College of American Pathologists
Printable Version

  Feature Story


cap today

Pulling out all the stops in the LIS marketplace

May 2003
Eric Skjei

Thinking about upgrading or replacing your laboratory information system?
Take note: it’s not your father’s LIS marketplace anymore.

Yes, it’s a whole new market out there. New forces and factors range from the need to accommodate genomics testing to the pressure to build your outreach business, support Web-based order entry and results reporting, handle multi-site operations, integrate robotics, and, oh yes, manage your costs more closely than ever.

Once you go shopping, you’ll find many good systems, all of which provide the basics and more—but few of which can do it all. You’ll find a marketplace dominated more and more by big players, one that is steadily tilting in favor of integrated, single-source solutions rather than interfaced, heterogeneous, best-of-breed systems. A market in which, in this era of uncertain economics and turbulence, virtually every vendor understands the importance of building strong long-term relationships with clients.

Most vendors agree that the LIS market is flat. New sales are, by and large, replacement sales, sometimes prompted by the vendor’s decision to sunset an older product.

“There seems to have been a decline in new lab system buying over the past year,” says Sidney Goldblatt, MD, lab director at Conemaugh Health System, Johnstown, Pa., and former CEO of Sunquest, now Misys Healthcare Systems. “It could be a cyclical phenomenon related to the general austerity in the health care industry. But I think it’s also sort of a breather in which clients are stepping back and assessing their strategic direction.”

As recently as 10 years ago, the person at the center of a search for a new LIS probably would have been the laboratory manager. Today, that assignment is at least as likely to fall to the chief information officer. “Even from the LIS point of view, the sell is now at least in part, if not in a very dominant way, to the CIO,” says Bruce Friedman, MD, professor of pathology and director of pathology data systems, University of Michigan Health System, Ann Arbor.

This shift to the primacy of the CIO favors large, unified solutions over best-of-breed alternatives. With the CIO in charge, Dr. Friedman says, the upgrade or purchase decision is less likely to be driven by functionality than by enterprise-wide concerns for ease of integration among clinical areas, notably radiology, pharmacy, and the lab, and possibly such computer-intensive departments as radiation oncology and cardiology. (Conformance with standards and consolidation along the lines of locally favored hardware and operating systems also play a major role in decisionmaking, Dr. Friedman says.)

Because best-of-breed options typically imply interfacing rather than tight integration for data exchange, and because interfacing can be an area of slow or problematic data exchange, these options may be subjected to greater scrutiny than ever before, even if they do offer superior functionality. “One could argue that having to sell to the CIO inhibits the best-of-breed approach for what is sometimes called the ‘good-enough’ approach to the LIS purchase,” Dr. Friedman says.

The good-enough strategy is in many ways a “common denominator” approach, Dr. Friedman adds. “The CIO wants to deal with a small number of vendors, with large, established companies, and from a business rather than a technology perspective, and this bias again favors the larger organizations, like Siemens, McKesson, and Cerner,” he says.

By contrast, focusing on best-of-breed means focusing on functionality. And rightly so, from the lab’s perspective, because a laboratory operating within a health system is going to be held accountable for its financial success and for how well it operates—and how well it operates will be highly dependent on its IS tools. Moreover, best-of-breed solutions are more likely to come from smaller, privately held vendors who are often newer to the market—all traits guaranteed to raise the CIO’s eyebrows.

“The CIO is interested in a whole suite of products, in how well they interoperate,” and in making sure they are from a dependable vendor, says Dr. Friedman. The CIO is looking for “a vendor that is publicly traded, on whom he or she can obtain financial information, and with whom he or she can negotiate large, complex contracts.” In general, the CIO will not dig deeply into the functionality and workflow requirements of any given department as a primary consideration.

Dr. Goldblatt concurs. “Many institutions are now committed to an integrated solution for all information technologies in their clinical area, and they want to do it from a single vendor,” he says. Admittedly, the success of HL7 is a mitigating factor in this process. Virtually all systems have achieved a high level of data integration through the use of HL7, and that level is improving with the use of XML. Nonetheless, Dr. Goldblatt says, “many institutions are increasingly anxious to harmonize all their clinical data in a single vendor’s system.”

Dr. Friedman cautions labs that favor the best-of-breed option against being deprived of functionality that they consider essential to achieving key operational and other performance goals because they will be held responsible for those goals. “When all is said and done, in aggressive organizations, the lab is still going to be assessed on operating cost per test, among other measures,” says Dr. Friedman. “And if you’re hobbled by a less-than-efficient LIS, ultimately you’re going to be called to task.”

Blurred boundaries
One of the larger trends shaping the LIS market during the last few years has been consolidation among providers and the related formation of integrated delivery networks, or IDNs. This consolidation has created what might be called a “hybridization” issue, which laboratories continue to encounter.

