Key Words: Work of Breathing, Mechanical Ventilation, Ventilator Weaning, Intensive Care Units, COVID- 19

Results and discussion

Definitions

The literature highlights that although respiratory drive, inspiratory effort, and work of breathing are intuitive concepts, even distinguished authors disagree on their definitions [22] (Table 1). Under healthy conditions, respiratory drive, inspiratory effort, and work of breathing are closely related, but this may be untrue in pathological states [23]. This background justifies analyzing definitions of the parameters addressed in this review (Figure 1).

Table 1. Definitions of respiratory drive, inspiratory effort, and work of breathing identified in the literature.

Figure 1. Graphical definition and alternatives for monitoring respiratory drive, inspiratory effort, and work of breathing in intensive care ventilators.

Definition of respiratory drive

The primary objective of respiration is to maintain balanced gas exchange. In other words, a balance between oxygen intake and carbon dioxide (CO₂) production. This goal is achieved by tight control of respiration through regulation of the respiratory center. The intensity and speed of the response depend on feedback from four input sources: biochemical, mechanical, emotional, and inflammatory [23].

The first and foremost input source is biochemical, which corresponds to reflex feedback mediated by input originating from central and peripheral chemoreceptors [24]. This control mechanism minimizes arterial carbon dioxide levels and pH fluctuations and avoids hypoxemia [25]. Peripheral chemoreceptors in the carotid bodies also regulate the respiratory drive by modifying central chemoreceptors' sensitivity and excitatory threshold [26].

The second input corresponds to information from mechanoreceptors at the lungs, costal grid, airway, and respiratory muscles. Muscle spindle use and slowly adapting stretch receptors are examples of this, as they provide information on stretch and lung volume through their vagal inputs. These receptors are responsible for the Hering-Breuer reflex, which terminates inspiration and facilitates expiration at high tidal volumes [27].

The third input corresponds to suprapontine feedback. Fear, anxiety, pain, and delirium transmit sensory feedback to the respiratory centers through the cerebral cortex and hypothalamus modulations.

The fourth input is mediated by inflammatory responses [23] from different pathological states, including sepsis and acute respiratory failure. Inflammatory mediators (histamine, bradykinins, and prostaglandins) elicit a vagal stimulus that increases respiratory rate and decreases tidal volume [25], triggering a rapid and shallow respiratory pattern.

Once the information from all four inputs is integrated, a response is elaborated by the respiratory center, known as the respiratory drive. Thus, the respiratory drive is defined as the intensity of the respiratory efference generated in the central nervous system [18],[23],[28]. Several authors agree that the drive does not necessarily reflect changes in the inspiratory effort since it is an electrical signal and not directly a mechanical force. The respiratory drive is originated from an interneuron group (respiratory center) at the brainstem [27]. The amplitude of the signal emerging from this neuronal group determines the mechanical response of the respiratory muscles.

In patients with mechanical ventilation, a diminished respira-tory drive – either due to over- assistance or sedation – will con-tribute to diaphragmatic weakness [29]. A weak drive will also be responsible for the loss of synchrony between the patient and the ventilator. This asynchrony may cause a high drive that damages the diaphragm due to overload [30]. Likewise, increased respiratory drive leads to a rapid and shallow respiratory pattern, with the recruitment of accessory musculature, increased dyspnea, and, consequently, failure of spontaneous breathing trials [31]. Another effect of an excessive respiratory drive is patient- ventilator asynchrony. During assisted ventila-tion, parameter adjustement that is not following the ventila-tory demands can lead to flow starvation, premature cycling, or ineffective triggering [32],[33]. Therefore, the drive (and its inputs) is an antecedent to respiratory effort and work. Its intensity and magnitude determine muscle activity and rhythm.

Definition of inspiratory effort

The definition of inspiratory effort seems intuitive; however, it has been little explored in the literature and could be confused with work of breathing [22],[34]. Moreover, in certain situations, such as air trapping (PEEPi) or respiratory muscle weakness, its conceptualization can be confusing whenever effort and work are assumed to be the same. We should keep in mind that respiratory drive controls inspiratory effort to maintain a balance with ventilatory demands. Thus, inspiratory effort depends on the intensity of the efference from the higher centers and the viscoelastic properties of the respiratory system [19],[34].

