Introduction
SARS-CoV-2 (COVID-19)-associated pneumonia has generated a global pandemic, causing substantial morbidity and mortality in patients with respiratory failure.1 Ventilatory support, including mechanical ventilation, is essential to improve gas exchange and allow patient recovery. Despite its benefits, mechanical ventilation may induce further lung injury and promote an exacerbated systemic inflammatory response, a phenomenon known as ventilator-induced lung injury (VILI).2 Several strategies such as the limitation of tidal volume and driving pressure (DP) and the use of prone positioning and neuromuscular blockers have been shown to prevent VILI and have significantly improved the prognosis of patients with ARDS.3-7 Respiratory rate is a fundamental ventilatory parameter related to the energy applied to the lung parenchyma (ie, mechanical power) and constitutes a central determinant of minute ventilation and respiratory homeostasis. Even though experimental studies have shown that high frequency may aggravate lung injury,8,9 the effect of reducing frequency while maintaining CO2 and pH within safe limits remains unclear and has not been formally tested in clinical trials.
In the present study, we assessed whether a specific protocol for decreasing respiratory rate is feasible in subjects with COVID-19-associated ARDS and evaluated the impact of reducing frequency on hemodynamics, respiratory mechanics, and mechanical power.
Methods
We performed a randomized crossover trial at the Clinical Hospital of the Pontificia Universidad Católica de Chile. The study was approved by the Research Ethics Committee of the Pontificia Universidad Católica de Chile (protocol ID: 180813016).
We included subjects with COVID-19–related ARDS within the first 48 h of mechanical ventilation, PaO2/FIO2 < 200 mm Hg, and requiring deep sedation and neuromuscular blockade. Main exclusion criteria were previous chronic respiratory disease; hypercapnic respiratory failure; concomitant severe metabolic acidosis; or catastrophic respiratory failure, defined as PaO2/FIO2 < 80 despite optimization of ventilatory parameters; or need for extracorporeal membrane oxygenation. All subjects were intubated and mechanically ventilated in volume-controlled mode based on accepted protocols: tidal volume initially set at 6 mL/kg ideal body weight and PEEP and FIO2 adjusted according to the ARDS Network table and corrected to achieve a DP ≤ 15 cm H2O. If PaO2/FIO2 was < 150 mm Hg, subjects were positioned in the prone position. All subjects used active humidification.
The duration of the protocol was 24 h. This protocol included 12 h with a low frequency and 12 h with a high frequency, where its sequence was randomly selected. To achieve the study’s aim, an approximate difference of 8–10 breaths/min between the high- and low-frequency periods was considered appropriate. These target frequencies were selected based on a designed nomogram that considers baseline frequency, pH, and PaCO2 as well as maintaining pH and PaCO2 within safety limits (pH 7.20–7.45 and PaCO2 35–65 mm Hg); see Figure 1. The nomogram was built using the following equations: PaCO2(b)=PaCO2(a)−2[frequency(b)−frequency(a)]; and pH(b)=pH(a)+.013[frequency(b)−frequency(a)], where (a) and (b) correspond to the observed and expected value, respectively.
Algorithm for setting target frequency. High and low frequency was set according to baseline frequency, pH, and PaCO2 variables. The first step is to locate pH and PaCO2 values obtained at baseline in the corresponding frequency column (circles). Then low and high target frequencies correspond to the very end positions where pH and PaCO2 are safe (horizontal lines and dashed circles). Finally, target frequencies are defined when both pH and PaCO2 are maintained within the safe limits and a minimum of 8-point difference is present. In this example, in a patient with a baseline frequency of 20 breaths/min, pH 7.28, and PaCO2 52 mm Hg, the low and high target frequency selected were 18 and 28 breaths/min, respectively.
Changes from baseline frequency were made in a stepwise manner either by increasing or decreasing frequency by 4 points each h until the target frequency was obtained. Inspiratory-expiratory ratio was maintained constant during the whole period, and arterial blood gases were taken at baseline, 2, 6, 12, 14, 18, and 24 h as safety measurement. Hemodynamics variables and respiratory mechanics were assessed at baseline and the end of each period (ie, high frequency and low frequency). The latter was measured using end-inspiratory and end-expiratory occlusion maneuvers as previously described.10 Mechanical power was estimated based on the original formula,11 and no recruitment maneuvers were performed.
