Prediction and estimation of pulmonary response and elastance evolution for volume-controlled and pressure-controlled ventilation

Highlights

• A physiologically relevant, predictive lung mechanics virtual patient model.

• Nonlinear elastance evolution is captured and predicted with the newly proposed function.

• Pressure–volume loops and peak pressures/volumes are accurately predicted.

• The model is validated for both pressure controlled and volume controlled mechanical ventilation modes.

• The basis function approach used is computationally fast and can be implemented for real-time.

Abstract

Mechanical ventilation (MV) is a core treatment for patients suffering from respiratory disease and failure. However, MV settings are not standardized due to significant inter- and intra- patient variability in response to care, leading to variability in outcome. There is thus a need to personalize MV settings. This research significantly extends a single compartment lung mechanics model with physiologically relevant basis functions, and uses it to identify patient-specific lung mechanics and predict response to changes in MV settings. Nonlinear evolution of pulmonary elastance over positive end expiratory pressure (PEEP) is modelled by a newly proposed, physiologically relevant and simplified compensatory function to enable prediction of pulmonary response for both volume-controlled ventilation (VCV) and pressure-controlled ventilation (PCV), and identified as patient-specific using each patient’s data at a baseline PEEP. Predictions at higher PEEP levels test the validity of the proposed models based on errors in predicted peak inspiratory pressure (PIP) in two VCV trials and volume (PIV) in one PCV trial. A total of 210 prediction cases over 36 patients (22 VCV; 14 PCV) yielded absolute predicted PIP errors within 1.0 cmH₂O (2.3%) and 3.3 cmH₂O (7.3%) for 90% cases in VCV, while predicted PIV errors are within 0.073 L (16.8%) for 85% cases in PCV. In conclusion, a novel deterministic virtual patient model is presented, able to offer accurate prediction of pulmonary response across a wide range of PEEP changes for the two main MV modes used clinically, enabling predictive decision support in real-time to safely personalize and optimize care.

Introduction

Acute lung injury (ALI) and acute respiratory distress syndrome (ARDS) have detrimental impacts on lung stiffness and oxygenation resulting in high morbidity and mortality [1], [2], [3], [4]. Mechanical ventilation (MV) is the core treatment for patients suffering from life-threatening respiratory failure in the intensive care unit (ICU). The primary goal of MV is to minimise the work of breathing, ensure adequate gas exchange, and recruit and hold open lung volume to enable recovery [5], [6], [7], [8], [9]. However, suboptimal MV settings can lead to over-distension and ventilator induced lung injury (VILI), both of which increase morbidity and mortality [1], [6], [8], [9], [10], [11], [12]. To avoid these harmful effects, protective MV settings have been proposed [1], [4], [13], [14], [15].

Low tidal volume is clinically well-accepted in MV to mitigate VILI, but can lead to alveolar de-recruitment [16], [17], [18]. Thus, a protective ‘open lung’ approach uses positive end-expiratory pressure (PEEP) during breathing to prevent alveolar collapse combined with low tidal volume ventilation, thus maintaining an open lung at the end of expiration to ensure sufficient oxygenation and pressure support [6], [8], [9], [18], [19], [20], [21]. Staircase recruitment maneuvres (RMs), comprising a series of incremental and decremental PEEP steps, have been used as one important component in clinical care for recruiting lung volume for lung protective strategies and assessing the choice of PEEP [6], [15], [22], [23], [24].

RMs with subsequent PEEP changes can be effective in improving oxygenation, while minimising harm [20], [25], [26], [27]. However, optimal RM and PEEP settings remain patient-specific, time-varying, and thus not standardized [11], [12], [18], [28], [29], [30], [31]. The ‘best’ setting can be different between patients, as well as varying over different conditions and time [5], [14], [32], [33], [34], [35]. An suboptimal PEEP setting can lead to excessive or insufficient support for patients, inducing VILI and leading to higher morbidity and mortality [36], [37]. Thus, it is critical to to provide clinicians with better information to monitor patient-specific pulmonary state and forecast the influence of new PEEP settings on pulmonary response for each patient, to improve and personalize care, minimize risk, and maximize care and safety [8], [19], [34], [35]. Therefore, accurate, predictive and patient-specific MV strategies are a major need in advancing care and minimising MV-associated injury [5], [19], [34].

Furthermore, two different MV modes are both widely used, volume controlled ventilation (VCV) and pressure controlled ventilation (PCV) [38], [39], [40]. VCV allows clinicians to control tidal volume directly, eliminating volutrauma, but they need to be alert to the resulting peak inspiratory pressure (PIP) and barotrauma. Conversely, PCV controls pressure, but risks volutrauma from too large a peak inspiratory volume (PIV) [38]. Both limitations may lead to unexpected VILI [39], [41]. To date, no noticeable clinical outcome differences have been seen comparing VCV and PCV [42], [43]. Thus, the decision on MV strategies relies on clinician preference, patient characteristics, or patient comfort [38], [39]. Therefore, accurate, model-based, and patient-specific pulmonary response prediction is necessary.

In the last two decades, several complex models have been proposed and can effectively capture a large range of nonlinear pulmonary dynamics [34], [44], [45], [46], [47], [48], [49], [50], [51]. However, their complexity means they suffer poor or non-identifiability [34], [52], or are too complex to identify or apply at the bedside [53], [54], [55], [56], [57], [58], [59], thus limiting or eliminating their potential for clinical application. Far simpler black box models can be created, but require large amounts of data to train and may lack the ability to capture or describe all physiological features in various situations [58], [60], [61]. In addition, physiological relevance is important because it supports clinical confidence and use and provides further insight to clinical end-users [52], [62], but such black-box models cannot offer physiological relevance. Finally, some models of all types capture lung mechanics well with good personalization of parameters, but are poor in predicting the response to changes in care [63], [64], [65], lacking the means to offer guidance to clinicians, and suggesting the identified parameters may not be correct. However, accurate prediction is a major need in guiding and improving the safety and efficacy of clinical MV treatment [66], [67], [68]. Thus, there is a need for simpler, identifiable, physiologically relevant models which can offer accurate prediction to changes in care by capturing the evolution of lung mechanics as MV parameters change [34], [52].

Currently, while several models can identify data [9], [11], [44], [69], [70], [71], the authors are aware of only one approach able to accurately predict outcomes from changes in MV care for resulted airway pressure for VCV and tidal volume for PCV[8], [19], [72]. These approaches use physiologically relevant basis functions to define respiratory mechanics over all possible pressures and volumes seen in MV, which is uncommon [35], [58]. However, while they predict well, the basis functions proposed are independent of known PEEP levels and changes, creating complexity in understanding and implementation, especially for PCV predictions. Thus, if accurate prediction can be obtained with a simpler model where elastance evolution as PEEP changes is captured as a function of PEEP, it could provide an easier, more intuitive, and clinically applicable approach in clinical use.

This research presents physiologically relevant, simpler, basis functions to estimate elastance and resistance evolution as MV parameters change using the same clinically validated single compartment lung mechanics model as previous studies [59], [73]. It is validated by assessing prediction error when made patient specific using data from one single PEEP level to predict pressure and flow at higher PEEP levels. Such a predictive model would offer the ability to quantify the trade-off or compromise between increasing basis function simplicity and improving clinical utility.