Background: The Focused Assessment with Sonography for Trauma (FAST) examination using conventional ultrasound has limited utility for detecting solid organ injury. Therefore, this systematic review and meta-analysis compares the performance of contrast-enhanced ultrasound (CEUS) to conventional ultrasound when used as the initial assessment for abdominal trauma prior to computed tomography (CT) imaging. Methods: A systematic literature search of major databases was conducted of human studies investigating the diagnostic accuracy of conventional ultrasound and CEUS occurring prior to CT imaging for abdominal trauma. The study followed the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) statement. The quality of studies was evaluated using the QUADAS-2 (Quality Assessment of Diagnostic Accuracy Studies 2) tool. Paired pooled sensitivity and specificity between conventional ultrasound and CEUS were compared using data extracted from the eligible studies. Diagnostic odds ratio, number needed to diagnose values, and likelihood ratios were also determined. Results: Ten studies were included. More than half of the included studies demonstrated low risk of bias. Using McNemar's test to assess for paired binary observations, we found that CEUS had statistically higher sensitivity (0.933 vs. 0.559; two-tailed, P < 0.001) and specificity (0.995 vs. 0.979; two-tailed, P < 0.001) than conventional ultrasound in the setting of abdominal trauma, respectively. When divided into particular findings of clinical interest, CEUS had statistically higher sensitivity than conventional ultrasound in screening for active bleeding and injuries to all abdominal solid organs. CEUS also had superior diagnostic odds ratios, number needed to diagnose values, and likelihood ratios than conventional ultrasound. Conclusion: The diagnostic value of CEUS was higher than that of conventional ultrasound for differentiating traumatic abdominal injuries when used as the initial assessment in the emergency department.
Keywords: Contrast-enhanced ultrasound; CEUS; Trauma; Emergency medicine
Supplementary Information The online version contains supplementary material available at https://doi.org/10.1186/s12873-023-00771-4.
The Focused Assessment with Sonography for Trauma (FAST) examination using conventional ultrasound has been an integral part of the evaluation of trauma patients for over 20 years [[
To date, there are numerous studies that investigated whether the use of contrast-enhanced agents with ultrasound, also known as contrast-enhanced ultrasound (CEUS), improved the sensitivity and specificity for detecting abdominal traumatic lesions. Utilizing small intravenous boluses of inert gas-filled microbubbles with a phospholipid shell, acute solid organ lesions can be depicted in real time through all the vascular phases [[
Currently, there are multiple studies that report the benefits of using CEUS for the identification of abdominal injuries following trauma [[
We performed a systematic review and meta-analysis of studies that compared the performance of CEUS to conventional ultrasound when used as the initial assessment for abdominal trauma prior to CT imaging. This study followed the guidelines in the "Cochrane Handbook for Systematic Reviews of Diagnostic Test Accuracy" [[
Studies were considered eligible for this systematic review and meta-analysis if they fulfilled the following criteria: 1) human studies investigating the diagnostic accuracy of conventional ultrasound, CEUS, and CT for abdominal trauma, with the reference standard clearly defined; 2) both conventional ultrasound and CEUS diagnostic tests must have been performed prior to assessments by CT scan; 3) all diagnostic tests must have been completed within 3 hours of the patient's presentation to the emergency department; and 4) both prospective and retrospective studies were eligible. We excluded studies when they met one of the following criteria: 1) experimentation with animals; 2) reviews, commentary, and case reports; and 3) non-traumatic conditions.
A standardized search was done in PubMed, OVID MEDLINE, Embase, and Web of Science, using the following search terms: ("contrast enhanced ultrasound" OR "contrast enhanced sonography" OR CES OR CEUS) AND trauma. The search was done on January 1, 2022 with no language restrictions.
