Reverse triggering is a frequently under-recognized form of patient–ventilatory asynchrony in which the patient’s respiratory center is activated in response to a passive insufflation of the lungs. It can be detected at the bedside by observing airway pressure and airflow waveforms.

Figure 1a and video 1 (in the electronic supplementary material, ESM) show airway pressure and flow waveforms from a patient ventilated in volume-control mode. The breaths are almost certainly machine triggered because no perturbation in airway pressure can be seen at the beginning of the breath. At first sight, the records seem quite normal. However, a closer look shows that breaths marked with arrows are slightly different from other breaths. The airway pressure waveform shows a fall during the inspiratory plateau. Provided there are no air leaks, the amount of gas inside the respiratory system is constant (both the inspiratory and expiratory valves are closed), and a decrease in airway pressure denotes an increase in compliance. In controlled mechanically ventilated patients, the only explanation for this increase in compliance is recruitment of previously closed alveoli or airways. In spontaneously breathing mechanically ventilated patients, this increase in compliance can also be explained by the activation of the patient’s inspiratory muscles.

Fig. 1
figure 1

Traces corresponding to airway pressure, flow, and volume where evidence of reverse triggering can be found. a (video 1 ESM) and b (video 2 ESM) are records in volume-control ventilation without and with double triggering, respectively. c (video 3 ESM) and d (video 4 ESM) show the same phenomena in patients in pressure-control ventilation. To avoid drifts, mechanical ventilators usually reset the end-expiratory lung volume to zero at the beginning of a new breath. However, as there is no expiration between double-triggered breaths, tidal volume and end-inspiratory volume substantially increase (b, d), raising concerns about the risk of lung injury. In airflow tracings in a and video 1, expiratory flow is still present at end expiration suggesting some degree of intrinsic PEEP. The presence of intrinsic PEEP should contribute to the inability of the patient to trigger the ventilator. During flow triggering, a relatively small negative deflection in the airway pressure tracing should be observed even in the presence of intrinsic PEEP. In the absence of airway pressure deflections at the beginning of inspiration, the breath is considered triggered by the ventilator [1]. The ESM includes video clips and details of the patients shown in the figures

Full size image

The activation of inspiratory muscles cannot be ruled out even in machine-triggered breaths (Fig. 1 ESM). The respiratory control system (RCS) is a servo-regulated oscillatory system located in the medulla. The RCS’s phasic output (motor neural fibers that activate respiratory muscles) is modified by information from several sources (mechanical and chemical signals). Increasing alveolar ventilation (keeping constant tidal volume) can lead to a decrease in PaCO2 and an inhibition of the RCS. However, the machine’s breaths can also activate the RCS (entrainment). Thirty 30 years ago, Petrillo et al. [2] and Graves et al. [3] showed that a passive insufflation of the lungs can initiate a ventilatory effort in cats and normal humans, respectively. In other words, the machine triggers the patient. More recently, Akoumianaki et al. [4] found that a mechanically ventilated critically ill patient can behave in the same way, and they call this phenomenon “reverse triggering” (RT).

During entrainment, a phase-locking phenomenon (a constant ratio between machine breaths and patient efforts) can be observed. Usually, the patterns are short lived, as they are interrupted by irregular patterns every 7–15 respiratory cycles [3]. The 1:1 ratio is not only the most frequent, but also the most stable pattern [4]. Graves et al. [3] systematically studied respiratory system entrainment in anesthetized humans. They found that changes in respiratory rate and tidal volume could lead to a variety of regular and irregular patterns of coupling between respiratory system output and passive insufflations of the lungs. The locked condition is less easy to reach when PaCO2 increases or when the level of anesthesia decreases [5]. For this reason, entrainment is far more common in deeply sedated (low conscious) patients.

The delay between the start of the machine-triggered breath and the start of the patient’s effort is a fairly constant fraction of the total ventilatory time (phase angle) [4]. Interestingly, apart from passive insufflations of the lungs, several stimuli can entrain the RCS. For instance, entrainment of respiratory and locomotor rhythms is well documented [6–10].

So, the fall in airway pressure shown in this machine-triggered breath could be explained by either recruitment or RT. As during RT the patient’s inspiratory effort starts a given amount of time after the start of the machine’s breath, the patient’s inspiratory effort usually persists beyond the end of the machine’s breath. Hence, the patient’s inspiratory muscles are still active at the beginning of expiration, impeding the elastic recoil of the respiratory system from increasing alveolar pressure and thus aborting the peak expiratory flow. The amputation of the peak expiratory flow seen in this record suggests that RT rather than recruitment explains the event. If deep enough and long enough, the persistent effort could even produce a fall in airway pressure that can lead to double triggering (Fig. 1b and video 2 ESM) [11].

Figure 1c and video 3 ESM show the same phenomenon during pressure-control ventilation. Once again, the breaths remain machine-triggered. Inspiratory flow in machine-triggered pressure-control breaths is expected to present an exponential decay: as the alveolar compartment fills with gas, the pressure gradient between the airway and alveoli falls, reducing flow between compartments and the flow the machine needs to insufflate into the airways to keep airway pressure at the desired level. In this record, however, a second inspiratory peak flow can be seen. This increase in airway flow denotes an increase in airway–alveolus pressure gradient and can only be explained by an increase in compliance (an opening of previously closed units or the activation of the patient’s inspiratory muscles). Again, the perturbation in expiratory flow waveform at the beginning of expiration suggests that RT with persistent activation of inspiratory muscles is the most plausible explanation for these findings. Once more, if deep and long enough, the patient’s effort can lead to double triggering (Fig. 1d and video 4 ESM).

Entrainment appears to be produced mainly by the stretching of the slowly adapting receptors and a sustained activation of the vagally mediated Hering–Breuer reflex. In fact, in animals, the abolition of the Hering–Breuer reflex by cooling or section of vagal nerves impedes entrainment [2, 5]. In humans, however, the phenomenon can also be seen in transplant patients even when the frequencies of mechanical inflations that can entrain the ventilatory system are clearly narrowed [12].

The prevalence of RT has yet to be addressed. However, subtle effects on flow and pressure waveforms will most likely make it hard to recognize and avoid unless special attention is paid [13, 14]. Moreover, although patient–ventilator asynchronies are in general worrisome [15], the consequences of RT in particular are unknown: while RT producing double-triggering can lead to lung injury, RT could potentially also be used to promote patient–ventilator synchrony [16]. In the meantime, it is worth trying to detect and understand RT.