Prehospital Antibiotics for Presumed Sepsis

Washington University Emergency Medicine Journal Club– March 16th, 2022


You are moonlighting in a large community ED one afternoon when EMS brings you a 65-year-old male with a remote history of lymphoma, now in remission, with a chief complaint of fevers and cough. En route, EMS noted the patient to be febrile to 38.4 ℃ with a blood pressure of 108/68, heart rate 114, respiratory rate 18, and normal oxygen saturation on room air. After a bolus of 500 mL of LR the patient remained febrile, and his heart rate improved to 95. By EMS protocol, the paramedics then drew blood cultures from the initial IV site and administered 2 grams of ceftriaxone for presumed bacterial sepsis.

The patient arrives in stable condition, is ultimately diagnosed with metapneumovirus with a clear chest x-ray and is discharged home. While you understand the concerns about early therapy in presumed bacterial sepsis (having read multiple prior journal club write-ups on early goal directed therapy in sepsis) you wonder if this EMS protocol isn’t a bit overly aggressive, and likely to overdiagnose sepsis in patients with non-infectious SIRS and viral syndromes. In addition to increased cost, you also worry about breeding antibiotic resistance to one of our most frequently used parenteral antibiotics. As you presume this protocol is evidence-based, you email your Washington University EMS and critical care gurus to see if they know what the evidence really shows…

PICO Question

Population: Adult patients with suspected or possible sepsis being transported to
the hospital by EMS

Intervention: Administration of antibiotics in the prehospital setting for presumed
sepsis, utility of screening tools to assess for possible sepsis

Comparison: Usual care (fluid resuscitation as needed during transport to the ED or
definitive care)

Outcome: Time to initial antibiotics, mortality, hospital length of stay, ICU length of
stay, need for and duration of vasopressors/mechanical ventilation/renal
replacement therapy

Search Strategy

Article 1: Varney J, Motawea KR, Kandil OA, Hashim HT, Murry K, Shah J, Shaheen A, Akwari J, Awad AK, Rivera A, Mostafa MR, Swed S, Awad DM. Prehospital administration of broad-spectrum antibiotics for sepsis patients: A systematic review and meta-analysis. Heal Sci Rep . 2022;5. Answer Key.

Article 2:  Jouffroy R, Gilbert B, Tourtier JP, Bloch-Laine E, Ecollan P, Bounes V, Boularan J, Léguillier T, Gueye-Ngalgou P, Vivien B. Impact of prehospital antibiotic therapy on septic shock mortality. Prehosp Emerg Care. 2021;25(3):317-24. Answer KEy.

Article 3: Alam N, Oskam E, Stassen PM, Exter PV, van de Ven PM, Haak HR, Holleman F, Zanten AV, Leeuwen-Nguyen HV, Bon V, Duineveld BAM, Panday RSN, Kramer MHH, Nanayakkara PWB, Alam N, Nanayakkara PWB, Oskam E, Stassen PM, Haak HR, Holleman F, Nannan Panday RS, Duineveld BAM, van Exter P, van de Ven PM, Bon V, Goselink J, De Kreek A, van Grunsven P, Biekart M, Deddens GJ, Weijschede F, Rijntjes N, Franschman G, Janssen J, Frenken J, Versluis J, Boomars R, de Vries G, den Boer E, van Gent A, Willeboer M, Buunk G, Timmers GJ, Snijders F, Posthuma N, Stoffelen S, Claassens S, Ammerlaan H, Sankatsing S, Frenken J, Alsma J, van Zanten A, Slobbe L, de Melo MM, Dees A, Carels G, Wabbijn M, van Leeuwen-Nguyen TTH, Assink J, van der Honing A, Luik P, Poortvliet W, Schouten WEM, Veenstra J, Holkenborg J, Cheung TC, van Bokhorst J, Kors B, Louis-Wattel GH, Roeleveld T, Toorians A, Jellema W, Govers A, Kaasjager HAH, Dekker D, Verhoeven MAM, Kramer MHH, Flietstra T, Roest L, Peters EJG, Hekker TAM, Ang W, van der Wekken W, Ghaem Maghami P, Kanen B, Wesselius H, Heesterman L, Zwietering AN, Stoffers J, on behalf of the, PHANTASi Trial Investigators and the ORCA (Onderzoeks Consortium Acute Geneeskunde) Research Consortium the Netherlands, PHANTASi Trial Investigators and the ORCA (Onderzoeks Consortium Acute Geneeskunde) Research Consortium the Netherlands. Prehospital antibiotics in the ambulance for sepsis: A multicentre, open label, randomised trial. Lancet Respir Med. 2018;6(1):40-50. Answer Key.

