Multi-center analysis of point-of-care ultrasound for small bowel obstruction: A systematic review and individual patient-level meta-analysis

The study aimed to assess the diagnostic accuracy of point-of-care ultrasound (POCUS) in identifying small bowel obstruction (SBO) and to investigate the impact of clinician experience level and body mass index (BMI) on POCUS performance for diagnosing SBO in the Emergency Department.

We systematically searched PubMed and Cochrane databases from January 2011-2022. We performed a meta-analysis using individual patient-level data from prospective diagnostic accuracy studies from which we obtained data from the corresponding authors. Overall test characteristics and subgroup analysis across clinician experience levels and a range of BMI were calculated. The primary outcome was SBO as the final diagnosis during hospitalization.

We included Individual patient data from 433 patients from 5 prospective studies. Overall, 33% of patients had a final diagnosis of SBO. POCUS had 83.0% (95%CI 71.7%-90.4%) sensitivity and 93.0% (95%CI 55.3%-99.3%) specificity; LR+ was 11.9 (95%CI 1.2-114.9) and LR- was 0.2 (95%CI 0.1-0.3). Residents had exhibited a sensitivity of 73.0% (95%CI 56.6%-84.9%) and specificity of 88.2% (95%CI 58.8%-97.5%), whereas attendings had demonstrated a sensitivity of 87.7% (95%CI 71.1%-95.4%) and specificity of 91.4% (95%CI 57.4%-98.8%). Among those patients with BMI<30 kg/m2, POCUS showed a sensitivity of 88.6% (95%CI 79.5%-94.7%) and a specificity of 84.0% (95%CI 75.3%-90.6%), while patients with BMI ≥ 30 kg/m2 exhibited a sensitivity of 72.0% (95%CI 50.6%-87.9%) and specificity of 89.5% (95%CI 75.2%-97.1%).

POCUS correctly identified those patients with SBO with high sensitivity and specificity. Diagnostic accuracy was slightly reduced when performed by resident physicians and among patients with a BMI ≥ 30 kg/m2.

PROSPERO registration number: CRD42022303598.