Methods
This study was conducted at a 1,154-bed quaternary hospital in the eastern United States with 65,000 ED visits and 16,000 hospital admissions per year. The study was implemented within the hospital’s adult general medicine units (268 beds) and excluded intensive care, step-down, and cardiac care units. Approximately 50% of beds on these units are equipped with telemetry. The patients on these units are managed by the internal medicine residents and internal medicine faculty. The study captured 9,881 patient encounters with telemetry.
We used a before and after study design. The study involved two interventions. The first intervention, or the nurse-discontinuation protocol, allowed nurses to discontinue telemetry at 24 or 48 hours depending on the practice standard-recommended duration for the specific indication entered upon placement of the order. After daily assessment of telemetry data using a specified protocol based on the 2004 AHA Telemetry Practice Standards, nurses consulted with the resident to discontinue telemetry if appropriate. Residents were able to renew the order if telemetry was still required. The second intervention, or the physician-discontinuation protocol, which occurred immediately after cessation of the nurse-discontinuation protocol, residents were notified by a best practice advisory (BPA) if a patient’s telemetry order had gone beyond the recommended duration according to the Practice Standards and recommended discontinuation of the order.
The study was implemented using both Allscripts (Chicago, IL) and Epic Systems (Madison, WI), as the hospital transitioned electronic medical record (EMR) systems mid-way through the study, at the start of the physician-discontinuation protocol. Data collection spanned eight months following the implementation of the nurse-discontinuation protocol and twelve months following the physician-discontinuation protocol, with data from the six months before the nurse-discontinuation protocol serving as the pre-intervention control.
The primary outcome was the total time spent on telemetry per patient per month. The secondary outcome was the percentage of patients placed on telemetry each month. All data were obtained from the electronic medical record. Analyses of average total time spent on telemetry every month were calculated using a t-test and, if data did not follow a normal distribution, were confirmed with a Wilcoxon Rank-Sum Test. Statistical computations were performed using SAS v9.4 software (SAS Institute, Cary NC).