Article, Endocrinology

Blood sugar changes and hospital mortality in multiple trauma

a b s t r a c t

Objectives: Hyperglycemia with unknown mechanism plays a Predictive role in determining the prognosis of Multiple trauma patients. The exact time of Blood sugar measurement and the role of blood sugar changes in the monitoring of these patients have not been well established.

Methods: This follow-up study was done on multiple trauma patients (N 18 years) with an injury severity scores (ISS) N 16. These patients didn?t have any history of diabetes, underlying disease, or drug or alcohol use. Data col- lection was done by the questionnaire (checklist), and the patients were followed by the medical records. Cox regression was used to measure the effect of independent variables on the patients’ hospital mortality.

Results: Of a total of 963 patients, 280 patients were enrolled. Of those, 202 were male (72.1%) and 78 were female (27.9%). Hospital mortality was 18 (6.4%). Cox regression analysis suggested that those who had high blood sugar 3h after admission had higher hospital mortality (P = 0.04). Changes in blood sugar, ?BS (BS 3 h after admission – BS on admission), in these patients was also significantly correlated with hospital mortality (P b 0.001). The mul- tivariate model using the backward conditional method showed that ?BS (P b 0.001), international normalized ra- tion (INR) (P b 0.001), and heart rate (P = 0.036) were significantly correlated with hospital mortality.

Conclusions: In multiple trauma patients, blood sugar changes in the early hours of admission to the emergency de- partment may help predict hospital mortality, but further studies are needed. Blood sugar monitoring in these pa- tients during this time frame may be helpful in predicting these patients’ outcomes. In addition, coagulopathy and tachycardia were significantly associated with hospital mortality.

(C) 2017


Multiple trauma refers to life-threatening injuries in single or multiple organ(s), defined as the severity of injury N 16 in the Injury Severity Score (ISS). Multiple trauma is one of the major and com- mon causes of death each year, leading to the deaths of 8.5 million people in the world. This issue has become a serious health problem in Developing countries [1-3]. Given that about 80% of deaths occur in the first four hours, rapid diagnosis and proper management of these patients are very important. Hence, the outcome determina- tion for these patients is considered an important part of trauma care [2,3]. Trauma scoring systems, anatomically or physiologically or combined, have been designed to predict multiple trauma

Abbreviations: BE, base excess; BS, blood sugar; BS 0, blood sugar on admission; BS 3, blood sugar 3 h after admission; ?BS, BS 3 h after admission – BS on admission; DBP, diastolic blood pressure; HR, heart rate; Hb, hemoglobin; HCT, hematocrit; INR, international normalized ratio; SBP, systolic blood pressure; ISS, injury severity score; RTS, Revised Trauma Score.

* Corresponding author.

E-mail address: [email protected] (M. Torabi).

patients’ outcomes [4-7]; however, given the complexity of these systems, laboratory tests may provide a proper predictor for mortal- ity [3].

Hyperglycemia after trauma (stress-induced hyperglycemia) is one of these laboratory abnormalities. Patients who refer after trauma with hyperglycemia, or critically ill patients, have a poor prognosis. This hyperglycemia can even be a predictor of hemorrhagic shock in these patients. However, in various studies, paradoxical results for the prognostic effect of Admission hyperglycemia have been reported. The timing of measurement of blood sugar seems to play an important role in determining its prognostic effect in these patients [8-14].

The aim of this study was to investigate blood sugar (BS) levels in mul- tiple trauma patients on admission, 3 h after admission (BS 3), and the dif- ference between them (?BS), and their relationship with hospital mortality. Also, this study investigates the role of International normalized ratio , Serum lactate level, base excess , the ISS, and revised trau- ma score (RTS) for the prediction of hospital mortality in these patients.


