Delta shock index predicts injury severity, interventions, and outcomes in trauma patients: A 10-year retrospective observational study

Abstract

BACKGROUND

Most trauma occurs among young male subjects in Qatar. We examined the predictive values of the delta shock index (DSI), defined as the change in the shock index (SI) value from the scene to the initial reading in the emergency unit (i.e., subtracting the calculated SI at admission from SI at the scene), at a Level 1 trauma center.

AIM

To explore whether high DSI is associated with severe injuries, more interventions, and worse outcomes [i.e., blood transfusion, exploratory laparotomy, ventilator-associated pneumonia, hospital length of stay (HLOS), and in-hospital mortality] in trauma patients.

METHODS

A retrospective analysis was conducted after data were extracted from the National Trauma Registry between 2011 and 2021. Patients were grouped based on DSI as low (≤ 0.1) or high (> 0.1). Data were analyzed and compared using χ² and Student’s t-tests. Correlations between DSI and injury severity score (ISS), revised trauma score (RTS), abbreviated injury scale (AIS), Glasgow coma scale (GCS), trauma score-ISS (TRISS), HLOS, and number of transfused blood units (NTBU), were assessed using correlation coefficient analysis. The diagnostic testing accuracy for predicting mortality was determined using the validity measures of the DSI. Logistic regression analysis was performed to identify predictors of mortality.

RESULTS

This analysis included 13212 patients with a mean age of 33 ± 14 years, and 24% had a high DSI. Males accounted for 91% of the study population. The trauma activation level was higher in patients with a high DSI (38% vs 15%, P = 0.001). DSI correlated with RTS (r = -0.30), TRISS (r = -0.30), NTBU (r = 0.20), GCS (r = -0.24), ISS (r = 0.22), and HLOS (r = 0.14) (P = 0.001 for all). High DSI was associated with significantly higher rates of intubation, laparotomy, ventilator-associated pneumonia, massive transfusion activation, and mortality than low DSI. For mortality prediction, a high DSI had better specificity, negative predictive value, and negative likelihood ratio (77%, 99%, and 0.49%, respectively). After adjusting for age, emergency medical services time, GCS score, and ISS, multivariable regression analysis showed that DSI was an independent predictor of mortality (odds ratio = 1.9; 95% confidence interval: 1.35-2.76).

CONCLUSION

In addition to sex-biased observations, almost one-quarter of the study cohort had a higher DSI and were mostly young. High DSI correlated significantly with the other injury severity scores, which require more time and imaging to be ready to use. Therefore, DSI is a practical, simple bedside tool for triaging and prognosis in young patients with trauma.

Key Words: Delta shock index; Trauma; Injury severity scores; Interventions; Outcomes

Core Tip: The delta shock index (DSI) is defined as the change of SI value from the scene to the initial reading in the emergency unit (i.e., subtracting calculated SI at admission from SI at the scene). Among young trauma patients, high DSI is associated with severe injuries, more interventions, and worse outcomes (i.e., blood transfusion, exploratory laparotomy, ventilator-associated pneumonia, hospital length of stay, and in-hospital mortality). Apart from gender-biased observation, almost one-quarter of the study cohort had a higher DSI and were mostly young. High DSI correlated significantly with the other injury severity scores that need more time and imaging to be ready to use. Therefore, DSI is a practical, simple bedside tool for triaging and prognosis in young, injured patients.

INTRODUCTION

Traumatic injuries have a substantial impact on morbidity and mortality rates worldwide, particularly in younger patients[1]. The initial severity of injury and quality of care significantly influence the outcomes of trauma patients in emergency settings[2-4]. Furthermore, bleeding remains the primary preventable factor leading to mortality among trauma victims[5,6]. Therefore, to improve survival based on an early goal-directed resuscitation approach, prompt identification of physiological derangement and efficient hemorrhage control are paramount[7].