Two authors screened and selected studies independently based on the criteria described above, with disagreements resolved by consensus together with a third author. Studies identified from different databases were de-duplicated after screening. Articles that passed the initial screening were reviewed for the full text. Studies with data available on true negative, true positive, false negative, and false positive results were included for the meta-analysis. This study followed the Preferred Reporting Items for a Systematic Review and Meta-analysis of Diagnostic Test Accuracy (PRISMA) [[
For the included studies, individual data of sample size, number of true negative, true positive, false negative, and false positive results per imaging modality in each study were extracted. If only partial information was available, outcomes were calculated using results from CT imaging as reference standard. Based upon how findings were presented in the included studies, datasets were nominally categorized according to identified injured organs or findings. These organs included the 1) liver, 2) kidneys and adrenals, 3) spleen, and 4) pancreas. The findings included 5) abdominal free fluid, 6) active bleeding, 7) any solid organ injury, or 8) a combination of solid organ injury, abdominal free fluid and/or active bleeding (composite). Dichotomous findings of positive or negative for each were compared against CT findings as the gold standard.
Information on location, study design, setting, mechanism and site of injury, contrast and dose, technical aspects of the ultrasound machine, and characteristics of the sonographer and retrospective reviewer of ultrasound clips were also retrieved.
The quality of each study was appraised with the Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS-2) tool, structured into patient selection, index test, reference standard, and flow and timing, structured as a list of 13 items and and qualified as "yes," "no," or "unclear" for an individual study. Each domain was evaluated for the risk of bias and the first three in terms of applicability. The answers were used to judge whether the risk of bias and concern for the applicability of the research is low, high, or unclear. Two reviewers independently judged the quality of each study, with disagreements resolved by consensus with additional input from a third.
Subgroup analyses for all categories of abdominal injury were conducted using only studies that provided paired data for both conventional ultrasound and CEUS. Articles that did not provide data for both modalities were excluded from this portion of the study. Sensitivity, specificity, diagnostic odds ratios, number needed to diagnose, and likelihood ratios with the associated 95% confidence intervals were calculated from true negative, true positive, false negative, and false positive cases with a 0.5 continuity correction for zero events. Equations are found in the Supplemental Material. Since the patients underwent both conventional ultrasound and CEUS consecutively prior to obtaining CT imaging, McNemar's chi-square test was chosen to compare paired pooled sensitivity and specificity. All P values were two-sided, and any P value < 0.05 was considered statistically significant. Forest plots were generated for individual studies according to solid organ injury or finding along with summary estimates for each category and overall estimates. Heterogeneity was assessed, whereby P < 0.05 for Cochran's Q and Higgin's I
Eight hundred sixty-eight studies were identified in our search. After assessing the titles and abstracts, 57 full texts were screened, as shown in Fig. 1. On the basis of our selection criteria, 47 of those studies were excluded, two of which consisted of CEUS studies with no conventional ultrasound diagnostic tests performed [[
Graph: Fig. 1 Study profile
All studies were published after the year 2000 and four studies were published after 2010 [[
The most common organ investigated was the liver [[
