RESUMO
OBJECTIVE: To evaluate the accuracy of abdominal radiography (AXR) for the detection of residual cocaine packets by comparison with computed tomography (CT). METHODS: Over a 1-year period unenhanced CT was systematically performed in addition to AXR for pre-discharge evaluation of cocaine body packers. AXR and CT were interpreted independently by two radiologists blinded to clinical outcome. Patient and packet characteristics were compared between the groups with residual portage and complete decontamination. RESULTS: Among 138 body packers studied, 14 (10 %) had one residual packet identified on pre-discharge CT. On AXR, at least one reader failed to detect the residual packet in 10 (70 %) of these 14 body packers. The sensitivity and specificity of AXR were 28.6 % (95 % CI: 8.4-58.1) and 100.0 % (95 % CI: 97.0-100.0) for reader 1 and 35.7 % (95 % CI: 12.8-64.9) and 97.6 % (95 % CI: 93.1-99.5) for reader 2. There were no significant patient or packet characteristics predictive of residual portage or AXR false negativity. All positive CT results were confirmed by delayed expulsion or surgical findings, while negative results were confirmed by further surveillance. CONCLUSION: Given the poor performance of AXR, CT should be systematically performed to ensure safe hospital discharge of cocaine body packers. KEY POINTS: ⢠Both abdominal radiography and computed tomography can identify gastrointestinal cocaine packets. ⢠Ten per cent of body packers had residual packets despite two packet-free stools. ⢠Seventy per cent of these residual packets were missed on AXR. ⢠No patient or packet characteristics predicted residual packets or AXR false negativity. ⢠CT is necessary to ensure safe medical discharge of body packers.