Uncontrolled bleeding in trauma patients results in hypovolemic shock, which may be further worsened by the lethal triad of coagulopathy, lactic acidosis, and hypothermia[8]. Therefore, early restoration of hemostasis with exploratory laparotomy or interventional radiology is crucial for treating shock, in addition to massive blood transfusion[9]. Furthermore, it has been suggested that evaluating the initial physiological characteristics of trauma patients could aid in promptly identifying patients at risk of bleeding and with the worst outcomes[10].

Notably, the assessment of initial vital signs offers a simple and practical approach to the early determination of physiological derangement in trauma patients; nevertheless, relying only on individual vital signs for prognosis is often inaccurate[1]. As a result, various shock indices (SI) have been suggested for predicting outcomes based on vital signs obtained at the scene and in the emergency department (ED)[2]. The delta SI (DSI), which is the difference between the SI obtained at the accident scene and that in the ED (DSI = ED SI – Scene SI), has shown a better predictive power for outcomes in adult patients[2,10-12]. This can be explained by the fact that the DSI considers time-dependent variations in vital signs and SI. Although earlier studies have demonstrated that ED SI exhibits better performance among patients with various levels of injury severity[13,14], we thought that the changes in hemodynamic status at admission (Delta) could be more practical and informative than the initial individual reading. Considering the contrasting evidence in the literature, it is necessary to investigate whether the DSI can be regarded as a predictive tool in the trauma population. We hypothesized that among young trauma patients, high DSI is associated with the need for more interventions and worse outcomes (i.e., receiving more blood transfusions, requiring exploratory laparotomy, having more ventilator-associated pneumonia (VAP), prolonged hospital length of stay (HLOS), and higher in-hospital mortality). Therefore, this study aimed to analyze the correlation between DSI and injury severity and to assess the predictive value of DSI for the need for blood transfusion, interventions, and outcomes in trauma patients.

MATERIALS AND METHODS

A retrospective cohort analysis was conducted to include trauma patients admitted to a national level I trauma center, Hamad General Hospital (HGH), Doha, Qatar, between January 2011 and June 2021. The study included all trauma patients who required hospital admission and had initial vital signs taken at the scene and repeated in the ED. Patients with missing vital signs, either at the scene or in the ED, and those who died before arrival at the hospital were excluded from the study. Data were retrieved from the Qatar National Trauma Registry (QNTR) database at HGH and from electronic medical records [Cerner Corporation (North Kansas City, Missouri, United States)].

Data variables extracted for the study included demographics (age and sex), emergency medical services (EMS) time, scene time, team level activation, vital signs at the scene and in the ED (pulse, systolic blood pressure, diastolic blood pressure, respiratory rate), Glasgow coma scale (GCS) at the scene and in the ED, ethanol levels, cardio-pulmonary resuscitation, ED disposition, revised trauma score (RTS), trauma score and injury severity score (TRISS), injury severity score (ISS), abbreviated injury scale (AIS), focused assessment with sonography in trauma (FAST), and interventions [intubation, exploratory laparotomy, extracorporeal membrane oxygenation (ECMO), and resuscitative endovascular balloon occlusion of the aorta (REBOA)]. Information regarding in-hospital complications (sepsis, VAP, acute respiratory distress syndrome, thromboembolic events, abdominal compartment syndrome, multiorgan failure, and disseminated intravascular coagulation), discharge disposition, need for prehospital fluid administration, overall blood transfusion, massive blood transfusion, length of hospital and intensive care unit (ICU) stay, ventilatory days, and mortality was also collected. Based on the vital signs seen at the scene and during ED triage, we computed the SI as the ratio of heart rate to systolic blood pressure[14]. DSI was calculated by subtracting the SI at the scene from the SI in the ED[10,11]. The Medical Research Center (Institutional Review Board, MRC-01-21-990) approved the study protocol, with a waiver of informed consent for the retrospective chart review of anonymous data from the QNTR database. This study follows the recommendations of the strengthening the reporting of observational studies in epidemiology.

