Defining rules for the identifcation of critical ventilatory events

Abstract

The automated recognition of critical clinical events by physiological monitors is a challenging task exacerbated by a lack of standardized and clinically relevant threshold criteria. The objective of this investigation was to develop consensus for such criteria regarding the identification of three ventilatory events: disconnection or significant leak in the anesthesia circuit, decreased lung compliance or increased resistance, and anesthetic overdose from inhaled anesthetics. We individually administered a structured interview to five expert anesthesiologists to gain insight into the cognitive processes used by clinicians to diagnose ventilatory events and to determine the common heuristics (rules of thumb) used in clinical practice. We then used common themes, identified from analysis of the structured interviews, to generate questions for a series of web-based questionnaires. Using a modified Delphi technique, iterative questionnaire administration facilitated rapid consensus development on the thresholds for the specific rules used to identify ventilatory events. A threshold for 75% agreement was described for each scenario in a healthy ventilated adult. A disconnection or significant leak in the anesthesia circuit is diagnosed with peak airway pressure (< 5 cm H2O or change of 15 cm H2O), ETCO2 (0 mmHg, 40% drop, or value below 10 mmHg for a duration of 20 sec), and inspired-expired volume difference (300 mL). Increased resistance or decreased lung compliance is diagnosed with high peak airway pressure (40 cm H2O or a 20 cm H2O change), asymmetry of capnogram, and changes in measured compliance or resistance. Anesthetic overdose from inhaled anesthetics is diagnosed with high end-tidal anesthetic agent concentration (2 MAC in a patient less than 60 yr of age or 1.75 MAC in a patient over 60 yr of age), low systolic blood pressure (below 60 mmHg), and low modified electroencephalogram (bispectral index or entropy). This investigation has provided a set of consensus-based criteria for developing rules for the identification of three critical ventilatory events and has presented insight into the decision heuristics used by clinicians.