Chinese Journal of Traumatology 2009; 12(2):1-4 .1.
Diagnostic accuracy of CT scan in abdominal blunt trauma
Javad Salimi*, Khadyjeh Bakhtavar, Mehdi Solimani, Patricia Khashayar, Ali Pasha Meysamie and Moosa Zargar
Objective: To evaluate the sensitivity and specificity were calculated in each case.
of CT-scan findings in patients with blunt abdominal trauma Results: CT scan had the highest sensitivity for detect-
admitted to the university hospital. ing the injuries to liver (100%) and spleen (86.6%). The speci-
Methods: All the patients with blunt abdominal trauma ficity of the method for detecting retroperitoneal hematoma
admitted at a tertiary teaching trauma center in Iran between (100%) and injuries to kidney (93.5%) was higher than other
2005 and 2007 were enrolled in this study. In the absence of organs. The accuracy of CT images to detect the injuries to
any clinical manifestations, the patients underwent a diag- spleen, liver, kidney and retroperitoneal hematoma was re-
nostic CT scan. Laparatomy was performed in those with ported to be 96.1%, 94.4%, 91.6% and 91.6% respectively.
positive CT results; while others were observed for 48 hours Conclusions: The findings of the present study reveal
and discharged in case no problem was reported; otherwise that CT scan could be considered as a good choice espe-
they underwent laparatomy. Information on patients’ de- cially for patients with blunt abdominal trauma in teaching
mographic data, mechanism of trauma, indication for CT hospitals where the radiologic academic staff is not present
scan, CT scan findings, results of laparotomy was gathered. in the hospital in the night shifts.
The sensitivity, specificity and accuracy of the CT-scan Key words: Wounds, nonpenetrating; Adbominal
images in regard with the organ injured were calculated. injuries; Tomography, X-ray computed; Diagnosis
The sensitivity, specificity and accuracy of the CT scan
Chin J Traumatol 2009; 12(2):1-4
lunt abdominal trauma may represent a life- plaints or external signs of abdominal injury on admis-
threatening condition requiring rapid diagnosis sion to hospital.3, 4 Abdominal examination is often com-
and treatment. The diagnosis of significant in- pounded by different factors including fractures of lower
tra-abdominal injury is a challenge in the management chest ribs, contusion and abrasions of the abdominal wall,
of such patients.1, 2 Failure to recognize and treat oc- presence of fractured lumbar vertebrae with retroperito-
cult hemorrhage is a common mistake. The clinical his- neal hematoma, and reduced level of consciousness.
tory and physical examination alone are usually Rapid assessment and appropriate treatment of poten-
unreliable, as nearly half the patients may have no com- tially life-threatening conditions is therefore essential.5
The optimal method of evaluating abdominal trauma re-
mains controversial. A combination of a sensitive and
Associate Professor of Surgery, Sina Trauma & Sur- specific screening test may be a safe and efficient ap-
gery Research Center, Sina Hospital, Medical Sciences/ proach to it. Diagnostic peritoneal lavage (DPL) is an
University of Tehran, Tehran, Iran (Salimi J) acceptable method for detecting the intra-peritoneal
Assistant Professor of Radiology, Sina Hospital, Medical
fluid. It has sensitivity greater than 90% for the detec-
Sciences/ University of Tehran, Tehran, Iran (Bakhtavar K)
tion of hemoperitoneum, but it is not specific and not
General Surgeons, Sina Hospital, Medical Sciences/
University of Tehran, Tehran, Iran (Solimani M) reliable for the assessment of retroperitoneal injuries.
