Outcomes of Trauma Patients Present to the Emergency Department with a Shock Index of ≥1.0.

Introduction: The study aimed primarily to evaluate the association between the initial shock index (SI) ≥1.0 with blood transfusion requirement in the emergency department (ED) after acute trauma. The study's secondary aim was to look at the outcomes regarding patients' disposition from ED, intensive care unit (ICU) and hospital length of stay, and deaths.
Methods: It was a retrospective, cross-sectional study and utilized secondary data from the Saudi Trauma Registry (STAR) between September 2017 and August 2020. We extracted the data related to patient demographics, mechanism of injuries, the intent of injuries, mode of arrival at the hospital, characteristics on presentation to ED, length of stay, and deaths from the database and compared between two groups of SI <1.0 and SI ≥1.0. A P < 0.05 was statistically considered significant.
Results: Of 6667 patients in STAR, 908 (13.6%) had SI ≥1.0. With SI ≥1.0, there was a significantly higher incidence of blood transfusion in ED compared to SI <1.0 (8.9% vs. 2.4%, P < 0.001). Furthermore, SI ≥ 1.0 was associated with significant ICU admission (26.4% vs. 12.3%, P < 0.001), emergency surgical intervention (8.5% vs. 2.8%, P < 0.001), longer ICU stay (5.0 ± 0.36 vs. 2.2 ± 0.11days, P < 0.001), longer hospital stays (14.8 ± 0.61 vs. 13.3 ± 0.24 days, P < 0.001), and higher deaths (8.4% vs. 2.8%, P < 0.001) compared to the patient with SI <1.0. Conclusions: In our cohort, a SI ≥ 1.0 on the presentation at the ED carried significantly worse outcomes. This simple calculation based on initial vital signs may be used as a screening tool and therefore incorporated into initial assessment protocols to manage trauma patients. Copyright:

INTRODUCTION

Trauma or physical injury is a significant cause of mortality and disability worldwide.[1] Road traffic accidents (RTAs) are one of the most common reasons for major trauma.[23] The RTAs are responsible for approximately 50 million injuries each year including the deaths of over 1.25 million people.[4] It is estimated that injuries from RTAs rank as the ninth cause of death globally and are assumed to be the seventh leading cause of death by 2030. In Saudi Arabia, nearly 80% of all trauma admissions are due to RTAs and cost the economy around 21 billion US dollars every year.[56] Moreover, 4.7% of all deaths are caused by RTAs, representing nearly 11% of total deaths in 2010.[678] RTAs rank as the leading cause of disability also.[46] In 2013, there were 526,400 accidents, resulting in 7661 deaths, and every day there are approximately 25 deaths as a result of RTAs.[69] It is estimated that between 35 and 38 per 100,000 people die or are disabled due to RTAs each year.[10]

Massive hemorrhage accounts for 50% of deaths during the first 24 h after hospital admission following major trauma.[11] Studies demonstrate an association with improved outcomes when massive transfusion is started in the emergency department (ED). The failure to do so is an independent predictor of mortality.[12]

The emergency evaluation of trauma victims can be challenging when vital signs and physical examination findings do not reflect severe injuries. The use of the shock index (SI), defined as the ratio of heart rate (HR) to systolic blood pressure (SBP), has been correlated with the degree of shock, decreased tissue oxygenation, and left ventricular performance.[13] A SI value >0.9 has helped to identify the ED patients requiring immediate therapy, admission, and intensive care unit (ICU) admission.[14] Massive transfusion initiation is inconsistent among providers, and up to 25% of “potentially preventable” trauma deaths did not receive a massive transfusion because of failure to identify the need.[15]

This study aimed primarily to evaluate the association between initial SI ≥1.0 with blood transfusion in the ED after acute trauma. The study's secondary aim was to look at the outcomes regarding patients’ disposition from ED, ICU and hospital length of stay, and deaths.

METHODS

Setting: The King Saud Medical City (KSMC) is one of the largest hospitals in Saudi Arabia, with 1400 inpatient beds. The KSMC ED is the busiest in the Kingdom.[16] The annual patient visit at the ED of KSMC is more than 246,000, of which more than 23,000 are trauma patients, which makes KSMC the largest in the country.[17] Saudi TraumA Registry (STAR) became operational at KSMC in September 2017. It is a comprehensive trauma management software system to document patient injuries, the care provided, patient outcomes, and system performance, which will link to all MOH trauma centers’ registry system. All trauma patients with a principal diagnosis of injury and one of the following were included in the registry: death in the ED due to injury, inpatient admission or transfer to KSMC, inpatient death following injury, and admission to the intensive care unit (ICU). On the other hand, superficial injury and/or amputation of single fingers and toes only, length of stay <3 calendar days apart from death and/or admission to the ICU, burns <10% total body surface area, and injury date more than three calendar days before admission to the first hospital were excluded in the registry.[18]

Design: It was a retrospective, cross-sectional study and utilized the secondary data from the STAR between September 2017 and August 2020. The data related to patient demographics, mechanism of injuries, the intent of injuries, mode of arrival at the hospital, baseline (on presentation to ED) characteristics, length of stay, and deaths were extracted from the STAR database. The data were compared between two groups of SI <1.0 and SI ≥1.0. Blood transfusion requirement at ED with SI ≥1.0 was the primary outcome variable. The secondary outcome variables were patients’ disposition from ED, ICU and Hospital length of stay, and deaths.

