Ultrasound assessment of the rectus femoris and diaphragm for predicting extubation outcomes in mechanically ventilated patients: a prospective observational study

Critically ill patients receiving invasive mechanical ventilation often experience rapid loss of muscle mass and function, increasing the risk of extubation failure. Ultrasound enables real-time, noninvasive assessment: the diaphragmatic thickening fraction (DTF) quantifies inspiratory muscle contraction, and the rectus femoris (RF) reflects muscle mass and nutritional status. We longitudinally monitored diaphragm and RF ultrasound parameters to characterize their trajectories and evaluate associations with extubation outcomes. We conducted a prospective observational study in a tertiary teaching-hospital Intensive Care Unit (ICU). Ultrasound examinations were performed on study days 1, 3, 5, and 7 and on the planned extubation day, measuring RF thickness (RF-Thickness), RF cross-sectional area (RF-CSA), and DTF. Day-1 values served as baselines to compute variation rates (Δ%). We assessed correlations between RF and diaphragm parameters and their predictive performance for successful extubation. Subgroup analyses were prespecified by sedation depth. 92 patients were enrolled (extubation success, n = 60; failure, n = 32). Compared with the success group, the failure group exhibited significantly greater declines in RF-Thickness and RF-CSA from Day 3 onward, reflected by more negative RF-Thickness Δ% and RF-CSA Δ% values, while DTF was consistently lower at all time points (all P < 0.05). DTF correlated positively with RF variation rates ( ρ ranged from 0.29 to 0.43; P < 0.05). On the extubation day, AUCs for DTF, RF-Thickness Δ%, and RF-CSA Δ% were 0.828 (95% CI 0.743–0.912), 0.748 (95% CI 0.643–0.854), and 0.760 (95% CI 0.646–0.875), respectively (all P < 0.05). The combined model achieved an unadjusted AUC of 0.864 (95% CI 0.792–0.936); internal validation via 1000-iteration bootstrapping yielded a bias-corrected AUC of 0.845 (95% CI 0.759–0.911). In deep-sedation strata, DTF distinguished outcomes at every time point; under non-deep sedation, the success group showed better preservation of RF-CSA. RF ultrasound parameters partially mirror diaphragmatic function and broaden monitoring of respiratory-muscle performance. RF-Thickness Δ% and RF-CSA Δ% are reliable indicators of extubation outcomes; combined with DTF, discrimination improves substantially. Across sedation strategies, diaphragm and RF ultrasound retain predictive value and can inform bedside extubation decisions.