General Practitioner, Research and Development In addition, like any invasive procedure, DPL carries
Center, Sina Hospital, Medical Sciences/ University of some risks. 6 The use of Focused Assessment
Tehran, Tehran, Iran (Khashayar P) Sonography for Trauma (FAST) for unstable patients
Assistant Professor of Social Medicine, Medical Sci-
with intra-abdominal hemorrhage rather than injury has
ences/ University of Tehran, Tehran, Iran (Meysamie AP)
dramatically changed the methods for diagnosing blunt
Professor of Surgery, Sina Trauma & Surgery Research
Center, Sina Hospital, Medical Sciences/ University of abdominal trauma.7 CT scan is now considered as a
Tehran, Tehran, Iran (Zargar M) rapid means of assessment for the early detection of
*Corresponding authors: Tel & Fax: 98-21-66735018, intra-peritoneal injury, which is not dependent on the
E-mail :firstname.lastname@example.org operator.8 As a result, this study was conducted to evalu-
.2. Chinese Journal of Traumatology 2009; 12(2):1-4
ate the specificity and sensitivity of CT-scan findings in injury if present, and was calculated by dividing the
patients with blunt abdominal trauma in order to verify number of true positive CT scan by the number of posi-
the accuracy of the method as a non-operator depen- tive laparotomies. The specificity was defined as the
dant method which could be used by residents in the ability of CT scan to rule out an injury none existed,
teaching hospitals where academic staff are absent and was calculated by dividing the number of true nega-
specially in the night shifts. tive CT scans by the number of patients without injury.
Finally, the accuracy of the CT scan was determined
METHODS by dividing the sum of true positive and true negative
CT scans by total number of CT scans performed. The
After approval by the Institutional Ethical Board data were entered in SPSS version 13. Student’s t-test
Committee, all the patients with blunt abdominal trauma and Chi-square analysis were used to analyze the
admitted at a tertiary teaching trauma center in Iran variables.
between 2005 and 2007 were evaluated. Those who
underwent a CT scan were enrolled in this study. Indi- RESULTS
cations for abdominal CT scan were 1) abdomen /pel-
vic pain/tenderness, 2) visible abdomen /pelvis trauma, A hundred patients were enrolled in this study, 87
3) history of homodynamic instability (SBP<80), 4) of whom were male and 13 were female. The peak age
hematuria, 5) abnormal pelvis x-ray, 6) alcohol incidence was 20-40 years (43%) and followed by pa-
intoxication, 7) unreliable examination secondary to tients aged more than 40 years (24%).Patients aged
distracting injuries, unresponsiveness and neurological 10-19 years accounted for 21% of the admissions.
injury, 8) decreased mental state (GCS<9), loss of
consciousness, 10) mechanism of injury (high speed/ Accident (63%) and fall (22%) were the most fre-
energy accident/extensive damage to vehicle/fatalities quent mechanism of trauma in these patients. The vital
involved), 11) intubations and 12) pelvic fractures. signs were revealed to be unstable at the admission
time in 77% of the patients.
CT scan images were reviewed by resident of
radiology, blind to the patients’ condition and unaware Thirty six of the patients underwent laparotomy;
of the objectives of the study. CT scan was considered sixty were discharged following close observation in 48
positive if one of the following criteria was met: 1) trauma hours and three were referred to other centers and one
of solid organs (liver, spleen, kidney, and pancreas), 2) patient died before operation.
trauma of small intestine, colon, mesentery, diaphragm
or bladder, 3) detectable amount of free fluid with hem- Totally 7 patients died during the study period, 6 of
orrhage in abdomen and pelvis. whom were operated upon. Laparotomy revealed spleen
rupture (4 cases), liver rupture (2 cases), duodenal and
Laparotomy was done for all the patients with any pancreatic rupture (2 cases), small intestine, kidney
evidences of intra-abdominal and pelvic injuries. In the and mesenteric rupture (3 cases) and retroperitoneal
absence of any clinical manifestations, the patients hematoma (2 cases). One death occurred in the emer-
underwent a diagnostic CT scan. Laparatomy was per- gency department while others happened in ICU. The
formed in those with positive CT results; while others mean time of death was reported to be 26 hours follow-
were observed for 48 hours and were discharged in case ing admission.
no problem was reported; otherwise they underwent
laparotomy. No specific findings were reported in the CT-scan of
29 of the patients. Among them, surgery was performed
Information on patients’ demographic data, mecha- in one case (3.4%). Two of these patients were referred
nism of trauma, indication for CT scan, CT scan findings, to other centers and 26 were discharged. Of the 71 pa-
results of laparatomies was gathered. The sensitivity, tients who had positive CT scan, 1 (1.4%) was referred
specificity and accuracy of the CT-scan image in re- and 1 died. Thirty-six underwent an operation, while 33
gard with the organ injured were calculated. The sensi- (44%) were discharged without being operated upon.
tivity was defined as the ability of CT scan to detect an
Chinese Journal of Traumatology 2009; 12(2):1-4 .3.
