Working with customers and sister divisions helps instrument developers resolve integration, consolidation, and ease-of-use issues.
|Cass Grandone is vice president, systems development, for the diagnostics division of Abbott Laboratories (Abbott Park, IL).
He can be reached at firstname.lastname@example.org.
A current trend for today's IVD companies is to integrate—both physically and virtually—different testing formats, while consolidating pre- through postanalytical activities to conserve elements of labor. Customers are beginning to expect complete solutions to address pre- through postanalytical workflow, and to demand access to information in real time. IVD manufacturers face the challenge of providing instruments to meet these needs, while keeping those instruments user-friendly.
To get an idea of how one market leader overcomes challenges in instrumentation development, IVD Technology editor Richard Park spoke with Cass Grandone, division vice president, systems development, for Abbott Laboratories (Abbott Park, IL). In this interview, Grandone also discusses the never-ending quest for quicker turnaround and accuracy of result, and synergies between point-of-care and routine laboratory testing.
IVD Technology: What have been the biggest technological advances in instrumentation development during the past few years, and what are the latest trends in that area?
Cass Grandone: One very significant trend is workstation integration and consolidation. Physically and virtually, there's been an effort by most large IVD companies to consolidate or integrate systems and data or information. Abbott Diagnostics focused a lot, within the Architect product line, on bringing chemistry and immunoassay into an integrated work cell that we call the ci8200. Integrating the two allows the laboratory a single user interface that can introduce a single sample tube and get results across multiple technologies. It's a much better utilization of time in terms of training and setup, and I see a lot of IVD companies—Roche, Bayer, and others—doing the same thing because of the cost pressure to reduce the labor component of the work flow.
Is this an attempt to make things better, faster, cheaper? Is it the layman's attempt at system integration?
I see a lot of IVD companies trying to consolidate more of the analytical, and pre- and postanalytical, activities. At Abbott, we're focusing on integration of classic clinical chemistry and immunoassay, both physically and virtually, to save elements of labor. That can include everything from reviewing the data in a single consolidated location, to routing the sample tubes in an automated fashion, to having a common and consistent user interface so that the instruments are maintained and set up in a single fashion.
Based on what you see and what you've heard in your interactions with your colleagues industrywide, is this something that is going on at many IVD companies? Is it something that we're going to see more of?
I think the big companies are moving in this direction. I think all of us are looking at ways to make the testing work flow much more efficient. I think bigger companies will continue to try to do more and more in that regard, even integrating hematology into the consolidation equation. And I see the smaller companies partnering up with bigger companies.
What new or unique technical challenges does the consolidation and integration trend present to you and your engineers?
I'll give you two examples. One pretty interesting challenge is when you have to physically integrate different types of testing formats that have vastly different processing times. When you connect things with tracks and automation, you have to think a lot more about optimizing throughput and queuing theory. For instance, a clinical chemistry analyzer will run much faster than an immunoassay analyzer. As you're routing a single tube across both of those platforms, you have to really consider the significant differences in throughput. You can't route tubes to the slower analyzer first, or you will choke down the throughput.
Another challenge is dealing with differences between the analyzers when it comes to protecting the integrity of the sample. In chemistry testing, there's not as much of an emphasis on carryover as there is in immunoassay testing. When you want to route a single tube to different analyzers, you have to make sure that you can protect the integrity of that tube and not contaminate it. Different companies do it different ways. Some aliquot the primary tube, which is probably the easiest and safest way. But the downside of that is more complexity and longer processing times. At Abbott, we spend quite a bit of time optimizing and proving that we can rinse and clean the sample probe as it accesses the primary tube, so that we can get the same kind of performance on the chemistry platform as on the immunoassay system.
What are the primary challenges that IVD companies encounter when designing and developing their laboratory instruments?
The biggest challenge is that hospitals and private laboratories everywhere are looking for companies that can offer complete solutions to address everything from the preanalytical to the postanalytical part of the work flow. I think the days are gone when a company—whether it's Abbott, Bayer, Roche, or Beckman—can offer a single analyzer. Customers are much more savvy. They are looking for single solutions for comprehensive lab work-flow issues, so IVD companies have to offer the complete solution.