This follow-up study was carried out on patients with multiple trau- ma that were referred to Bahonar academic Hospital, Level II Trauma 0735-6757/(C) 2017

M. Torabi et al. / American Journal of Emergency Medicine 36 (2018) 816819 817

Center, in southeastern Iran. Multiple trauma patients with ISS N 16 and older than 18 years from 1 September 2015 to 1 September 2016 were enrolled in this study. Patients referred after an hour with a history of chronic lung, kidney, heart, or liver diseases or diabetes, anticoagulant medication consumption, drug or alcohol intoxication, or shock (except hemorrhagic shock) were excluded from the study. After evaluating pa- tients in accordance with advanced trauma life support (ATLS) guide- lines, up to 20 min after arrival blood samples were sent to the lab. BS on admission, BS 3 h after admission, ?BS, INR, serum lactate, and arte- rial blood gas (ABG) for evaluating base excess (BE) were measured. Also, according to the pre-set questionnaire, age, sex, vital signs, ISS, and RTS were calculated and recorded for all patients. Before completing the questionnaire, informed consent was taken orally from patients or their relatives. Finally, hospital mortality rate as an outcome for these patients was investigated by using medical records. Cox regression was used to analyze the effect of independent variables on the patients? outcomes. To assess the Proportional Hazard (PH) assumption, the time interaction test was applied. We selected variables with univariate P– value b 0.05 to offer them to the multifactorial model. However, because of the low number of mortality events and multicollinearity between in- dependent variables, we only constructed the multifactorial model using expert knowledge. One of the aims of this study was to explore the Predictive ability of ?BS in our multifactorial model; we only adjust- ed the effect of four established risk factors including heart rate, hemat- ocrit, INR, and BE. Statistical analysis was performed by SPSS 20 software and using survival analysis.


A total of 963 trauma patients were admitted within one year. Of them, 683 patients were excluded from the study and 280 were en- rolled (Fig. 1). A total of 18 patients (6.4%) died during hospitalization (Table 1). No departure from the PH assumption was seen. Using the Cox proportional hazard regression model, subjects who had high blood sugar levels on admission did not have higher mortality than pa- tients with normal glucose at baseline, but blood sugar 3 h after admis- sion and ?BS (BS 3 h after admission – BS on admission) was significantly correlated with hospital mortality. In particular, an in- crease of one unit in ?BS was associated with a 7% increase in the hazard of mortality. In addition, one unit increase in blood sugar 3 h after ad- mission corresponded to a 5% increase in the risk of death. In the univar- iate model, increased heart rate, INR, lactate, and ISS had aggressive effects, and high systolic and diastolic blood pressure, hematocrit, BE, and RTS had a protective effect on hospital mortality (Table 2). In the final model and multivariate analysis by using the backward conditional method, only the variables ?BS (P b 0.001), INR (P b 0.001), and heart

Fig. 1. Flow chart showing enrollment of patients.

Table 1

Patients’ characteristics

Variables Number (%)


Male 202 (72.1)

Female 78 (27.9)

Age(y),Mean +- SD

33.93 +- 12.54

ISS,Mean +- SD

26.47 +- 8.49

RTS,Mean +- SD

7.42 +- 0.95

Hospital Mortality

18 (6.4)

rate (P = 0.036) had a significant relationship with hospital mortality (Table 3).

As a complementary analysis, a decision tree was constructed. Here, all independent variables were offered to the tree. These models are at- tractive as they provide pictorial evidence that assists patient manage- ment. The decision tree demonstrated that for patients with a ?BS higher than 32.5, 93.8% had hospital mortality (15 out of 16). Among pa- tients with a ?BS lower than 32.5, only 3 out of 264 cases experienced death, corresponded to mortality rate of 1%.


There are several limitations in this study. One of the limitations is the fact that it is a unicenter study. Some patients or their relatives did not consent to enrollment, or some patients were referred with delay. Unfortunately, in some cases, there is missing data. Also, some tests–including lactate–were not measured during some hours of the day. This study was limited to patients with multiple trauma N 18 years with an ISS N 16 and without any history of diabetes, under- lying disease, or drug or alcohol use, which made it impossible to inves- tigate the other patients.