Table 1 Characteristics of included studies
Author Study design, setting N Mechanism of trauma; injury Mean Age Contrast, dose; protocol (scanning time) Transducer; mechanical index Sonographer; ultrasound image interpreter Catalano [ Sing. Pro., ED. 120 Blunt; Spleen Abd. Fluid Act. Bleed 28 Sonovue, 4.8 mL; Left upper quad. (3–4 min total) 2.5–5 MHz or 5.5–10 MHz curved; 0.06–0.07 Radiologist on duty; Two radiologists, third for disagreements Catalano [ Sing. Retro., ED. 87 Blunt or Pen.; Liver Abd. fluid 33 SonoVue, 2.4–4.8 mL; 1. Left kidney 2. Right kidney & liver 3. Spleen (1–5 min total) 3.5 MHz curved; 0.06–0.08 Radiologist on duty; Two radiologists, third for disagreements Catalano [ Mult. Pro., ED. 156 Blunt or Pen.; Liver Kidneys & adrenal gl. Spleen Comp. 39 SonoVue, 2.4 mL (two doses); 1. Right sided organs (right kidney, adrenal, liver) (1–3 min) 2. Left sided organs (left kidney, adrenal, pancreas, spleen) (3–4 min) 2.5 MHz or 1–5 MHz curved; 0.05–0.10 Radiologist on duty; Same Lv [ Mult. Retro., ED. 22 Blunt; Pancreas 29 SonoVue, 2.5 mL/kg; Liver & spleen (NA min) 2–4.5 MHz or 1–5 MHz curved; Low Two ultrasound specialists (5 years exp.); Same Menichini [ Sing. Retro., ED. 73 Blunt; Liver Kidneys & adrenal gl. Spleen Act. Bleed Solid org. Comp. 9 SonoVue, 1.2 mL (two doses); 1. Right & left kidneys 2. Liver 3. Pancreas 4. Spleen (3 min total) Curved and linear; Low Three radiologists (10+ years exp.); Same Miele [ Sing. Retro., ED. 203 Blunt; Liver 36 SonoVue, 2 mL (two doses); Liver (10–15 min total) Curved; 0.2 NA; NA Regine [ Sing. Pros., ED. 277 Blunt; Kidneys & adrenal gl. NA SonoVue, 2.4 mL; 1. Liver 2. Spleen, 3. Kidneys, 4. Supra- & submesocolic peritoneal recesses (NA min) NA; Low NA; NA Sessa [ Sing. Retro., ED. 256 Blunt; Liver Kidneys & adrenal gl. Spleen Abd. Fluid Act. Bleed 41 SonoVue, 2.4 mL (two doses); 1. Right sided organs (right kidney, liver) (1–3 min) 2. Left sided organs (left kidney, spleen) (3–4 min) 4 MHz curved; 0.15–0.19 Multiple radiologists (5+ years exp. In emergency radiology & expertise in trauma imaging); NA Valentino [ Sing. Pro., ED. 32 Blunt; Liver Kidneys & adrenal gl. Spleen Abd. Fluid Solid org. Comp. 38 SonoVue, 2.4 mL (two doses); 1. Left upper quad. (left kidney & adrenal gland, spleen) 2. Right upper quad. (right kidney & adrenal gland, liver, pancreas) (4–6 min total) 3.5 MHz or 2–5 MHz curved; Low Multiple sonographers (5+ years exp.); Independent expert sonography and radiologist Valentino [ Sing. Retro., ED. 133 Blunt; Liver Kidneys & adrenal gl. Pancreas Spleen Abd. Fluid Solid org. Comp. 40 NA, 2.4 mL (two doses); 1. Right upper quad. (right kidney, liver, pancreas) 2. Left upper quad. (left kidney, spleen) (NA min) 2–5 MHz curved; Low One radiologist; Two radiologists not involved with examination
Abbreviations: Sing. single-center, Mult. multi-center, Pro. prospective, Retro. retrospective, ED. emergency department, Pen. penetrating, Exp. experience, NA not available, Adrenal gl. adrenal glands, Abd. Fluid abdominal free fluid, Act. Bleed active bleeding, Solid org. Any solid organ injury, Comp. solid organ injury, abdominal free fluid, and/or active bleeding, Quad. quadrant
Table 2 depicts the risk of bias assessment using the QUADAS-2 tool, with visual representation in Fig. 2. Less than half of all studies (
Table 2 Risk of bias assessment performed with the Quality of Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) tool
