EBM Review Series: Are they shocky?
VACEP Evidence-Based Medicine for General Emergency Physicians Series
Authors: Megyn Christensen, DO & Martin D. Klinkhammer, MD, MPH, FACEP | Eastern Virginia Medical School at Old Dominion University
Reviewers: Peter Victoria, MD & Joshua Easter MD, MBA, MSc, MHA | UVA Health
Editor: Winston Wu, MD | Virginia Tech Carilion School of Medicine
The VACEP Evidence-Based Medicine Review Series allows Virginia emergency medicine residents and attendings to share and analyze a recent peer-reviewed clinical study. You can also read the full article, “Shock index identifies compensated shock in the ‘normotensive’ trauma patient.” from the Injury Journal, Volume 56 Issue 9.
THE CASE
You are working at a community hospital when you receive a radio call from EMS about a young female involved in a motor vehicle crash. The report states she was going ~60 mph when she struck a tree trying to avoid a deer on the road. Medics report positive loss of consciousness, no blood thinner use, and the patient complains of headache, neck pain, and back pain. EMS reports positive airbag deployment and moderate vehicle damage. Medics state she has some minor abrasions, a small head laceration with bleeding controlled, no major deformities/injuries, and the patient was ambulatory on the scene. Initial vitals include heart rate 121, blood pressure 108/64, respiratory rate 20, pulse ox 98 on room air. The nearest trauma center is over 1 hour away, so the patient is brought to you.
Is this trauma patient in shock and will she require early transfusion?
BACKGROUND
Shock index (SI), defined as heart rate (HR) divided by systolic blood pressure (SBP), was first described in 1967 as an approximation of hemodynamic status. Currently, shock index can be used as a prediction tool and risk stratification to help determine critically ill patients from a variety of medical causes. There are now several variations of shock index including the original SI (HR/SBP), modified SI (HR/mean arterial pressure), Age SI (age x HR/SBP), and shock index pediatric adjust (SIPA) (Koch, 2019). Shock indices have been used to help guide care in the pre-hospital setting, triage, and in-hospital management of patients.
There is a significant amount of literature on using shock index as a predictive tool in trauma. Prior studies have found that an abnormal shock index is associated with increased mortality, hospital length of stay (LOS), injury severity score (ISS > 16), intensive care unit (ICU) admission rate, and incidence of blood transfusion (Cannon, 2009; Bruijns 2013; Liao, 2024). However, these studies all included patients who were frankly hypotensive (SBP < 90), and so the findings may have been influenced by the patients that were in obvious shock. The authors of our paper sought to focus on individuals potentially in hemorrhagic shock but not hypotensive, in order to determine if shock index was a useful triage tool among these patients (Lin, 2025).
STUDY SUMMARY
This study looked at the utility of SI in identifying trauma patients with compensated shock, using early transfusion as a surrogate marker for shock. The authors hypothesized that an abnormal SI (> 0.7) in non-hypotensive (SBP > 90) trauma patients would be associated with early blood transfusion (Lin 2025). This was a retrospective study, performed at a Level 1 trauma center from January 2016 - March 2023. Patient demographics, injury characteristics, vitals upon hospital arrival, blood products administered during the first hour following arrival, and hospital disposition were gathered. Exclusion criteria were individuals with a SBP < 90, age < 18, pregnancy, interfacility transfers, and individuals who died within 60 minutes of arrival. SI was calculated by arrival HR divided by arrival SBP. The SI was then stratified into thresholds of: ≤ 0.7, 0.7-0.9, > 0.9 to 1.1, > 1.1 to 1.3, and > 1.3. They used the odds of transfusion within 1 hour of arrival for each SI category with ≤ 0.7 as the referent and calculated using contingency tables from the cross-tabulation function and reported with 95% confidence intervals. P-values < 0.05 were considered statistically significant. Receiver operating characteristic curve (ROC) analysis and the Youden index criteria were used to determine a meaningful SI value to raise clinical concern for early transfusion in the “normotensive” early compensated shock trauma population.
About the EBM Review Series
This is a literature review series started by the University of Virginia’s Josh Easter, MD, MSc, a VACEP board member working to connect the academic community in Virginia. We invite each residency in Virginia (and D.C.) to create a faculty/resident team to submit and review articles. Sign up to submit one.
Goals
Provide a brief monthly synopsis of a high yield article germane to the practice of emergency medicine for distribution to all VACEP members
Provide an opportunity for a peer reviewed publication and invited presentation for faculty and trainees
Foster an academic community focused on evidenced based medicine for emergency medicine residency programs in the region
STRENGTHS and LIMITATIONS:
Strengths:
A key strength of the study is that it specifically examined only normotensive trauma patients, helping to clarify the role of shock index in compensated shock rather than including those patients that present frankly hypotensive. It also gives some predictive values to various presenting SIs and the likelihood of these patients requiring transfusion. Finally, the sample size of this study was fairly large, even if it was from a single trauma center.
Limitations:
One limitation of this study was that transfusions were a surrogate measure and one cannot be certain if blood transfusion truly represented shock in all study patients. Additionally, since it was a retrospective study, the authors could not determine if the decision to transfuse was influenced by the SI of the patient (information bias); or if other factors such as repeat vitals, the patient’s clinical condition, or imaging studies were the predominant factors causing the clinicians to decide to transfuse (though the authors state that this is not the institutions typical practice). This study also only focused on the first hour of patient arrival and associated transfusions as the primary outcome, and did not look into hemorrhage control procedures or ICU admission. Lastly, excluding patients who died in the first hour may have impacted the results, though as they didn’t look at these patients it is impossible to know if this would have led to a greater or lesser association between SI and transfusion.
