Article, Radiology

FAST for blunt abdominal trauma: Correlation between positive findings and admission acid-base measurement

a b s t r a c t

Purpose: This study aimed to determine any association between positive findings in ultrasonography examina- tion and initial BD value with regard to diagnosis of intra-abdominal bleeding following Blunt abdominal trauma. Methods: A prospective, multi-center study of consecutive adult patients was performed from April to September 2015. Demographics, initial vital signs and arterial BD were evaluated with respect to presence of any association with intra-abdominal bleeding and in-hospital mortality. FAST study was performed to find intra-abdominal bleeding. receiver operating characteristic curves tested the ability of BD to identify patients with intra- Abdominal hemorrhage and probable mortality.

Results: A total of 879 patients were included in final analysis. The mean (SD) age was 36.68 (15.7) years and 714 patients (81.2%) were male. According to multivariable analysis, statistically significant association was observed between negative admission BD and both intra-abdominal bleeding (OR 3.48, 95% CI 2.06-5.88, p b 0.001) and in- hospital mortality (OR 1.55, 95% CI 1.49-1.63, p b 0.001). ROC curve analysis demonstrated sensitivity of 92.7% and specificity of 22.1% for the best cut-off value of BD (-8 mEq/L) to diagnose internal hemorrhage. Further, a cut-off value of -7 mEq/L demonstrated significant Predictive performance, 94.8% sensitivity and 53.6% spec- ificity for in-hospital mortality.

Conclusion: This study revealed that arterial BD is an early accessible important marker to identify intra-abdom- inal bleeding, as well as to predict overall in-hospital mortality in patients with blunt abdominal trauma.

Introduction

Mortality and morbidity caused by blunt abdominal trauma is one of the most important concern and common health problems for all age groups. The prevalence of intra-Abdominal injury after abdominal trau- ma has been reported to range from 12% to 15% [1,2]. Rapid determina- tion of the extent of injury and following ongoing blood loss is an important lifesaving step in treatment process, particularly for patients with unstable Hemodynamic state. Failure to early recognition of occult bleeding leads to a prolonged hypoperfusion condition and increased risk for multi-organ failure [3,4].

focused assessment with sonography for trauma has been established as a standard and reliable technique over the past decades

* Corresponding author at: School of Medicine, Shahid Beheshti University of Medical Sciences, Evin, Tehran, Iran.

E-mail addresses: [email protected], [email protected] (S. Asadollahi).

for detecting intra-abdominal hemorrhage during primary workup of trauma patients [5,6]. This traditional method of ultrasonography has gained considerable clinical application because of its ease of perfor- mance, availability, inexpensive and noninvasive nature. This imaging modality lacks ionizing feature and can be serially performed. The goal of FAST study in an emergency setting is to assist physicians to promptly detect suspected life-threatening states after an abdominal Traumatic event [7]. Thus, it is generally accepted in trauma management as an ad- ditional modality of physical examination [5]. Nevertheless, FAST is re- stricted to identifying free fluid present in intraperitoneal or pericardial spaces, which is supposed to be blood in the setting of trau- ma. Therefore, FAST results should be analyzed in parallel with other clinical and laboratory findings as part of a Diagnostic approach for the affected patients [8].

Among Laboratory measurements, base deficit (BD) has been broad- ly used in clinical practice as a sensitive indicator of reduced tissue per- fusion (hypoperfusion and hypovolemia) in critically ill trauma patients [9]. Impaired Tissue oxygenation causes conversion to anaerobic

metabolism and metabolic acidosis. A great degree of negative BD as a marker of metabolic imbalance and poor tissue perfusion can predict final related complications and mortality in the patients [10].

Previous investigations have evaluated the use of BD in blunt trauma

[3,11-14]; however, to our knowledge, this study is the first study to as- sess the correlation between positive findings in FAST examination and BD measurement with regard to diagnosis of intra-abdominal bleeding following blunt abdominal trauma. We also aimed to analyze the pre- dictive value of these factors and clinical variables on overall in-hospital mortality of the affected patients.

Methods

Study design and setting

We performed a prospective multicenter study at four urban, uni- versity-affiliated level 1 trauma centers. Our academic institutions are referral teaching hospitals with approximately 35 000 emergency de- partment visits per year. The study was carried out according to the principles of the Declaration of Helsinki [15]. The study was reviewed and approved by the local review board or ethics committee at each in- vestigative site, and the final study protocol was approved by the Ethics Committee of Shahid Beheshti University of Medical Sciences (Iran). Written informed consent was obtained from all participants.