The second biggest challenge is that instrument systems are getting more and more complex. There's much more computerization and much more software involved in data information flow. So companies have to do a lot more to make very complex systems seem simple to use. And because of the shortage of talented medical technologists and technicians, we have to design systems with very simple user interfaces and very simple daily maintenance procedures—and make sure that they are even self-diagnosing their faults and problems—so that it doesn't require the same level of technical talent to run them.
How do IVD companies go about overcoming these challenges?
The most significant thing we are doing is developing platform families. Back in the 1980s and 1990s, when we developed a new system, it would have a different user interface and a different way of setting up and running. Now we're designing common user interfaces, so that when you get an Abbott system, it has a similar operational flow. This means people don't have to retrain on different instruments all the time.
We're also providing a product called AbbottLink, which we can interface to our AxSYM, Aeroset, and Architect systems. This software allows the system to be in constant communication with Abbott and can be used to remotely determine any problems that are going on in the system in real time. When AbbottLink is connected back to Abbott, our field service or customer service folks can more quickly diagnose symptoms that technologists can't explain over the phone. To address HIPAA and patient privacy issues, we've worked through logistics to make sure that none of the data that are transmitted back to Abbott are confidential or patient related.
What other issues have emerged and will continue to emerge?
I think ease of use is a significant area of focus. When Microsoft introduced Windows 95 in September of 1995, it changed the whole paradigm of the desktop user interface rather dramatically. Going from Windows 3.1 to Windows 95 fundamentally changed the interaction of the average person with a computer. We feel that whether it's with connectivity or with the software user interfaces that are local to the specific analyzers, this notion of ease of use and human factors will be a fairly significant focus for quite a while on systems.
However, we continue to focus heavily on reducing turnaround times, so that treatment decisions can be made more quickly. It's the never-ending search for the Holy Grail. The other thing we focus on, which doesn't get a lot of publicity, is the importance of accuracy of result. We do everything we can during what we call the chain of custody—from the time we get access to a sample until the point that that result is released to the practitioner—to make sure that the result is as accurate as possible. There's a lot of documented evidence of laboratory error throughout the work flow—preanalytic, analytic, postanalytic—and I think ensuring accuracy of result is going to continue to be a major focus, because that's probably one of the most important things we can do.
When developing lab instruments, do IVD companies try to come up with completely new instruments, or do they tend to build on their current instrumentation offerings through upgrades and modular editions?
I see a lot of what I call platforming, which means you develop a family of systems and define certain common elements within that family. For Abbott, one of those common elements is the software. We focus on ease of use and a common look and feel for our Architect software, and then that single software application will run on six to eight different platform members.
A second example is our sample-handling device, which we call the retest sample handler. It's a novel approach that we developed to maximize throughput, provide continuous access to loading and unloading of samples, and provide a way of introducing samples in a stat mode. We believe it represents a significant improvement in terms of throughput and turnaround time, so we implement that across the board. It's a way for us to establish our value proposition and our brand identity.
On the analytic side, we tend to innovate without having to change something completely every time we develop a new platform. I think this is the case with any company. If you start over from scratch every time you launch a new system, there's a tremendous amount of technical risk in introducing a new reagent system, optical detection technology, or software architecture. So we tend to lay out our platform architectures years in advance and plan for innovations, enhancements, or product refreshes in a very organized fashion. We introduce new innovations or focus technology updates without having to start all over.
We measure technological advances against fundamental customer needs, which tend to revolve around ease of use, time savings, and accuracy of result. So advances have to first measure up to the red-face test. Then we spend a lot of time very early on, years in advance, doing basic research and feasibility studies in reducing the overall technical and implementation risk. We put a lot of our activities through what's called a phase gate process, where you start from the invention phase and reduce the risk over time all the way up until the implementation phase. A number of these innovations or technological advances are flushed out, sometimes years in advance.
With the emergence of the Internet and integrated logistics support (ILS), what role has connectivity played in instrumentation development?
It hasn't played a big role until the last two to three years. When you integrate chemistries and technologies, you're physically integrating them and you're virtually integrating them. The virtual integration is really that connectivity. Historically, labs have relied on their LIS systems to integrate data management, but increasingly customers want more- portable software.
You've probably heard the term middleware, meaning software solutions that can connect analyzers and collate the data. They allow the laboratory technician to view the data in a single, consolidated area with a much more easy-to-use interface than many traditional LIS interfaces.
