Examination of non-clinical factors affecting tracheostomy practice in an academic surgical intensive care unit

Objective: To gain insight into nonclinical factors potentially influencing tracheostomy practice and determine whether a specialized consultation form impacts tracheostomy utilization. Design: Prospective, observational. Setting: Surgical intensive care unit (SICU). Patients: Patients requiring mechanical ventilatory support. Data abstracted from the Project Impact administrative database served as a practice benchmark. Interventions: Prospective data collection, completion of online survey, and implementation of specialized tracheostomy consultation form. Measurements and main results: Data were prospectively collected on 539 patients and 13 attending intensivists. Our SICU tracheostomy rate (54.2%) exceeded that of 18 comparable ICUs participating in Project Impact (13.9%, p <.001). We attempted to identify factors that might account for liberal tracheostomy use. In 41.5% (±0.6%) of patients undergoing tracheostomy, extubation had not occurred despite successful completion of spontaneous breathing trial on 1 occasion, a rate that varied significantly among attending intensivists responsible for decision making for this procedure (p <.001). Attending intensivists and postgraduate surgical trainees with SICU experience were surveyed to better understand perceptions of tracheostomy practice. Most respondents (96.1%) reported relying on spontaneous breathing trial to guide decision for extubation, 72.6% estimated that ≤25% of patients successfully completed spontaneous breathing trial but did not proceed to immediate extubation, 86.3% estimated that ≤25% of such patients undergo tracheostomy, and 58.8% felt an acceptable benchmark for this practice was ≤10%. In most survey domains, respondents' perceptions underestimated actual practice. Implementation of a specialized tracheostomy consultation form did not impact tracheostomy utilization. Conclusions: We identified variation among clinicians with respect to tracheostomy practice as well as discrepancies between perceptions of this practice and actual utilization. These factors may underlie the liberal use of this procedure in our SICU. Processes for providing accurate physician feedback may assist in optimizing tracheostomy use.