Statistical analysis

Data are presented as proportions, median, or the mean ± SD, as appropriate. Differences in categorical and continuous variables were analyzed using the χ² test and Student’s t-test, as appropriate. The Yates corrected χ² test was used for categorical variables if the expected cell frequencies were less than 5. Patients were grouped based on DSI as low (≤ 0.1) or high (> 0.1). Correlations between DSI and ISS, RTS, AIS, Glasgow coma scale (GCS), TRISS, HLOS, and NTBU were assessed using Pearson correlation coefficient (r) analysis. Receiver operating characteristic curves and area under the curve were used to identify the predictive power of DSI for mortality and need for massive transfusion protocol activation (MTP) and exploratory laparotomy. Furthermore, the diagnostic testing accuracy of DSI for predicting mortality and need for MTP and exploratory laparotomy was identified in terms of sensitivity, specificity, negative predictive value, and negative likelihood ratio. Multivariable regression analyses were performed to determine the predictors of mortality, massive transfusion, and exploratory laparotomy, using the most relevant covariates (age, sex, ISS, GCS ED, DSI, EMS time, and FAST). Data were expressed as odds ratios (OR) and 95% confidence intervals (CIs). A two-tailed P value of < 0.05 was considered statistically significant. All data analyses were performed using Statistical Package for the Social Sciences version 21 (SPSS, Inc., Chicago, IL, United States).

RESULTS

During the study period, a total of 13212 trauma patients with a mean age of 33 ± 15 years were included in this study, of which 91% were male and 24% had a high DSI. The overall injury severity score was 12.4 ± 9.3, and GCS at ED was 15 (3-15), 18% required blood transfusion, and massive transfusion protocol was activated in 3.2% of cases. Table 1 compares the demographic and clinical characteristics based on DSI. Patients with a high DSI had prolonged scene time (P = 0.001) and markedly impaired vital signs at the accident scene (P = 0.001) and upon arrival at the hospital (P = 0.001), including the mean GCS score (P = 0.001 for both), as compared to the low-DSI group. The need for cardio-pulmonary resuscitation (5.8% vs 1.3%, P = 0.001) and trauma activation level (TT1) was also higher in patients with a high DSI (38% vs 15%, P = 0.001). Moreover, patients with a high DSI were more likely to be transferred from the ED to the operating room (P = 0.001) and intensive care unit (P = 0.001) than those in the low-DSI group.

Table 1 Demographics and clinical characteristics based on delta shock index in trauma patients (n = 13212), n (%)/mean ± SD.

	DSI ≤ 0.1 (n = 10039; 76%)	DSI > 0.1 (n = 3173; 24%)	P value
Age (years)	33.6 ± 14.5	32.2 ± 15.1	0.001
Males	9113 (90.8)	2860 (90.1)	0.28
EMS time (min)	70.6 ± 27.9	70.8 ± 28.3	0.67
Scene time (min)	24.2 ± 15.8	26.9 ± 18.1	0.001
Level 1 trauma activation code	1350 (15.4)	1058 (37.6)	0.001
Level 2 trauma activation code	7177 (82.0)	1712 (60.9)	0.001
Vital signs at scene
Pulse rate	94.4 ± 19.8	88.1 ± 21.5	0.001
Systolic blood pressure	128.7 ± 21.9	138.1 ± 25.2	0.001
Diastolic blood pressure	80.1 ± 16.7	86.7 ± 21.8	0.001
Respiratory rate	19.3 ± 4.3	19.7 ± 5.4	0.001
GCS (mean, 95%CI)	14.30 (14.26-14.34)	13.03 (12.90-13.15)	0.001
Vitals at TRU
Pulse rate	87.3 ± 16.8	103.4 ± 22.4	0.001
Systolic blood pressure	131.2 ± 18.9	115.4 ± 21.6	0.001
Diastolic blood pressure	79.4 ± 13.4	72.5 ± 16.6	0.001
Respiratory rate	19.4 ± 3.2	20.3 ± 4.3	0.001
GCS (mean, 95%CI)	14.08 (14.03-14.14)	12.22 (12.05-12.38)	0.001
Cardio-pulmonary resuscitation	131 (1.3)	184 (5.8)	0.001
ED disposition
Operating room	1698 (16.9)	720 (22.7)	0.001
Intensive care unit	1917 (19.1)	1065 (33.6)	0.001
Step-down unit	107 (1.1)	21 (0.7)	0.001
Floor	5735 (57.1)	1261 (39.7)	0.001
High dependency unit	214 (2.1)	45 (1.4)	0.001
Transfer to another facility	54 (0.5)	6 (0.2)	0.001
Home	305 (3.0)	41 (1.3)	0.001
Died	9 (0.1)	14 (0.4)	0.001
Revised trauma score	7.6 ± 0.8	6.9 ± 1.5	0.001
TRISS	0.97 ± 0.08	0.90 ± 0.19	0.001
Injury severity score	11.3 ± 8.2	15.9 ± 11.3	0.001
Head AIS	3.3 ± 0.9	3.6 ± 1.0	0.001
Chest AIS	2.6 ± 0.7	2.7 ± 0.7	0.001
Abdomen AIS	2.5 ± 1.2	2.6 ± 1.0	0.04
Pelvis AIS	2.1 ± 0.5	2.3 ± 0.7	0.001