In many parts of the world, the LIS system is still a very difficult interface. It's analogous to the old DOS format, where the information is not represented in a graphical or easy-to-use format. Customers want to get information in real time, accessible to anybody who's running the systems, in very close proximity to the analyzer. They want to be able to see where a sample is among four or six different analyzers that the sample tube is being routed to. So we're seeing a much more significant emphasis in the area of software, but also in connecting systems with tracks. Connectivity is really emerging as a fairly significant driver in the laboratory work flow.
How are HIPAA, privacy rights, and security issues handled by IVD manufacturers like Abbott?
The real issue is whether patient-specific data get outside the lab. There's a huge emphasis on protection of patient rights. In our systems, we know the various pieces of information that are private. We spend time with customers doing usability tests early on, getting their feedback on how we can mask the data, and how we can validate or provide them ways of ensuring that the patient data are not being sent out. It can be a very sensitive issue with customers, so we have to involve them in the design phase, and implement design techniques and validation means so they can see for themselves that the data that are private aren't being sent.
How is point-of-care testing affecting instrumentation development?
There are a couple of significant variables there. One is that the instrument must be portable, because you have to get the testing to the point of care. The second is that you have to be able to get the results very, very quickly with a fairly low level of technical training. When you bring the test to the bedside, it requires a whole different paradigm in terms of simplicity and turnaround time. We've worked with our point-of-care businesses, looking at the techniques they use and bringing those into the routine laboratory. We're looking at quite a few synergies between our point-of-care technology and our routine testing.
Point-of-care testing, for instance, typically uses whole blood as a testing medium. You can imagine that in the routine laboratory, where you have these larger analyzers, testing whole blood and not having to worry about gel separation tubes and centrifugation would have a huge impact on the sample-preparation element of the work flow.
Another example is sensor testing using potentiometric means. Our sister division has quite a bit of expertise in that area that we don't have. They can give us a lot of good feedback on developing new sensor-based detection technologies that we can implement, for instance, on our Architect c8000 or our AxSYM analyzers. They also have core competencies in designing for costs, miniaturization, and simplicity, which we don't necessarily have.
How have developments in molecular diagnostics affected instrumentation development?
I see molecular diagnostics evolving much like routine clinical chemistry testing did in the 1970s and 1980s, and like immunoassay testing did in the 1980s and 1990s. These technologies were reduced to practice, and made much more cost-effective and reliable, through full and complete automation of the various manual steps. Right now in molecular diagnostics, they do a lot of the pieces of the automations separately. So it's just a matter of time before companies find cost-effective ways of fully integrating lysing, thermocycling, and optical detection. From a detectability perspective, molecular diagnostics represents a quantum leap in technology. But it's not necessarily the most cost-effective approach in some laboratory environments. Automation will continue to evolve as it becomes more cost-effective. And as IVD companies reduce the technical risk and make it more reliable, I think it will go the way that clinical chemistry testing and immunoassay testing did.
What sort of things will IVD manufacturers have to do to address current cost, reliability, and technical concerns with molecular diagnostics so that it can become fully integrated?
The first challenge is in simplifying sample preparation, which is the most significant technical risk in molecular diagnostics. If you don't take extraordinary steps to contain the sample when you're amplifying it, you can create contamination issues that can compromise the result. These issues are far more devastating than traditional liquid-based contamination issues, because you're talking about DNA and RNA amplification.
The second challenge is finding ways of multitasking so that you can combine, for instance, the thermocycling and the optical read—whether concurrently or sequentially—in a way that is fast and simple to do from a reliability perspective.
From what you know and what you've heard, is biodefense being addressed by the IVD industry?
I think there are small start-up companies focusing in these areas. Systems tend to be much more manual, though. The level of automation is just not there. This is a niche market. But there has to be an element of consistency. My personal feeling is that the consistency in terms of the need isn't there yet, so there are a number of smaller companies that have developed more manual assays to address various markers and bioorganisms. But I'm not aware of the same level of investment in big hardware and automation.
What challenges will emerge in instrumentation development in the next couple of years?
We'll continue to focus a lot on physical and virtual consolidation and integration, and combining data management. We'll augment that with additional capabilities on tools like AbbottLink, and we'll be involved in working with partnerships to develop some of these preanalytic and postanalytic solutions. These days, as an IVD company, you have to really understand the complete work flow. You can't just focus on the test. You've got to focus on everything from the phlebotomist to when the result is transmitted back to the practitioner.
Copyright ©2005 IVD Technology