According to the results of the present study, CT DISCUSSION
scan had the highest sensitivity for detecting the inju-
ries to liver (100%), spleen (86.6%), kidney (80%) and In the past 20 years, CT has proved to be an excel-
stomach (80%). On the contrary, the specificity of the lent imaging modality for diagnosing and managing
CT-scan images for detecting retroperitoneal hematoma hemodynamically stable patients with abdominal injuries.9
(100%) and injuries to kidney (93.5%) and mesentery However, sonography is largely used as the preferred
(87.5%) was higher than other organs (Table 1). screening technique in many trauma centers. The value
of sonography in revealing an organ injury varies greatly
The accuracy of CT images to detect the injuries to with the location of the lesion.10, 11
spleen, liver, kidney and retroperitoneal hematoma were
reported to be 96.1%, 94.4%, 91.6% and 91.6% Many studies have stated that CT plays a critical
respectively. In general, the sensitivity, specificity and role in the identification of suspected bowel and me-
accuracy were calculated as 63.9%, 82.1% and 77.2% senteric injuries. It is worthy to note that the accuracy
respectively. of this method in diagnosing the very injuries has been
questioned.12 On the contrary, the present study docu-
The highest positive predictive value (PPV) was for mented CT scan to be more sensitive for liver, kidney
liver (83.3%) and spleen (81.2%) and overall 56.8%. and spleen injuries. According to these results, this
Similarly, liver had the highest negative predictive value diagnostic method has a low sensitivity in patients with
(NPV) (100%). Stomach (94.7%) and colon (91.3%) pancreas, mesentery and colon injury. In a study per-
were the organs with the next highest NPV. Overall formed in 2003, it was also shown that 13% of patients
NPV was 86%. Table 2 shows the PPV and NPV values. with perforated small bowel injury had a normal CT scan
Table 1. Sensitivity and specificity of CT scan preoperatively.13
Organs Sensitivity(%) Specificity(%)
Liver 100 66.6 Others have documented CT as the reference tech-
Kidney 80 93.5
nique for identification of retroperitoneal hematoma and
kidney injuries.14 Similarly, the present study supported
Spleen 86.6 85.7
the abovementioned results.
Pancreas 45.4 80
Stomach 80 58 Moreover, many researchers believe that CT scan
Duodenum 71.4 72.4 could be reserved for patients with negative findings on
Small intestine 75 75 sonography and clinical suspicious of injury. Marco et
Colon 50 81.2 al11 in their study have reported an acceptable accu-
racy for fluid and liver, splenic and renal injuries in pa-
Mesentery 25 87.5
tients with major trauma. In other words, they stated
Retroperitoneal hematoma 62.5 100
that CT had a high diagnostic accuracy for single le-
Table 2. Positive and negative predictive values of CT scan sions in both major and minor trauma. They suggested
Organs PPV(%) NPV(%) that the false negative reports in their study represented
Liver 83.3 100 lack of detail of whole body protocols and artifacts due
to patients’ condition. On the contrary, Livingston et
Kidney 66.6 86.1
al15 believe that patients with suspected abdominal in-
Spleen 81.2 90
jury should be observed in hospital for at least 24 hours
Pancreas 50 74.9
even when there is no signs of injury in CT scans
Stomach 23.5 94.7
Duodenum 38.4 91.3 It could be concluded that compared with diagnos-
Small intestine 60 85.7 tic methods such as ultrasonography, CT scan is es-
Colon 25 92.8 pecially for solid organs. On the other hand, it is not an
invasive method compared with DPL. In regard with the
Mesentery 71.4 48.2
high specificity of CT images in the absence of injury,
Retroperitoneal hematoma 100 90.3
no laparotomy is required when the CT scan images
.4. Chinese Journal of Traumatology 2009; 12(2):1-4
are reported to be normal. 7. Radwan MM, Abu-Zidan FM. Focused assessment
sonograph trauma (FAST) and CT scan in blunt abdominal trauma:
The findings of the present study reveal that CT scan surgeon’s perspective. Afr Health Sci 2006; 6(3): 187-190.
