The Infrastructure Behind Virtua Health's Push Toward 1.5 Million Studies A Year

By John Oncea, Chief Editor, Clinical Tech Leader

Laura Patrone has watched Virtua Health’s imaging volume roughly double in the years she’s managed its cardiology and neurology informatics from around half a million studies a year to close to a million now, with a newly onboarded orthopedic practice projected to push that figure toward 1.5 million. The systems built to support half a million studies a year don’t hold up unmodified at three times that volume, and Patrone’s team has spent the past several years rebuilding the infrastructure underneath that growth piece by piece, often under pressure from vendors that haven’t kept pace themselves.

A PACS Migration Driven By Security, Not Ambition

Virtua is currently mid-migration on its picture archiving and communication system, the platform that stores every medical image the health system generates, moving to the cloud and switching vendors entirely. The driver isn’t feature parity or cost; it’s security. “We’re in the middle of a PACS migration; we’re going to the cloud, we’re going to be switching our PACS vendors because the current system we have is not keeping up with the security,” Patrone said. “And it’s a challenge for us to keep our data safe because of the lack of continuing to keep up.”

That gap isn’t confined to the PACS platform itself. Patrone pointed to the imaging equipment vendors as a broader pattern: “It seems like partners like GE and Phillips, that they’re always behind with security when it comes to CAT scan machines and MRI.” Her read on the cause is candid rather than diplomatic: “I don’t know if it’s because they buy too many products and they’re spread too thin,” but the practical effect is that her team’s biggest ongoing challenge isn’t building new capability; it’s keeping vendor-side security current enough to trust with rising imaging volume.

What Scale Actually Looks Like

The volume growth isn’t an abstraction; it’s spread across an unusually wide footprint. Patrone’s team supports imaging and cardiology technology across five hospitals plus a sixth “virtual” hospital-at-home program (the first in South Jersey), five EEG departments, seven emergency departments, 11 imaging centers, three advanced cardiovascular testing centers, 16 cardiac testing centers, and a mobile mammography van that brings screenings into underserved communities. Every one of those touchpoints feeds imaging data back into the same core systems her team maintains.

The pace of change has accelerated to match. Virtua used to run about four major IT projects a year; it now runs about four a quarter. “Technology and the AI and just the rate at which things are coming out,” Patrone said, pointing to shifting reimbursement models, changing patient acuity, and a growing push toward remote patient monitoring as the forces driving that pace.

The Epic Integration Nobody Gets For Free

Underneath the volume and the vendor churn sits Epic, which took Virtua roughly two years to fully implement after coming from a “disparate” systems environment – separate EMRs for the ED, for radiology, for inpatient data, which required staff to log into seven or eight different systems to assemble a full clinical picture for a single patient. “It was bumpy at first,” Patrone said, “because you’re finding all the workflows and kinks, and as much as you test things, it’s not the same as the people who are using it day-to-day.”

The most persistent friction point turned out to be cardiac stress testing, where the workflow has to execute in exact sequence, or the resulting PDF fails to route into Epic automatically, forcing a manual attachment on the back end. Fixing it wasn’t a systems change; it was retraining technologists on why the sequence couldn’t be skipped, short of a genuine clinical emergency requiring the patient to come off the treadmill. Virtua now runs Epic upgrades twice a year plus an annual disaster recovery event in June, on top of that baseline maintenance discipline.

Interoperability Gets Better, But It’s Not Automatic

Where systems are well established, Patrone said integration has gotten genuinely good, often through HL7 or FHIR. Virtua was an early adopter of MUSE-NX with FHIR Integration, an EKG platform embedded directly in Epic, so a cardiologist can read an EKG without leaving the record to open a separate application. But “pretty good at making things work” is Patrone’s own phrase for describing the state of interoperability generally, and it implies exactly what it sounds like: tweaks and refinements, not plug-and-play compatibility, even with established vendors.

That extends to disagreements with Epic itself. On a vascular reporting project involving carotid artery stenosis, Virtua’s lead vascular sonographer found that Epic’s AI was pulling stenosis measurements from a different, less-validated data methodology than the one Virtua’s own team uses. Rather than a one-way directive, the discrepancy triggered a meeting: Epic wanted to understand why Virtua’s approach differed, a working relationship reinforced by Virtua’s regular participation in Epic’s development immersion trips, where sites help shape upcoming EHR functionality before it ships.

Where This Is Headed

The next wave of infrastructure Patrone is building isn’t primarily about imaging volume at all; it’s about extending the same integrated, single-record philosophy into the home. She’s currently running two to three cardiology projects centered on remote patient monitoring, including migrating toward a standardized bio-button sensor usable both in-hospital and at home, and a heart-failure monitoring program where implanted-device patients rest on a sensor pillow daily so a nurse can review their vitals remotely and flag early risk before a hospitalization becomes necessary.

For health systems watching their own imaging volumes climb, Virtua’s experience suggests the constraint isn’t usually the imaging AI itself; it’s whether the PACS, the EHR integration, and the equipment vendors underneath it can keep pace with both the data volume and the security expectations that come with it.