RESUMO
AIM: This meta-analysis evaluates the diagnostic accuracy of ultrasound (US) for bone fractures over the past 47 years, comparing it to established imaging standards. MATERIAL AND METHODS: We adhered to PRISMA 2020 guidelines to search Medline, EMBASE, and the Cochrane Library using tailored search strategies. The primary outcome, US diagnostic performance, was analyzed across various subgroups including clinical relevance, patient age, and anatomical considerations. The QUADAS-2 tool was employed to assess study quality and minimize bias. RESULTS: From 5,107 initially identified studies, 75 met the inclusion criteria, encompassing 7,769 participants and 3,575 diagnosed fractures. The majority of studies were prospective (79%) and compared US primarily with plain radiography (76%) and CT scans (19%). Of these, 61 studies were amenable to systematic analysis, revealing US to have a sensitivity and specificity of 91% (95% CI: 90%-92%) and 91.3% (95% CI: 90.5%-92.1%), respectively. Likelihood ratios were favorable, with a positive value of 9.955 and a negative value of 0.087, and an odds ratio of 132.67. The area under the curve stood at 0.9715, indicating high diagnostic accuracy despite significant heterogeneity (I²=81.3% for sensitivity, 89.3% for specificity). CONCLUSION: The evidence supports US as a highly accurate diagnostic tool for bone fractures, rivalling standard imaging methods like CT and radiography. Its notable diagnostic efficacy, combined with advantages in reducing pain, wait times, and radiation exposure, advocates for its broader application. Further validation in large-scale, randomized trials is essential to integrate US more fully into clinical guidelines for fracture management.