Alberto F. Goldszal, PhD, University Radiology Group; Robert E. Epstein, MD; Murray D. Becker, MD, PhD

In today’s demanding healthcare environment, it becomes fundamentally important to increase radiologist productivity and efficiency through the appropriate use of information technologies that streamline workflow, aggregate clinical data, and optimize workload distribution. Within a radiology setting, these goals can be achieved by the implementation of information systems that consolidate all the reporting function and imaging viewing (both priors as well as new imaging studies), independent of imaging acquisition site, into a single interpretation platform.

Our previous research has shown that up to 20% of imaging cases have relevant prior images (i.e., same anatomy, any modality) archived in 3rd-party PACS.[1] Most of these priors are not available for comparison at the point of interpretation because they are archived outside the radiologist’s own environment. Therefore, for those radiologists (or radiologist groups) that serve multiple (and often unaffiliated) healthcare facilities, the only mechanism available to obtain a comprehensive view of the patient’s imaging history is through the use of disparate, multiple hospital-centric RIS and PACS systems. Obviously, this model can severely limit the radiologist’s productivity and efficiency and, many times, simply impractical.

Our workflow model proposes a centralized interpretation suite where order information and imaging data, sent from multiple facilities, are aggregated, matched, organized, and presented for interpretation within a single “cockpit.” The appropriate matching and exchange of electronic data enabled by our workflow orchestration architecture can overcome institutional borders allowing the free flow of imaging data, orders, and reports. Therefore, by aggregating prior images and reports from unaffiliated organizations, radiologists can now deliver a comprehensive interpretation, taking into account the overall patient history and, as a consequence, improving the diagnostic accuracy.[2,3]

In this work, we have implemented a reporting-driven workflow orchestration that consolidates radiological studies, acquired at multiple unaffiliated healthcare organizations, into a single interpretation platform. This consolidation is possible due to the development and implementation of HL7-based orders-results interfaces between our interpretation system and the radiology information systems of the imaging acquisition sites. In addition, based on the order and patient information drawn from the RIS at each facility, we are able to automatically pre-fetch the corresponding new and old digital diagnostic imaging studies into our PACS. Furthermore, we link the patient’s identity across unaffiliated healthcare facilities, using probabilistic matching algorithms, enabling an all-encompassing access to relevant prior images stored across regional healthcare organizations.

The data flow starts with an order being placed at the imaging acquisition site utilizing the local RIS. This order, in turn, is securely sent via HL7 to our interpretation site where radiologists use speech recognition to create a report. The report and an updated status (e.g., prelim, final) are sent back into the ordering (acquisition) site, also via HL7, to close the order. On the imaging side, we take full advantage of the demographic information presented in the RIS order to drive pre-fetching of new and prior diagnostic imaging studies. In addition, with the appropriate consents and permissions, we automatically and electronically query other regional healthcare providers and organizations for any relevant prior imaging data. The patient id, matching across unaffiliated organization is performed using a probabilistic algorithm that employs name, date of birth, gender, and other data attributing to weight in the matching criteria.

At the end of the process, the radiologist is presented with a wealth of historical clinical information about the patient, exam status (e.g., STAT), the new imaging study, and related priors. Some of the prior imaging sets are retrieved from the local hospital PACS, whereas others are retrieved via the regional imaging exchange highlighted above. From a productivity stand point, radiologist efficiency is increased, due to the use of a single cockpit to drive the interpretation process, regardless of data acquisition site. From a patient care perspective, diagnostic accuracy is increased, due to the availability of prior imaging history at the time of interpretation.

In order to increase radiologist productivity, in particular where radiologists are covering multiple sites operating with disparate RIS, PACS, and dictation systems, the use of a centralized, reporting-driven workflow seems to yield the right benefits. Rather than utilizing multiple information systems to accomplish their jobs, we believe radiologists can be better served by a consolidated imaging and reporting infrastructure that brings the relevant clinical data to them.

This consolidation is accomplished by leveraging on the HL7 and DICOM standards that facilitate the data transport from organization to organization, independent of vendor-specific technologies. The aggregation of the ordering information is complemented on the imaging side by pre-fetching new and prior images from the acquisition site, as well as other healthcare facilities that may hold relevant patient data. We believe retrieving related priors stored at other healthcare facilities is an important component of our model, as the availability of prior imaging has been associated with improved diagnostic specificity. We have also observed that, in many situations, patients move within a region to obtain care at different facilities (e.g., driven by insurance coverage or specialty care). Therefore, in order to promote optimal patient care, we believe that bridging the technologies that aggregate imaging and report data across unaffiliated facilities is essential.

We have implemented a reporting-driven workflow orchestration and automated imaging exchange across several healthcare facilities in the central New Jersey region, including six hospitals and eight imaging centers. The workflow engine allows our practice to receive RIS orders from multiple facilities and to send results back to each appropriate site. This direct interfacing with regional hospitals also allows us to fetch the new and prior studies from the acquisition site and intelligently pre-fetch related priors from the regional healthcare facilities. This model has proven successful in providing efficient teleradiology and sub-specialty coverage to our affiliated organizations, as well as enabling the entire patient history to contribute to and weight in the new exam diagnosis.

[1] Lakhani P, Menschik ED, Goldszal AF, Murray JP, Weiner MG, Langlotz CP. Development and Validation of Queries Using Structured Query Language (SQL) to Determine the Utilization of Comparison Imaging in Radiology Reports Stored on PACS. J Digital Imaging. March 2006;Vol. 19:1:52-68.
[2] White K, Berbaum K, Smith WL. The role of previous radiographs and reports in the interpretation of current radiographs. Invest Radiol. 1994;29:263-5.
[3] Aideyan UO, Berbaum K, Smith WL. Influence of prior radiologic information on the interpretation of radiographic examinations. Acad Radiol. 1995;2:205-8.