can be considered as a good choice especially for pa- 8. Otimo Y, Henmi H, Mashiko K, et al. New diagnostic
tients with blunt abdominal trauma in teaching hospi- peritoneal lavage criteria for diagnosis of intestinal injury. J Trauma
tals where the radiologic academic staff is not present 1998; 44(6): 991-999.
in the hospital in the night shifts. Many believe the 9. Wintermark M, Mouhsine E, Theumann N, et al. Thora-
sonography performed during these hours lack the ac- columbar spine fractures in patients who have sustained trauma:
ceptable accuracy because it is not performed by an depiction with multi-detector row CT. Radiology 2003; 227(3):
expert radiologist. Using CT scan, the residents are 681-689.
able to have the images with an acceptable sensitivity 10. Valentino M, Serra C, Zironi G, et al. Blunt abdominal
and specificity to report the abdominal injuries. trauma: emergency contrast-enhanced sonography for detection
of solid organ injuries. AJR Am J Roentgenol 2006; 186(5): 1361-
Acknowledgment: The authors of this article are indebted 1367.
to the Research and Development Center of Sina Hospital for its 11. Marco GG, Diego S, Giulio A, et al. Screening US and CT
cooperation. for blunt abdominal trauma: a retrospective study. Eur J Radiol
2005; 56(1): 97-101.
REFERENCES 12. Brofman N, Atri M, Hanson JM, et al. Evaluation of
bowel and mesenteric blunt trauma with multidetector CT.
1. Nagy KK, Roberts RR, Joseph KT, et al. Experience with Radiographics 2006; 26(4): 1119-1131.
over 2500 diagnostic peritoneal lavages. Injury 2000; 31(7): 479- 13. Fakhry SM, Watts DD, Luchette FA, et al. Current diag-
482. nostic approaches lack sensitivity in the diagnosis of perforated
2. Healey MA, Simons RK, Winchell RJ, et al. A prospective blunt small bowel injury: analysis from 275, 557 trauma admis-
evaluation of abdominal ultrasound in blunt trauma: is it useful? J sions from the EAST multi-institutional HVI trial. J Trauma 2003;
Trauma 1996; 40(6): 875-885. 54(2): 295-306.
3. Chiguito PE. Blunt abdominal injuries. Diagnostic perito- 14. Scaglione M, de Luto di Castelguidone E, Scialpi M, et al.
neal lavage, ultrasonography and computed tomography scanning Blunt trauma to the gastrointestinal tract and mesentery: is there a
injury 1996; 27(2): 117-124. role for helical CT in the decision making process? Eur J Radiol
4. Boulanger BR, Brenneman FD, McLellan BA, et al. A pro- 2004; 50(1): 67-73.
spective study of emergent abdominal sonography after blunt 15. Livingston DH, Lavery RF, Passannante MR, et al. Ad-
trauma. J Trauma 1995; 39(2): 325-330. mission or observation is not necessary after a negative abdominal
5. Brown CK, Dunn KA, Wilson K. Diagnostic evaluation of computed tomographic scan in patients with suspected blunt ab-
patients with blunt abdominal trauma: a decision analysis. Acad dominal trauma: results of a prospective, multi-institutional trial.
Emerg Med 2000; 7(4): 385-396. J Trauma 1998; 44(2): 273-280.
6. Blow O, Bassam D, Butler K, et al. Speed and efficiency in
the resuscitation of blunt trauma patients with multiple injuries: (Received August 25, 2008)
the advantage of diagnostic peritoneal lavage over abdominal com- Edited by SONG Shuang-ming
puterized tomography. J Trauma 1998; 44(2): 287-290.