In this follow-up study on multiple trauma patients admitted to the emergency department, patients with higher difference between blood sugar 3 h after admission and on admission (?BS) were found to have higher hospital mortality. Thus, blood sugar levels on admission may not play a predictive role in the mortality of patients, but delay in mea- surement of blood sugar or blood sugar changes may play a predictive role. Also, other parameters such as prolonged INR and tachycardia can be helpful in predicting hospital mortality in multiple trauma patients.

In multiple trauma patients, stress-induced hyperglycemia could be helpful as an independent predictive of mortality. Hyperglycemia can result from the release of catecholamines and the physiological stress response to trauma [8]. Also, hepatic gluconeogenesis is activated after trauma and glucose metabolism is under the influence of counter-regu- latory hormonal responses and signals from the nervous system.

Table 2

Univariate Cox regression analysis and their association with mortality

Variables Mortality (Mean +- SD) OR (95% CI) P value




90.7 +- 16.88

132.8 +- 15.07

1.03 (1,1.05)


SBP(mm Hg)

121.02 +- 14.99

97.22 +- 18.23

0.97 (0.94, 0.99)


DBP(mm Hg)

72.78 +- 11.14

59.89 +- 14.03

0.96 (0.92, 0.99)



12.91 +- 1.72

10.25 +- 2.02

0.77 (0.61, 0.97)



38.84 +- 4.76

32.69 +- 4.70

0.88 (0.78, 0.97)


BS 0 (mg/dl)

101.4 +- 21.67

145.3 +- 17.02

0.99 (0.96, 1.03)


BS 3 (mg/dl)

108.86 +- 23.48

183.83 +- 11.67

1.05 (1.02, 1.09)



7.44 +- 6.51

38.50 +- 11.39

1.07 (1.03, 1.11)



1.09 +- 0.20

1.80 +- 1.55

2.24 (1.48, 3.38)


Lactate (mmol/lit)

22.63 +- 4.02

35.61 +- 6.7

1.12 (1.03, 1.21)


BE (mmol/lit)

-1.79 +- 3.38

-11.48 +- 4.37

0.87 (0.77, 0.99)



25.04 +- 6.60

47.22 +- 5.16

1.07 (0.99, 1.14)



7.61 +- 0.58

4.65 +- 1.03

0.38 (0.22, 0.61)


818 M. Torabi et al. / American Journal of Emergency Medicine 36 (2018) 816819

Table 3

Multivariate Cox regression analysis



OR (95% CI)

P value


1.09 (1.03, 1.11)



3.17 (1.48, 3.38)



1.04 (1, 1.05)


Following this, an impaired glucose uptake in heart and skeletal muscles occurs, which is called “diabetes of injury” [15]. But there is controversy in estimating when blood sugar levels should be measured as a prog- nostic factor. One study has shown that blood sugar on admission is in- effective in prediction of mortality in emergency patients [14]. However, some studies have shown that the measurement of blood sugar on admission has been effective not only in determining mortality [16], but also in prediction of morbidity in trauma patients [17]. Even blood suger control is recommended to improve outcomes of these pa- tients [11]. Laird et al. showed that early measurement of blood sugar during the first 24 to 48 h of admission was effective in determining mor- tality of trauma patients [18]. Wahl et al. showed that more severe hyper- glycemia causes higher hospital mortality in these patients [13]. Egi et al. reported that mean blood sugar is effective in predicting mortality in crit- ically ill patients [19]. Hence, there is controversy about the timing of blood sugar measurement to determine outcomes. Blood sugar measure- ment is a convenient, available, and cheap test, and monitoring changes in critically ill trauma patients can be helpful in determining hospital mortality. Therefore, blood sugar monitoring could be considered for pa- tients with multiple trauma [16]. Given that most deaths occur in the early hours of admission, our study examines blood sugar changes on ad- mission and 3 h after admission. It showed that measurement of blood sugar 3 h after admission and also ?BS (the difference between blood sugar 3 h after admission and on admission) have a predictive effect on hospital mortality and introduce it as an independent risk factor.

