Objective: To compare the ease and efficacy of two commonly used methods for choosing optimal positive end-expiratory pressure (PEEP) in patients with acute respiratory distress syndrome: a static pressure-volume curve to determine the lower inflection point (P(flex)) and the "best PEEP" (PEEP(best)) as determined by the maximal compliance curve.
Design: Prospective study.
Setting: Medical and respiratory intensive care units of university-associated tertiary care hospital.
Patients: Twenty-eight patients on mechanical ventilation with acute respiratory distress syndrome.
Interventions: A critical care attending physician or fellow and an experienced respiratory therapist attempted to obtain both static pressure-volume curves and maximal compliance curves on 28 patients with acute respiratory distress syndrome by using established methods that were practical to everyday use. The curves then were used to determine both P(flex) and PEEP(best), and the results were compared.
Measurement and main results: Our results showed at least one value for optimal PEEP was obtained in 26 of 28 patients (93%). P(flex) was determined in 19 (68%), a PEEP(best) in 24 (86%), and both values in 17 (61%). In patients who had both P(flex) and PEEP(best) determined, there was a close concordance (+/-3 cm H2O) in 60%. When the values of P(flex) and PEEP(best) were interpreted by two additional investigators, there was unanimous agreement on the P(flex) (+/-3) only 64% of the time. There was agreement on the value of PEEP(best) 93% of the time.
Conclusions: Our data show that optimal PEEP, as determined by a pressure-volume curve and a maximal compliance curve, are sometimes unobtainable by practical means but, when obtained, often correspond. A maximal compliance is more often identified, has less interobserver variability, and poses less risk to the patient. We conclude that determining optimal PEEP by maximal static compliance may be easier to measure and more frequently obtained at the bedside than by using a static pressure-volume curve.