Study population

Between April and September 2015, a consecutive sample of adult individuals with diagnosis of blunt abdominal trauma was screened for the study. Patients were excluded from the final analysis if they sustained head trauma, isolated intra-abdominal organ injury (organ injury without hemoperitoneum), or comorbid medical conditions (his- tory of chronic renal failure, diabetes mellitus, and current episode of di- arrhea, alcohol intoxication and septic state). We also excluded pregnant patients and who were transferred from other facilities or were dead on presentation.

Measurements

All trauma patients underwent care management according to Ad- vanced Trauma Life Support (ATLS) principles [16]. The emergency phy- sicians who performed the Primary evaluation completed a standardized data forms on admission to the trauma emergency depart- ment. The following data were recorded for all participants: demo- graphics, baseline Revised Trauma Score [17] and hemodynamic signs (heart rate [HR] and systolic blood pressure [SBP]), history of inju- ry mechanism and arterial serum BD.

An arterial blood sample was collected from each patient upon pre- sentation to the emergency department. Arterial BD was measured by use of (AVL Compact 3 Blood gas analyzer, Roche Diagnostics, Mann- heim, Germany). Serum BD was defined as the required amount of base, in milliequivalents per liter (millimoles per liter), to restore the pH of one liter of whole arterial blood to 7.4, with the sample fully oxy- gen saturated at 37 ?C with a PaCO2 of 40 mm Hg [18]. All patients were then followed until discharge or in-hospital death.

Technique

Bedside FAST study was performed by board-certified attending physicians. The emergency physicians had at least 10 years of clinical experience with emergency ultrasound. For re-evaluation and quality assurance, images were collected and saved for additional review. Twenty-four-hour sonography coverage was provided in our study and the expert sonographers were always available to perform the FAST examination. All participants underwent ultrasonography exam minutes of presentation concurrently with the Initial resuscitation.

The examinations had been conducted by use of phased-array or convex linear 2.5-5.0-MHz transducers, depending on patient size. The FAST scans were obtained while the patients were in the supine position to assess four standard anatomic areas for the presence/absence of any amount of free fluid. The following four parts were evaluated; the hepatorenal space (Morison’s pouch), splenorenal space, pericardial area and the pelvic [19,20]. With regard to our study purpose, a specific evaluation of the intra-abdominal organs was not performed to diag- nose or rule out visceral injury. However, limited sonographic scans of the solid organs were obtained if there was related free fluid. The re- search assistants, administering physicians and sonographers were all blinded to the study purposes.

End points

The primary endpoint for this study was to find any correlation be- tween arterial BD and FAST examination to reveal the presence of intra-abdominal hemorrhage in patients with blunt abdominal trauma. The secondary end point studied the predictive value of admission clin- ical factors, FAST findings and arterial BD for in-hospital mortality in those patients.

Statistical analysis

A descriptive analysis of demographic and clinical characteristics of patients presenting with blunt abdominal injury was performed. Pa- tients were divided into two groups: FAST-negative (FAST-: Patients without intraperitoneal fluid) and FAST-positive (FAST+: Patients with any amount of intraperitoneal fluid). Comparisons between the study groups for continuous data were explored by univariable analysis using Student’s t-test or nonparametric test (Mann Whitney U), and by chi-squared analysis in the case of dichotomous variables. Continuous data were reported as means +- standard deviation. Categorical vari- ables were presented as percentages. Multivariable logistic regression modeling was performed based on stepwise selection to determine sig- nificant predictors of intraperitoneal bleeding and in-hospital mortality. Adjusted odds ratios (OR) and corresponding two-sided 95% confidence interval (CI) associated with each demographic and baseline clinical variables and BD level were calculated by forward logistic regression technique. All test characteristics; sensitivity, specificity, negative pre- dictive value (NPV) and positive predictive value (PPV) were also calcu- lated for admission arterial BD. Additional analyses were completed to determine cutoff points for initial measurements of BD discriminating between survivors and nonsurvivors. For this evaluation, a receiver op- erating characteristic (ROC) curve was created and area under the curve (AUC) was estimated. Youden index [J = sensitivity + specificity – 1] was also calculated to determine the best cutoff points of BD. All P– values were based on two-tailed tests and a probability value <= 0.05 were considered to be statistically significant. All statistical analyzes were performed using SPSS version 22.0 (SPSS Inc., Chicago, Ill).