The respiratory muscle pump is responsible for maintaining a balance according to metabolic demands and comprises different skeletal muscles that act in a coordinated manner to support alveolar ventilation [22]. The diaphragm carries out the inspiratory muscle pump function [35]. However, when the respiratory load increases and the diaphragm alone cannot respond to the demand, accessory muscles are recruited to contribute to inspiration. Sternocleidomastoid, parasternal, and external intercostal muscles are the main contributing muscles [36]. Therefore, during spontaneous breathing, the fall in the pleural pressure depends on the contraction of the inspiratory muscles (muscle pressure) and the pressure gradient generated over the chest wall [22],[37].

According to the literature, the inspiratory effort is defined as any consumption of energy by the respiratory muscles to conduct respiration [22], which considers the time and type of muscle contraction. However, isometric contractions consume energy and constitute an effort but do not translate into a volume change. This aspect is crucial because it differentiates inspiratory effort from the work of breathing. Not every contraction of the respiratory muscles results in a change of volume, but it always translates into energy consumption. In healthy subjects, inspiration begins at the equilibrium point of the respiratory system. That is, where the collapsing tendency of the lungs is equal to the tendency toward rib cage expansion [38]. At this point, the action of the inspiratory muscles (muscle pressure) causes a fall in pleural pressure and, consequently, generates a volume change.

Definition of work of breathing

In physics, work refers to the force applied to an object and the displacement generated by it [34]; in other words, it corre-sponds to force times displacement (W= F × d). In the respiratory system, work of breathing (WOB) is calculated as the product of the integral of pressure and volume:

WOB = ∫ pressure × volume

I.e., the area in a pressure- volume diagram [13]. Conceptually, it is defined as the work required to cope with the load imposed by the respiratory system. This work can be carried out by the respiratory muscles when breathing is spontaneous, exclusively by the mechanical ventilator if breathing is artificially controlled, or by both in the case of assisted ventilation [39]. To understand the components of the work of breathing, we need to mention the equation of motion. This equation shows that the airway pressure has an elastic component (necessary to overcome the elastance of the parenchyma), a resistive component (the circulation of a flow is constantly opposed by a resis-tive load), and an inertial component (negligible for respiratory frequencies lower than one Hertz) [40]:

P_aw = Ers × V + R_aw × Q + PEEPtot

Where, P_aw: airway pressure, Ers: elastance of the respiratory system, V: volume, R_aw: airway resistance, Q: gas flow, and PEEPtot: positive end-expiratory pressure.

From this equation, it can be deduced that the work of breath-ing must overcome resistive and elastic components and the air trapped at the end of expiration (positive end- expiratory pres-sure) [13].

The concept of work of breathing can be confusing because the classical definition of work states there must always be a change in volume following a change in pressure. However, the consumption of energy from the isometric component of respiratory muscles in healthy and diseased individuals is not included in the classical concept [34].

Non-invasive maneuvers of drive, effort, and work of breathing available in intensive care ventilators

Within the maneuvers available in ventilators, we identified the airway occlusion pressure of the first 100 milliseconds (P_0,1), the expiratory occlusion pressure difference (ΔP_occ), and the pressure muscle index (PMI). In addition, variables calculable with ventilator information such as muscle pressure, pressure- time product, and patient work of breathing are also considered (Table 2).

Table 2. Non-invasive maneuvers to monitor respiratory drive, inspiratory effort and/or work of breathing during pressure support ventilation available in intensive care ventilators used in Chile.

