Dynamic Radiology Workflow Allocation System
within the Air Force Medical Service
 
Authors:
Rasu Shrestha, MD, MBA, University of Pittsburgh Medical Center; Goran Momiroski, MSE; Carlos Bentancourt; Aaron Yanuzo, MBA; Thomas Coast
 
Background:
Air Force Medical Service (AFMS) and Military Health Service (MHS) have initiated several programs to address improvements in health care through information technology. One of these initiatives commissioned UPMC Innovative Medical and Information Technology (IMITS) Center to develop a prototype solution for Radiology Dynamic Workload Allocation (DWA) addressing radiology workflow deficiencies. Current AFMS staffing constraints, limited system capabilities, and a mobile patient population requires a sophisticated load-balanced distributed imaging workflow model and supporting infrastructure. These continually changing circumstances within the military healthcare community have identified the need for a sophisticated workflow model that supports a true enterprise view. The prototype solution results in increased productivity and enhanced patient care across the AFMS regardless of physician staffing constraints, systems capabilities, and patient location. The infrastructure supports a load-balanced distributed workflow model across multiple Major Commands (MAJCOMS) and within a multiple Picture Archiving Communication Systems (PACS) environment. The solution allows dynamic bi-directional transmission of clinical studies and optimal workflow load-balancing to effectively leverage resources irrespective of location, PACS, or particular local workload demands. The DWA solution, which is uniquely intelligent and dynamic, employs automated distributed algorithms to allocate radiology cases throughout AFMS based on radiologist availability, specialty/subspecialty requirements, as well as modality and location. The solution also provides the relevant patient history to the radiologist, thus enabling an accurate diagnosis. With regard to workload distribution, these capabilities allow maximum workload flexibility during radiologist deployments, TDY, on-call support, and the development and availability of subspecialty expertise.
 
Evaluation:
A uniquely intelligent and dynamic distributed workflow allocation solution was developed. Some of the characteristics of the solution are:

1. It is an automated solution that employs distributed algorithms to allocate radiology cases throughout AFMS based on radiologist availability and specialty/subspecialty requirements, as well as modality and location.

2. It supports transmission of diagnostic studies and distribution workflow between multiple Major Commands (MAJCOMS), is a PACS neutral solution, and supports multiple PACS systems available at the AFMS.

3. The solution provides a distinct and configurable capability of obtaining the relevant patient history, both relevant prior studies and previous radiology reports. There are several “levels” of obtaining the relevant prior studies ranging from “near study” location to procedure code match.

4. It was developed using widely accepted industry standards such as DICOM and HL7, and is based on Service Oriented Architecture (SOA).

The solution provides a number of benefits compared to the current workflow model in use at the AFMS. Some of the benefits are:

1. Improved radiology resource utilization at the AFMS level. That also addresses the issues of radiology stuff deployment and TDY.

2. Improved ability for specialty/subspecialty development of the AFMS radiology staff

3. Improved access to the historical patient data, such as prior studies and radiology reports.
 
Discussion:
The DWA system has been developed as a prototype system. When implemented, the system has the potential of significantly improving the patient care, the performance, and the effectiveness of the radiologists at the AFMS. Additionally, it provides a framework for the enterprise radiology and can provide a uniform view of the workflow across the AFMS enterprise.
 
Conclusion:
Given the dynamic and mobile nature of the Air Force patient and clinician population, as well as the fact that the network pipelines are more and more available each and every day, the initial experience shows that the AFMS would benefit from a dynamic workflow allocation solution.