Results

Characteristics of study subjects

A total of 1279 patients with abdominal trauma were admitted to the ED. Of all, 879 patients met the inclusion criteria and were analyzed in the study. Four hundred patients were excluded for various reasons (n): penetrating trauma (98), head trauma (76), isolated visceral injury (74), associated medical conditions (73), transfer from another center (41), and pregnancy (13). Twenty-five patients refused consent for par- ticipation in the investigation. The study aimed to comply with the Stan- dards for Reporting of Diagnostic Accuracy (STARD) initiative and the flow of participants through the study is shown in Fig. 1.

The mean age was 36.68 +- 15.7 years with 81.2% of the included pa- tients being male. The causes of blunt injury were as follows: motor

Fig. 1. STARD flow diagram: diagnostic accuracy of admission arterial base deficit determination in predicting 30-day mortality after blunt abdominal trauma. STARD The Standards for Reporting of Diagnostic Accuracy, BD base deficit.

vehicle collision (56.0%), Motorcycle collisions (20.5%), fall from height (18.4%) and assault (5.1%). Seven hundred eighty-six patients (89.4%) arrived to the ED by ambulance and 93 patients (10.6%) arrived by pri- vate transportation. The time interval between trauma and ED admis- sion was 16.24 +- 10.6 min.

Main results

FAST examination

One hundred sixty-eight subjects (19.1%) had positive FAST exami- nation on presentation. The study population was divided into the main study outcome variable, FAST+ vs. FAST-. No correlation was found regarding age at injury, sex, duration from injury to hospital ad- mission and type of transport between the two study groups. The

results from the univariable analysis of demographics and baseline clin- ical characteristics are shown in Table 1.

Admission physiologic parameters were statistically different be- tween the two groups. Mean SBP (97.56 vs. 120.61 mm Hg) and HR on arrival (105.26 vs. 94.07 beats per minute) were different between the FAST+ and FAST- patients (p b 0.001). Additional analysis evaluat- ed the relationship between arterial BD and the presence of posttrau- matic intra-abdominal hemorrhage according to positive finding on initial FAST study. Accordingly, mean BD was significantly larger in FAST+ group compared with FAST- group (-4.66 vs. -2.19 mEq/L, p b 0.001).

Multivariable analysis of potential FAST+ predictors after adjust- ment for all indicators revealed that low admission SBP (OR 0.89, 95% CI 0.87-0.91, p b 0.001) could predict intra-abdominal bleeding follow- ing blunt trauma. Moreover, we observed a significant predictive value

Table 1

Demographics, clinical and laboratory characteristics of patients with blunt abdominal trauma by FAST findings.

and negative arterial BD (OR 1.742, 95% CI 1.692-1.794, p b 0.001) proved to be outcome predictors. No correlation was found between positive FAST finding and in-hospital mortality of the included patients.

Demographic and clinical characteristics FAST+

FAST-

P-value

The ROC curve analysis demonstrated that posttraumatic arterial BD

	N = 168	N = 711	is a significant discriminator of in-hospital mortality with an AUC of 0.83
Age, y, mean (SD)	38.17	36.31	0.170 (95% CI 0.79-0.86, p ? 0.001). The optimal cutoff point for this metabolic
	(16.8)	(15.4)	parameter was -7 mEq/L providing a sensitivity of 94.8% with specific-
Sex			ity of 53.6% (Fig. 3).
Male	129 (76.8)	585 (82.3)	0.714
Female	39 (23.2)	126 (17.7)
Time duration from injury to admission,	21.28	15.69	_b0.001? 4. Discussion
minutes, mean (SD)	(12.7)	(17.1)
Type of transport			Traumatic events are considered as one of the most common causes
Ambulance	153 (91.1)	633 (89.0)	^0.439 of death and significant economic and global health problem. The third
Private transportation Mechanism of trauma	15 (8.9)	78 (11.0)	most frequently injured body region and injury type is the abdominal

0.652

Motor collision	93 (55.3)	399 (56.1)	0.858	trauma [21]. Biomarkers have been shown to be of great utility in help-
Motorcycle collision	21 (12.5)	159 (22.4)	0.004?	ing the physician to determine the presence of blood loss and hypoper-
Fall	33 (19.6)	129 (18.1)		fusion in affected patients [22]. BD is believed to be an accepted and

Assault 21 (12.5) 24 (3.4) b0.001?