Coaglupathy is common immediately after trauma. There is a relation- ship between impaired coagulation tests and mortality in these patients [20,21] due to adrenalin release in blood which increases endothelial damage and activates fibrinolysis [22]. Also the activation of protein c pathway appears to be involved in its release. Therefore, coagulopathy immediately after trauma may indicate critical conditions such as shock or need for blood transfusion, and can be helpful in determining these pa- tients’ outcomes [23]. Our study demonstrated a significant relationship between prolonged INR on admission and hospital mortality.

In multiple trauma patients, tachycardia may be an ominous sign. Tachycardia plays a considerable role in determining trauma patients’ outcomes [24]. However, the absence of tachycardia cannot rule out a critical condition [25]. Viewing tachycardia in combination with the physiologic trauma score is helpful in determining trauma patients’ out- comes [26]. This study indicates that patients with increased heart rate on admission have higher hospital mortality.

This study found a significant relationship in the univariate model between hypotension, decreased hematocrit, increased lactate, de- creased BE, RTS, and increased ISS and hospital mortality, but this rela- tionship was not significant in the multivariate model. Several reasons may account for these findings: according to this study sample size, the event was low and the correlation between variables was high, sug- gesting that the role of these variables in determining the mortality of patients may be very faint. In the other words, the role of ?BS may be significant, and may diminish the role of other variables with which they have strong correlation. Therefore, another study is required with a larger sample size and the proper event in this respect.


Blood sugar measurement on admission in multiple trauma patients may be helpful in predicting their outcomes, but our study did not

confirm this. ?BS (difference between blood sugar 3 h after admission and on admission) can be a better indicator in determining outcomes for multiple trauma patients referred to the emergency department. We also found that coagulopathy and tachycardia are helpful in predicting hospital mortality. Further studies with larger sample sizes in this field are required.

Conflict of interest

None declared.


This study was supported by Clinical Research Center of Bahonar Ac- ademic Hospital [grant number 96000011], Kerman University of Med- ical Science, Kerman, IRAN.