LISs traditionally have fallen into one of two categories: a hospital LIS, focused on the patient, and a reference LIS, focused on the customer (more likely to be the office-based physician than the patient) and the specimen. Consolidation and the attendant formation of core labs require some of both product areas.

“In the last five years or so, the line has blurred between what’s occurring
in the commercial reference lab business and what’s happening in hospitals,” says Al DeStefano, CEO of SIA. “As hospital labs, for example, have moved into the outpatient business, they have had to face a whole new set of issues.”

Outpatient billing requirements, for instance, differ from inpatient charge capture. Labs also find they must manage courier systems and produce reports formatted in significantly different ways to meet new client expectations. The results are mixed. “Systems designed to handle reference labs and systems designed to handle hospital labs’ needs don’t cross over into one another’s area very well,” DeStefano says.

The core lab concept that arose out of consolidation still serves hospitals, but doing so in a multi-site facility or integrated delivery network often means taking on much of the operational role of a reference lab, DeStefano notes. “The multi-site issues are being handled, so the lab can tell where specimens are coming from and can get the results back to the correct facility, but problems remain,” he says. Simply having to produce patient-centered cumulative reports for an inpatient environment as well as client-centered reports for the outpatient and outreach environments can be tricky. “Most systems don’t do this particularly well,” he says. “They’re either patient-oriented or specimen-oriented.”

Labs of all sizes and types, whether they operate alone or in a multi-site system, are often pressured to become more entrepreneurial and to generate revenues that can help offset costs. That often means building an outreach business.

This need has been prominent long enough that most LISs can address it. “Almost every LIS vendor now has a relatively strong outreach solution,” says Hal Weiner, president of Weiner Consulting, “at least from the perspective of recognizing that the enterprise handles more than patient beds, that there is a need to market services outside the institution.” At the
same time, Weiner adds, there is a relative dearth in the market of total solutions that accommodate everything needed in the outreach environment, what is in effect a reference lab within the walls of a hospital-based enterprise.

In outreach, the virtues of integration can be particularly striking. J.P. Fingado, enterprise vice president for lab solutions at Cerner, says the only ones who are going to win in the outreach marketplace are the suppliers who can offer a single physician system that supports all of the ancillaries in one place. “You’ve got to be able to have a single physician desktop that has all the specialty lab functionality for ordering and reporting, as well as specialty radiology and pharmacy, so physicians can get their results all at once,” Fingado says. What physicians do not want, he adds, is to have to log on to one application for lab results, another for radiology orders, and a third for pharmacy information. “That’s not going to fly,” he says.

Even best-of-breed vendors need to understand and accept this underlying business reality, concurs Bill Blair, vice president of sales and marketing for SIA. “The outreach business is after the ‘whole doc,’” he says. A hospital lab must understand first, that the hospital is competing hard for the physician’s business, and second, that from the physician’s point of view, the lab is too closely allied to radiology, pharmacy, and several other areas of the hospital to allow it to fly solo. “The doc needs to be able to hit one icon on the desktop and have all the ancillary services of that delivery system available to him or her, rather than have to separately log on to get access to the lab, then log off, log on to radiology, and so on,” he says.

Bringing the lab online
Many labs need to make better use of the Internet to create a stronger relationship with clients, says Dr. Goldblatt. This often can be done without upgrading or replacing the LIS. Many new lab portal vendors have rushed to serve this niche. In a sense, these third-party products extend the life of an incumbent LIS, particularly by facilitating online ordering and reporting and by offering better, less costly ways of delivering reports.

“Would the need to make better use of the Internet drive someone to replace their lab system?” Dr. Goldblatt asks. “My guess is no, because there are a number of vendors that offer a solution in this area and because those solutions are adaptable to existing lab systems.”

This new category of LIS/lab portal vendor, Dr. Friedman says, is carving out a set of functions that can be grouped under the general heading of customer relationship management, or CRM. Developed in other business contexts, CRM is intended to provide software tools that organize the interaction between supplier and client and make the process more manageable and efficient. In a lab environment, CRM focuses on handling and presenting test results to customers—primarily private-office physicians—for reference labs and within hospital systems, via a Web
front end.

Dr. Friedman speculates that the appearance of a group of vendors serving a largely CRM function for the lab signals the beginning of a divergence between the classic LIS and a newer form of “disaggregated,” or decentralized, LIS. As this divergence unfolds, the classic LIS will persist, representing as it does many decades invested in expertise and talent, but it will steadily move into the background of the enterprise’s operations. The classic LIS will become less apparent to customers and more dedicated to such highly specialized lab functions as specimen tracking, creation of phlebotomy drawing lists, and quality control. “What’s going to be apparent to the customer are these Web-based, order-entry, result-reporting engines, the so-called lab portals,” he says.

The classic LIS was designed around the work and information that flow through a hospital or complex reference lab environment, Dr. Friedman says. He refers to this as a centralized LIS, or C-LIS. The new, emerging LIS, destined to parallel and coexist with the C-LIS, is the decentralized LIS, or D-LIS, he says, the fundamental architecture of which is Web-based. The D-LISs—the first wave of which are products from the lab portal vendors—will be designed to handle data generated by POC testing, home health care testing, home kit testing, and even direct access testing and results reporting, he adds.