ED: Emergency department; AIS: Abbreviated injury score; TRISS: Trauma score and injury severity score; GCS: Glasgow coma scale; EMS: Emergency medical services time; DSI: Delta shock index.

Patients with a high DSI sustained significant injuries, as indicated by higher mean ISS (P = 0.001), greater head AIS (P = 0.001), chest AIS (P = 0.001), abdominal AIS (P = 0.04), and pelvis AIS (P = 0.001), and lower TRISS (P = 0.001) and RTS (P = 0.001) scores than those with a low DSI.

Table 2 compares the management and outcomes of the patients with trauma. The need for intubation (P = 0.001), exploratory laparotomy (P = 0.001), ECMO (P = 0.001), REBOA (P = 0.001), prehospital fluid administration (P = 0.001), and blood transfusion (P = 0.001) was significantly associated with high DSI compared to low DSI. Moreover, patients with a high DSI frequently required intubation at the scene, whereas those with a low DSI were more likely to be intubated in the ED (P = 0.001). In addition, the rates of FAST positivity (P = 0.001) and in-hospital complications, such as VAP (P = 0.001), acute respiratory distress syndrome (P = 0.001), thromboembolic events (P = 0.001), abdominal compartment syndrome (P = 0.002), multiorgan failure (P = 0.001), and disseminated intravascular coagulation (P = 0.01), were significantly higher in the high-DSI group. Furthermore, patients with a high DSI were found to have a longer hospital stay (P = 0.001) and critical care unit stay (P = 0.001) and required prolonged mechanical ventilation (P = 0.001). The overall in-hospital mortality rate was 2.8%, and patients with a high DSI (7.3% vs 1.4%, P = 0.001) had a 5-fold higher risk of mortality than those with a low DSI. Patients with a low DSI were more likely to be discharged home (88.3% vs 75.1%; P = 0.001), whereas those with a high DSI frequently required rehabilitation (8.0% vs 4.6%; P = 0.001).

Table 2 Comparison of management and outcome based on delta shock index, n (%)/mean ± SD/ mean, 95%CI.

