Diagnostic accuracy of tests for assessing readiness for liberation from mechanical ventilation in adults: an overview of reviews

TL;DR

This paper reviews the accuracy of various tests used to determine if adult patients are ready to stop using mechanical ventilation, finding that some tests are more reliable than others.

Contribution

The study provides a comprehensive overview of the diagnostic accuracy of 23 tests for ventilator weaning readiness in adults.

Findings

Lung ultrasound score, diaphragmatic rapid shallow breathing index, venous oxygen saturation, and brain natriuretic peptide showed high to moderate diagnostic accuracy.

Tests like cuff leak test, cough peak flow, and RSBI demonstrated weak diagnostic accuracy.

Physicians should combine multiple tests to improve decision-making for ventilator weaning.

Abstract

To summarize the evidence on the accuracy of tests evaluating readiness for liberation from mechanical ventilation in the adult population.

Searches were conducted in MEDLINE, Embase, CENTRAL, and CINAHL, with additional publications identified through conference proceedings and contact with experts. Systematic reviews (SRs) were independently assessed for inclusion, data extraction, and risk of bias, without language or date restrictions. Included SRs focused on adults diagnosed with ventilatory failure requiring invasive support for more than 24 h. Successful weaning was considered being alive in absence of ventilatory support 72 h following liberation from mechanical ventilation.

Ten SRs examining the diagnostic accuracy of 23 readiness tests were included. These tests were conducted before, during, or after spontaneous breathing trials using various methods, such as pressure support, T-piece and continuous positive airway pressure. Among these, lung ultrasound score (sensitivity 0.94, 95% CI 0.59–0.99; specificity 0.87, 95% CI 0.62–0.97), diaphragmatic rapid shallow breathing index (sensitivity 0.84, 95% CI 0.76–0.90; specificity 0.87, 95% CI 0.79–0.92), venous oxygen saturation (sensitivity 0.83, 95% CI 0.74–0.90; specificity 0.88, 95% CI 0.83–0.92), and brain natriuretic peptide (sensitivity 0.88, 95% CI 0.83–0.92; specificity 0.82, 95% CI 0.73–0.89), showed high to moderate diagnostic capacity for ruling in and ruling out weaning failure. The remaining tests (e.g., cuff leak test, cough peak flow, P0.1, RSBI, MIP, DE, DTF, DTF–RSBI, EeDT, and EiDT) demonstrated weak diagnostic accuracy. In the SRs, risk of bias ranged from low to high.

The accuracy of tests used to assess readiness for withdrawing ventilatory support in adults varies considerably. Physicians should integrate the results of physiological, ultrasound, and paraclinical measures to minimize uncertainty in deciding which patients should progress toward ventilator weaning, always considering individual patient needs to ensure personalized healthcare.

The online version contains supplementary material available at 10.1186/s40560-026-00848-9.

Figures