Not All Breaths That Follow a Ventilator Cycle Are Reverse Triggering

To the Editor:

We read with great interest the article by Shimatani and colleagues (1), in which the authors sought to describe the frequency of reverse triggering (RT) and its associated risk factors among pediatric patients with acute respiratory distress syndrome. The authors have conducted a colossal work, analyzing the patient–ventilator synchronization in 36 patients. Their main conclusion is that RT is highly prevalent in this population, as it was observed in 41.6% of patients and associated with breath stacking in 25% of the cases. This finding is quite surprising, as only one case of RT has been previously reported in children (2).

We completely agree that patient–ventilator asynchrony was very frequent in their series, in line with previous findings (3). However, we respectfully disagree with the diagnosis of RT in this study and, therefore, its prevalence. As initially described by Akoumianaki and colleagues (4), the notion of RT implies that the patient’s effort is triggered by the ventilator, with the patient’s respiratory drive being entrained by the ventilator rate. Differentiating a patient’s spontaneous breath from a breath triggered by the ventilator is difficult. To allow this distinction, important criteria have been proposed (4): the RT breaths should occur according to a stable and repetitive pattern, with minimal variability (as assessed by the coefficient of variation) of both the neural respiratory time and the phase difference, and at least five breaths with a fixed mechanical/patient effort ratio (1:1, 1:2, or 1:3) should be present.

Shimatani and colleagues did not use these discriminant criteria. All patients’ breaths occurring after the beginning of insufflation were defined as RT breaths, with no criteria related to the phenomenon’s regularity and predictiveness (see Table E1 in the online supplement). They observed a single patient with a clear repetitive pattern of RT (1:1), which is a worthy and significant result, as descriptions of RT in pediatrics are scarce. However, there is no evidence that the “RT” breaths observed in the 14 remaining patients were really triggered by the ventilator. No regular pattern was observed in these patients, in contradiction with recent adult studies in which the absence of an identified pattern was very rare (5) or simply excluded by definition.

Except for the patient with a 1:1 entrainment pattern, our interpretation is that most asynchronous breaths observed after the beginning of the ventilator cycle are not RT but instead reflect severe patient–ventilator asynchrony, with a complete “dissociation” between the ventilator’s and patient’s rates. This is well illustrated in their Figure 3, in which patient B exhibits regular spontaneous breaths. It seems quite probable that the first and sixth breaths (considered RT) were not triggered by the ventilator but rather occurred at this time fortuitously. In the absence of a convincing demonstration of entrainment by the ventilator, we believe that this type of asynchronous event should rather be classified as “premature triggering” (3) or complete desynchronization.

Is this distinction important? As pointed out by Shimatani and colleagues, a theoretical risk of ventilation-induced lung injury exists in cases of breath stacking with increased tidal volume. We fully agree that detecting these events is primordial, whether they are considered RT breaths or not. However, we believe maintaining a rigorous definition and differentiating RT from other types of asynchrony is essential because the pathophysiology and management likely differ. Adjusting the ventilator settings to the patient’s breathing effort or innovative ventilation modes such as neurally adjusted ventilatory assist can markedly improve severe asynchrony, whereas the management of RT is not well established. Observing the patient’s response to a prolonged expiratory hold can help facilitate the distinction between RT episodes and simple asynchronous effort (6).