Article 4: Lane DJ, Wunsch H, Saskin R, Cheskes S, Lin S, Morrison LJ, Scales DC. Screening strategies to identify sepsis in the prehospital setting: A validation study. CMAJ. 2020;192(10):E230-9. Answer Key.

Bottom Line:

Sepsis is a time critical diagnosis (TCD) with mortalities (sepsis 11%, septic shock
26%) meeting or exceeding those of other TCDs such as STEMI (Seymour 2017,
Vincent 2017, Jollis 2022). Delay to antibiotic administration has been shown to
increase absolute mortality per hour both for septic shock (1.8-7.6%/hr) and sepsis
(3%/hr) (Kumar 2006). Increased mortality with community acquired sepsis has
been associated with delays both in ED arrival to antibiotic start and \irst medical
contact (FMC) to antibiotic start. As with other FMC-based TCDs, EMS can play a
major role in in\luencing downstream care by improving earlier identi\ication,
management, and mobilizing hospital resources.
Earlier recognition of sepsis requires prehospital identi\ication and this is best
achieved through standardized criteria. Traditionally for healthcare settings,
infection plus some type of organ dysfunction has been used to characterize “sepsis.”
In addition to these criteria, CMS also requires ≥2 SIRS criteria for severe sepsis or a
provider diagnoses of severe sepsis or septic shock (CMS Expert Panel Evaluation of
Measures). SEP-3 on the other hand does not require any SIRS criteria to be present,
but recommends infection plus new or acute change of SOFA ≥2 to be classi\ied as
sepsis (Singer 2016). Assessing the full array of CMS or SEP-3 SOFA lab-based organ
dysfunctions in the prehospital setting is not feasible. However, some physiologic
organ dysfunction criteria used by CMS (i.e. SBP < 90 or MAP < 65) and SOFA/
modi\ied SOFA (MAP/vasopressors, GCS, 02 requirement) could be used as
screening tools (Yealy 2021). Unfortunately, determining the presence of “infection”
can be variable and prone to bias based on risk factors, clinical gestalt, exam, or lab/
imaging \indings. EMS personnel would therefore require relatively reliable and
consistently reproducible criteria in addition to signs of physiologic organ
dysfunctions to identify sepsis. This Journal Club therefore sought to evaluate the
use of objective standardized sepsis tools in the prehospital setting as well as the
utility of prehospital antibiotics in sepsis, severe sepsis, and septic shock.
The \irst study reviewed assessed the performance of 21 prehospital screening tools
for identi\ication of sepsis (Lane 2020). While none of these tools had both high
sensitivity and high speci\icity for classifying sepsis, 3 scoring tools had good
predictive ability and the greatest range in probability of sepsis diagnosis from their
minimum to maximum scores. These 3 scoring tools are MEWS (uses 5 criteria), CIP
(6 criteria), and qSOFA (3 criteria). Interestingly, besides temperature (≥38.5C or
<35C) used in MEWS, none of the 3 tools required diagnosis of infection and were
driven solely by physiologic parameters (i.e. age, vital signs, mental status). Their
predictive abilities to diagnose sepsis at maximum scores is likely coincidental due
to the fact that sepsis is one of the most prevalent causes of acute deterioration or
critical illness unless an alternative etiology is readily apparent.
Of these scoring tools, the simplest to utilize would be qSOFA, as it only requires 3
criteria (SBP, RR, GCS) in addition to suspected infection. If the patient meets all 3
qSOFA criteria their probability of sepsis is 87%. This offers the consideration of a
graded treatment approach with a score of 1 triggering a sepsis alert to the hospital
(to expedite organ dysfunction evaluation) and initiation of \luids if needed. A qSOFA
