The lessons of laboratory information system products for IVD manufacturers

Diagnostic Insight, Winter 1995/96, Vol. 16, No. 1, pp. 4

The lessons of laboratory information products have been "hard learned" by in vitro diagnostics manufacturers, said Robert Bauer, at the 13th Annual Symposium on Automated Information Management in the Clinical Laboratory, held at the University of Michigan Medical School. In the early 1980s, Bauer said, IVD manufacturers had "less than motivating" experiences when they developed their own mini-LIS systems and more recently as they tried to develop plug-and-play interfaces for third-party LIS systems. Bauer is a leading consultant to the IVD industry and managing director of Dallas-based CaseBauer & Associates. The symposium, directed by Bruce Friedman, MD of the University of Michigan Medical School, included a seminar sponsored by CAP Today examining the issue of instrument and LIS integration. As participants pointed out, laboratory informatics presents a special challenge to manufacturers in the in vitro diagnostics field. Symposium moderator and BMA President-elect Robert McGonnagle declared that "In our increasingly capitated, cost sensitive healthcare system, the care delivery enterprise wants to be able to view the laboratory as a black box. A 'virtual instrument,' one that can deliver timely and accurate results regardless of its internal processes and its configuration." The lab that proactively addresses the problem and meets the challenge, suggested McGonnagle and other speakers, will gain competitive ad vantage and enjoy significantly greater control over its own destiny.

IVD Manufacturers' Contributions to Informatics

"Many IVD manufacturers initially looked to data management and informatics products to provide a bridge during these longer product development cycles," Bauer said, "a glue-and-fix solution that manufacturers were willing to supply at very low prices, sometimes even at cost, to maintain customer favor and stretch out product life cycles while they developed their next generation product." The realities of clinical laboratory LIS needs were an early challenge that IVD manufacturers often failed to fully understand, Bauer said. First, the clinical laboratory market is highly fragmented, populated by thousands of unique operations, each with its own requirements. Second, since as much as 50 percent of a typical hospital's information transactions involve laboratory data, the information management task in the lab is not trivial, and IVD manufacturers often underestimated the scope and complexity of the LIS task. Third, the level of support clinical laboratories required was often beyond the scope of the glue-and-fix budgets of IVD manufacturers. In short, for the IVD manufacturer, the task far exceeded the benefit, according to Bauer, a fact that, for many, came to light only after the LIS products were developed and installed in laboratories. Bauer feels that another challenge for the IVD industry was LIS vendors' business imperatives. While the two industries share a common customer, IVD and LIS vendors operate from a different set of business priorities. To an IVD manufacturer, the LIS interface is not a revenue source, but a product feature that facilitates an instrument system placement. Instrument and reagent revenue is what drives the IVD industry. LIS vendors view this same interface differently. Post-installation consulting fees are important to many LIS businesses and can represent 40 percent or more of an LIS vendor's annual revenue, and the importance of these consulting fees is growing as the number of new system installations declines.

To move toward solutions, Dr. Friedman highlighted the need to get beyond our current dualistic, "us and them" models and seek "co-opetition," that is, solutions that bring a Hitachi, Abbott, or Beckman to the table with a Cerner, Sunquest, or CHC. Clearly, "there's all kinds of interaction going on between these two camps now," he pointed out. One way to frame the opportunity, Dr. Friedman suggested, is to think in terms of dynamically linked "up stream" and "downstream" integration of data. "Up stream" integration, which includes, but is not limited to, IVD-LIS interfaces, has a better chance of happening if the two parties can establish a common agenda. Success upstream will, in turn, facilitate the downstream process of delivering value to physicians, patients, and the entire healthcare enterprise.

