University Radiology: Building an IT platform that grows with the practice

Through both organic growth and merger-and-acquisition activity, the New Brunswick, NJ-based mega-practice University Radiology has increased in size from 61 to 96 radiologists in just six years. The task of technologically knitting all practice and service sites together into one integrated whole has fallen to practice CIO Alberto Goldszal, PhD.

It is a task Goldszal has managed with uncommon success, constructing not only an elastic infrastructure that accommodates multiples sites, but also one that provides the practice’s radiologists with access to priors, no matter where the study was generated in the coverage area. He attributes this achievement in part to good technology choices, but primarily to his understanding from the moment he was hired that the practice was bent on growth and any technology had to be scalable.

“I had the privilege to start the selection process knowing full well that I would have to scale up anything I was bringing in in 2007 and thereafter,” Goldszal explains. “The model that was delineated for me is the model we have been following since then.”

From a cost and contracting perspective, the good news for University Radiology and other large practices is that scaling up is easier than scaling down. “With all of the things you have to do today in radiology, when you have a small volume it can be very costly on a per-unit basis,” he says. “You want to have an understanding with your partners that as you do more volume, the unit cost must go down.”

Off the shelf

The tools he used to implement the plan were a series of well-chosen applications, including Synapse PACS from FUJIFILM, layered onto an off-the-shelf (OTS) hardware infrastructure that has since been virtualized and made available through a private cloud.

“Everything we did here in our organization was done using OTS equipment,” Goldszal says. “We didn’t customize anything, and that’s the reality. You would be surprised at how much custom workflow you can generate today with over-the-counter software and equipment.”

In fact, University Radiology doesn’t have a development branch, preferring instead to work with its vendors to add necessary features and functionality, which in turn become part of the standard application. 

“One thing that we wanted to avoid—and that is perhaps the take-home message here—is to become so customized that we would be virtually impossible to be supported,” he explains. “The more customized we are, the further away we are from the core competence of the supporting organization.”

A practice is better off picking the partners that best fit its growth model, Goldszal advises. It’s an easier and less painful alternative than trying to fix issues with a mismatched system as the practice grows.

For the highly subspecialized University Radiology, which went from five to eight hospital contracts and eight to 15 imaging centers within six years, the IT infrastructure had to provide a workflow that was identical across all sites in order to leverage its subspecialized organization.

“Even though all of the hospitals we deal with have their own RIS and PACS, we provide a central place where all of the images come to, all orders are received and all results are generated,” he describes. “If you ask me what has been central to our success to integrate all of these hospitals and accommodate all of this growth rather quickly, it is that all of the reporting is done through a central location.”

Legacy inertia

Goldszal operated under the understanding that one of the greatest contributions he could make as CIO of a growing practice was to limit the number of information systems that the organization deploys. This requires both discipline and diplomacy.

“Over the long haul, you are probably better off constraining the number of information systems that you have to serve your enterprise, so you operate in a less costly manner with fewer systems,” he says.

Benefits of limiting applications range from economies of scale to simplifying support, enough to justify the approach from a financial perspective. The primary purpose, however is to leverage the aforementioned subspecialization across the organization.

University Radiology has several radiologists with pediatric skill sets, but a single hospital may not have enough cases to sustain a pediatric radiologist. “If that person is on the cloud, serving multiple hospitals, then the hospital benefits from the specialty care, and we can run the business efficiently by bringing enough cases to that specialist,” he explains.

That said, when University Radiology merges with another practice, IT is always a point of negotiation. “As partners, we don’t want to just impose our wish on others,” he explains. “We want to eventually have a discussion and move altogether in what we define as best practices. If there are features and functionality or processes and workflows that make sense, then we should adopt those solutions and flows for everyone.”

Goldszal calls the reluctance to convert to a new IT system “legacy inertia” and believes many organizations give short shrift to gaining buy-in for a new system. This is a process that requires communication and training.

“At the end of the day, it is not the technical component that will hold you from migrating systems and moving forward, it is the knowledge of the workforce with the new tools and the challenges of the transition,” he says. “We have seen that the path to a successful consolidation often gets impeded by knowledge. Knowledge, to us, is the ultimate frontier—the more people know, the better they become with the systems and the less resistance to adoption.”

Getting the technology right from the gate

Nonetheless, the right technical choices made at the outset can help smooth a practice’s growth over time. Goldszal opted for lighter, web-based applications that are more easily deployed and for which support is focused more on the central server.

“From the beginning, we wanted a virtualized self-healing, duplicated system, because we run clinical applications that require the utmost uptime,” he explains.

The option of deploying Synapse PACS as a software solution was instrumental in the PACS choice made by University Radiology. “Of course we sought their guidance in what has worked best for other clients out there, so we definitely leveraged on their knowledge,” Goldszal says. “But, we were free to choose and configure our workflows, processes and hardware farm.” 

He said a key feature of the software figured prominently in the PACS choice: The ability to look at the data in a patient-centric fashion rather than a site-centric way while handling multiple data sources originating from multiple, unaffiliated sites with overlapping medical records and accession numbers. “Our business requires us to see a holistic view of the patient, independent of where the data was acquired,” he says.

“If the patient presents herself in multiple institutions that we serve, we want to see that history that the patient has scattered around these organizations into a single view, a patient-centric view, independent of location,” he continues. “That was the reason back then and, if you believe it, is still the reason today that a lot of companies out there choose Fuji Synapse, because of their elegant handling of multi-source data sets.”

Plan the work, work the plan

The biggest problem with mergers and other forms of expansion is the lack of a cohesive plan and proper communication with all stakeholders, Goldszal says. Communicating where the organization is going as a merged, new entity is critical to success.

“Plan the work and work the plan, preferably in that order,” he says. “One would think that the technical issues would be the dominant and the most challenging ones, but what we find is that we are fine on the technical side. A lot of the issues we face are related to lack of communication, lack of a clear roadmap and, ultimately, lack of proper training and knowledge transfer.”

Organizations fail over and over on the knowledge transfer, Goldszal says, which dovetails with the lack of planning and inadequate training.

“You want people firing on all cylinders and peeling every layer of our critical apps, utilizing these information systems to their full capacity,” he says. “You don’t want to just survive another day, you want to thrive and truly, positively impact patient care. Most of our company’s success has hinged on proper planning, plenty of communication and detail-oriented project execution.”