score of 2 denotes an overall 10% mortality and strong consideration for timely
antibiotics given the risk of hourly incremental increases in mortality if infection is
ruled in and antibiotics were delayed. Patients meeting all 3 qSOFA criteria are
obviously sicker which matches other studies reviewed in this journal club
indicating the sicker sepsis patients in more advanced stages of sepsis are more
likely to bene\it from prehospital intervention.
Multiple previous studies indicate that interventions are feasible in the prehospital
setting, including both IV hydration and IV antibiotics. Not surprisingly, multiple
articles included in a meta-analysis on prehospital antibiotics for sepsis (Varney
2022) demonstrated that prehospital initiation of antibiotics signi\icantly reduces
the time to administration of antibiotic. While this does not necessarily correlate
with improved outcomes, additional studies reviewed in this journal club indicate
that sicker sepsis patients are the ones most likely to bene\it. While both the metaanalysis
and an observational retrospective review (Jouffrey 2021) revealed a
mortality bene\it (pooled relative risk 0.81, 95% CI 0.68-0.97; hazard ratio 0.56,
95% CI 0.35-0.90), the single randomized controlled trial we reviewed did not show
any mortality bene\it (Alam 2018). Unfortunately, this RCT has been criticized for
many methodological limitations including prehospital provider bias (opening
envelopes until they got one recommending treatment with antibiotics)
and treatment violations (40 patients in the usual care group were given antibiotics
and 12 patients in the intervention group did not receive antibiotics). The
retrospective review looked speci\ically at patients with septic shock, indicating a
potential bene\it in patients with more advanced stages of sepsis. It should be
noted, however, that patients in this retrospective review were treated by physicians
in the prehospital setting, which is not currently a part of the U.S. prehospital model.
Bottom line: Timely antibiotics for septic shock and sepsis can reduce mortality. The
reliability of determining de\inite/probable infection outside of timely con\irmative
diagnostics remains dif\icult in both the ED/hospital, as it will in the prehospital
setting. We will need to identify a reliable prehospital approach or scoring tool to
identify sepsis and more speci\ically “infection” in a reliable manner. qSOFA is an
appealing tool because it only requires evaluation of 3 criteria, is simple, and
meeting all 3 criteria (plus infection suspected) indicates a very high probability of
sepsis. In addition, prehospital medicine could explore adopting the ’21 Surviving
Sepsis Campaign guidelines and make modi\ications applicable to the prehospital
setting. This would likely involve using FMC as time zero and working out logistics to
achieve <1 hour antibiotics for any possible septic shock or high likelihood sepsis
(Figure 1, Evans, 2021).
Recommendations at this time:
1.Develop FMC-based metrics for septic shock and sepsis TCDs.
2.Consider a tiered prehospital approach:
a.Sepsis alert without antibiotics: possible or de\inite/probable infection
and no prehospital physiologic organ dysfunction (i.e. qSOFA <2 or modi\ied
SOFA<2 [Table 3: MAP, GCS, 02 requirement]7)
b.Sepsis alert with FMC to antibiotics <1hr (Administer antibiotics by
EMS and/or ED if ED can ensure FMC to antibiotics <1hr if alerted):
a.Any possible septic shock (possible or de\inite/probable infection +
SBP<90 or MAP<65)
b.High likelihood normotensive sepsis (de\inite/probable infection +
organ dysfunction: qSOFA ≥2 or SOFA ≥2 [Table 3: GCS≤12 or 2L NC to keep