Risk of bias Applicability concerns Domain 1: Patient selection Domain 2: Index text Domain 3: Reference standard Domain 4: Flow and timing Patient selection Index test Reference standard 1. Was a consecutive or random sample of patients enrolled? 2. Was a a case-control design avoided? 3. Did the study avoid inappropriate exclusions? 1. Were the index test results interpreted without knowledge of the reference standard? 2. If a threshold was used, was it pre-specified? 1. Is the reference standard likely to correctly classify the target condition? 2. Were the reference standard results interpreted without knowledge of the results of the index? 1. Was there an appropriate interval between index test and reference standard? 2. Did all patients receive the same reference standard? 3. Were all patients included in the analysis? Catalano [ No No Yes Yes Yes Yes Yes Yes Yes Yes High Low Yes Catalano [ No No Yes Yes Yes Yes Unclear Yes Yes Yes High Low Yes Catalano [ Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Low Low Yes Lv [ Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Low Low Yes Menichini [ Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Low Low Yes Miele [ Yes Yes Yes Yes Yes Yes Unclear Yes Yes Yes Low Low Yes Regine [ Yes Yes Yes Yes Yes Yes Unclear Yes Yes Yes Low Low Yes Sessa [ Yes Yes Yes Yes Yes Yes Unclear Yes Yes Yes Low Low Yes Valentino [ No No Yes Yes Yes Yes Yes Yes Yes Yes High Low Yes Valentino [ No No Yes Yes Yes Yes Unclear Yes Yes Yes High Low Yes
Graph: Fig. 2 Quality assessment of diagnostic accuracy studies 2 (QUADAS-2) finding per domain for included studies in the systematic review. Green represents low level of bias, red represents high level of bias, while orange represents unclear level of bias
A total of 30 pairwise comparisons between conventional ultrasound and CEUS using results from CT imaging as standard reference were used for subgroup analyses, producing a combined sample size of 3877. Forest plots of all studies are shown in Figs. 3 and 4. Sensitivity ranged from 0.030 to 0.976 for conventional ultrasound and 0.500 to 0.994 for CEUS. CEUS sensitivities for all individual studies were equal or superior to conventional ultrasound results. Specificity ranged from 0.500 to 0.998 for both modalities. CEUS specificities for all individual studies except for two [[
Graph: Fig. 3 Forest plots of sensitivities of all studies that performed contrast-enhanced CT as the reference test. Black represents conventional POCUS. Red represents CEUS. n/N: number of studies/total diagnostic tests
Graph: Fig. 4 Forest plots of specificities of all studies that performed contrast-enhanced CT as the reference test. Black represents conventional POCUS. Red represents CEUS. n/N: number of studies/total diagnostic tests. NA: not available
When paired datasets were examined, as presented in Figs. 3 and 4, we found that CEUS had significantly higher overall sensitivity of 0.933 (95% CI 0.917–0.948) versus 0.559 for conventional ultrasound (95% CI 0.527–0.591) (two-tailed, P < 0.001). The specificity of pooled datasets for CEUS of 0.995 (95% CI 0.993–0.998) was also significantly higher than conventional ultrasound at 0.979 (95% CI 0.974–0.985) (two-tailed, P < 0.001). Further subgroup analyses revealed that CEUS had statistically higher sensitivity when evaluating the liver (two-tailed, P < 0.001), kidneys and adrenals (two-tailed, P < 0.001), spleen (two-tailed, P < 0.001), pancreas (two-tailed, P < 0.01), the presence of active bleeding (two-tailed, P < 0.001), any solid organ injury (two-tailed, P < 0.001), and in the composite subgroup (two-tailed, P < 0.001) when compared to conventional ultrasound. When subgroup data was paired to conventional ultrasound, CEUS demonstrated statistically significant superiority with regards to specificity for the spleen (two-tailed, P < 0.05), the presence of active bleeding (two-tailed, P < 0.01), solid organ injury (two-tailed, P < 0.05), and in the composite subgroup (two-tailed, P < 0.001). While only one study yielded results for the sonographic examination of the pancreas [[
The diagnostic odds ratio ranged from 21.0 to 29,568.0 for CEUS versus 0.8 to 285.3 when utilizing conventional ultrasound depending on the pathologic subgroup. When pooled, the diagnostic odds ratio for all noted pathologies carried an odds ratio of 3036.9 (95% CI 3036.3–3037.5) for CEUS in comparison to 60.4 (95% CI 60.1–60.7) for conventional ultrasound. Values are presented in Table 3.