Statistical analysis

The data were analyzed using SPSS 25.0 (IBM SPSS Statistics for Windows, Version 25.0. Armonk, NY: IBM Corp.). The patients were classified based on presenting SI ≥1.0 versus <1.0 in the ED. The nominal variables were presented in frequency and percentage. The association between the demographic and risk factors was tested by the Mann–Whitney test, Chi-square test, and Wilcoxon signed-rank test based on the number of categories. The diagnostic tools sensitivity, specificity with receiver operating characteristic (ROC) curves, and the odds ratio (OR) with 95% confidence interval (CI) were provided. The Kaplan–Meier survival curves were used to predict the probability of survival among the patients with SI ≥1.0 versus <1.0. All the tests were performed at a 5% level of significance.

RESULTS

In total, 6667 patients were included for analysis from the STAR database during the study period. Of these, 908 (13.6%) patients had SI ≥1.0 [Figure 1].

Figure 1

Sample

The overall mean age was 32.8 (standard deviation 0.22) years and the majority (5574 [83.6%]) were men. The most common mechanism of injury was blunt (6096 [91.4%]) trauma. Of these, 3909 (58.6%) patients sustained motor vehicle collisions and 2187 (32.8%) fell. A trauma team was activated in 680 (10.2%; 95% CI: 9.5%–10.9%) cases. The majority of patients (5358 [80.8%]) ISS was <15 [Table 1]. A total of 221 (3.3%; 95% CI: 2.8%–3.7%) received a blood transfusion in the ED. Overall, 238 (3.6%; 95% CI: 3.1%–4.0%) patients required emergency surgical intervention and were transferred to the operating room directly from the ED. A total of 950 (14.2%; 95% CI: 13.4%–15.1%) patients were admitted to ICU from ED, and overall mortality was 237 (3.6%; 95% CI: 3.1%–4.0%) among the cohort [Table 2]. The comparison between demographics, injury mechanisms, the intent of injuries, and baseline (on presentation to ED) characteristics between two groups of SI <1.0 and ≥1.0 are described in Table 1. The GCS (14.4 vs. 13.5; P < 0.001) and PH (7.4 vs. 7.3; P < 0.001) were significantly lower, and the base deficit was significantly higher (0.90 vs. 2.6; P < 0.001) in SI ≥1.0 group. Furthermore, there was a significantly higher proportion (12% vs. 28.7% P < 0.001) of mechanically ventilated patients in the SI ≥1.0 group [Table 1].

Table 1

The comparison between demographics, mechanisms, intent of the injuries, and baseline (on presentation to emergency department) characteristics

Variable	Total (n=6667), n (%)	Shock index <1.0 (n=5759; 86.4%), n (%)	Shock index ≥1.0 (n=908; 13.6%), n (%)	P
Age (years), mean±SD	32.8±0.22	32.8±0.23	33.2±0.61	<0.001*
Male sex	5574 (83.6)	4865 (84.5)	709 (78.1)	<0.001*
Injury mechanism
Blunt	6096 (91.4)	5325 (92.5)	771 (84.9)	<0.001*
MVC	3909 (58.6)	3428 (59.5)	481 (53.0)	0.243
Fall	2187 (32.8)	1897 (33.0)	290 (31.9)	0.030*
Penetrating	261 (3.9)	223 (3.9)	38 (4.2)	0.666
Stab	217 (3.2)	190 (3.3)	27 (3.0)	<0.001*
Firearm	44 (0.7)	33 (0.6)	11 (1.2)	<0.001*
Burn	299 (4.5)	204 (3.5)	95 (10.5)	<0.001*
Others/unknown	11 (0.2)	7 (0.1)	4 (0.4)	0.025*
Injury intent
Nonintentional	6367 (95.5)	5501 (95.5)	866 (95.4)	0.893
Intentional	32 (0.5)	26 (0.5)	6 (0.7)	0.439
Assault	240 (3.6)	209 (3.6)	31 (3.4)	0.763
By police and military	5 (0.1)	4 (0.1)	1 (0.1)	1.000
Unknown	23 (0.3)	19 (0.3)	4 (0.4)	0.617
Mode of arrival
Government	1707 (25.6)	1452 (25.2)	255 (28.0)	0.072
Red crescent	2049 (30.7)	1875 (32.6)	174 (19.2)	<0.001*
Private	1879 (28.2)	1635 (28.4)	244 (26.9)	0.351
Others	1032 (15.5)	797 (13.8)	235 (25.9)	<0.001*
Trauma team activation	680 (10.2)	511 (8.9)	169 (18.6)	<0.001*
Systolic blood pressure (mmHg), mean±SD	126.5±0.28	130.5±0.27)	101.1±0.57)	<0.001*
Heart rate (beats/min), mean±SD	93.7±0.23	89.2±0.19	122.3±0.61	<0.001*
GCS, mean±SD	14.3±0.03	14.4±0.03	13.5±0.11	<0.001*
PH, mean±SD	7.3±0.00	7.4±0.00	7.3±0.00	<0.001*
Base deficit, mean±SD	1.1±0.03	0.90±0.03	2.6±0.12	<0.001*
INR, mean±SD	1.1±0.00	1.1±0.00	1.1±0.01	1.000
Mechanical ventilation	950 (14.2)	689 (12.0)	261 (28.7)	<0.001*
ISS
<15	5385 (80.8)	4736 (82.2)	649 (71.5)	<0.001*
15-25	923 (13.8)	738 (12.8)	185 (20.4)	<0.001*
>25	359 (5.4)	285 (5.0)	74 (8.1)	<0.001*