Initial clinical and laboratory parameters

Revised trauma score, mean (SD) 7.21 (0.6) 7.57 (0.4) b0.001?

validated marker, which is readily available by venous or arterial draws. The marker represents the combined sum of lactic acid and all

Heart rate, beats per minute, mean (SD) 105.26

(23.4)

Systolic blood pressure, mmHg, mean (SD) 97.56

(14.8)

94.07

(22.0)

120.61

(20.2)

b0.001? b0.001?

other metabolic acids released during posttraumatic hemorrhage and shock, as such, is recognized as a reliable marker of tissue malperfusion [23]. Beside this marker, the presence of free intraperitoneal fluid,

Capillary refill >= 2 s 84 (50.0) 216 (30.3) b0.001?

which mostly is considered as blood can be easily diagnosed by abdom-

Arterial base deficit, mEq/L, mean (SD) -4.66

(4.5)

-2.19

(4.4)

b0.001?

inal ultrasound [7]. In the current investigation designed as a prospec- tive study with trauma patients, we aimed to find any correlation

FAST Focussed Assessment with Sonography for Trauma, SD Standard Deviation, mEq/L milliequivalents per liter.

Numbers in parentheses are percentage unless otherwise indicated.

* P b 0.05.

for decreasing BD in terms of determining intra-abdominal bleeding on FAST exam (OR 3.48, 95% CI 2.06-5.88, p b 0.001) (Table 2).

We performed ROC curve analysis to assess the diagnostic perfor- mance of BD in detecting FAST+ finding (Fig. 2). The AUC for admission BD was 0.67 (95% CI 0.63-0.72). This area was considerably different from that of the unity line (p ? 0.001). We also calculated the optimal cutoff point for BD value demonstrating the highest sensitivity and specificity for discriminating between patients with FAST+ and FAST- exam. The cutoff point of -8 mEq/L was obtained with sensitiv- ity, specificity, PPV and NPV of 92.7%, 22.1%, 21.9%, and 92.7%, respectively.

In-hospital mortality

The rate of in-hospital mortality for the entire study population was 21.8% (n = 192). When comparing the group of subjects discharged alive to those who died during the hospital stay, nonsurvivors had lower admission SBP (p b 0.001) and higher HR (p b 0.001) (Table 3). The patients of nonsurvived group were also older than those who lived (39.15 vs. 35.95 years, p = 0.013). Nonsurvivors also had worse initial arterial BD (- 7.24 vs. – 1.41 mEq/L, p b 0.001) compared to the survived group. Multivariable analysis also demonstrated similar re- sults as shown in Table 4. Accordingly, admission SBP (OR 0.028, 95% CI 0.009-0.041, p = 0.001), HR (OR 1.040, 95% CI 1.026-1.048, p b 0.001)

Table 2

Multivariable logistic regression model predicting intra-abdominal hemorrhage in pa- tients with blunt abdominal trauma.

Predictor Adjusted OR^a 95% CI P-value Trauma factors

Assault 5.391 1.927-15.081 0.001

Initial clinical and laboratory factors

Systolic blood pressure 0.892 0.873-0.911 b0.001?

Arterial base deficit 3.484 2.063-5.885 b0.001?

OR odds ratio, CI confidence interval.

^a Odds ratios after adjustment for all covariates listed in the table.

* P b 0.05.