Reducing Information Float

Dr. Friedman believes that there are opportunities in data handling to reduce information "float" in the lab. Drawing on the analogy of travelers' checks, which are paid for at purchase and redeemed later, being, in effect, a loan, Dr. Friedman defined the laboratory information float as the time period beginning when a clinician orders a test and ending when the clinician "knows" the result. In an age of time-based competition; laboratory competitiveness and success will increasingly derive from the ability to reduce information float. What we are seeing in our deployment of sophisticated work stations, automation of front end processes, innovation of POC testing, and refinement of the LIS with rules, reflex testing, long-term archiving, and clinically specific reporting, as well as streamlined IVD-LIS interfaces, is nothing more than better ways to do just that.

Lack of Standardization and Cooperation as Impediments to Progress

Calling the separate evolution of instruments and the LIS "interesting and sad," the Cleveland Clinic Foundation's Dr. David Chou went on to outline the progress that has been made in interface technology. Not much more than a decade ago, Dr. Chou pointed out, it was not uncommon for a laboratory to buy a major analyzer only to find it had no LIS interface capabilities at all. When minicomputers first began to appear in the lab, in the 1960s and early 1970s, the typical analyzer's "interface" capabilities were merely an extension of the ability to print hardcopy chart records. Only in the late 1970s did interfaces based on the soon-to-be ubiquitous RS232 emerge, and even then labs typically had to pay extra for them. However widespread it may have become, the RS232 interface is rapidly becoming obsolete, Dr. Chou said. Its electrical and connector specifications are limited. The maximum data-transmission rates are too slow. The requirement that instruments can be no more than 50 feet from the computer is almost laughable in an environment where a data center may be miles from the laboratory. By the 1980s, newer, unidirectional and bidirectional interface technology had emerged, and by the late 1980s barcode technology was also playing a growing role in the lab and thus in the interface picture. In the 1990s, Dr. Chou said, we have entered an era of standards advancement and promulgation, most notably under the aegis of the American Society of Testing and Materials. While we are not likely to see true "plug and play" interfaces soon, we are making progress toward a more comprehensive, stream lined instrument-LIS interface environment. Remaining barriers pertain to the inherently slow nature of standards development, depending as the process does on volunteer expertise and lacking as it does the involvement of a clearly dominant vendor - the clinical laboratory equivalent of a Microsoft or an Intel who can drive the process by sheer weight of influence and vested interest. Also, there does not yet exist independent means to verify interface protocols, Dr. Chou added. Standards do not adequately address the special demands of high volume analyzers. Delays are inherent in system design and FDA approval. For all its potential benefits, point-of-care technology has "set back interface technology 10 years," Dr. Chou said, since few if any POC devices fully support data transmission to the LIS.

IVD Manufacturers Becoming Better Equipped to Expand their Role in Informatics

Robert Bauer also observed that IVD interest in laboratory information management is on the rebound, driven by the manufacturers' desire for more product differentiation in a market where product offerings are increasingly reaching parity. One result is the wave of new workstation management systems from Abbott, Johnson & Johnson, Beckman, and others, which are filling voids that LIS systems and vendors have not been able to address rapidly, namely, lowering the cost of interfaces, providing modern information management tools such as e-mail to older LIS systems, and optimizing the workflow across two or more related instrument systems. Will the IVD role in lab information systems become more significant? Yes, Bauer predicted. All indications point in this direction. First, unlike the LIS, which made more sense to develop on a laboratory-wide basis, combination LAS/LIS solutions are likely to develop and be more affordable on a workstation basis rather than across the en tire laboratory. Second, IVD manufacturers are continually broadening the definition and scope of their core products to establish competitive advantage. Front-end automation is a developing trend. Third, the consolidation of IVD manufacturers begets larger organizations with a broader breadth of instrumentation that can be integrated with out the complexities of multi company coordination.

