Shafiqul Abedin, Presenter: Nothing to Disclose

Jules Henry Sumkin DO, Abstract Co-Author: Scientific Advisory Board, Hologic, Inc

Judith Marie Joyce MD, Abstract Co-Author: Nothing to Disclose

Matthew Thomas Heller MD, Abstract Co-Author: Nothing to Disclose

Bethany Uphold Casagranda DO, Abstract Co-Author: Nothing to Disclose

Scheduling over 200 Radiologists and 50 Residents for the Department of Radiology at the University Of Pittsburgh Medical Center (UPMC)is an onerous task and decreasing the effort to properly schedule and increasing radiologist productivity are essential in this demanding medical setting. An experimental platform (Oceanetta), was developed to deal with the stochastic nature of the schedule in a real operational environment. We view this presentation has having three contributions: (a) Provide an overview of challenges influencing design of the medical scheduling system, (b) Describe approaches taken to overcome those challenges and (c) Provide an auto-scheduling guideline and implementation mechanism catered towards the radiology community to improve tracking, sharing, automation and efficiency during staff scheduling.

After evaluating numerous Usability Engineering (UE) approaches during a literature search, we started with the Spiral Model but ultimately resorted to the Star life cycle for successful software development. We discovered specific challenges and solved (i.e. Intra scheduling conflict, Post Call conflict, Multiple schedule integration) for the abstract interface design. We took a “guided automation” approach to solve the combined task of manual and automated scheduling. The schedules were pre filled with special cases, after which, the automation populated the rest by maximizing the utility function for each physician. We did a NASA workload and satisfaction survey to evaluate the performance of the application.

The satisfaction survey result revealed significant advantage in usability & tracking. The automation algorithm showed promise as demonstrated by one of our largest subspecialty divisions with 14 services and 30 Physicians when only an average of 14.74% changes were necessary compared to the manual process.

We have successfully designed and deployed a versatile scheduling system using the guidelines we set. Guided automation is a viable approach to complex physician scheduling. Utilizing our approach may be helpful to other radiology practices, especially with the increasing need for 24/7 subspecialized radiologist coverage.

Abedin, S, Sumkin, J, Joyce, J, Heller, M, Casagranda, B, Computerized Large Scale Radiology Personnel Scheduling (Oceanetta): Challenges, Resolution & Automation.  Radiological Society of North America 2013 Scientific Assembly and Annual Meeting, December 1 - December 6, 2013 ,Chicago IL. http://archive.rsna.org/2013/13013296.html