Study selection and characteristics

Initially, 3412 articles were obtained from searching databases including PubMed, Scopus, and Web of Science. 2751 records were excluded based on the title and abstract screening or being duplicates before assessing the full text. 764 out of 861 studies were filtered after elimination of non-English articles, reviews, non-available abstracts, and irrelevant to the main subject. The full texts of the 97 remaining articles were fully assessed, and 70 more studies were excluded due to unclear or insufficient data (n = 46) and low quality (n = 24). After full-text screening of the remaining articles, 27 articles were eligible and included in the meta-analysis (Fig. 1).

Adhering to PRISMA standards and conventions of meta-analysis [23], 68,838 patients were studied among the selected studies. Overall, 67.5% of the presented trauma patients studied were male. The mean age of presentation was 45.0 ± 24.7 years. Twenty studies introduced and monitored Injury Severity Score among their patients/databases as a correlational measure of positive prediction value; within these studies, the mean ISS was 22.49 ± 11.53.

Mechanisms of injury (MOI)

Detailed categorization of trauma cases from 13 studies revealed the frequency of the mechanisms of injury among cases in the emergency room [5, 14, 28, 30, 31, 35,36,37,38,39,40,41,42]. Motor vehicle accidents (car, bicycle, scooter, pedestrian, etc.) accounted for 6167 injuries (80.0%), thereby being the leading cause of trauma among patients, followed by falls (4027, 20.0%). Penetrating and assault-related injuries were less common and counted for other mechanisms of injury 1518 (5.0%) (Table 1 and 2).

Table 1 Baseline characteristics in studies included in this meta-analysisTable 2 Statistical analysis of the studied literatureLength of stay in the hospital (LOS)

LOS of 33,146 patients (mean age: 44.45 ± 25.65, ISS: 21.98 ± 11.53) within the hospital was gathered from 11 studies [5, 24,25,26, 29, 35, 40, 42,43,44,45,46]. The mean LOS for patients going through iWBCT upon arrival to the hospital was 19.26 ± 24.88, 16 for immediate total body scan (iTBCT), and 6.9 for pan-scan. Other diagnostic techniques (MR, SCT, ultrasound, and radiograph) corresponded to a mean LOS of 16.01 ± 21.49.

Mortality rate

The mortality rate within the hospital was gathered from 10 WBCT/pan-scan studies and 1 TBCT with a total of 57,680 patients (mean age, 46.99 ± 28.79; ISS, 25.0 ± 12.2) [5, 26, 29, 35, 39, 41,42,43,44, 46, 47]. The mean value for patients undergoing WBCT/pan-scan immediately after admission to the hospital was 12.37% and 14.35%, respectively. In contrast, other diagnostic techniques corresponded to a mean mortality rate of 14.10%. The sample size was more significant for the WBCT (34,427) than non-WBCT/pan-scan (23,310). A mortality odd ratio of 0.94 (95% CI, 0.83–1.06; I2 = 40.1%) was calculated to compare people who underwent WBCT/pan-scan and those not using these imaging scans (Fig. 9).

Whole-body CT/pan-scan findings

A cumulative total of 33,790 injuries were found using the studied modalities of WBCT. After presentation to the emergency department (ED) within the studies, these injuries were broken down and reported based on differing body regions. Data from these studies were pooled into head/neck and chest, respectively. Overall, thoracic injuries were the leading immediate WBCT/pan-scan diagnosed injury, followed by abdominal, head, neck, face, and extremity injuries (thorax 11,028; head, neck, and face 10,516; abdomen 5600; extremities: 5203). The remaining were other injuries within the studies. Findings were then gathered further region-specific to provide a specific analysis of each’s prevalence in traumatic injury WBCT/pan-scan findings.

Head injuries

Twelve studies were filtered based on their presentation of data on head findings [25,26,27, 29,30,31, 36, 37, 43, 46, 48]. 14,168 patients with a mean age of 43.65 ± 19.67 years and an ISS of 24.2 were included. Our analysis found that 0.44 (95% CI, 0.28–0.60; I2 = 99.8%) of patients had head injuries diagnosed by WBCT/pan-scan (Fig. 2).

Neck injuries

Six studies were filtered based on their presentation of data on neck findings [36, 43, 46, 48,49,50]. 8,417 patients with a mean age of 45.05 and an ISS of 23.71 were included. Our analysis found that 0.06 (95% CI, 0.02–0.09; I2 = 97.2%) of patients had neck injuries diagnosed by WBCT/pan-scan (Fig. 3).

Facial injuries

Five studies were filtered based on their presentation of face findings [14, 36, 37, 40, 43]. 6543 patients with a mean age of 39.46 and an ISS of 21.65 were included. Our analysis found that 0.09 (95% CI, 0.05–0.13; I2 = 97.1%) of patients had face injuries diagnosed by WBCT/pan-scan (Fig. 4).

