Choosing the First Ventilator Mode in Critically Ill Patients

FOR RELATED ARTICLE, SEE PAGE 912

Invasive mechanical ventilation (IMV) is a routine treatment for critically ill patients in the ICU. Although indications are various, IMV remains associated with high hospital mortality rates (up to 40%), notably in large, unbiased epidemiologic cohort studies.¹ While lifesaving, inappropriate ventilator settings can induce or worsen lung injuries, broadly known as ventilator-induced lung injury.²

Since Vesalius’s first description of positive pressure ventilation in the mid-16th century and after centuries of neglect, negative pressure ventilation (“iron lung”) was used briefly from the late 19th century to the 1950s. Modern positive pressure ventilation gained momentum during the 1952 polio epidemic, which led to a significant reduction in mortality rates through tracheostomy and positive pressure ventilation.³

However, widespread positive pressure ventilation adoption soon revealed adverse effects, especially pneumothorax, in both experiments and clinical practice. A key experimental study by Dreyfuss et al⁴ showed the harmful effects of excessive tidal volumes, later termed “volutrauma.” The landmark ARDS Network's Lower Tidal Volumes (ARMA) trial⁵ marked a shift toward lung-protective strategies, focusing on minimizing tidal volumes and distending pressures, which included plateau and driving pressures.

Although efforts focused on protective strategies with low tidal volumes and plateau pressures, less attention was given to determining the optimal ventilation mode, despite the growing variety of ventilator modes.

In this issue of CHEST, Seitz et al⁶ conducted a pragmatic, unanonymized, cluster-randomized, cluster-crossover trial that compared volume control (VC), pressure control (PC), and adaptive pressure control (APC) modes in critically ill adults whose condition required IMV in an academic ICU. Over 9 months, all consecutive patients in the ICU who needed IMV were enrolled. The primary outcome was ventilator-free days through day 28, according to mode allocation: VC, PC, or APC. Secondary outcomes included feasibility (eg, time spent in the assigned mode during the first 72 hours), safety (eg, proportion of breaths exceeding the 8 mL/kg predicted body weight [PBW] tidal volume target), and clinical outcomes (eg, in-hospital mortality at day 28). Across groups, all patients targeted a tidal volume of 6 mL/kg PBW and plateau pressure < 30 cm H₂O. The primary outcome was reanalyzed with adjustments for prespecified covariates and included patients from washout periods.

Concerning the primary outcome, the number of ventilator-free days through day 28 did not significantly differ among the 3 study groups, with a median of 23 days (interquartile range [IQR], 0 to 26) in the VC group, 22 days (IQR, 0 to 26) in the PC group, and 24 days (IQR, 0 to 26) in the APC group (P = .60). This was not different after adjustment for confounders. Concerning exploratory outcomes, although the median tidal volume was similar within the 3 groups, the percentage of breaths larger than 8 mL/kg of PBW differed among VC control (median, 4.0%; IQR, 0.0 to 14.1), PC (10.6%; IQR, 0.0 to 31.5), and APC (4.7%; IQR, 0.0 to 19.2) groups (P < .001). Besides, the mean peak inspiratory pressure also differed significantly among groups, with higher peak pressures in the VC group than the PC and APC groups (P = .006). Other secondary outcomes that assessed safety (eg, pneumothorax) and clinical outcomes (eg, in-hospital mortality rates at day 28) did not differ significantly across the 3 allocation groups.

How can we translate these findings into daily clinical practice?

At first glance, one might argue that only the outcome truly matters. However, a closer look at the methods reveals important nuances, particularly how ventilator alarms were configured depending on the ventilation mode. For instance, in the VC and APC groups, the high peak pressure alarm was set at 40 cm H₂O; in the PC group, the low minute ventilation alarm was set 2 L/min below the patient’s current minute ventilation. This setup carries implications: In the PC group, if the high tidal volume alarm is not adjusted precisely, tidal volumes can exceed the target range without triggering an alert. Conversely, in the VC and APC groups, plateau pressures can surpass the recommended threshold of 30 cm H₂O without reaching the high peak pressure alarm, particularly in cases of low respiratory system resistance. In short, you can effectively control only what is precisely configured, be it volume or pressure. Besides, although the clinical impact of higher tidal volume cannot be assessed directly in this study, such volumes may be particularly harmful in patients with more impaired lung compliance.⁷ Also, physiologic measurements of respiratory system compliance and patient-ventilator synchrony were not investigated. However, the authors underlined that scarce data^8,9 have been published that properly evaluate the effects on clinical outcomes of volume-controlled mode vs pressure-controlled mode. These trials preclude the adoption of protective lung strategies, daily interruption of sedation, rigorous spontaneous breathing trial, and the use of adaptive pressure-controlled ventilation. Recently published, a large French multicenter randomized open-label trial (n = 700) has compared a conventional volume assist-control strategy with a pressure-controlled strategy that allowed nonsynchronized, unassisted, spontaneous breathing in patients with ARDS.¹⁰ Although the primary outcome (60-day in-hospital mortality rate) did not differ significantly among groups, some differences concerning secondary outcomes (eg, need for sedation and adjunctive therapies of hypoxemia) were observed. Together, these studies are essential to validate and support the use of different ventilation modes in various clinical scenarios. They give to the clinicians the ability to choose the most appropriate ventilation mode for each patient while respecting lung-protective ventilation strategies.

Financial/Nonfinancial Disclosures

The authors have reported to CHEST the following: C. G. received consulting fees from Xenios FMC and Ambu companies outside the submitting work, reports a relationship with Xenios AG that includes consulting or an advisory position, and reports a relationship with Ambu Inc that includes consulting or an advisory position. None declared (D. B.).