Airway occlusion pressure

Occlusion pressure at the first 100 milliseconds is a non-invasive measurement for drive monitoring, defined by the airway pressure drop during the first 100 milliseconds (0.1 seconds) against an end-expiratory occlusion. It was initially described by Whitelaw et al. in 1975 in healthy individuals [41], later began to be used in critically ill patients [42],[43], and is still a subject of study [44],[45],[46]. It is also a tool used to predict weaning from mechanical ventilation and to adjust respiratory support, even in patients on extracorporeal life support [47]. Occlusion pres-sure in the first 100 milliseconds ranges from 0.5 to 1.5 centimeters of water in healthy individuals and 1.3 to 3.5 centimeters of water in critically ill patients [18]. An occlusion pressure at the first 100 milliseconds greater than 3.5 centimeters of water has a sensitivity of 92% and a specificity of 89% for detecting excessive inspiratory effort [48]. Modern ventilators continuously display occlusion pressure at the first 100 milliseconds by estimating it according to the airway pressure drop during the activation phase or by a brief occlusion maneuver at the end of expiration [49]. Occlusion pressure at the first 100 milliseconds increases proportionally to alveolar carbon dioxide pressure during assisted ventilation and is unaffected by respiratory muscle weakness [47],[50]. As the breath-to-breath variability of occlusion pressure at the first 100 milliseconds is significant, it is recommended to consider an average of three to four measurements [18].

Expiratory occlusion maneuver

With an expiratory occlusion maneuver in assisted ventilation, we can obtain a difference in the expiratory occlusion pressure, i.e., the difference between the baseline pressure (commonly the positive end-expiratory pressure) and the most negative value during the pressure drop after a mechanical occlusion at the end of expiration (usually a five- second expiratory pause) [51]. In simple terms, it is the continuation of the occlusion pressure maneuver to the first 100 milliseconds considering the most negative value.

The expiratory occlusion pressure difference maneuver was first described by Bertoni et al. in 2019 [51] as a non-invasive method to estimate inspiratory effort during assisted ventilation. Expiratory occlusion is commonly used to measure posi-tive intrinsic end-expiratory pressure in controlled ventilation or peak inspiratory pressure (PIP) during maximal inspiratory efforts. Inspiratory effort can also be assessed by airway pressure oscillation during a brief randomly applied end-expiratory occlusion maneuver without maximal voluntary effort. Indeed, the airway pressure oscillation during occlusion correlates with the pleural pressure oscillation [37], as long as this single, brief, unexpected occlusion does not majorly modify the respiratory drive [52].

Considering the differences in esophageal pressure oscillation between occluded and dynamic conditions, it has been pro-posed to multiply the result of the expiratory occlusion pressure difference by 0.75 to obtain a predicted muscle pressure [20]. That is:

P_mus = −0.75 × ΔP_occ

This corresponds to the global force generated by the inspira-tory muscles (https://rtmaven.com). The expected muscle pressure for a patient with adequate inspiratory effort is between 5 and 10 centimeters of water [53]. For expiratory occlusion, pressure differences are between 7 and 15 centime-ters of water [20]. Alternatively, using the expiratory occlusion maneuver, the predicted dynamic transpulmonary driving pres-sure (ΔPL) can be calculated as:

ΔP_L = (peak P_aw − PEEP) −0.66 × ΔP_occ

This equation indicates overdistention with values over 16 to 17 centimeters of the water. The expiratory occlusion maneuver can detect potential lung damage and excessive or absent respi-ratory effort, and it can also be used to differentiate various forms of patient- ventilator asynchronies. This maneuver can unmask ineffective efforts, identify auto triggers, and differentiate double triggers from reverse triggers (in which case there will be apnea during the expiratory occlusion) [54]. Suppose the ventilator can measure the start and end of inspiratory effort during the expiratory pause. In that case, one could estimate the patient’s neural inspiratory time and observe the morphology of the pressure drop given by the slope between the occlusion pressure at the first 100 milliseconds and the difference of the expiratory occlusion pressure. A notable limitation of the expiratory occlusion maneuver is that dynamic hyperinflation and intrinsic positive end-expiratory pressure can underestimate the effort, especially if the intrinsic positive end-expiratory pressure has not equilibrated with the occluded airway [54].

Inspiratory occlusion maneuver

A brief end- inspiratory occlusion maneuver is widely used to measure plateau pressure (P_plat) and driving pressure (ΔP) in controlled mechanical ventilation. Bellani et al. suggested that a brief inspiratory occlusion maneuver allows reliable plateau pressure measurements even in assisted mechanical ventilation [55]. During an inspiratory occlusion in mechanically assisted ventilation, patients relax the inspiratory muscles, increasing airway distending pressure, which is detectable on the ventilator waveform.