“Direct access testing, where patients sign on to a Web site and order tests for themselves and pay for them out of pocket, has been enabled by the Web, which can push the order engine directly into patients’ homes,” says Dr. Friedman. “Going retail” like this is one of the few ways in which labs have gained access to a new market.

Grappling with genomics
For larger teaching facilities, one of the hottest forces driving upgrades to the LIS, as well as the purchase of new systems and third-party products, is genomics testing. “A small but growing number of labs are frustrated by the inability to integrate their new genomics testing into the current lab system,” says Dr. Goldblatt, “which after all was not designed for this kind of data, and which in many instances takes a different form from the existing array of lab tests.”

Because genomics information is data-intensive, one area where LISs are likely to prove deficient is the database. “As we move toward personalized patient databases, where the human genome has been specifically mapped to that patient,” says Weiner, “we’re going to be monitoring and testing predispositions to particular disease states and, from the pharmacy side, seeking to provide patient-specific drugs and creating individualized drug regimens.” As they move futher into this area, labs will need enormous databases. That’s something legacy systems—as well as many newer LISs—were not designed for and are not likely to be able to handle.

Managing genomics information will also require developing standards for presenting and sharing this data. “In many cases this data is really pattern data, and the image may be the best way to view, store, and share it,” Dr. Goldblatt says. Lab systems can’t easily accommodate this type of data, and doing so will require new subsystems, he adds.

Cerner’s Fingado makes a strong case for the benefits of integration: “When labs start to do molecular testing, they’re going to want to tie that work into pharmacy data,” to refine drug treatment, among other things, he points out. To meet this need, he says, Cerner is building a complex “bioinformatics layer” tied closely to the application layer in the Cerner system architecture and accessible by pharmacy, radiology, and the laboratory.

Robotics and relationships
In general, lab information system vendors are still working to improve the automation capabilities of their LISs. “Most legacy LISs did not have a good way to handle the automation process,” Weiner says. Newer LISs, by contrast, integrate more of those capabilities, especially in the area of specimen management and tracking.

“The products that are currently out there are beginning to address this problem,” says SIA’s DeStefano, “but I don’t think anyone has really good answers yet.” DeStefano believes that labs are just beginning to understand the need to automate, and that the need has not yet been addressed effectively in the marketplace.

Automation poses many challenges. “For example, most LISs don’t take error messages directly from the instrument,” says DeStefano. If your analyzer is running out of reagent, the LIS needs to understand the alerts. Otherwise, he says, “here you are chugging along automatically, the analyzer is facing 300 glucoses, and all of a sudden it is running out of reagent.” Without automatic communication between the analyzer and the LIS, time-consuming and costly human intervention will be needed.

Another challenge is the system’s ability to assign and track unique identification numbers, particularly for specimens. As simple as this sounds, this capability isn’t universally available, and when it is available, it’s not well-integrated into basic LIS functionality. “Unless your LIS has true unique, specific specimen ID capabilities, it is very hard for the lab to smoothly and fully track where the specimen and all daughter tubes came from and where they are at any point in time,” says DeStefano. Being able to locate specimens can make the difference between efficient and inefficient automation.

To underscore the automation differences between European and U.S. labs, SIA’s Blair recalls a conversation he had with a pathologist in Germany. “When I asked him to describe the difference between a reference lab in Germany and a lab in the United States, his response was: ‘Here in Germany we get the sample, test it, process it, and send out the results. In the United States you make friends with the specimen—you relabel it, sometimes more than once, you handle it repeatedly, you move it around.’”

In many cases labs and vendors, including instrument vendors, have been forced to create what is, in effect, a mini-LIS to handle automation-related functions that the primary LIS isn’t up to. But the results are often less than ideal. “Unless the system’s basic architecture is designed for unique specimen ID and related functions,” Blair says, “solving automation challenges by adding a layer on top of an architecture that does not support that capability probably just makes the process more complicated.”

As new as automation and many of the other forces shaping the LIS market are, in many ways the dynamic between client and supplier is returning to that of an earlier era, with an emphasis on trust and integrity.

“We’re getting back into the integrity game,” Cerner’s Fingado says. “People really want to do business with folks they trust and with whom they are comfortable. They want to work with a supplier who can understand their long-term strategy, who takes the time to listen, and who can actually be there to deliver for them.”

Fingado is reminded of the conditions that shaped the LIS market when, over a decade ago, he was an executive with leading anatomic pathology system vendor CoPath. “This is like it was when we started to automate the AP business 12 years ago,” he says. And, he adds, it’s something of a refreshing change from “the race for functionality, the race for technology, and the hype of the Internet.”

Eric Skjei is a writer in Stinson Beach, Calif.