RESULTS:
In total, there were 11,703 trauma activations, however with the exclusion criteria applied the sample size was 5,958. Of the 5,958, 211 were transfused with PRBCs within 1 hour of hospital arrival. Massive transfusion protocol (MTP) was initiated for 91 of these 211. There were no pre-hospital transfusions in this population. Of the transfusion products, the median number of PRBCs transfused in the first hour was 4. Of the 211 who received transfusion, 133 also received pre-thawed plasma, 92 received platelets, and 62 received cryoprecipitate. Of the 56.1% (n=3,343) patients with an SI ≤ 0.7, 42 (1.3%) were transfused within one hour. As the shock index increased, the proportion of patients receiving transfusions rose, reaching 42.9% among those with an SI greater than 1.3. As SI increased, the proportion of patients in each SI category decreased linearly, down to 0.9% with a SI > 1.3.
There was a 3-fold increase in the odds ratios of receiving blood with each increase in SI category. The authors additionally created a similar model only using patients with SBP > 110 and found similar findings, with higher shock index associated with increased likelihood of transfusion. Among this group there was still a roughly 3-fold increase in odds ratio of receiving blood with each successive SI category. They also looked at HR and SBP in place of SI to test these variables as predictors of early transfusion. While decreasing blood pressure and increasing HR were both associated with increased odds of receiving blood, neither performed as well as the SI. The study also observed that the average admission lactic acid level and proportion of patients who went from the ED to OR or IR suite for hemorrhage control also had a respective increase in each SI category.
The receiver operating characteristic curve (ROC) analysis of SI predicting early transfusion was 0.77 (0.73-0.81 P < 0.0001) while the Youden index criteria was > 0.825. The Youden index optimizes sensitivity and specificity to give the optimal cutoff value of a diagnostic test. This would indicate that a SI of greater than 0.825 is of clinical importance. However, the authors advocate using a SI of 1 which has a sensitivity of 36.0% (95% CI 29.5-42.9), a specificity of 95% (95% CI 94.4-95.5), a positive likelihood ratio of 7.1 (95% CI 5.8-8.8), and a negative likelihood ratio of 0.7 (95% CI 0.6-0.8). Using a cutoff SI of 1 also has a positive predictive value of 22.8% (95% CI 17.5-24.5) and a negative predictive value of 97.6% (95% CI 97.3-97.8), indicating that a patient with a SI greater than 1 is roughly 10 times more likely to receive a transfusion than a patient with a SI of less than 1. Using a shock index of 1 has the added simplicity that it doesn’t require math in the trauma bay, if the HR is greater than the SBP then the SI is greater than 1 and should raise concern for compensated shock.
CASE CONCLUSIONS:
Shock index is an easy prediction tool and serves as a marker of hemorrhagic shock in trauma patients. This study showed that a SI greater than 1 is significantly associated with the likelihood of requiring blood. Using SI is a more useful clinical metric than either SBP or HR alone in identifying early compensated shock. In normotensive trauma patients, a shock index of one or higher should prompt concern for compensated hemorrhagic shock and early consideration of transfusion and hemorrhage control.
TAKE-HOME DISCUSSION:
Back to our patient, given her initial vital signs with a HR of 121 and a SBP of 108 (again, the HR is greater than the SBP) you recognize immediately that she has a high likelihood of developing decompensated shock and requiring early transfusion. Given this, you do a FAST exam, identify intraperitoneal blood, and while hanging blood make a call to your nearest trauma center for expedited transfer.
Reviewer questions and author responses
Effect of prehospital interventions on shock index
Question: Were there any distinctions regarding patients receiving prehospital interventions that might affect shock index, such as intravenous fluids or analgesics
Author response: The authors did not collect or analyze data on prehospital interventions.
Use of age adjusted or modified shock indices
Question: Was there any consideration of whether age adjusted or modified shock indices such as age shock index or modified shock index would improve prediction in normotensive trauma patients
Author response: The authors did not evaluate alternative shock indices. They examined heart rate and blood pressure individually but did not calculate or compare other derived indices.
Relationship between shock index and massive transfusion protocol activation
Question: Was there any consideration of using shock index to predict activation of a massive transfusion protocol rather than any early transfusion?
Author response: Approximately half of the patients who received blood, 91 of 211, had a massive transfusion protocol activated. The authors considered analyzing this subgroup separately but ultimately did not, in part because Table 2 suggests that once transfusion was initiated, patients were similarly likely to receive multiple units across shock index categories. They infer that the association between shock index and massive transfusion protocol activation would likely mirror the relationship between shock index and any early transfusion, without adding additional discriminatory value.
Application of shock index beyond trauma
Question: Could shock index be applied beyond trauma, for example in sepsis or gastrointestinal bleeding, to detect early compensated shock
Author response: Other studies have evaluated shock index in nontrauma populations such as sepsis and gastrointestinal bleeding, and have generally found it useful. However, most of these studies included patients who were already hypotensive. The authors see an opportunity for future work that applies a similar approach to normotensive patients with nontraumatic shock.