The Impact of Laboratory Automation on Laboratory Information Management

Possibly half of all clinical labs will embark on automation projects within the next five years, predicted Dr. Rodney Markin. In the process, this change will radically alter our current paradigm of information development and management in the lab. Dr. Markin is professor of pathology and microbiology, Departments of Pathology and Microbiology, at the University of Nebraska Medical Center. In Dr. Markin's view, arriving at that new paradigm will require improving our under standing of what happens in the laboratory. That in turn will mean reevaluating laboratory processes from several perspectives and on several levels, from local work-bench to very large, multi-organizational healthcare information net work. What we will find in this process may surprise us, Dr. Markin suggests. Most labs, he believes, do not fully understand the extent to which specimen flow differs from information development and management. Automating and integrating them will, among other things, force us away from our current view of the laboratory as being composed of many smaller, relatively independent sub sections. As a result of this change, we will increasingly find our selves thinking in terms of an LAS—a laboratory automation system—as well as an LIS, Dr. Markin declared. Moreover, that LAS will be come not just an important element in effective laboratory design but its essential starting point. "Ideally, if we were going to step back, look at the processes in a laboratory, and design a lab information system from scratch," Dr. Markin observed, "we would start with LAS and then move on to instrument interfaces and the LIS."

Does the Clinical Laboratory Industry Have the Motive, Means, and Opportunity?

Given that the clinical laboratory is becoming more physically dispersed, no longer contained within four walls or the "lab-as-factory assembly-line" model, the concept of a "virtual lab" has become more relevant, Bauer added. "With developments like point-of-care testing, with nurses running tests, the lab is becoming harder to define in a physical and organizational sense," he said. But, according to Friedman, making the virtual lab a working reality will require further evolution of interface protocols and standards, a better understanding of the complex interdependencies among specimen movement, information management, and workflow, and a much clearer grasp of the process control issues that the introduction of robotics and front-end automation is now raising. It will also require new initiatives, new business and technology arrangements between two industries accustomed to operating at an arm's length from one another, speakers agreed. "Aligning all these competitive objectives is not unlike asking Federal Express in the 1970's to invent the fax machine," states Bauer, "the virtual laboratory is going to happen... its just not clear yet who is going to lead the process."



	Information Now Exists that can Redefine the Laboratory: But is the IVD Industry Ready?

