Study objective: To investigate patient-ventilator trigger asynchrony (TA), its prevalence, physiologic basis, and clinical implications in patients requiring prolonged mechanical ventilation (PMV).
Study design: Descriptive and prospective cohort study.
Setting: Barlow Respiratory Hospital (BRH), a regional weaning center.
Patients: Two hundred consecutive ventilator-dependent patients, transferred to BRH over an 18-month period for attempted weaning from PMV.
Methods and interventions: Patients were assessed clinically for TA within the first week of hospital admission, or once they were in hemodynamically stable condition, by observation of uncoupling of accessory respiratory muscle efforts and onset of machine breaths. Patients were excluded if they had weaned by the time of assessment or if they never achieved hemodynamic stability. Ventilator mode was patient triggered, flow control, volume cycled, with a tidal volume of 7 to 10 mL/kg. Esophageal pressure (Peso), airway-opening pressure, and airflow were measured in patients with TA who consented to esophageal catheter insertion. Attempts to decrease TA in each patient included application of positive end-expiratory pressure (PEEP) stepwise to 10 cm H2O, flow triggering, and reduction of ventilator support in pressure support (PS) mode. Patients were followed up until hospital discharge, when outcomes were scored as weaned (defined as >7 days of ventilator independence), failed to wean, or died.
Results: Of the 200 patients screened, 26 were excluded and 19 were found to have TA. Patients with TA were older, carried the diagnosis of COPD more frequently, and had more severe hypercapnia than their counterparts without TA. Only 3 of 19 patients (16%), all with intermittent TA, weaned from mechanical ventilation, after 70, 72, and 108 days, respectively. This is in contrast to a weaning success rate of 57%, with a median (range) time to wean of 33 (3 to 182) days in patients without TA. Observation of uncoupling of accessory respiratory muscle movement and onset of machine breaths was accurate in identifying patients with TA, which was confirmed in all seven patients consenting to Peso monitoring. TA appeared to result from high auto-PEEP and severe pump failure. Adjusting trigger sensitivity and application of flow triggering were unsuccessful in eliminating TA; external PEEP improved but rarely led to elimination of TA that was transient in duration. Reduction of ventilator support in PS mode, with resultant increased respiratory pump output and lower tidal volumes, uniformly succeeded in eliminating TA. However, this approach imposed a fatiguing load on the respiratory muscles and was poorly tolerated.
Conclusion: TA can be easily identified clinically, and when it occurs in the patient in stable condition with PMV, is associated with poor outcome.