The inspiratory occlusion maneuver in assisted ventilation is not new. Foti et al. first described it in 1997 as a method to assess the pressure developed by the inspiratory muscles in patients with acute lung injury during pressure support ventilation [56]. When the patient is over-assisted, and the respiratory effort is low, the airway pressure falls below the plateau pressure during inspiratory occlusion in assisted ventilation. On the contrary, when the airway pressure rises above the plateau pressure value, it could indicate a high respiratory effort or under-assistance, indicated by the magnitude of the relaxation of the inspiratory muscles. The difference between airway and plateau pressure during an inspiratory occlusion is known as PMI, P_{musc index}, or pressure muscle index [56]. During pressure support ventilation, the muscle pressure index is calculated as:

PMI = P_plat − (PEEP + PS)

Where the P_plat is the pressure reached during inspiratory occlusion. Values of pressure muscle index greater than three centimeters of water indicate an elevated inspiratory effort [56]. Part of the reasoning behind the pressure muscle index was described by Prinianakis et al. in 2008 while studying the cause of the increased ventilator pressure above the end- inspiratory pressure support during pressure support ventila-tion [57]. Since inspiratory muscle relaxation precedes expiratory muscle contraction, the increase in ventilator pres-sure using pressure support ventilation may imply the end of neural inspiration. However, the technique used to measure pressure muscle index is not without limitations. Because the pressure is obtained under quasi- static conditions, this mea-surement may be affected by Pendelluft mechanisms, espe-cially in patients with acute respiratory distress syndrome [58]. Moreover, the airway pressure tracing may be unstable during occlusion, which could confound inspiratory muscle relaxation with abdominal muscle activity that may rapidly increase air-way pressure in the onset of neural expiration during inspira-tory occlusion [59].

Calculations derived from ventilator information

Current ventilators provide estimated values of inspiratory effort and work of breathing variables without needing a probe or additional equipment, including the pressure-time product and the patient’s work of breathing parameters. The methods available in ventilators to approximate these parameters – derived initially from esophageal balloons [60] – are not entirely accurate. For this reason, they should be taken with caution, as each ventilator has different pressurization characteristics [61]. Some ventilators deliver the pressure- time product or patient’s work of breathing based on estimates of the area of the pressure-time curve.

Particularly, proportional assist ventilation with adjustable load was initially described by Magdy Jones in 1991 as a ventilatory mode that delivered support according to patient demand, considering the mechanical properties of the respiratory system [62]. Since 2007, proportional assist ventilation with load- adjustable gain factors (PAV+) delivers ventilator and patient’s elastic and resistive work of breathing values from a brief 300- millisecond inspiratory occlusion. From this analysis, we can calculate the plateau pressure and, consequently, the main components of the equation of motion – i.e., compliance and resistance [63],[64]. When using proportional assist ventilation with load- adjustable gain factors, the expected values for a good work of breathing range from 0.3 to 0.7 Joules per liter. Since 2013, proportional assist ventilation with load- adjustable gain factors permits calculating muscle pressure and pressure- time product using Carteaux et al. formulas [53] without an esophageal balloon. Therefore, muscle pressure can be calculated as:

Pmus = (P_peak – PEEP) × (% support ÷ % patient)

Where % support is the percentage of assistance set in the ventilator and % patient is the percentage of work performed by the patient (https://bit.ly/3kj6oTz). Consequently, by obtain-ing the muscle pressure, it is possible to calculate:

PTP = [(P_mus x inspiratory time) × respiratory frequency] ÷ 2.

Estimation of the pressure-time product (PTP) was only possi-ble with proportional assist ventilation with load- adjustable gain factors until the expiratory occlusion pressure difference was described using the expiratory occlusion maneuver, from which muscle pressure is obtained. However, – although it has not yet been studied – it is now possible to estimate the pressure- time product during ventilation with pressure support using the muscle pressure calculation with the expiratory occlusion maneuver.

