Clinical Management Strategies for Airway Pressure Release Ventilation: A Survey of Clinical Practice

Discussion

Our survey results showed that APRV was the first choice for most respondents for patients failing conventional mechanical ventilation. However, there was only modest consensus for initial settings used during APRV among those surveyed. In addition, respondents' initial management strategies often deviated from published guidelines.^10,11

APRV offers some interesting physiologic features that may have utility in the management of patients with acute respiratory failure. However, APRV is not an on-off switch; it requires manipulation of 4 key variables. At the present time, the optimal way to configure these variables for various clinical scenarios is largely based on physiologic concepts, small trials, and clinical experience rather than on results from large randomized trials. As a consequence, interpreting trials and making recommendations about APRV is problematic.

The most marked variability in survey responses and deviation from APRV published guidelines was in the choice of T low. Both published guidelines^10,11 recommend using expiratory flow analysis to ensure that the set T low creates significant auto-PEEP. However, 39% of our respondents used an arbitrary set time instead. The use of auto-PEEP rather than set PEEP to maintain alveolar recruitment is controversial. Whereas set PEEP distributes to all alveolar units, auto-PEEP is most prominent in alveolar units with slow emptying (high-compliance/high-resistance units) and is least prominent in alveolar units with fast emptying (low-compliance/low-resistance units).¹³ Conceptually, this would seem opposite of the clinical goals for PEEP in patients with alveolar inflammation, flooding, or collapse. Clinical trials are clearly needed to sort this out.

A high percentage of respondents appeared to expose patients to potentially harmful end-inspiratory and tidal lung distention. Current standards for lung-protective conventional ventilation are to limit end-inspiratory P_plat to < 30 cm H₂O and tidal volume to < 8 mL/kg.^14,15 The 2 published APRV guidelines^10,11 do not specifically address these parameters. Instead, the only recommendation is to limit P high to < 30–35 cm H₂O, a limit that 36% of our respondents were willing to exceed. Importantly, limiting P high is not the same as limiting P_plat, and the harm from exceeding P high may be worse than that from exceeding P_plat. Specifically, P_plat, by definition, is measured in the absence of patient effort and under no-flow conditions. In the setting of nearly normal chest wall mechanics, P_plat is an approximation of the end-inspiratory transpulmonary pressure, the actual stretching pressure on lung tissue. In contrast, APRV encourages patient effort and patient-generated flow (often augmented with pressure support) that can take the end-inspiratory trans-pulmonary pressure well above the set P high. Targeting a P high of < 30–35 cm H₂O may thus underestimate the risk of overdistention during APRV. We believe that this important point is often underappreciated by clinicians managing patients receiving APRV.

The published APRV guidelines^10,11 pay little attention to monitoring tidal volumes, a parameter increasingly appreciated as an independent risk factor for ventilator-induced lung injury.^14,15 Reasonable data support generally limiting tidal volumes to < 8 mL/kg (ideal body weight).^14,15 However, 44% of our respondents were willing to exceed this value for the release volume. Importantly, the effective tidal distention of the lung during APRV is not only the release volume but also whatever volume is added during spontaneous breathing (with or without pressure support).

Although we did not specifically address driving pressure in our survey, this is another parameter that could reflect regional tidal alveolar overdistention during APRV.¹⁴ The driving pressure during conventional ventilation is the P_plat minus the measured PEEP. However, this is not equivalent to P high minus P low in APRV. This is because of the potential differences in set P high versus a true P_plat noted above and because the actual total PEEP in the lungs is the set P low plus the unevenly distributed intentional auto-PEEP. Thus, determining a driving pressure during APRV is clearly problematic and in need of considerable further study before being incorporated into APRV management algorithms.

Our survey has important limitations, and thus results should be interpreted cautiously. The sample size of 60 is small, and the sampling method used carries self-selection bias because only members of the AARConnect Adult Acute Care Section were surveyed. The small sample size also prevented any analysis of regional differences in clinical practice. However, respondents were probably well-versed in APRV, had a special interest in APRV, and were motivated to contribute to our understanding of APRV. Our survey results therefore would seem to provide exploratory data opening a window into the clinical management of APRV.

Another important limitation of our study is simply the nature of a voluntary survey. Questions must be short and few in number to obtain a reasonable response level. Details of the clinical scenario surrounding a question are thus limited, and nuances cannot be explored. Finally, the terminology surrounding APRV is confusing, a situation worsened by manufacturers insisting on proprietary names and non-standardized ways to display settings and monitors.