Data are expressed as median and 25–75% interquartile range. The paired Wilcoxon signed-rank test was used to compare continuous variables in high- and low-frequency periods. All statistical analyses were performed by R software version 3.3.2 (R Foundation for Statistical Computing, Vienna, Austria). A 2-sided P < .05 was considered statistically significant.
Results
We enrolled 10 proned subjects (10 males, age 56 [50–66] y, PaO2/FIO2 112 [79–136], and Acute Physiology and Chronic Health Evaluation score 13 [11–15]). The median period from the onset of symptoms to study enrollment was 2 [2–3] d. Six subjects were hypertensive; two had coronary artery disease, and two had diabetes mellitus. The frequency modification was feasible, and target frequency was achieved within 2–6 h with a significant difference in median set frequency between study periods (27 [26–32] breaths/min vs 18 [16–22] breaths/min, P < .001). PaCO2 and pH were kept within recommended limits in both study sequences except for one subject who had excessive PaCO2 at hour 6. As compared with high-frequency period, levels of PaCO2 were significantly higher when reducing frequency (52 [46–60] mm Hg vs 38 [35–41] mm Hg, P < .001). The evolution of frequency, PaCO2, and pH values throughout the study period are shown in Figure 2. Minute ventilation and the energy applied to the respiratory system were significantly lower when using low frequency versus high frequency (7.7 [6.4–8.3] L/min vs 10.9 [9.8–13.2] L/min, P < .001; and 15 [13–18] J/min vs 22 [20–31] J/min, P = .002, respectively). No significant changes in hemodynamic and respiratory mechanics parameters were observed. All subjects finished the protocol, and no additional intervention was needed within the study period. One subject died within 28 d.
Evolution of respiratory rate, PaCO2, pH, and minute ventilation over time. Modification in frequency was feasible throughout the study period, and a significant difference was observed in frequency between the low- and high-frequency periods in both randomization sequences. A: Changes in PaCO2; B: respiratory rate; C: pH; and D: minute ventilation observed during the study protocol.
Discussion
In the present short report, we showed that lowering the frequency and, therefore, the energy applied to the lungs was feasible in subjects with COVID-19-associated ARDS. Results from this study reveal that the decrease in frequency, tolerating moderate hypercapnia and in the context of low tidal volume ventilation, was not associated with hemodynamic or respiratory mechanics changes as compared with a high respiratory rate strategy.
Recent studies have shown that it is possible to protect the lungs, even against a high tidal volume, by decreasing the frequency and thus the energy applied to the lungs.11 In our study, we made modifications to frequency while maintaining a low tidal volume, and we did not observe major hemodynamic or respiratory mechanics changes with the low or high frequency. The 29% decrease in minute volume that we observed with the low frequency, along with a 32% decrease in mechanical power, may denote a more protective ventilation, but this should be balanced against the physiological and clinical impact of mild or moderate hypercapnia.12 For instance, hypercapnic acidosis has been associated with increased mortality in subjects with traumatic brain injury, and potential for bacterial proliferation in sepsis has been proposed.13 Moreover, changes in minute ventilation do not impact PaCO2 efficiently when high dead space is present as in late COVID-19 ARDS. Hence, the prolonged effects of reduction of frequency may be further studied, assessing not only its impact on the respiratory system but in clinical outcomes.
In this feasibility study of subjects with COVID-19-associated ARDS, we could lower the frequency, while maintaining adequate pH and PaCO2 values, and thus decrease minute ventilation and mechanical power on the lungs. Whether this protocol reducing frequency has an impact on VILI development (eg, inflammatory cytokines) and consequently on outcomes remains to be elucidated.
Footnotes
- Correspondence: Guillermo Bugedo MD, Departamento de Medicina Intensiva, Facultad de Medicina, Pontificia Universidad Católica de Chile, P.O. Box 114D, Diagonal Paraguay 362, 6° piso, Santiago, Chile. E-mail: gbugedo{at}gmail.com
The authors have disclosed no conflicts of interest.
The study was funded by Fondo Nacional de Desarrollo Científico y Tecnológico (Fondecyt), number 1191315-2019.
- Copyright © 2022 by Daedalus Enterprises