*Statistically significant at 5% level. MVC: Motor vehicle collision, INR: International normalized ratio, ISS: Injury severity score, SD: Standard deviation, GCS: Glasgow Coma Scale, PH: It is a measure of how acidic/basic the blood is.

Table 2

Outcome analysis

Variable	Total (n=6667), n (%)	Shock index <1.0 (n=5759; 86.4%), n (%)	Shock index ≥1.0 (n=908; 13.6%), n (%)	P	OR (95% CI)
Blood transfusion in ED	221 (3.3)	140 (2.4)	81 (8.9)	<0.001*	3.9 (3.0-5.2)
Disposition from ED
Ward	5291 (79.4)	4853 (84.3)	438 (48.2)	<0.001*	10.1 (8.6-11.9)
ICU	950 (14.2)	710 (12.3)	240 (26.4)	<0.001*	2.6 (2.2-3.0)
Operating room	238 (3.6)	161 (2.8)	77 (8.5)	<0.001*	3.2 (2.4-4.3)
Others	188 (2.8)	35 (0.6)	153 (16.9)	<0.001*	33.1 (22.7-48.2)
ICU length of stay (days), mean±SD	2.6±0.11	2.2±0.11	5.0±0.36	<0.001*	-
Hospital length of stay (days), mean±SD	13.6±0.22	13.3±0.24	14.8±0.61	0.023*	-
Death	237 (3.6)	161 (2.8)	76 (8.4)	<0.001*	3.2 (2.4-4.2)

*Statistically significant at 5% level. SD: Standard deviation, ED: Emergency department, ICU: Intensive care unit, OR: Odds ratio, CI: Confidence interval

The performance of SI to predict blood transfusion in ED is illustrated in Figure 2.

Figure 2

Receiver operator characteristics curve for penetrating and blunt injury (n = 6667)

Overall AUROC for the entire cohort was 0.712 (P = 0.000, 95% CI: 0.676–0.748). The patients sustained blunt trauma with AUC = 0.712 (P = 0.000, 95% CI: 0.674–0.750) performed better than the patients sustained penetrating trauma with AUC = 0.693 (P = 0.002, 95% CI: 0.571–0.816) [Figure 2].

With SI ≥1.0, a significantly higher incidence of blood transfusion in ED was observed compared to SI <1.0 (8.9% vs. 2.4%, P < 0.001, OR 3.9, and 95% CI: 3.0–5.2) in our cohort. Furthermore, SI ≥1.0 on presentation was associated with significant ICU admission (26.4% vs. 12.3%, P < 0.001, OR 2.6, 95% CI: 2.2–3.0), emergency surgical intervention (8.5% vs. 2.8%, P < 0.001, OR 3.2, 95% CI: 2.4–4.3), longer ICU stay (5.0 ± 0.36 days, P < 0.001), longer hospital stay (14.8 ± 0.61 days, P < 0.001), and higher deaths (8.4% vs. 2.8%, P < 0.001, OR 3.2, 95% CI: 2.4–4.2) compared to the patient with SI <1.0 [Table 2].

The mean survival time of patients with SI ≥1.0 was 94 ± 8 days compared to those with SI <1.0 was 200 ± 10 days (Mantel Hazel log-rank P = 0.000). The probability of survival among the patients with SI ≥1.0 versus SI <1.0 is shown in Figure 3.