between admission BD and positive FAST exam relating to the presence

of intra-abdominal fluid. Accordingly, a considerable positive associa- tion was found between serum BD and positive FAST study in our pa- tient population. As the serum BD decreases more abnormally, the probability of intra-abdominal bleeding increases. In line with our find- ings, a BD of -8 mEq/L or lower offered a probability of N 92% for the presence of intra-abdominal hemorrhage. Further, decreased SBP and increased HR suggested that there is a significantly enhanced risk for intra-abdominal hemorrhage. Thus, decreased BD and/or abnormal he- modynamic parameters were highly sensitive for identifying cases with internal bleeding, which is in accord with earlier reports. Previously, Mofidi et al. [9] prospectively studied 400 patients with blunt abdomi- nal trauma and demonstrated that BD had a remarkable diagnostic per- formance in identifying intra-abdominal bleeding. The investigators reported that a BD <=-6 mEq/L had sensitivity of almost 88.2% for de- tecting free intraperitoneal fluid in the enrolled subjects. Earlier, Davis et al. [24] reviewed the records of 209 trauma patients with serial arte- rial analysis. They found that the required volume for fluid resuscitation was greater in the groups with more severe BD (<=-6 mEq/L). Their re- port suggested that the initial BD is a reliable indicator of the extent of volume deficiency in patients with intracavitary (thorax and abdomen) hemorrhage.

We found the optimal cutoff value for initial BD to be associated with

an acceptable sensitivity of 92.7%, but a low specificity of 22.1%. A possi- ble explanation for the low specificity may be high rates of false positive conditions and concomitant abdominal injuries which might not be di- agnosed by FAST examination. This is due to the fact that ultrasonogra- phy is not likely to detect most occult injuries such as retroperitoneal or pelvic injuries [25]. Free fluid detected by FAST was only considered as a true positive finding in our investigation. We also did not exclude pa- tients with coexistent Intra-abdominal injuries such as solid organ (liver, spleen, or kidney) injuries. With regards to our findings, a normal BD does not exclude intra-abdominal injury in blunt trauma patients, but the presence of a BD equal or less than -8 mEq/L should be consid- ered a strong indication for abdominal evaluation.

Decreased admission BD has consistently been shown to be an inde- pendent predictor of mortality in the trauma patient population. Several studies have revealed this marker at arrival to be a reliable predictor of mortality. Our results support prior data reported by Ouellet et al. [14] who reviewed 2269 consecutive adult trauma admissions. The overall in-hospital mortality was 17.1%. They found lower SBP (p = 0.0004)

Fig. 2. Receiver operating characteristic curve for initial arterial base deficit as a discriminant between FAST+ and FAST- patients. FAST Focussed assessment with sonography for trauma, AUC area under the curve, CI confidence interval, mEq/L milliequivalents per liter, NPV negative predictive value, PPV positive predictive value.

and lower arterial pH at admission (p b 0.0001) to be associated with higher rate of in-hospital mortality. The investigators also showed that non-survivors had worse median BD compared to survived group (-10 vs. – 5 mEq/L, p b 0.0001). This correlation between initial BD value and mortality was also confirmed in a prospective, multi-center study by Rixen et al. [26]. They included 1264 Multiple trauma patients to assess the prognostic significance of post-trauma BD documented on

hospital admission. The patients were stratified into five categories ac- cording to BD value. They reported that in-hospital mortality was 14% for patients with BD <=- 2 mEq/L. Generally, it is concluded that more negative BD implies worse injury and higher mortality.

Despite the practical applicability of FAST examination, lack of clarity and sufficient evidence regarding any definite contribution to patient survival still remain. Previous literature mainly focused on the efficacy

Table 3

Demographics, clinical and laboratory characteristics of patients with blunt abdominal trauma by in-hospital mortality.

Demographic and clinical characteristics	In-hospital mortality Survivors	Non-survivors	P-value
	N = 687	N = 192
Age, y, mean (SD)	35.95 (15.3)	39.15 (17.0)	0.013?
Sex
Male	576 (83.8)	138 (71.8) b 0.001?
Female	111 (16.1)	54 (28.1)
Time duration from injury to admission, minutes, mean (SD)	15.19 (11.5)	20.24 (10.3)	0.001?
Type of transport
Ambulance	609 (88.6)	177 (92.1)	0.158
Private transportation	78 (11.3)	15 (7.8)
Mechanism of trauma Motor collision	366 (53.3)	126 (65.6)	0.002?
Motorcycle collision	162 (23.6)	18 (9.4)	b 0.001?
Fall	120 (17.5)	42 (21.9)	0.164
Assault	39 (5.6)	6 (3.1)	0.498
Initial clinical and laboratory parameters Revised trauma score, mean (SD)	7.61 (0.4)	7.12 (0.5)	b 0.001?
Heart rate, beats per minute, mean (SD)	91.74 (20.1)	111.82 (24.2)	b 0.001?
Systolic blood pressure, mmHg, mean (SD)	115.95 (17.4)	100.28 (23.9)	b 0.001?
Capillary refill >= 2 s	216 (31.4)	84 (43.7)	0.001?
Arterial base deficit, mEq/L, mean (SD)	-1.41 (3.5)	-7.24 (5.1)	b 0.001?
Positive FAST^a	111 (16.1)	57 (29.7)	b 0.001?

