Introduction The objective of this study was to compare the accuracy of Focused Assessment with Sonography for Trauma (FAST) with Local Wound Exploration (LWE) in detecting intra-abdominal injuries for patients with abdominal gunshot and stab wounds by calculating the sensitivity, specificity, predictive values and likelihood ratios of each diagnostic modality.
Methods This is a cross-sectional study conducted from June 2009 to June 2011 at a tertiary government level II trauma center in Manila. Included in the study were the medical records of all patients treated for abdominal gunshot and/or stab wounds. The following data were obtained from the records: demographic profile; mechanism of injury, whether gunshot or stab; location of injury, whether anterior, posterior abdomen (back); diagnostic procedures and results, whether FAST, local wound exploration, or both; management, whether surgical or conservative; and intraoperative findings, including organ/s injured, if any. For the gold standard, the subjects were categorized as positive or negative injury based on the laparotomy findings or, in the absence of surgical intervention, recovery without sequelae. The results of FAST and LWE were compared to the results of the gold standard by calculating the sensitivity, specificity, positive predictive value, negative predictive value and likelihood ratio with 95% confidence interval, for each modality.
Results The study included 95 subjects with a mean age of 31 years. Sixty eight (68) patients (71.6%) underwent FAST while 62 (65.3%) underwent LWE. There were 35 patients (36.8%) who underwent both the FAST and LWE. A total of 78 patients had anterior abdominal stab and gunshot wounds. FAST was more specific (97%) than sensitive (36.5%); LWE, however, was more sensitive (95%) than specific (50%). The positive predictive value of FAST was 100% while that of LWE was 77.6%. FAST had a low negative predictive value at 32.7% while LWE had a high negative predictive value of 84.6%. FAST had a higher positive likelihood ratio than that of negative likelihood ratio (12 versus 0.635). LWE, on the other hand, had a lower positive likelihood ratio of 1.9 but a high negative likelihood ratio of 0.1. The confidence intervals of the accuracy estimates were moderately precise.
Conclusion This study showed that LWE is an accurate diagnostic procedure in detecting traumatic injuries, while FAST is accurate in ruling out significant injury in patients with abdominal stab and gunshot wounds. For patients with suspected intra-abdominal injury, LWE should be done first as a screening test followed by FAST as a confirmatory test.