	Diagnostic Insight, Winter 1995/96, Vol. 16, No. 1, pp. 4 The lessons of laboratory information products have been "hard learned" by in vitro diagnostics manufacturers, said Robert Bauer, at the 13th Annual Symposium on Automated Information Management in the Clinical Laboratory, held at the University of Michigan Medical School. In the early 1980s, Bauer said, IVD manufacturers had "less than motivating" experiences when they developed their own mini-LIS systems and more recently as they tried to develop plug-and-play interfaces for third-party LIS systems. Bauer is a leading consultant to the IVD industry and managing director of Dallas-based CaseBauer & Associates. The symposium, directed by Bruce Friedman, MD of the University of Michigan Medical School, included a seminar sponsored by CAP Today examining the issue of instrument and LIS integration. As participants pointed out, laboratory informatics presents a special challenge to manufacturers in the in vitro diagnostics field. Symposium moderator and BMA President-elect Robert McGonnagle declared that "In our increasingly capitated, cost sensitive healthcare system, the care delivery enterprise wants to be able to view the laboratory as a black box. A 'virtual instrument,' one that can deliver timely and accurate results regardless of its internal processes and its configuration." The lab that proactively addresses the problem and meets the challenge, suggested McGonnagle and other speakers, will gain competitive ad vantage and enjoy significantly greater control over its own destiny. IVD Manufacturers' Contributions to Informatics "Many IVD manufacturers initially looked to data management and informatics products to provide a bridge during these longer product development cycles," Bauer said, "a glue-and-fix solution that manufacturers were willing to supply at very low prices, sometimes even at cost, to maintain customer favor and stretch out product life cycles while they developed their next generation product." The realities of clinical laboratory LIS needs were an early challenge that IVD manufacturers often failed to fully understand, Bauer said. First, the clinical laboratory market is highly fragmented, populated by thousands of unique operations, each with its own requirements. Second, since as much as 50 percent of a typical hospital's information transactions involve laboratory data, the information management task in the lab is not trivial, and IVD manufacturers often underestimated the scope and complexity of the LIS task. Third, the level of support clinical laboratories required was often beyond the scope of the glue-and-fix budgets of IVD manufacturers. In short, for the IVD manufacturer, the task far exceeded the benefit, according to Bauer, a fact that, for many, came to light only after the LIS products were developed and installed in laboratories. Bauer feels that another challenge for the IVD industry was LIS vendors' business imperatives. While the two industries share a common customer, IVD and LIS vendors operate from a different set of business priorities. To an IVD manufacturer, the LIS interface is not a revenue source, but a product feature that facilitates an instrument system placement. Instrument and reagent revenue is what drives the IVD industry. LIS vendors view this same interface differently. Post-installation consulting fees are important to many LIS businesses and can represent 40 percent or more of an LIS vendor's annual revenue, and the importance of these consulting fees is growing as the number of new system installations declines. To move toward solutions, Dr. Friedman highlighted the need to get beyond our current dualistic, "us and them" models and seek "co-opetition," that is, solutions that bring a Hitachi, Abbott, or Beckman to the table with a Cerner, Sunquest, or CHC. Clearly, "there's all kinds of interaction going on between these two camps now," he pointed out. One way to frame the opportunity, Dr. Friedman suggested, is to think in terms of dynamically linked "up stream" and "downstream" integration of data. "Up stream" integration, which includes, but is not limited to, IVD-LIS interfaces, has a better chance of happening if the two parties can establish a common agenda. Success upstream will, in turn, facilitate the downstream process of delivering value to physicians, patients, and the entire healthcare enterprise. Reducing Information Float Dr. Friedman believes that there are opportunities in data handling to reduce information "float" in the lab. Drawing on the analogy of travelers' checks, which are paid for at purchase and redeemed later, being, in effect, a loan, Dr. Friedman defined the laboratory information float as the time period beginning when a clinician orders a test and ending when the clinician "knows" the result. In an age of time-based competition; laboratory competitiveness and success will increasingly derive from the ability to reduce information float. What we are seeing in our deployment of sophisticated work stations, automation of front end processes, innovation of POC testing, and refinement of the LIS with rules, reflex testing, long-term archiving, and clinically specific reporting, as well as streamlined IVD-LIS interfaces, is nothing more than better ways to do just that. Lack of Standardization and Cooperation as Impediments to Progress Calling the separate evolution of instruments and the LIS "interesting and sad," the Cleveland Clinic Foundation's Dr. David Chou went on to outline the progress that has been made in interface technology. Not much more than a decade ago, Dr. Chou pointed out, it was not uncommon for a laboratory to buy a major analyzer only to find it had no LIS interface capabilities at all. When minicomputers first began to appear in the lab, in the 1960s and early 1970s, the typical analyzer's "interface" capabilities were merely an extension of the ability to print hardcopy chart records. Only in the late 1970s did interfaces based on the soon-to-be ubiquitous RS232 emerge, and even then labs typically had to pay extra for them. However widespread it may have become, the RS232 interface is rapidly becoming