SD Standard Deviation, mEq/L milliequivalents per liter, FAST Focussed Assessment with Sonography for Trauma. Numbers in parentheses are percentage unless otherwise indicated.

^a Defined as the presence of any amount of free fluid in any of the hepatorenal, splenorenal, pericardial and pelvic windows.

* P b 0.05.

Table 4

Multivariable logistic regression model predicting in-hospital mortality in patients with blunt abdominal trauma.

Predictor Adjusted OR^a 95% CI P-value Demographics

Age 1.027 1.004-1.050 0.020?

patients with debilitating clinical state (positive FAST and unstable he- modynamics) are prone to in-hospital mortality.

Although this study reported important findings with regard to the reliability of arterial BD in trauma patients, some limitations exist and need to be considered. First, our findings are only applicable to adult pa- tients with isolated blunt abdominal trauma, but not the pediatric pop-

Trauma factors

* ulation. Because reference levels of BD for healthy children are different

Time duration from injury to admission 1.031 1.014-1.048 b0.001 Motor collision 1.246 1.128-1.474 b0.001?

Initial clinical and laboratory factors

Revised trauma score, mean (SD)	0.154	0.084-0.283	b0.001?
Heart rate	1.029	1.012-1.046	0.001?
Systolic blood pressure	0.035	0.021-0.050	b0.001?
Arterial base deficit	1.557	1.491-1.633	b0.001?

OR odds ratio, CI confidence interval.

^a Odds ratios after adjustment for all covariates listed in the table.

* P b 0.05.

of this method as a diagnostic tool and in changing the management of patients [27]. Our analysis confirmed that positive FAST examination at arrival correlated significantly with mortality by univariate method. A positive FAST exam should make the ED physicians and trauma sur- geons suspicious for serious injury requiring laparatomy. Accordingly,

from adult population [28,29]. Second, BD is also altered in several con- ditions leading to acid-base imbalance. Although those patients with concomitant abdominal injuries and medical conditions were excluded, this aspect has not been examined separately and directly in this study and should be considered in future studies. Categorizing patients with blunt abdominal trauma by the amount of free fluid and evaluating the relationship with BD level is an important method of assessment, which was not considered in our investigation. Moreover, additional analysis focusing on the results of computed tomography and laparatomy of critical patients in relation to BD value was not included. However, this specific area was not the main purpose of the current study and warrant future investigation.

Based on this study’s findings, admission BD had the diagnostic abil- ity to determine the presence of free intra-abdominal bleeding in pa- tients with blunt abdominal trauma. In addition, we have determined that BD is a more reliable predictor of in-hospital mortality compared to positive FAST result. The finding of low level of serum bicarbonate

Fig. 3. Receiver operating characteristic curve for initial arterial base deficit as a discriminant between survivors and non-survivors (in-hospital mortality). AUC area under the curve, CI confidence interval, NPV negative predictive value, PPV positive predictive value, mEq/L milliequivalents per liter.

and larger BD has been also reported in several studies as a predictor of mortality in the trauma patient. It is of interest in this context that the association between BD and mortality may not be completely linear as the patients with severe excess of serum bicarbonate and significant BD (e.g. chronic obstructive pulmonary disease) also may have a higher mortality. To our knowledge, this is the first large study to find the asso- ciation between FAST exam and admission BD in screening the patients with internal hemorrhage following blunt abdominal trauma. This ex- perience serves as a reminder of an important point. Considering initial arterial findings such as BD as a marker of blood loss after trauma in ad- dition to FAST examination is highlighted. Therefore, using these indica- tors and parameters of clinical status during evaluation and management of critical trauma patients is highly suggested.

Conflict of interest

Authors disclose no conflict of interest.

Acknowledgements

This study was supported by a grant (#SB-3318) from Shahid Beheshti University of Medical Sciences (Tehran, Iran).

References

AJEM

FAST for blunt abdominal trauma: Correlation between positive findings and admission acid-base measurement

a b s t r a c t

Introduction

Methods