Temporal changes in parasternal intercostal muscle thickening in predicting weaning outcomes: A prospective observational study.

OBJECTIVES

To evaluate whether temporal changes in parasternal intercostal muscle thickening fraction during spontaneous breathing trials provide superior diagnostic accuracy for predicting outcomes of ventilator weaning compared to single time‑point measurements.

DESIGN

Prospective observational study.

SETTING

Mixed intensive care unit.

PATIENTS OR

PARTICIPANTS

Sixty‑four adult patients aged 18-65 years who required invasive mechanical ventilation for more than 24 h and were deemed ready for a spontaneous breathing trial. Seventeen patients were excluded based on predefined criteria.

INTERVENTIONS: Parasternal intercostal muscle thickness was measured at baseline and after 30 min of a spontaneous breathing trial. The temporal change in thickening fraction was calculated and compared between patients with successful and failed weaning.

MAIN VARIABLES OF INTEREST: Clinical and physiological parameters, parasternal intercostal muscle thickening fraction at baseline and 30 min, and the temporal change in thickening fraction. Predictive accuracy was assessed using receiver operating characteristic analysis and logistic regression.

RESULTS

Of the 64 patients, 42 (65.6%) achieved successful weaning, while 22 (34.4%) failed, including 15 with failed SBT and 7 requiring reintubation. Failed patients had longer mechanical ventilation duration, higher APACHE II scores, and significantly elevated RSBI values. Ultrasound revealed higher PIC‑TF at baseline and 30 min in failed patients, with PIC‑TFΔ showing positive increases compared to negative changes in successful cases. ROC analysis demonstrated that PIC‑TFΔ provided superior diagnostic accuracy compared to baseline measurements, with AUC values of 0.77-0.81 for failed weaning and 1.00 for reintubation. PIC‑TFΔ demonstrated a non‑significant trend toward predicting weaning failure in the multivariate model (OR 2.28, 95% CI 0.95-5.48, p = 0.067). Larger studies with greater statistical power are required to confirm its independent predictive value.

CONCLUSION

Both PIC‑TF₃₀ and PIC‑TFΔ demonstrated superior diagnostic accuracy compared to baseline PIC‑TF₀, highlighting the added value of dynamic or end‑trial measurements over static baseline assessment. While PIC‑TF₃₀ provided the highest prognostic performance, PIC‑TFΔ offered complementary mechanistic insight into evolving respiratory muscle load. PIC‑TFΔ demonstrated a preliminary signal for reintubation risk. However, this finding is hypothesis‑generating and requires validation in larger, multicenter trials before clinical incorporation can be recommended.