	DSI ≤ 0.1 (n = 10039; 76%)	DSI > 0.1 (n = 3173; 24%)	P value
Intubation	1315 (13.1)	1108 (34.9)	0.001
Intubation location
Scene	433 (33.0)	539 (49.0)	0.001
Trauma resuscitation unit	550 (42.0)	397 (36.1)	0.001
Operating room	226 (17.2)	128 (11.6)	0.001
Intensive care unit	72 (5.5)	27 (2.5)	0.001
Floor	12 (0.9)	3 (0.3)	0.001
Referring facility	18 (1.4)	6 (0.5)	0.001
FAST done	9570 (95.3)	3047 (96.0)	0.09
Negative	8786 (93.9)	2638 (88.5)	0.001
Positive	445 (4.8)	304 (10.2)	0.001
Undetermined	125 (1.3)	39 (1.3)	0.001
Exploratory laparotomy	380 (3.8)	281 (8.9)	0.001
ECMO	8 (0.1)	11 (0.3)	0.001
REBOA	1 (0.0)	3 (0.1)	0.01
Prehospital fluid administered	3626 (36.1)	1392 (43.9)	0.001
Prehospital fluid (mL)	466 ± 399	585 ± 472	0.001
Blood transfusion	1358 (13.5)	1039 (32.7)	0.001
Total blood units transfused	3 (1-73)	4 (1-68)	0.001
Massive transfusion protocol	155 (1.5)	269 (8.5)	0.001
In-hospital complications
Sepsis	98 (1.0)	87 (2.7)	0.001
Ventilator-associated pneumonia	263 (2.6)	226 (7.1)	0.001
Acute respiratory distress syndrome	56 (0.6)	51 (1.6)	0.001
Thromboembolic events	46 (0.5)	39 (1.2)	0.001
Abdominal compartment syndrome	3 (0.05)	7 (0.2)	0.002
Multiorgan failure	6 (0.1)	12 (0.4)	0.001
Disseminated intravascular coagulation	1 (0.01)	4 (0.1)	0.01
Hospital length of stay	5 (1-505)	7 (1-360)	0.001
ICU length of stay	3 (1-155)	5 (1-161)	0.001
Ventilatory days	3 (1-180)	5 (1-115)	0.001
Mortality	141 (1.4)	231 (7.3)	0.001
Death on arrival	1 (0.02)	6 (0.2)	0.001
Discharge disposition
Home	8840 (88.3)	2376 (75.1)	0.001
Rehabilitation	461 (4.6)	254 (8.0)	0.001
Long-term care	72 (0.7)	61 (1.9)	0.001
Transfer to another hospital	499 (5.0)	242 (7.6)	0.001
Died	141 (1.4)	231 (7.3)	0.001

FAST: Focused assessment with sonography in trauma; ECMO: Extracorporeal membrane oxygenation; REBOA: Resuscitative endovascular balloon occlusion of the aorta; DSI: Delta shock index.

Table 3 shows the univariate correlation between the DSI and the other relevant covariates. DSI was significantly correlated with ISS (r = 0.223), GCS (r = -0.239), TRISS (r = -0.297), RTS (r = -0.301), head AIS (r = 0.166), chest AIS (r = 0.065), blood transfusion (r = 0.20), and HLOS (r = 0.14) (P = 0.001 for all).

Table 3 Univariate correlation between delta shock index and other relevant covariates.

Variable	Pearson correlation (r)	P value
Age	0.002	0.79
Injury severity score	0.223	0.001
Glasgow coma scale	-0.239	0.001
TRISS	-0.297	0.001
Revised trauma score	-0.301	0.001
Blood transfusion units	0.205	0.001
Head AIS	0.166	0.001
Chest AIS	0.065	0.001
Hospital length of stay	0.142	0.001
ICU length of stay	0.109	0.001

TRISS: Trauma score and injury severity score; AIS: Abbreviated injury score; ICU: Intensive care unit.

Figure 1A shows the receiver operating characteristic curve analysis. For mortality prediction, high DSI had better specificity, negative predictive value, and negative likelihood ratio (77%, 99%, and 0.49, respectively). Similarly, high DSI had better specificity, negative predictive value, and negative likelihood ratio (77.3%, 98.5%, and 0.47%, respectively) for predicting the need for MTP (Figure 1B) and exploratory laparotomy (Figure 1C). Figure 2 shows the multivariate regression analysis for the predictors of mortality and need for massive blood transfusion and exploratory laparotomy. After adjusting for age, sex, EMS time, GCS score, ISS, and FAST results, DSI was found to be an independent predictor of mortality (OR = 1.931; 95%CI: 1.326-2.813; P = 0.001) (Model-a), massive blood transfusion (OR = 5.259; 95%CI: 3.787-7.303; P = 0.001) (Model-b), and exploratory laparotomy (OR = 2.947; 95%CI: 2.134-4.070; P = 0.001; Model-c).