Table 3 Pooled diagnostic odds ratio, number needed to diagnose, and likelihood ratios
Pooled diagnostic odds ratio (95% CI) Pooled Numbers needed to diagnose (95% CI) Positive likelihood ratio (95% CI) Negative likelihood ratio (95% CI) Conventional ultrasound OVERALL 60.4 (60.1–60.7) 1.297 (1.256–1.342) 20.55 (20.14–20.96) 0.41 (0.36–0.45) Liver 285.3 (284.5–286.2) 1.126 (1.076–1.181) 82.51 (80.53–84.48) 0.29 (0.17–0.40) Kidneys and adrenals 122.4 (121.2–123.6) 1.199 (1.086–1.338) 49.07 (47.34–52.80) 0.67 (0.61–0.72) Spleen 143.7 (142.8–144.7) 1.189 (1.119–1.269) 51.41 (49.43–53.40) 0.43 (0.33–0.53) Pancreas 2.7 ([−1.4]–6.7) 21.450 ([−2.674]–2.141) 1.45 (0.78–2.13) 0.55 ([−0.80]–1.89) Abdominal free fluid 92.0 (91.8–92.9) 1.038 (1.001–1.077) 14.68 (14.04–15.33) 0.21 (0.02–0.41) Active bleeding 0.8 ([−2.1]–3.7) 51.723 ([−13.523]–8.880) 1.42 ([−2.28]–5.12) 0.98 (0.84–1.13) Solid organ injury 50.9 (49.9–51.9) 1.545 (1.388–1.743) 26.22 (24.77–27.66) 0.52 (0.40–0.63) Composite 5.5 (5.1–6.0) 2.918 (2.368–3.801) 2.70 (2.44–2.96) 0.50 (0.32–0.67) Contrast-enhanced ultrasound OVERALL 3036.9 (3036.3–3037.5) 1.038 (1.030–1.046) 125.68 (124.99–126.37) 0.07 ([−0.04]–0.19) Liver 4926.4 (4924.8–4927.9) 1.029 (1.009–1.049) 378.43 (375.38–381.47) 0.06 ([−0.22]–0.34) Kidneys and adrenals 6195.5 (6193.4–6197.5) 1.026 (1.005–1.048) 759.97 (745.46–765.47) 0.10 ([−0.10]–0.30) Spleen 7142.2 (7140.5–7143.9) 1.025 (1.004–1.046) 201.14 (199.07–203.21) 0.03 ([−0.45]–0.50) Pancreas 21.0 (16.6–25.4) 3.225 (1.89–11.013) 253.00 (246.44–259.56) 0.04 ([−0.78]–0.86) Abdominal free fluid 29,568.0 (29,564.9–29,571.1) 1.012 (0.994–1.030) 613.47 (606.97–619.98) 0.03 ([−0.55]–0.62) Active bleeding 350 (347.1–352.9) 1.113 (0.961–1.321) 733.62 (720.3–746.93) 0.35 (0.24–0.46) Solid organ injury 6330 (6327.9–6332.1) 1.034 (1.022–1.047) 205.16 (203.04–207.28) 0.03 ([−0.70]–0.76) Composite 233.3 (232.5–234.2) 1.207 (1.172–1.244) 16.07 (15.61–16.53) 0.08 ([−0.45] –0.61)
As shown in Table 3, the number needed to diagnose values for conventional ultrasound ranged from 1.038 to 51.723, in contrast to CEUS examinations that ranged from 1.012 to 3.225. A comparison of the two modalities revealed that two conventional ultrasound diagnostic categories (pancreas and the presence of active bleeding) were not statistically significant (two-tailed, P > 0.05), whereas all CEUS categories were statistically significant (two-tailed, P < 0.05).
The positive likelihood ratio ranged from 1.42 to 82.51 for conventional ultrasound POCUS in comparison to 16.07 to 759.97 for CEUS, while the negative likelihood ratio ranged from 0.19 to 0.98 for conventional ultrasound versus 0.03 to 0.35 for CEUS. Identification of pancreatic injury and the presence active bleeding were statistically non-significant (two-tailed P > 0.05) for conventional ultrasound imaging, whereas all all CEUS categories were statistically significant (two-tailed P > 0.05). Results are found in Table 3.