Waveforms and asynchronies

The presence of asynchronies in pressure- time and flow- time curves can alert high respiratory drive and high respiratory effort, particularly in flow starvation, premature cycling, and double triggering [18]. In general, the interpretation of the ventilatory waveform may help identify a mismatch between neural and mechanical inspiratory time, which is common under pressure support ventilation [65]. Just as the slopes of pressure and flow versus time waveforms can help measure the ’stress index' and ineffective efforts, they can also help estimate the level of inspiratory effort.

Recently, a new method based on the concavity of inspiratory flow waveform called 'flow index' was described [66]. Although this method still requires more studies on clinical applicability without additional equipment, it allows approaching the inspi-ratory effort concept using the analysis and interpretation of the ventilatory waveform to adjust pressure support ventilation according to the patient’s effort level and tidal volume [63].

Future directions

It is often assumed that deep sedation minimizes or abolishes spontaneous inspiratory effort ("respirolysis") [67]. Future management of mechanical ventilation aims to reduce deep sedation as routine clinical practice by including other alterna-tives for inspiratory effort control [68]. First, by developing innovative ventilator features that improve patient- ventilator synchrony, such as neurally adjusted ventilatory assist. Second, by using extracorporeal support in patients with severe acute respiratory distress syndrome to eliminate excess carbon diox-ide as an undesirable stimulus to the respiratory center [67],[68]. Finally, by approaching current mechanical ventilation settings based on respiratory muscle demand to mitigate asynchronies without excessive sedation.

Conclusions

The differences between respiratory drive, inspiratory effort, and work of breathing concepts are essential to understanding the interaction between the patient’s respiratory muscle pump and the ventilatory support.

These three concepts are co-dependent and can be monitored noninvasively using maneuvers or tools from intensive care mechanical ventilators.

Nevertheless, not all ventilators can measure respiratory drive, inspiratory effort, or work of breathing variables using pressure support ventilation. The most commonly available maneuvers are the occlusion pressure in the first 100 milliseconds and expiratory occlusion pressure. These maneuvers have growing sci-entific evidence and could be used to monitor drive and inspiratory effort in more than 60% of ventilators used in Chile. Moreover, the accessibility of these maneuvers is even more relevant in pandemic times due to the large volume of critically ill patients.

Future studies should explore the accuracy and relevance of the non- invasive maneuvers identified in this review.

Notes

Contributor roles
FR- C: conceptualization, methodology, research, data curation, preparation of the original manuscript, review and editing of the article, visualization, and project management. FG- S: conceptualization, methodology, research, data curation, original manuscript preparation, manuscript review and editing, visualization, and project management. JM: conceptualization, manuscript review and editing, and supervision.

Acknowledgments
We thank the following institutions and collaborators during the field review of the intensive care ventilators: Hospital Metropolitano, Santiago (Nicolás Aguirre and Javier Leiva), Clínica INDISA, Santiago (Felipe Castillo Merino and Juan Enrique Lee Goic), Hospital del Salvador, Santiago (Antonieta Rosales and Andrea Fuentes), Hospital Eduardo Pereira, Valparaíso (Sebastián Castro), Clínica Bicentenario, Santiago (Nicolás Piñeiro), Hospital San Pablo, Coquimbo (Eduardo González), Critical Patient Department, Clínica Alemana, Santiago (Agustín Camus- Molina) and DTS - Neyün, Santiago (Roger Burgos, Patricia Espinosa, Michael Espinoza).

Competing interests
The authors declare no competing interests with this work.

Funding
The authors declare that this work was not financed.

Ethics
This article is exempt from approval by an accredited ethics committee because it is a narrative review article.

Provenance and peer review
Not commissioned. Externally peer- reviewed by three reviewers, double- blind.

Language of submission
Spanish

Table 1. Definitions of respiratory drive, inspiratory effort, and work of breathing identified in the literature.

Figure 1. Graphical definition and alternatives for monitoring respiratory drive, inspiratory effort, and work of breathing in intensive care ventilators.

Table 2. Non-invasive maneuvers to monitor respiratory drive, inspiratory effort and/or work of breathing during pressure support ventilation available in intensive care ventilators used in Chile.

This work is licensed under a Creative Commons Attribution 4.0 International License.