Figure 1 Receiver operating characteristic curve analysis. A: Receiver operating characteristic curve analysis yielded an area under the curve of 0.711 (95%CI: 0.676-0.746; P = 0.001) when the cutoff point of delta shock index for predicting mortality was 0.10 (sensitivity: 62.1%; specificity: 77.1%; negative predictive value: 98.6%; and negative likelihood ratio: 0.49); B: Receiver operating characteristic curve analysis yielded an area under the curve of 0.725 (95%CI: 0.692-0.758; P = 0.001) when the cutoff point of delta shock index for predicting the need for massive transfusion protocol activation was 0.10 (sensitivity: 63.4%; specificity: 77.3%; negative predictive value: 98.5%; and negative likelihood ratio: 0.47); C: Receiver operating characteristic curve analysis yielded an area under the curve of 0.604 (95%CI: 0.578-0.629; P = 0.001) when the cutoff point of delta shock index for predicting the need for exploratory laparotomy was 0.10 (sensitivity: 42.5%; specificity: 77.0%; negative predictive value: 96.2%; and negative likelihood ratio: 0.75). ROC: Receiver operating characteristic.

Figure 2 Multivariate regression analysis for the predictors of mortality and need for massive blood transfusion and exploratory laparotomy. A: Mortality; B: Massive blood transfusion; C: Exploratory laparotomy. FAST: Focused assessment with sonography in trauma; ISS: Injury severity score; SI: Shock index; GCS: Glasgow coma scale; ED: emergency department; EMS time: Emergency medical services time.

DISCUSSION

This study highlighted that DSI based on a nationally representative population could accurately predict the need for massive blood transfusion and exploratory laparotomy in severely injured patients. Therefore, a great difference between the SI in the ED and at the accident scene in trauma patients indicates the likelihood of the need for blood transfusion and surgical intervention. Additionally, a high DSI is associated with an increased risk of mortality in patients with injuries. Our findings suggest that an increase in heart rate at the same systolic blood pressure level in trauma patients signifies an increased risk of poor outcomes. In line with our observations, recent literature has highlighted the accuracy of DSI in predicting mortality and hemodynamic instability in adult[11,12,15] and pediatric[10] trauma patients. We also demonstrated that the predictive accuracy of DSI for the studied outcomes was significantly higher in our cohort.

Numerous adjusted scoring systems have been proposed to improve the accuracy of outcome predictions in patients[7,16-18]. However, only a limited number of studies in trauma patients[1,2,11,12] have explored the prognostic implications of DSI, which could be an effective early triage tool to identify minimal bleeding compared to conventional vital signs as well as SI at the scene and in the ED. Shock measurement that accounts for temporal variation is anticipated to be better than shock measures obtained at a single time point post-injury[19].

In our study, patients with a high DSI had prolonged scene times and markedly impaired vital signs, including a lower GCS score. Such an association between DSI and initial physiological parameters was foreseeable as severely injured patients at the accident scene require more time to stabilize before being transported to the ED for definitive resuscitation and urgent lifesaving interventions[20]. Similarly, an earlier retrospective observational study by Funabiki et al[21] of 139242 elderly trauma patients demonstrated an association between high DSI and emergency hemostatic resuscitation, especially among patients presenting with hypertension and a low GCS score on hospital arrival. Furthermore, in our study, the proportion of TT1 was higher in patients with a high DSI, and these patients had a greater chance of being transferred to the operating room and ICU directly from the ED. The high TT1 in patients with elevated DSI suggests greater urgency and severity of trauma. This would prompt a quicker decision-making to initiate massive blood transfusion, emergency interventions, and the need for ICU admission[22]. Supporting these observations, Kim et al[1] showed that a high DSI (> 0.1) was associated with an increased risk of mortality, need for ICU admission, and massive transfusion in patients with torso or extremity injuries. Another study supports the utility of DSI in appropriately assigning patients, specifically for ICU admission, in geriatric patients with blunt trauma[21]. Based on these observations, DSI may be a valuable tool for effective triage decision-making, activating the trauma team and ED disposition, and initiating blood transfusion, contributing to more timely and targeted interventions for patients with severe injuries. Nevertheless, validating triage decisions based on the DSI in a larger and more diverse cohort of trauma patients is crucial to enhance generalizability before considering it as a triage tool in routine patient care.

