Diaphragm Excursion-Time Index: A new parameter using ultrasonography to predict extubation outcome, by Atul Palka et al. 
Predicting successful weaning from mechanical ventilation isn’t the easiest task in the world. Extubation failure increases mortality, infections, ICU length of stay and costs [2,3]. And although in the past few years we came with some fairly reasonable predictors, the uncertainty is always there.
If the underlying cause of intubation is resolved, the gas exchange and hemodynamic are OK, the patient is metabolically compensated, and the level of consciousness is good, the primary remaining factor implied in success is how the respiratory muscles, especially the diaphragm, will deal with the load imposed by the work of breathing. Therefore, endurance is essential.
Knowing that, and how the respiratory mechanics is related to endurance, our friends from New York came with the idea if the excursion-time (E-T) index measured during spontaneous breathing trial, would predict successful extubation. The E-T index is the product between diaphragm excursion (DE) and inspiratory time (Ti). The diaphragm excursion was chosen as a surrogate for mean tidal pressure and tidal volume (both are determinant of the work performed by the diaphragm) and combined it with the inspiratory time (“a proxy of the work performed by the diaphragm in response to inspiratory loading during weaning”). Meaning, the longer the Ti and the greater the DE, the better; which results in a higher E-T index.
Between June 2015 and February 2017 they enrolled patients who were admitted to medical ICUs and were ready for extubation (decided by the ICU team). Patients were excluded if: <18y/o, MV for less than 24h, diaphragm paralysis, pneumothorax, use of NMBA during ICU stay and patients who underwent tracheostomy. Weaning failure was defined as re-intubation or use of non-invasive ventilation (really?) within 48h of extubation. It’s not clear if every patient with inclusion criteria was accessed for trial inclusion, and the statistical analysis plan is nonexistent (what a shame).
How they measured:
The DE was measured at three times: right before the SBT; following 30 minutes of SBT ( PSV: Pins 5cmH2O PEEP 5cmH2O); and between 4 and 24h after extubation. They also calculated the rapid shallow breathing index [RSBI = respiratory rate / tidal volume (in liters)] following 30 minutes of SBT. The ICU team was blinded to the measurements.
For the DE measurement: phased array probe below the right costal margin in the mid-clavicular line, longitudinal orientation. M-mode perpendicular to the movement of the posterior one-third of the diaphragm. The DE was measured on the vertical axis of M-mode tracing (beginning to the end of inspiration). The Ti was recorded for three breaths, and the mean was noted.
A total of 73 patients were included, in which 20 failed extubation (28%). The patients’ characteristics are shown in the table below.
Patients who failed extubation had greater RSBI, although the mean value (why not the median?) was 75.5 breaths/min/L which is far below the cutoff of 105 proposed by Yang and Tobin, of course, the standard deviation was 57.4 (small sample size sucks). Interpreting the data about ICU and hospital LOS is hard. Therefore, let’s stick to the generalities. The table below gives us a more comprehensive overview of the outcome (extubation failure).
Hence, of the 20 failures, 9 were re-intubations, and 11 were use of NIV. The reasons for MV were: respiratory failure (56.1%), shock (20.5%), neurologic (15.1%) and others (8.1%). The table below highlights the study results.
Both the DE and Ti was higher in the success group (during SBT and after extubation), and since the E-T index is a product of both, it was also higher. There was no difference in measurements before the SBT. The table below shows the mean percentage changes of measurements between groups.
A decrease in the E-T index during the transition of A/C to SBT of less than 3.8% has an AUC of 0.77 to predict successful extubation. The AUC in the Tobin article  was 0.89.
As I said, it isn’t easy to predict successful extubation. Few studies have shown that diaphragm excursion can help prediction successful weaning [5,6]. The idea of creating an index using ultrasound and Ti measurements is interesting. But if we look at the decrease in DE during the transition of A/C to SBT, with a cutoff of 16.5%, the AUC was 0.75. Therefore, the addition of the Ti in the product only increased the accuracy by 2%. Actually, the sensitivity of DE alone (84.9%) was better than the sensitivity of the E-T index (79.2%). The E-T index greater than 0.92 had sensitivity and specificity of 90 and 45%, respectively, while the AUC was 0.66. Nothing fancy here.
But I got a few questions? I don’t see any point taking all those measures while the patient is still in A/C. Also, calculating the delta A/C-SBT also seems pointless. Why not use the ABT values and compare both groups? The study has several limitations: only one investigator performed all measurements, high rates of extubation failure, small sample size, lack of statistical rigor, two patients were on non-invasive ventilation while the measurements were performed, and they considered the use of NIV as extubation failure.
Of course, this small study won’t change our practice, but it’s the epitome of a good idea put into practice, with some interesting results I might say. That old discussion that if you have a lot of instruments to predict something, none of them really works, is still true for extubation prediction, but again, since there are no good instruments, two are better than one.
1- Palkar A, Narasimhan M, Greenberg H et al. Diaphragm Excursion-Time Index: A new parameter using ultrasonography to predict extubation outcome Chest. 2018;
2- Thille AW, Harrois A, Schortgen F, Brun-Buisson C, Brochard L. Outcomes of extubation failure in medical intensive care unit patients* Critical Care Medicine. 2011; 39(12):2612-2618.
3- Epstein SK. Extubation failure: an outcome to be avoided. Crit Care. 2004; 8(5):310-.
4- Yang KL, Tobin MJ. A Prospective Study of Indexes Predicting the Outcome of Trials of Weaning from Mechanical Ventilation N Engl J Med. 1991; 324(21):1445-1450.
5- Kim WY, Suh HJ, Hong SB, et al. Diaphragmdysfunction assessed by ultrasonography: influence onweaning from mechanical ventilation. Critical caremedicine. 2011 Dec 1;39(12):2627-30.
6- Osman AM, Hashim RM. Diaphragmatic and lungultrasound application as new predictive indices for theweaning process in ICU patients. The Egyptian Journalof Radiology and Nuclear Medicine. 2017 Mar31;48(1):61-6