Why Labs Should Control Their Own IT

Contributing Lab Leader: Peter Gershkovich, MD

Operational

Article highlights:

Integrated clinical information systems are widespread, but "best-of-breed" lab systems can give labs flexibility.
Dedicated lab systems are easier to customize bridge "functionality gaps."
Labs can make a case to the C-suite for IT autonomy to achieve greater efficiency and less down-time.

As healthcare providers embrace electronic health records, it's tempting to try to bring as many functions as possible under the umbrella of one integrated system. Larger and more comprehensive clinical information systems incorporate modules for all major clinical departments, including nursing, radiology, pharmacy, surgery, oncology, cardiology, obstetrics, emergency, and, of course, laboratory departments.

The advantages can be significant: all the pieces work together (or at least, better than products purchased from different vendors), the provider doesn't have to pay for expensive interfaces between different systems, and system upgrades happen in a coordinated manner. But departments that want to purchase and maintain their own software often have a fight on their hands when they make their case to the CIO.

"Best of breed" vs. integrated system

But it's a fight worth having, says Peter Gershkovich, MD, Director of Pathology Informatics at Yale University Medical School. What a lab may gain in the ability to be part of a larger IT infrastructure, it may lose in functionality and the flexibility to pursue new ways of working. “Best of breed" laboratory information systems, supplemented by smaller systems either acquired or designed by the staff, can help maximize the lab's level of computerization and give it the ability to adopt new tests and technologies quickly and inexpensively.

Of course, the lab needs one or more staff members whose idea of fun is figuring out how to make computers do things—people like Gershkovich himself, who has worked extensively in medical software development and completed a fellowship in medical informatics at Yale. Not all labs will have the appropriate skill sets available, but those that do can adapt their own IT to changing circumstances, and much more quickly and easily than a central IT department can.

Every lab is different

“We need to realize that laboratories are very different," he says. “A lot of labs don't need autopsy services. Others don't need next-generation genetic sequencing. Some labs have multiple sites, and some have outreach clients." Best-of-breed, stand-alone laboratory information systems (LISs) are built to be flexible and highly customizable so they can address these diverse needs in ways that integrated clinical information systems with lab modules cannot.

Even a stand-alone LIS can't do everything, however, and that's where in-house software development comes in. Gershkovich's department has added many functions including a bar-coding system, a method for communicating pathology results to surgeons during an operation, and a system for annotating DNA variances found during genetic testing. One key to success is inexpensive or even free software tools often used to develop websites and internet-based services.

“We use the same technologies that Amazon, Facebook, and Google are using" to create new features on the fly, Gershkovich says. “The cost to operate them is negligible and that allows us to rapidly deploy our systems and close functionality gaps by adding ancillary systems around the core LIS. That ability and that cost would certainly be unattainable to us if we didn't have ownership of our system."

Moreover, Gershkovich says, an integrated clinical information system exposes the lab to all the delays and errors that can occur during an institution-wide IT upgrade. “How much does it cost to make an error, and how long will people be without a system if you deliver code that is broken?" he asks. The lab can also lose control of its ability to change the system's workflow to fit its needs. "That desire to have everybody have the same workflow will eventually break against the necessity to be different in various groups and have incremental improvements in their systems," Gershkovich says. “I hope the C-suite will realize that everybody's different."

On Everybody Being Different

Update Required To play the media you will need to either update your browser to a recent version or update your Flash plugin.

Peter Gershkovich, MD

Research Scientist, Director Pathology Informatics

Yale University Medical School

Contributing Lab Leader