Figure 3

Survival Probability Functions of the shock index (n = 6667)

DISCUSSION

Trauma leads to bleeding and can cause deaths if not controlled.[1920] Traditionally, hypovolemic shock is classified based on the percentage of circulatory blood loss.[21] Practically, accurate quantification of blood loss is not possible. Hence, physiological signs such as HR, SBP, pulse pressure, respiratory rate, urine output, and mental status have been correlated to the class of shock by the ATLS®. The clinical validity of shock classification by ATLS was recently questioned by data from the TARN registry and TraumaRegister DGU®, and physiological parameter base deficit was recommended as more appropriate to recognize the hypovolemic shock in trauma patients.[222324] The point of care testing of base deficit may not be readily available everywhere. The SI calculation to recognize hypovolemia due to bleeding in trauma patients is more practical and can quickly be done when the vitals are available.

The SI of ≥1.0 has been correlated with moderate (Class III) to severe (Class IV, SI ≥1.4) shock and blood and blood products transfusion needs.[2125] In our cohort, the patients with SI of ≥1.0 showed significantly higher blood transfusion requirements in the ED. This group of patients was also considerably acidotic and had a higher base deficit on our ED presentation. Vandromme et al. demonstrated on 8111 patients that a SI of more than 0.9 had increased the risk of massive blood transfusion by 1.5-fold. A further increase of SI more than 1.3 was associated with an eightfold risk of blood transfusion.[26] Zarzaur et al. showed overall trauma population, SI was most closely associated with a transfusion of four blood units or more within the first 48 h after admission to the hospital.[27]

There was a significantly higher patients’ proportion of injury severity scores (ISS) (15–25 and >25) in the SI of ≥1.0 group in our series. Mutschler et al. on a retrospective analysis of 21,853 trauma patients from TraumaRegister DGU® showed worsening of SI category was associated with increments in ISS, new ISS, and revised injury severity classification scores as well as higher percentages of chest, abdomen, and pelvic injuries.[25]

We found that the mortality was significantly higher, with an SI of ≥1.0, corresponding with the previously published research in the literature.[252829] Cannon et al. demonstrated increased mortality with SI >0.9 in 2445 trauma patients treated in an urban level I center.[28] They also concluded that a five-fold increase in mortality between the accident scene and the arrival at ED in patients increased SI by ≥0.3.[28] According to Zarzaur et al. the SI was also a significantly better mortality predictor for 48-h than SBP or HR alone.[29] In another study, Kim et al. on 628 patients showed that SI is a good predictor of mortality in polytrauma patients.[30] Although they emphasized “delta SI” defined as the difference between SI of at scene and arrival in ED is the best predictor, which was also supported by Schellenberg et al. in their study on 2591 patients.[31] Schellenberg et al. also showed delta SI >0.1 is associated with an increased need for blood transfusion and ICU length of stay.[31] In our study, SI of ≥1.0 was associated with a significantly higher portion of mechanically ventilated patients and required ICU admissions. Zampieri et al. support these findings in their study on 3140 patients.[32]

Liu et al. argued that diastolic blood pressure falls earlier than systolic pressure, hence modified SI, calculated by dividing HR by mean arterial pressure, may be a more accurate marker for assessing shock state and mortality.[33] Few studies supported age-SI as superior to SI and modified SI in predicting mortality.[3435]

Our study has several limitations. Being a retrospective registry-based study, it did not evaluate at scene SI, the number of units of blood transfusion, any other blood products transfusions in ED, reasons for surgical interventions, etc., Moreover, additional calculation of delta-SI, modified-SI, or age SI would have given our study more strength in outcome analysis. However, as the first report on approximately 7000 trauma patients over 3 years from a newly developed trauma registry-the STAR, would help the emergency health professional recognize the importance of SI in managing trauma patients. Carefully selecting and adding more variables in our trauma registry with a systematic collection of data on patients’ presentation and outcomes would be invaluable in the future.

CONCLUSIONS

Early recognition of shock is essential for early initiation of blood transfusion and lethal triad-hypothermia, acidosis, and coagulopathy arrest and improve outcomes. This study indicates that an easily calculated physiological variable, the SI, may identify blood transfusion prediction among adult trauma patients at high risk. In our cohort, a SI ≥1.0 on the ED presentation carried significantly worse outcomes, including increased emergency surgeries, ICU admissions, length of stay, and deaths among trauma patients. This simple calculation based on initial vital signs may be used as a screening tool and therefore incorporated into initial assessment protocols to manage trauma patients.

Research quality and ethics statement

The KSMC IRB approved the study with the reference number H1RI-22-Dec20-07. The authors followed applicable EQUATOR Network (http://www.equator-network.org/) guidelines during the conduct of this research project.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.