Diaphragm Ultrasound Beyond the Spontaneous Breathing Trial: Refining Residual Risk of Extubation Failure in Invasively Ventilated Adults—A Systematic Review and Bivariate Prognostic Test-Accuracy Meta-Analysis

Background/Objectives: Extubation failure remains clinically important despite apparent readiness for extubation. Previous diaphragm-ultrasound syntheses have largely addressed broader weaning or ventilator-liberation outcomes, leaving uncertainty about its role in refining residual extubation-failure risk. We evaluated whether diaphragm ultrasound can refine residual risk of extubation failure after standard readiness assessment in invasively ventilated adults. Methods: We performed a PROSPERO-registered systematic review and prognostic test-accuracy meta-analysis (CRD420261396412). PubMed/MEDLINE, Scopus, Web of Science Core Collection, Embase, and the Cochrane Library were searched on 18 May 2026. The primary evidence was organised by the diaphragm ultrasound index and clinically compatible extubation-failure endpoints; a broader cross-index synthesis was considered exploratory. Results: Of 1997 records, 1034 unique records were screened, 303 underwent full-text or potential eligibility assessment, and 201 were retained: 10 core DTA studies, 10 additional expanded-sensitivity records, 74 structured secondary records, and 107 narrative records. The core dataset included 729 patients and 161 events. Three studies evaluated DTF/TFdi (338 patients), five DE (225 patients), one DSF (82 patients), and one contraction velocity (84 patients). In the index-specific analyses, DTF/TFdi yielded sensitivity 0.808 (95% CI 0.661–0.900) and specificity 0.752 (0.245–0.966), whereas DE yielded sensitivity 0.831 (0.724–0.902) and specificity 0.911 (0.819–0.959). The exploratory 10-study cross-index synthesis yielded sensitivity 0.789 (0.712–0.849) and specificity 0.810 (0.620–0.918). Evidence of independent incremental value beyond SBT, rapid shallow breathing index, or usual readiness assessment was indirect. Conclusions: Diaphragm ultrasound indices provide potentially useful prognostic information about extubation failure; their accuracy requires index-specific interpretation. Current evidence does not justify stand-alone extubation decisions, universal thresholds, or established independent incremental clinical value.