Abstract
Background: In a healthcare delivery system, labor costs typically represent over 50 percent of a hospital's total operating expenses. For the typical respiratory therapy (RT) department, work is performed in a cyclical pattern throughout the shift, have seasonal variations, and labor is not necessarily matched to work. A reduction in labor metric swings is an opportunity from a labor management perspective. Methods: The Respiratory Therapy Department at Hershey Medical Center, the flagship hospital for the Penn State Health System, began using a web based scheduling system in 2013. The initial rollout allowed staff to view work schedules, submit requests for time off, and pick up extra shifts on any internet enabled device. In July of 2016, the department developed an RT activity volumes feed, based off of orders from our electronic health record to our scheduling and predictive analytics partner. The volume feed interfaced 5 times per day, with orders placed in to four categories; medications, ventilators/noninvasive devices, high flow oxygen therapy, and non-medication secretion management and hyperinflation therapy. The time associated with each activity is a hybrid of internal time studies and the American Association for Respiratory Care’s (AARC) Uniform Reporting Manual (URM). The weighted time of RT activity are converted into work hours and indicates the number of staff required for each area of the hospital in four hour blocks of time. The volume feed of orders translated into work hours and integrated into our staff scheduling system captures three things: a seamless snapshot of ordered RT activity throughout the organization, a more granular view of activity ordered by unit/floor for resource deployment, and historical data to predict low and high demands for RT resources, adjusting RT staffing needs based off the data. Results: For 12 months prior to July 2016, month to month average percent variance in worked hours/total primary units (WHPU) = 11.27 ± 9.09. For 30 months after the implementation of the volume feed, the variance = 7.51 ± 6.36; P = .184. For the first 10 months of our 2018-2019 fiscal year, the department was under budget on the WHPU metric by 1%. In both FY 2016/2017 and in 2017/2018 we experienced a 5% unfavorable variance. Conclusions: Aligning RT staffing to cyclical patterns of work is a challenge for any department. The use of technology and predictive analytics shows promise in closely aligning staff to when the work happens.
Footnotes
Commercial Relationships: None
- Copyright © 2019 by Daedalus Enterprises
