital recognition of severe sepsis: de- velopment and validation of a novel EMS screening tool. Am J Emerg Med 2015;33: 1119-25.
- Hunter CL, Silvestri S, Ralls G, Stone A, Walker A, Papa L. A prehospital screening tool utilizing end-tidal carbon dioxide predicts sepsis and severe sepsis. Am J Emerg Med 2016;34:813-9.
- Studnek J, Artho M, Garner C, Jones A. The impact of emergency medical services on the ED care of severe sepsis. Am J Emerg Med 2012;30:51-6.
- Guerra W, Mayfield T, Meyers M, Clouatre A, Riccio J. Early detection and treatment of patients with severe sepsis by Prehospital personnel. J Emerg Med 2013;44: 1116-25.