Furthermore, our research demonstrated that patients with a high DSI were more likely to sustain severe injuries, as evidenced by frequent anatomical injuries, greater ISS, and lower TRISS and RTS scores. Therefore, the degree of physiological derangement is associated with the severity of the injury, which mainly depends on the anatomical injury location[1,23]. Consistent with our findings, a previous study by Funabiki et al[21] reported a significantly higher rate of severe injuries (ISS > 15) in patients with high DSI than in those with low DSI. This indicates that DSI can serve as an indicator of evolving trauma severity, with an increase in DSI suggesting hypovolemic shock, continued bleeding, and inadequate resuscitation in the field[11,15,24]. Bardes et al[25] conducted a retrospective study including 549 adults with blunt chest or abdominal trauma who were transferred from the scene and 127 referred from other facilities. This study revealed a positive correlation between ISS and SI, with no observed correlation between ISS and DSI. Notably, variability in prehospital care, resuscitation approaches, existing health conditions, and the timing of interventions can affect the initial physiological condition of patients, injury severity, and outcomes.

As shown in Table 2, a high DSI was associated with a higher rate of interventions (exploratory laparotomy, intubation, ECMO, and REBOA), hemostatic resuscitation (prehospital fluid administration and blood transfusion), and in-hospital complications than a low DSI. Our results are consistent with a previous study involving adult trauma patients conducted by Joseph et al[11], which similarly showed that patients with a positive DSI were more likely to undergo exploratory laparotomy and had a higher risk of developing in-hospital complications and mortality than those with an unaltered or negative DSI. Earlier publications from our center showed the prognostic implications of SI in predicting the need for thoracic interventions and exploratory laparotomy in adult patients with blunt thoracic[26] and abdominal[27,28] trauma, respectively. We highlighted that high SI in the ED is associated with higher rates of mortality, complications, and need for exploratory laparotomy and intubation as compared to low SI in published studies based on anatomical region. This association aligns with the findings presented in the current study on DSI and the existing body of literature.

Kin et al[15] conducted a retrospective observational analysis of 628 adult patients with trauma who were presented to the ED. The study revealed no notable differences in length of hospital and ICU stay, amount of blood transfused, or duration of mechanical ventilation associated with DSI. In contrast, in our study, DSI significantly correlated with hospital and ICU stay, ventilatory days, and number of blood units transfused. This variability could arise because the previous study used different cutoff values for defining high DSI (≥ 0.3), as opposed to the present study (DSI > 0.1), leading to discrepancies in the observed associations.

Previous studies have shown that DSI is associated with the worst clinical outcomes in trauma patients[1,2,21,29]. Consistent with these observations, the overall in-hospital mortality rate was 2.8% in our study, and a high DSI was significantly associated with the risk of mortality. Funabiki et al[21] reported a significantly higher rate of in-hospital mortality in elderly trauma patients with a high DSI (OR = 1.27). Huang et al[2] demonstrated that a high DSI upon ED admission is associated with increased in-hospital and early mortality among critically ill trauma patients who were transferred to the ICU. Kim et al[1] reported a significantly higher rate of in-hospital mortality in the high DSI (> 0.1) group than in the low DSI (≤ 0.1) (2.0% vs 0.8%, P < 0.01). Similarly, a previous study involving trauma patients with moderate injury severity identified a correlation between DSI and 48-h mortality[29].

In this study, we performed multivariate regression analysis to identify predictors of mortality and the need for massive blood transfusion and exploratory laparotomy. After adjusting for potential confounding factors, it was found that DSI was an independent predictor of mortality [adjusted OR (aOR) = 1.931], massive blood transfusion (aOR = 5.259), and exploratory laparotomy (aOR = 2.947). An earlier study on pediatric trauma patients found an independent association between a positive DSI and early mortality (< 24 h) and in-hospital complications, and overall mortality[10]. Similarly, an independent association was reported between high DSI and the need for surgical intervention (aOR = 1.22) and mortality (aOR = 0.127) in elderly trauma patients. Hosseinpour et al[13] reported a significant independent association of prehospital SI, ED SI, and DSI with both 24-h and in-hospital mortality, need for blood transfusion, and length of hospital and ICU stay on multivariate analysis.

