Sci Prog. 2026 Oct-Dec;109(4):368504261503871. doi: 10.1177/00368504261503871. Epub 2026 Oct 30.
ABSTRACT
ObjectiveTo evaluate the diagnostic accuracy of emergency clinician-performed ultrasound for proximal lower-extremity deep vein thrombosis (DVT) and explore variation by scan protocol and operator group.MethodsThis systematic review and meta-analysis included studies of adults with suspected DVT in emergency settings. PubMed, Embase, the Cochrane Library, and Web of Science Core Collection were searched from inception to 8 April 2026, and a targeted literature update was conducted through 20 September 2026 during revision. A bivariate binomial random-effects model jointly estimated sensitivity and specificity. The primary analysis included one dataset per study to avoid duplicate weighting of paired cohorts. Risk of bias and applicability were assessed with QUADAS-2. Protocol and operator subgroup analyses were exploratory. The review was registered in PROSPERO (CRD420261361599).ResultsSixteen studies contributed quantitative data, providing 17 protocol datasets because one paired cohort was evaluated with both two-point and three-point protocols. Selecting the three-point result from that paired cohort for the primary overall model yielded 16 independent study datasets comprising 2,781 patient or limb/examination units. Pooled sensitivity was 0.91 (95% confidence interval [CI] 0.85-0.95), and specificity was 0.95 (95% CI 0.91-0.97). Sensitivity estimates for two-point and three-point/extended protocols were 0.87 and 0.92, respectively. Source-data uncertainties, including a reconstructed diagnostic-analysis table in Jahanian 2019 and a provisional Crowhurst input, were examined in sensitivity analyses. Koylu 2026 was added during the revision-stage targeted literature update.ConclusionsEmergency clinician-performed ultrasound may support proximal DVT assessment within a defined diagnostic pathway. Negative limited examinations require interpretation alongside clinical probability and examination adequacy. Heterogeneity and unresolved source data limit confidence in the pooled estimates. Exploratory subgroup findings do not establish the superiority of any protocol or operator group.
PMID:42961238 | DOI:10.1177/00368504261503871