3.4 Predictors of prolonged mechanical ventilation duration and weaning failure (Figure 4, 5, 6)
Age < 1.5 years can predict weaning failure with a sensitivity of 71% and specificity of 68%. The LUS scores and DE did not have any significant correlation with weaning failure or mechanical ventilation duration and ICU stay days. We found the baseline right diaphragm thickness value of < 24 mm could predict MV duration of > 13 hours with (70 % sensitivity and 70% specificity) and AUC of 0.68, p=0.026.. The left diaphragm thickening fraction on PSV with a cut-off value of 17.15% was found as a predictor for the weaning failure (sensitivity 85%, specificity 49%, AUC ROC 0.75, p= 0.032).
DISCUSSION
In this observational study we tried to identify the bedside ultrasonic predictors of weaning failure from mechanical ventilation in pediatric patients who underwent cardiac surgery under cardiopulmonary bypass. The pulmonary complications of CPB can be manifested by the presence of B-lines and diaphragm dysfunction. 24-29
We found the median LUS score to increase from baseline to PSV (post operatively) in concurrence with the previous literatures21,22,29,41.The significant elevation of the scores in the postextubation period in group 1 may be the reason for the weaning failure, whereas in the group 2 the scores remained the same postextubation. However, we did not find any correlation between LUS scores and the events of weaning failure. This can be attributed to the fact that the post cardiac surgery lung edema is of multifactorial origin. 12-14 The similar intraoperative management with a negative fluid balance after CPB and before extubation (table 1) and cardiac functions maybe the reason for the similar lung profiles for both the groups. The cardiogenic pulmonary edema did not contribute significantly to the LUS scores of our patients, and the LVEF and E/e’ did not differ between the two groups of patients. The fluid balance measured before extubation corelated significantly with MV duration (r2 =0.471, p=0.001) and ICU stay days (r2 =0.297, p=0.038), suggesting the patients with negative fluid balance have favorable outcomes.
We could establish the relevance of TDF value of less than 17.15 % (sensitivity of 85%) of the left side diaphragm during the PSV in predicting the weaning failure .The diaphragm contributes to 75% of respiratory effort in children, as evidenced by a DTF less than 17.15% with increased rates of reintubation. The DTF less than 17%- 21 % during SBT in children have been shown to be associated with extubation failure. 42 This study was conducted in medical ICU patients with longer times of intubation prior to giving an extubation trial , which could have resulted in greater diaphragm remodeling as compared to our subset of patients who were intubated for a short duration. These studies measured the diaphragm only on the right side which was easier to visualize. In our study, we found significant changes of the diaphragm on the left side only. This could be due to the fact that there is frequent handling of the phrenic nerve on the left side.
The diaphragm atrophy rate is 3.4% per day in children.43 There was a decrease of thickness values during the pressure support mode of ventilation in our patients. There are no defined reference values of normal diaphragm thickness in children, so we used the pre-operative baseline values as our reference. These values were lower than the values measured by Glau et al of 2cm (1.8-2.5) measured in children with acute respiratory failure , as our patients were critically ill cardiac patients who might have had compensated respiratory problems prior to surgery.
The diaphragm excursion is decreased after mechanical ventilation even four hours after extubation (group 1 :p-value -0.022;group 2 : p-value -0.00), suggesting the detrimental effects of neuromuscular blockers and mandatory ventilation on diaphragm functions. However, the diaphragm excursion has limited predictive ability for weaning outcomes.32, 40, 42 The excursion may be influenced by the pressure provided by the ventilator leading to a similar degree of excursion despite significantly different levels of muscle efforts , whereas thickness is influenced only by active contraction.
The DE and TDI have been used to assess the feasibility of extubation during SBT in the adult patients on prolonged mechanical ventilation; however its suitability to be used during the weaning process in pediatric patients after cardiac surgery has not been clearly established. 31-35
CONCLUSION
LUS scores cannot predict weaning failure whereas diaphragmatic thickening fraction during SBT <17.15% was found to be a predictor of weaning failure in pediatric post cardiac surgical patients.
Our study was limited by the fact that we had a short time course and we did not follow the patients in the post extubation period beyond 4 hours in non-reintubated patients. Our study had heterogeneous patient population undergoing variety of procedures having very different surgical procedures and their effect on the kids. So the cut –off values may not be applied generalized to all pediatric patients. We had a small sample size, which could have made the significant results appear non-significant. We did not evaluate the concurrent effects of lung compliance on the extubation outcomes. We did not evaluate EVLW or diaphragm dysfunction by other known methods like Trans diaphragmatic pressure movements, Fluoroscopy, Phrenic nerve conduction studies. In future, the study can be expanded to the learning of the long-term effects of CPB on mechanical ventilation. The study can be extended onto non-CPB surgery patients.