[Correlation analysis between mechanical power normalized to dynamic lung compliance and weaning outcomes and prognosis in mechanically ventilated patients: a prospective, observational cohort study].

Abstract

Objective: To explore the correlation between mechanical power normalized to dynamic lung compliance (Cdyn-MP) and weaning outcomes and prognosis in mechanically ventilated patients.

Methods: A prospective, observational cohort study was conducted. Patients who underwent invasive mechanical ventilation (IMV) for more than 24 hours and used a T-tube ventilation strategy for extubation in the intensive care unit (ICU) of Lianyungang First People's Hospital and Lianyungang Second People's Hospital between January 2022 and December 2023 were enrolled. The collected data encompassed patients' baseline characteristics, primary causes of ICU admission, vital signs and laboratory indicators during the initial spontaneous breathing trial (SBT), respiratory mechanics parameters within the 4-hour period prior to the SBT, weaning outcomes and prognostic indicators. Mechanical power (MP) and Cdyn-MP were calculated using a simplified MP equation. Univariate and multivariate Logistic regression analyses were utilized to determine the independent risk factors associated with weaning failure in patients undergoing mechanical ventilation. Restricted cubic spline (RCS) analysis and Spearman rank-sum test were employed to investigate the correlation between Cdyn-MP and weaning outcomes as well as prognosis. Receiver operator characteristic curve (ROC curve) was constructed, and the area under the ROC curve (AUC) was computed to evaluate the predictive accuracy of Cdyn-MP for weaning outcomes in mechanically ventilated patients.

Results: A total of 366 patients undergoing IMV were enrolled in this study, with 243 cases classified as successful weaning and 123 cases classified as failed weaning. Among them, 23 patients underwent re-intubation within 48 hours after the successful withdrawal of the first SBT, non-invasive ventilation, or died. Compared with the successful weaning group, the patients in the failed weaning group had significantly increased levels of sequential organ failure assessment (SOFA) score, body temperature and respiratory rate (RR) during SBT, and respiratory mechanical parameters within the 4-hour period prior to the SBT [ventilation frequency, positive end-expiratory pressure (PEEP), platform pressure (Pplat), peak inspiratory pressure (Ppeak), dynamic driving pressure (ΔPaw), fraction of inspired oxygen (FiO₂), MP, and Cdyn-MP], dynamic lung compliance (Cdyn) was significantly reduced, and duration of IMV, ICU length of stay, and total length of hospital stay were significantly prolonged. However, there were no statistically significant differences in age, gender, body mass index (BMI), smoking history, main causes of ICU admission, other vital signs [heart rate (HR), mean arterial pressure (MAP), saturation of peripheral oxygen (SpO₂)] and laboratory indicators [white blood cell count (WBC), albumin (Alb), serum creatinine (SCr)] during SBT of patients between the two groups. Univariate Logistic regression analysis was conducted, and variables with P < 0.05 and no multicollinearity with Cdyn-MP were selected for inclusion in the multivariate Logistic regression model. The results demonstrated that SOFA score [odds ratio (OR) = 1.081, 95% confidence interval (95%CI) was 1.008-1.160, P = 0.030], and PEEP (OR = 1.191, 95%CI was 1.075-1.329, P = 0.001), FiO₂ (OR = 1.035, 95%CI was 1.006-1.068, P = 0.021) and Cdyn-MP (OR = 1.190, 95%CI was 1.086-1.309, P < 0.001) within the 4-hour period prior to the SBT were independent risk factors for weaning failure in patients undergoing IMV. The RCS analysis after adjusting for confounding factors showed that as Cdyn-MP within the 4-hour period prior to the SBT increased, the risk of weaning failure in patients undergoing IMV significantly increased (P < 0.001). The Spearman rank correlation test showed that Cdyn-MP within the 4-hour period prior to the SBT was positively correlated with respiratory mechanical parameters including ΔPaw and MP (r values were 0.773 and 0.865, both P < 0.01), and negatively correlated with Cdyn (r = -0.587, P < 0.01). Cdyn-MP within the 4-hour period prior to the SBT was positively correlated with prognostic indicators such as duration of IMV, length of ICU stay, and total length of hospital stay (r values were 0.295, 0.196, and 0.120, all P < 0.05). ROC curve analysis demonstrated that, within the 4-hour period preceding the SBT, Cdyn-MP, MP, Cdyn, and ΔPaw possessed predictive value for weaning failure in patients undergoing IMV. Notably, Cdyn-MP exhibited superior predictive capability, evidenced by an AUC of 0.761, with a 95%CI ranging from 0.712 to 0.810 (P < 0.001). At the optimal cut-off value of 408.5 J/min×cmH₂O/mL×10^-3, the sensitivity was 68.29%, and the specificity was 71.19%.

Conclusion: Cdyn-MP is related to weaning outcomes and prognosis in mechanically ventilated patients, and has good predictive ability in assessing the risk of weaning failure.