Thoracic injuries

Nineteen studies were filtered based on their presentation of data on the chest. Their inclusion accounted for a total of 39,710 patients. Mean age was 42.28 ± 18.92 [14, 25,26,27,28,29,30,31, 36,37,38, 40, 44,45,46, 48,49,50,51]. The mean ISS was 22.36 ± 11.53. Our analysis found a 0.39 (95% CI, 0.28–0.51; I2 = 99.8%) prevalence of chest injuries using WBCT/pan-scan (Fig. 5).

Abdominal injuries

Sixteen studies were filtered based on their data on abdominal injury WBCT/pan-scan findings. This totaled 20,226 patients [14, 25,26,27,28,29,30,31, 37, 40, 45, 46, 48,49,50,51]. The mean age was 41.74 ± 18.92. ISS was 22.25 ± 11.53. The meta-analysis found a prevalence of 0.23 (95% CI, 0.03–0.43; I2 = 99.9%) abdominal injury findings using WBCT/pan-scan (Fig. 6).

Spinal injuries

Data on traumatic spine injuries was pooled from six studies [36, 40, 44, 46, 51, 52], totaling to a group size of 20,267 patients with a mean age of 37.00. The mean ISS was 19.53 ± 12.1. Spinal injuries were calculated to have a prevalence of 0.19 (95% CI, 0.11–0.27; I2 = 99.4%) through our analysis (Fig. 7).

Extremity injury

Data on extremity injuries were pooled from nine studies [25, 26, 28,29,30,31, 37, 46, 49], totaling to a group size of 15,195 patients with a mean age of 41.6. The mean ISS was 43.37 ± 12.2. The prevalence of upper and/or lower extremity injuries diagnosed by WBCT/pan-scan was calculated to be 0.33 (95% CI, 0.23–0.43; I2 = 99.2%) through our analysis (Fig. 8).

Pelvis

Data pertaining to pelvic injuries were pooled from five studies [43, 46, 48, 51, 53], totaling to a group size of 2579 patients with a mean age of 44.02 ± 19.9. The mean ISS was 26.51. Pelvic injuries diagnosed by WBCT/pan-scan were 0.11 prevalent (95% CI, 0.04–0.18; I2 = 97.4%) through our analysis (Fig. 9).

 Injuries Missed on non-WBCT

In four studies, missed injuries were studied in conventional imaging modalities that were later confirmed using WBCT/pan-scan, leading to a delayed course of treatment and increased morbidity/mortality [14, 24, 45, 46]. Weninger et al. and Yoong et al. provided the injuries as a definite area of missed injuries, while Hong et al. and James et al. found smaller studies identified each missed injury. An amalgamation of their data resulted in a dataset corresponding to a patient size of 1277 with a mean age of 44.8 ± 19.35. The mean ISS was 31.1 ± 10.3.

In addition to their primary data, two studies examined cases in which immediate WBCT/pan-scan findings were compared against selective CT (SCT) and whole-body MRI (WBMR). Kimura et al. [36] provided immediate SCT cases, while Raimann et al. [40] focused on WBMR. A total of 4456 injuries were found in selective CT, while WBCT found 2927 injuries within a controlled timeline in a Trauma Center in Japan. In contrast, WBCT found 307 injuries, while WBMR found 272 injuries within a controlled timeline in a trauma center in Germany.

Missed injuries in WBCT consisted of 3 bowel injuries, 1 internal iliac artery injury, 2 femur neck fractures, 4 rib fractures, 1 sub-capital femur fracture, 1 intertrochanteric femur fracture, 1 subtrochanteric femur fracture, 1 nasal bone fracture, 1 orbital roof fracture, 1 temporal bone fracture, 11 radial fractures, 10 carpal fractures, and 16 phalangeal fractures. This observed pattern of increased missed extremity injuries was attributed to the technique within the papers.

Mortality rate analysis

Odds ratio analysis performed from the data presented on mortality rates in 11 studies demonstrated an odds ratio of 0.94 (95% CI, 0.83–1.06, I2 = 40.1%). This initially did not represent a statistically significant report (P = 0.082) (Fig. 10).

Sensitivity analysis

The sensitivity analysis results revealed that any single study or cluster of studies with shared characteristics had minimal influence on the effect size and its corresponding 95%CI, indicating robustness in the findings. Sensitivity analysis rejected the null hypothesis of a single study or any cluster of studies with statistical outlier characteristics. All studies had minimal effect on effect size and 95% CI, confirming the robustness of the findings overall (Fig. 13).

Publication bias

Figure 12 shows Begg’s funnel plot based on applying WBCT in trauma patients into body regions. The interpretation of our Begg’s funnel plot (P = 0.484) and Egger test (P = 0.193) shows no publication bias in the included studies. Therefore, it is understandable that reports have been published with both positive and negative outcomes (Figs. 11 and12).