  1. Pfeifer R, Teuben M, Andruszkow H, Barkatali BM, Pape H-C. Mortality patterns in patients with multiple trauma: a systematic review of autopsy studies. PloS one 2016;11(2):e0148844.
  2. Aghajani MH, Haddadi M, Saadat S. Epidemiological pattern of injuries in Iran; a na- tionwide review of seven million emergency department admissions. Emergency 2017;5(1).
  3. Saad S, Mohamed N, Moghazy A, Ellabban G, El-kamash S. Venous glucose, serum lactate and Base deficit as biochemical predictors of mortality in patients with polytrauma. Ulus Travma Acil Cerrahi Derg 2016;22(1):29-33.
  4. Champion H. Trauma scoring. Scand J Surg 2002;91(1):12-22.
  5. Chawda M, Hildebrand F, Pape HC, Giannoudis PV. Predicting outcome after multiple trauma: which scoring system? Injury 2004;35(4):347-58.
  6. Mazandarani PD, Heydari K, Hatamabadi H, Kashani P, Danesh YJ. Acute physiology and chronic health evaluation (APACHE) III score compared to trauma-injury sever- ity score (TRISS) in predicting mortality of trauma patients. Emergency 2016;4(2): 88.
  7. Lam S, Lingsma H, Van Beeck EF, Leenen L. Validation of a base deficit-based trauma prediction model and comparison with TRISS and ASCOT. Eur J Trauma Emerg Surg 2016;42(5):627-33.
  8. Kreutziger J, Rafetseder A, Mathis S, Wenzel V, El Attal R, Schmid S. Admission blood glucose predicted haemorrhagic shock in multiple trauma patients. Injury 2015; 46(1):15-20.
  9. Martin J, Blobner M, Busch R, Moser N, Kochs E, Luppa PB. Point-of-care testing on admission to the intensive care unit: lactate and glucose independently predict mor- tality. Clin Chem Lab Med 2013;51(2):405-12.
  10. Martin W, Galligan J, Simpson S, Greenaway T, Burgess J. Admission blood glucose predicts mortality and length of stay in patients admitted through the emergency department. Intern Med J 2015;45(9):916-24.
  11. Bochicchio GV, Joshi M, Bochicchio KM, Pyle A, Johnson SB, Meyer W, et al. Early hy- perGlycemic control is important in critically injured trauma patients. J Trauma Acute Care Surg 2007;63(6):1353-9.
  12. Kerby JD, Griffin RL, MacLennan P, Rue IIILW. Stress-induced hyperglycemia, not di- abetic hyperglycemia, is associated with higher mortality in trauma. Ann Surg 2012; 256(3):446-52.
  13. Wahl WL, Taddonio M, Maggio PM, Arbabi S, Hemmila MR. Mean glucose values predict trauma patient mortality. J Trauma Acute Care Surg 2008;65(1):42-8.
  14. Allwell-Brown E, Afuwape O, Ayandipo O, Alonge T. Correlation of the association of serum lactate, random blood sugar, and revised trauma score as predictors of out- come in hemodynamically unstable Abdominal emergencies. Niger J Clin Pract 2016;19(2):196-200.
  15. Kreutziger J, Schlaepfer J, Wenzel V, Constantinescu MA. The role of admission blood glucose in outcome prediction of surviving patients with multiple injuries. J Trauma Acute Care Surg 2009;67(4):704-8.
  16. Kreutziger J, Wenzel V, Kurz A, Constantinescu MA. Admission blood glucose is an independent predictive factor for hospital mortality in polytraumatised patients. In- tensive Care Med 2009;35(7):1234-9.
  17. Bochicchio GV, Sung J, Joshi M, Bochicchio K, Johnson SB, Meyer W, et al. Persistent hyperglycemia is predictive of outcome in critically ill trauma patients. J Trauma Acute Care Surg 2005;58(5):921-4.
  18. Laird AM, Miller PR, Kilgo PD, Meredith JW, Chang MC. Relationship of early hyper- glycemia to mortality in trauma patients. J Trauma Acute Care Surg 2004;56(5): 1058-62.
  19. Egi M, Bellomo R, Stachowski E, French CJ, Hart G. Variability of blood glucose con- centration and short-term mortality in critically ill patients. J Am Soc Anesthesiol 2006;105(2):244-52.
  20. MacLeod JB, Lynn M, McKenney MG, Cohn SM, Murtha M. Early coagulopathy pre- dicts mortality in trauma. J Trauma Acute Care Surg 2003;55(1):39-44.
  21. Maegele M, Lefering R, Yucel N, Tjardes T, Rixen D, Paffrath T, et al. Early coagulop- athy in multiple injury: an analysis from the German trauma registry on 8724 pa- tients. Injury 2007;38(3):298-304.

    M. Torabi et al. / American Journal of Emergency Medicine 36 (2018) 816819 819

    Johansson PI, Stensballe J, Rasmussen LS, Ostrowski SR. High circulating adrenaline levels at admission predict increased mortality after trauma. J Trauma Acute Care Surg 2012;72(2):428-36.

  22. Brohi K, Cohen MJ, Davenport RA. Acute coagulopathy of trauma: mechanism, iden- tification and effect. Curr Opin Crit Care 2007;13(6):680-5.
  23. Bruijns SR, Guly HR, Bouamra O, Lecky F, Lee WA. The value of traditional vital signs, Shock Index, and age-based markers in predicting trauma mortality. J Trauma Acute Care Surg 2013;74(6):1432-7.
  24. Victorino GP, Battistella FD, Wisner DH. Does tachycardia correlate with hypotension after trauma? J Am Coll Surg 2003;196(5):679-84.
  25. Kuhls DA, Malone DL, McCarter RJ, Napolitano LM. Predictors of mortality in adult trauma patients: the physiologic trauma score is equivalent to the trauma and injury severity score. J Am Coll Surg 2002;194(6):695-704.