A rational framework for selecting modes of ventilation

Abstract

Mechanical ventilation is a life-saving intervention for respiratory failure and thus has become the cornerstone of the practice of critical care medicine. A mechanical ventilation mode describes the predetermined pattern of patient-ventilator interaction. In recent years there has been a dizzying proliferation of mechanical ventilation modes, driven by technological advances and market pressures, rather than clinical data. The comparison of these modes is hampered by the sheer number of combinations that need to be tested against one another, as well as the lack of a coherent, logical nomenclature that accurately describes a mode. In this paper we propose a logical nomenclature for mechanical ventilation modes, akin to biological taxonomy. Accordingly, the control variable, breath sequence, and targeting schemes for the primary and secondary breaths represent the order, family, genus, and species, respectively, for the described mode. To distinguish unique operational algorithms, a fifth level of distinction, termed variety, is utilized. We posit that such coherent ordering would facilitate comparison and understanding of modes. Next we suggest that the clinical goals of mechanical ventilation may be simplified into 3 broad categories: provision of safe gas exchange; provision of comfort; and promotion of liberation from mechanical ventilation. Safety is achieved via optimization of ventilation-perfusion matching and pressure-volume relationship of the lungs. Comfort is provided by fostering patient-ventilator synchrony. Liberation is promoted by optimization of the weaning experience. Then we follow a paradigm that matches the technological capacity of a particular mode to achieving a specific clinical goal. Finally, we provide the reader with a comparison of existing modes based on these principles. The status quo in mechanical ventilation mode nomenclature impedes communication and comparison of existing mechanical ventilation modes. The proposed model, utilizing a systematic nomenclature, provides a useful framework to address this unmet need.