obsolete, Dr. Chou said. Its electrical and connector specifications are limited. The maximum data-transmission rates are too slow. The requirement that instruments can be no more than 50 feet from the computer is almost laughable in an environment where a data center may be miles from the laboratory. By the 1980s, newer, unidirectional and bidirectional interface technology had emerged, and by the late 1980s barcode technology was also playing a growing role in the lab and thus in the interface picture. In the 1990s, Dr. Chou said, we have entered an era of standards advancement and promulgation, most notably under the aegis of the American Society of Testing and Materials. While we are not likely to see true "plug and play" interfaces soon, we are making progress toward a more comprehensive, stream lined instrument-LIS interface environment. Remaining barriers pertain to the inherently slow nature of standards development, depending as the process does on volunteer expertise and lacking as it does the involvement of a clearly dominant vendor - the clinical laboratory equivalent of a Microsoft or an Intel who can drive the process by sheer weight of influence and vested interest. Also, there does not yet exist independent means to verify interface protocols, Dr. Chou added. Standards do not adequately address the special demands of high volume analyzers. Delays are inherent in system design and FDA approval. For all its potential benefits, point-of-care technology has "set back interface technology 10 years," Dr. Chou said, since few if any POC devices fully support data transmission to the LIS. IVD Manufacturers Becoming Better Equipped to Expand their Role in Informatics Robert Bauer also observed that IVD interest in laboratory information management is on the rebound, driven by the manufacturers' desire for more product differentiation in a market where product offerings are increasingly reaching parity. One result is the wave of new workstation management systems from Abbott, Johnson & Johnson, Beckman, and others, which are filling voids that LIS systems and vendors have not been able to address rapidly, namely, lowering the cost of interfaces, providing modern information management tools such as e-mail to older LIS systems, and optimizing the workflow across two or more related instrument systems. Will the IVD role in lab information systems become more significant? Yes, Bauer predicted. All indications point in this direction. First, unlike the LIS, which made more sense to develop on a laboratory-wide basis, combination LAS/LIS solutions are likely to develop and be more affordable on a workstation basis rather than across the en tire laboratory. Second, IVD manufacturers are continually broadening the definition and scope of their core products to establish competitive advantage. Front-end automation is a developing trend. Third, the consolidation of IVD manufacturers begets larger organizations with a broader breadth of instrumentation that can be integrated with out the complexities of multi company coordination. The Impact of Laboratory Automation on Laboratory Information Management Possibly half of all clinical labs will embark on automation projects within the next five years, predicted Dr. Rodney Markin. In the process, this change will radically alter our current paradigm of information development and management in the lab. Dr. Markin is professor of pathology and microbiology, Departments of Pathology and Microbiology, at the University of Nebraska Medical Center. In Dr. Markin's view, arriving at that new paradigm will require improving our under standing of what happens in the laboratory. That in turn will mean reevaluating laboratory processes from several perspectives and on several levels, from local work-bench to very large, multi-organizational healthcare information net work. What we will find in this process may surprise us, Dr. Markin suggests. Most labs, he believes, do not fully understand the extent to which specimen flow differs from information development and management. Automating and integrating them will, among other things, force us away from our current view of the laboratory as being composed of many smaller, relatively independent sub sections. As a result of this change, we will increasingly find our selves thinking in terms of an LAS—a laboratory automation system—as well as an LIS, Dr. Markin declared. Moreover, that LAS will be come not just an important element in effective laboratory design but its essential starting point. "Ideally, if we were going to step back, look at the processes in a laboratory, and design a lab information system from scratch," Dr. Markin observed, "we would start with LAS and then move on to instrument interfaces and the LIS." Does the Clinical Laboratory Industry Have the Motive, Means, and Opportunity? Given that the clinical laboratory is becoming more physically dispersed, no longer contained within four walls or the "lab-as-factory assembly-line" model, the concept of a "virtual lab" has become more relevant, Bauer added. "With developments like point-of-care testing, with nurses running tests, the lab is becoming harder to define in a physical and organizational sense," he said. But, according to Friedman, making the virtual lab a working reality will require further evolution of interface protocols and standards, a better understanding of the complex interdependencies among specimen movement, information management, and workflow, and a much clearer grasp of the process control issues that the introduction of robotics and front-end automation is now raising. It will also require new initiatives, new business and technology arrangements between two industries accustomed to operating at an arm's length from one another, speakers agreed. "Aligning all these competitive objectives is not unlike asking Federal Express in the 1970's to invent the fax machine," states Bauer, "the virtual laboratory is going to happen... its just not clear yet who is going to lead the process."

	CaseBauer has 5 practices: Market Modeling, Strategy / Business Development, Benchmarking, Market Research and Business Services.