Another study reported that in-hospital mortality (aOR = 2.82), the need for massive transfusion (aOR = 5.24), embolization (aOR = 3.15), and surgery (aOR = 1.29) were independently associated with a high DSI (> 0.1)[1]. These results imply that high DSI levels could serve as an indicator for identifying patients in shock and those at a higher risk of mortality, necessitating prompt intervention[30].

Early identification of such patients based on DSI levels could enable emergency physicians to intervene promptly and implement appropriate treatments and interventions[31,32] to improve patient outcomes and increase the chances of survival following trauma.

Based on the sensitivity analysis for mortality and the need for massive transfusion and exploratory laparotomy, the present study determined a threshold value of 0.1. Consequently, a DSI > 0.1 is deemed clinically relevant for detecting minor physiological changes and identifying patients at risk of mortality and those requiring massive transfusion and emergency surgery. Our findings align with those of previous investigations[1,11,13,21], which also identified > 0.1 as the cutoff value for high DSI in trauma patients. Based on our findings, DSI > 0.1 appears to have moderate or modest predictive value for mortality and the need for massive transfusion in trauma patients; however, its predictive value for determining the necessity of exploratory laparotomy seems to be limited. Therefore, clinicians should consider additional factors and clinical assessments when making decisions regarding surgical intervention in patients with trauma.

This observational study has limitations, primarily owing to its retrospective and single-institution design. Second, the cutoff values for DSI may vary across different trauma systems; therefore, the generalizability of our findings may be limited to other regional trauma centers in rural settings. Third, we could not address the possible difficulty of recording prehospital vital signs by EMS providers, particularly in severely injured patients. Fourth, we could not ascertain the causal relationship between high DSI and the increased mortality rate observed within the cohort. Fifth, we lacked data on resuscitation involving interventions and drugs administered in the prehospital setting, which could influence heart rate and, subsequently, the DSI. Therefore, it was challenging to determine whether a high DSI resulted from physiological responses due to injury or inappropriate resuscitation. Sixth, the use of beta-blockers, which can lead to a lower heart rate, may influence the accuracy of the DSI in reflecting the true clinical status of patients, potentially underestimating their risk. However, in our study, the majority of trauma patients were young and less likely to be regularly on beta-blockers before the index admission for comorbidities. Finally, in contrast to abdominal or extremity trauma, which may be more easily examined and treated with hemostasis, respiratory deterioration in chest trauma caused by pneumothorax may have a different effect on vital signs, as we analyzed a variety of anatomical injuries in our study. Therefore, future research is needed to account for the anatomical diversity and focus on comparisons of the DSI with existing tools in different trauma settings and patient populations. Such studies will help determine the specific advantages or limitations of the DSI, especially in situations where factors such as medication use (e.g., beta-blockers), age, or comorbidities may affect its accuracy. Moreover, further investigations should focus on refining and validating the DSI in diverse clinical settings, including secondary and prehospital care.

CONCLUSION

In conclusion, in addition to sex-biased observations, almost one-quarter of the study cohort had a higher DSI among young trauma patients. DSI was an independent predictor of mortality, massive blood transfusion, and exploratory laparotomy in patients with trauma. Therefore, it could potentially serve as a tool for predicting unfavorable outcomes and the necessity for hemostatic intervention. In addition, a high DSI correlated significantly with the other injury severity scores, which require more time and imaging to be ready to use. Therefore, it is imperative to use the DSI and devise a treatment plan that considers the anatomical location and severity of the injury. Overall, the DSI is a practical, simple bedside tool for triaging and prognosis in young patients with trauma. Further research is necessary to validate these findings and to establish a more accurate cutoff point for DSI that correlates with the indications for exploratory laparotomy in patients with trauma.

ACKNOWLEDGEMENTS

The authors thank all the staff of the trauma registry database at the Trauma Surgery Section, Department of Surgery, Hamad General Hospital, Doha, Qatar.