[Predictive value of left ventricular diastolic dysfunction on mechanical ventilation weaning].

OBJECTIVE

To investigate the predictive value of left ventricular diastolic function on mechanical ventilation weaning in patients with left ventricular ejection fraction (LVEF) > 0.50.

METHODS

A retrospective case control study was conducted. Sixty-five patients with LVEF > 0.50 undergoing mechanical ventilation for more than 48 hours admitted to surgery intensive care unit (ICU) of China-Japan Friendship Hospital from June 2014 to December 2016 were enrolled. The clinical data and parameters of echocardiography before spontaneous breathing trial (SBT) were collected. The possible relationship between left ventricular diastolic function and the results of weaning was analyzed according to analysis of blood flow filling parameters of mitral valve orifice. According to the grading standard of left ventricular diastolic function, the patients were divided into normal, mild (level 1) and moderate to severe (level 2-3) groups, and the outcomes of weaning were compared among the groups. Then patients were also divided into two groups of weaning successfully and weaning failure, and the clinical data and left ventricular diastolic function parameters of patients were compared between the two groups. The predictive value of left ventricular diastolic function on results of weaning was evaluated with receiver operating characteristic curve (ROC).

RESULTS

Sixty-five patients were enrolled and 28 patients (43.1%) failed weaning, 22 patients failed the first SBT and 6 required reintubation within 48 hours, 31 of the patients presented normal left ventricular diastolic function, 9 of patients presented mild diastolic dysfunction, and 25 of them presented moderate to severe diastolic dysfunction. So with the gradual increase of the severity of diastolic dysfunction, the rate of weaning failure was gradually increased, which was 16.1%, 44.6% and 76.0% respectively (χ = 20.240, P = 0.001). Patients who failed weaning presented evidence of increased left ventricular filling pressures at pre-SBT, by demonstrating decreased deceleration time of E (DTE, s: 180.4±5.1 vs. 196.8±4.0, t = 2.567, P = 0.013), increased left ventricular mitral valve diastolic early and late filling velocity ratio (E/A: 1.47±0.08 vs. 1.14±0.05, t = 3.827, P = 0.000), increased lateral, septal and averaged left ventricular mitral valve diastolic early velocity and maximal velocity of mitral annulus in early diastolic velocity ratio (E/Em: 10.26±0.52 vs. 7.28±0.41, t = 4.535, P = 0.000; 10.08±0.58 vs. 8.16±0.40, t = 2.797, P = 0.007; 10.17±0.48 vs. 7.72±0.35, t = 4.231, P = 0.000), and the rapid shallow breathing index (RSBI) was also increased significantly (61.7±3.6 vs. 50.6±2.7, t = 2.507, P = 0.015). It was shown by ROC curve analysis that the basic left ventricular diastolic function at pre-SBT had the diagnostic performance in predicting the outcome of weaning from mechanical ventilation, especially E/A and lateral E/Em. Pre-SBT values of E/A greater than 1.2 and lateral E/Em greater than 7.9 predicted weaning failure with an area under the ROC curve (AUC), sensitivity, and specificity of 0.81±0.06 and 0.85±0.06, 82.6% and 91.3%, 81.4% and 80.7%, respectively, and the AUC was higher than RSBI (0.70±0.07). The AUC of combination of E/A > 1.2 and lateral E/Em > 7.9 predicting weaning failure was 0.86±0.05 with a sensitivity of 78.3% and a specificity of 93.6%.

CONCLUSIONS

The results suggest that left ventricular diastolic dysfunction is significantly associated with weaning outcome in critical patients with LVEF > 0.50. The combination of E/A ratio greater than 1.2 and E/Em ratio greater than 7.9 may identify patients at high risk of weaning failure.