Using the random effects model, the Cochran's Q for conventional ultrasound was 4.311 with 28 degrees of freedom (two tailed, P > 0.05) while for CEUS it was 0.221 with 28 degrees of freedom (two tailed, P > 0.05), indicating that the proportion of variability of measurements for both modalities are the same in the population when considering overall results. Subsequently, the Higgin's I
Both conventional ultrasound and CEUS produced outstanding discriminating abilities for all categories, as shown in Fig. 5. The AUC for overall datasets was 0.965 for conventional ultrasound, whereas CEUS was marginally higher at 1.000.
Graph: Fig. 5 SROC curve plotting sensitivity (true positive rate) against 1-specificity (false positive rate)
To our knowledge, this is the first systematic review and meta-analysis examining and directly comparing the performance of contrast to conventional ultrasound diagnostic tests in abdominal trauma. While both modalities performed equally well with identifying abdominal free fluid, CEUS performed significantly better than conventional ultrasound at ruling out every solid organ injury of interest in abdominal trauma. Our results suggest that in a direct comparison, CEUS is superior to conventional ultrasound as the initial screening test of patients with abdominal trauma and raises the question as to whether it should be integrated as common practice when performing the FAST examination. One of the aims of this systematic review was to assess whether CEUS can be the next logical progression of clinical ultrasound in the abdominal trauma setting.
First, the diagnostic accuracy of CEUS is considerably greater to conventional ultrasound for traumatic abdominal injuries. Previous systematic reviews have shown that for patients with abdominal trauma, a negative FAST exam could not reliably rule out injury given its low sensitivity [[
Second, utilization of CEUS during trauma resuscitation should achieve similar, if not better, survival outcomes than conventional ultrasound. The REACT-2 study, a large international, multicenter, randomized control trial published in the Lancet revealed no mortality benefit to performing an immediate total-body CT scan in patients with severe trauma than conducting standard radiological work-up, which included radiographs and FAST examination during primary survey, with selective CT scans following further assessment and resuscitation [[
Third, trauma teams must have the confidence that there will be sufficient time to administer contrast agent prior to the FAST exam. For the included studies of our systematic review, all authors reported using CEUS as a part of or even in place of a FAST examination. During the evaluation of nine hemodynamically unstable patients, Catalano et al. [[
Our study, nonetheless, had few limitations of note. First, our study did not investigate allergic complications that may arise with CEUS. Nonetheless, although prior research has supported an excellent safety profile when used for cardiology applications compared to agents such as low osmotic iodinated intravenous contrast media [[
In conclusion, we showed that the diagnostic value of CEUS was greater than that of conventional ultrasound in the abdominal trauma work up. CEUS exhibited strong sensitivity, which makes it poised to be an answer to the criticisms that conventional FAST examination faces today. Future studies should investigate whether the incorporation of CEUS as an initial assessment tool is diagnostically superior to the conventional bedside FAST examination during the evaluation of abdominal trauma and whether it influences morbidity, mortality, or other patient-centered outcomes.
None.
Research was conceived by EG and designed by BS. BS and EG selected and screened the articles. MK was the third reviewer for conflicts. Data collection and analysis were performed by BS. Data interpretation by EG. BS, EG, MK, and DA had access and verified the data. BS and EG wrote the manuscript with input from DA and MK. All authors had full access to all the data in the study and had final responsibility for the decision to submit for publication. The authors read and approved the final manuscript.
None.
The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.
Not applicable.
Not applicable.
The authors declare that they have no competing interests.
Graph: Additional file 1. PRISMA 2020 Checklist.
Graph: Additional file 2. Equations.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
By Bayu Sutarjono; Matthew Kessel; Dorian Alexander and Ekjot Grewal
Reported by Author; Author; Author; Author