To our knowledge, this is the first study that evaluates the DTF as a predictive index of weaning after a spontaneous breathing trial. The present study shows that a DTF >36% is associated with a successful weaning.
The important role of ultrasound in assessing diaphragm function has already been studied, but there are few data concerning the usefulness of diaphragm ultrasound evaluation as a weaning predictor.
In an elegant study, Lerolle and co-workers assessed the ultrasound criteria to determine diaphragmatic dysfunction after cardiac surgery [9]. They found that ultrasound determination of diaphragm excursion, expressed as the greater excursion of the muscle on a maximal inspiratory effort in patients requiring prolonged mechanical ventilation, may help identify those patients with severe diaphragmatic dysfunction. They also showed that their ultrasound measurement correlated well with trans-diaphragmatic pressure. However, this study was not designed to predict success of a weaning trial.
To assess the usefulness of diaphragmatic ultrasound in predicting successful weaning, Kim and co-workers studied the amplitude of diaphragm movements [11]. A diaphragm dysfunction was diagnosed by ultrasound if an excursion <10 mm or a paradoxical movement was observed. Diaphragmatic dysfunction was associated with weaning failure, and the authors conclude that ultrasound may identify patients at risk of difficult weaning.
Another ultrasound method, described since the last decade of the 1990s, is to evaluate the diaphragm thickness [15]. Ultrasound measures accurately the thickness of the muscle in the zone of apposition, with high reproducibility. Thickening ratio was reported to be a good indicator of diaphragm strength [15]. B-mode ultrasound may be used to assess the thickness of the muscle over a wide range of lung volumes from RV to TLC, as was demonstrated by Cohn and co-workers [16].
Diaphragm thickness may also be estimated in M-mode, although this method was criticized [16, 17]. Nevertheless, Vivier and co-workers concluded that diaphragm thickness evaluated in M-mode is a non-invasive and reproducible ultrasound method, useful to evaluate muscle function and its contribution to respiratory workload [12]. However, the great majority of diaphragm ultrasound studies have measured diaphragm thickness in B-mode. Some evaluated the variation of thickness at different lung volumes from RV to TLC in normal subjects [15, 18]. Another study measured diaphragm thickness in patients with diaphragm paralysis to monitor recovery of the muscle over time [19]. Interestingly, in this latter study, no thickening was observed by ultrasound in patients who did not recover from paralysis, thus providing useful information for both diagnosing diaphragm paralysis and indicating recovery.
We decided to measure diaphragm with B-mode ultrasound because this technique provides a greater anatomical definition of the muscle and its adjacent structures together with a more panoramic view in comparison with M-mode [16, 18].
Conventional methods used to assess diaphragm function are the measurement of trans-diaphragmatic pressure (Pdi) and phrenic nerve stimulation. Also, fluoroscopy and electromyography have been largely used. However, all these methods are invasive and uncomfortable or expose the patients to radiations.
B-mode diaphragm ultrasound is a simple, rapid, reproducible, and non-invasive test that can be repeated several times without any risk for patients and provides important information on its respiratory function. Our data suggest that DTF has a potential in predicting those patients who may fail a weaning attempt, similarly to other already established weaning parameters and tests.
Our study has some limitations. The first is that, except with PImax, we did not perform a comparison with other methods that may be considered a gold standard in the assessment of diaphragmatic function, to validate ultrasound. A previous study found a good correlation between trans-diaphragmatic pressure-time product and DTF [12]. Although trans-diaphragmatic pressure-time may be considered a gold standard in studies that evaluate new tests of diaphragmatic function, it is highly invasive and uncomfortable for the patient. Furthermore, other studies already concluded that diaphragm ultrasound is a reliable method to evaluate its respiratory function, because measurements correlated well with lung volumes and with PImax [15, 19]. Nevertheless, a future prospective trial comparing diaphragm ultrasound with trans-diaphragmatic pressure in patients with difficult weaning may be of high scientific interest and further validate this ultrasound method.
Another limitation of our study is the validity of the DTF cutoff that we obtained. We have studied a very selected population referred to our unit from the ICU of our hospital, after failure of a first gross attempt of weaning. This selection may only be obtained from the path specifically designed for our patients in our hospital. Our data cannot be automatically generalized to other populations of ordinary polyvalent ICUs. For this reason, other prospective studies should be designed to confirm the reliability of diaphragm US and to propose the most generalizable DTF threshold value.