Key Developments and Challenges in Laboratory Networks

Being part of a laboratory network offers many advantages to both labs and patients, including increased and improved data collection, as well as stronger relationships between different laboratories. However, although the technologies that allow these networks to grow and function have been around for a while, many hurdles to lab networks’ efficient operation remain, including how members of the network interface with one another.

Aaron Nichols, chief operating officer at Arkana Laboratories, an esoteric pathology laboratory, says his organization faced several challenges when searching for a network interface solution that fit their business model. They eventually found a vendor-neutral option from UK-based X-Lab (aka Labgnostic Inc.) that suited their needs. Nichols and Steve Box, global business development director at X-Lab/Labgnostic, share three key issues around laboratory networks to be aware of:

1. A lack of vendor-agnostic solutions for lab network interfaces

According to Box, although electronic medical record (EMR) and laboratory information system (LIS) vendors have long been creating interfaces to allow labs to form networks, most of these solutions only allow end users to connect with others who use the same vendor’s systems. Instead of having one vendor-agnostic network that all providers can connect to regardless of which EMR or LIS they use, most current lab networks in the US are tied to specific vendors such as Epic AURA or Cerner Reference Lab Network (RLN).

Box adds that once labs and other providers choose a vendor, it’s very difficult for them to switch to another LIS or EMR if they find a better solution, as they will lose their connections to members of their current network.

“If you decide to change EMRs, and you’ve got 60 interfaces tied to an EMR’s network, you’re basically going to lose all of that the minute you change, because you just locked into that model,” Box explains. “That’s true of a variety of EMRs and LISs.”

Another issue with current network interface solutions is that many vendors require customers to implement their full system offerings to gain access to such network connections, even if the lab doesn’t require all those features.

Nichols says his lab ran into this issue when searching for a network interface solution.

“There was no need for us to implement such a robust solution just to gain access for the network connection as there would be all sorts of additional costs associated with us doing that and very little added utility,” Nichols says, adding that many vendors don’t support pathology workflows. He explains that, even though Arkana is a leading renal pathology lab in the US that also offers a neuropathology service, they do a small volume of work compared to the lab industry as a whole. “The way that a lot of these interface systems or pricing models are structured is not very advantageous for us. It would only make financial sense for us for maybe our top five to seven customers around the country.”

Sharing such costs with customers also isn’t an option, Nichols adds, as these facilities often don’t have the financial resources to do so.

2. An uphill push for greater connectivity and more efficient data sharing

Despite the lack of vendor-neutral network options, the need for better connectivity and faster, easier sharing of test orders and results is only increasing. This is particularly true as hospitals have outsourced or eliminated their medical records departments to save costs, Nichols says.

“Almost all of our clients at one point or another have asked us about getting an interface,” he says, adding that many of these clients specifically want a solution that will automatically send lab test results to their electronic health records system and notify them when results are in. But many interface solutions that offer these abilities are, again, too expensive while providing “little utility in the context of pathology’s narrative data format,” Nichols says.

However, there is a push for more laboratory network interface solutions that aren’t tied to a specific LIS or EMR, which can be more cost-effective. Such solutions include X-Lab’s Labgnostic, which has recently started to expand into the US, with Arkana an early adopter.

Nichols says that Arkana decided to partner with Labgnostic as the company’s model made the most financial sense and Labgnostic’s system allows their lab to connect with all others on the network, regardless of which EMR or LIS they use.

“We pay for that connection and then every other provider facility laboratory that comes on to that system, we now have access to transact [with them] without going through the full process of setting up an additional interface with [those providers],” he says. It’s also less strain on Arkana’s small IT team: “Rather than maintaining 20 to 30 different interfaces or more, managing one interface that goes to 30 different providers is so much more efficient and cost effective.”

Nichols adds that such a vendor-agnostic solution has allowed them, in turn, to offer a more cost-effective interface arrangement to their own customers.

“Arkana is better able to assist its clients in establishing a connection to the network,” Nichols says. “Then they have the option once they’re on to find other vendors that use Labgnostic, or to encourage other vendors to utilize Labgnostic, because they already have that single connection established, and their added costs in the future would be significantly lower.”

However, with vendor-dependent network solutions such as Epic currently dominating the market, vendor-agnostic solutions may face a tough road to adoption.

“Epic has such a huge head start right now. It seems inevitable that they will continue to take that market share if there isn’t a workable agnostic way to connect to various systems,” Nichols says, adding that such an agnostic solution gives labs more options and increased competition on the system side.

Box says that, on the imaging side of health care, there are more vendor-neutral solutions available, likely driven by the adoption of AI as a diagnostic tool.

“To enable you to develop a good algorithm, you need a lot of data,” he explains. “And that means you need good connectivity and to be able to easily exchange images. I’m not aware of networks like that in the US, but in the UK there are networks forming where there’s a combination of academic centers, universities, and hospitals, which have come together to build vendor-agnostic, vendor-neutral archives to enable easier sharing of imaging.”

However, a shortage of informatics and information security staff across the healthcare industry is currently a major barrier for the formation and maintenance of lab networks, vendor-neutral or not.

“We have worked with health systems where a single information security officer reviews every single request for an organization of tens of thousands of employees,” Box says. “So, there’s this sequential queue. Some vendors get a response in weeks, some get a response in multiple months.”

Addressing this impediment will be key to improved lab network interfacing going forward.

3. Challenges with standardization efforts

Another key trend related to laboratory networks includes challenges around standardization. Box explains that in many countries, there are national efforts to standardize and harmonize either medical test codes (for example, SNOMED and LOINC), and/or the transport mechanism for healthcare data, such as HL7/FHIR. The key issue with these efforts, Box says, is that “there’s a gap between what’s actually discussed and designed at a national level, versus what actually happens on the ground.”

This means that a lack of standardization continues to hamper data exchange in health care. In particular, ordering and reporting can be a challenge in laboratory networks.

“For example, for report types, we could have a mixture of a PDF and discrete data. But then every single system may store that discrete data in different ways,” Box says. “That can be quite a practical implementation challenge if we’re reporting out to 20 different systems.”

He adds that even systems that are the same may consume results in different ways, posing another implementation difficulty as well as a challenge in discussions with clinical teams to understand how results should be reported.

Both Nichols and Box say that government intervention will likely be needed to address issues with standardization and other barriers to the efficient flow of data in healthcare, such as regulatory compliance.

“I do think that if the government is really serious about efficiency in health care and taking advantage of AI gains, they’re going to have put together some strategy where they can relax compliance or provide some sort of utility or protocol that can assist with [compliance],” Nichols says. He stresses that maintaining proper data security and privacy while making compliance easier will be critical.

Box adds that it’s also essential that medical laboratories have a say in any standardization and regulation efforts to ensure the resulting standards and regulations actually make sense for how labs operate:

“It’s really difficult to keep the two things aligned. How do you actually build something that can be adopted and add value in the real world? The only way that’s going to happen is if you’ve got key contributors from people in labs and people in the software vendor organizations—those who are living and breathing it—contributing and adding value to make sure it does stay on course.”