IN PERSON
 |
Larry J. Kricka, DPhil, is professor of pathology and laboratory medicine at the University of Pennsylvania and director of the general chemistry laboratory at the Hospital of the University of Pennsylvania. His research interests include analytical applications of bioluminescence/ |
As diagnostic possibilities expand with the advance of genomics and discovery of new disease markers, the burdens of central testing laboratories grow. Multivalent analyzers and the linking of instruments to an automation system can give labs a chance to provide the turnaround speed and throughput physicians and hospitals require.
IVD companies are developing the necessary new analyzers. Some are developing the automation tracks larger labs need. Still, the labs do not have everything they want, including, for some, sufficient options.
To learn more about thae state of laboratory testing today, and to find out what labs want, IVD Technology editor Richard Park spoke with Larry Kricka, DPhil, who is professor of pathology and laboratory medicine at the University of Pennsylvania and director of the general chemistry laboratory at the Hospital of the University of Pennsylvania. In this interview, Kricka discusses the important things to consider when planning for automation and describes the interplay between IVD companies and the labs. He focuses on trends, judiciously predicts future developments in lab automation, and presents his technology wish list.
IVD Technology: What are the principal factors taken into account when a laboratory is deciding which instrument to purchase?
Larry Kricka: I think I have to preface my answer by saying that our lab is fully automated. So we have a slightly different take on things in terms of instrumentation: we’re always looking for analyzers that are track compatible.
Nowadays, our choices of instrumentation are really determined by that. And that includes chemistry, hematology, immunoassay, and coagulation. In each case, we’re looking for analyzers that will attach to our track and become part of our total automation system. That’s often a key consideration for us. But, beyond that, I think there’s a fairly extensive list of criteria people think about when they choose an analyzer for a laboratory.
Do you want to say more about track compatibility?
Yes. I think that, with the development of different track systems, finding analyzers that can be attached to a particular track is not always easy. The companies making the tracks have gone two ways with this. Either they’ve tried to lock people into buying their tracks and into buying their equipment to attach to the track, or they have taken a more liberal attitude and encouraged other manufacturers to build the links between their analyzers and the track.
Nowadays, it seems to be trending more toward the former; that is, it’s a little harder now to get interfaces onto existing tracks, I think. The manufacturer often has the customer almost tied into its equipment going onto the track. So that tends to cut down the choice when it comes to selection of analyzers, especially if you’re going to have a track in the lab. Once a laboratory has decided which track it’s going to have, its scope of choice regarding analyzers becomes somewhat restricted.
Have IVD manufacturers responded to this increased demand for track compatibility from labs such as yours? And does the kind of limit on choice you just mentioned—if a lab buys one manufacturer’s track, it’s locked into using that manufacturer’s equipment long-term—cause the lab to have to consider very carefully before it makes a major investment in a lab analyzer?
I think it is increasingly the case that the vendors are locking you into their equipment if you buy their track. It’s just an unfortunate reality that to choose a particular type of track and then choose individual analyzers one by one and put them all together into one system is difficult, perhaps impossible, to do now. A lab has to make some concessions when it comes to analyzer choice if it has chosen a particular type of track.
I’m not sure there are any big trends toward openness of track systems currently. I’d say perhaps it’s the reverse.
That has certainly been our experience with attaching analyzers to tracks. We’re just in the process of putting in our second track system. When we put in our first track system, there were perhaps a few more choices. Now there seem to be fewer choices.
Do you see any indication that IVD manufacturers realize the limitation this imposes on clinical labs? Is there hope that these companies will be a little more flexible so that, if a lab wanted to get its clinical equipment from Company A but liked what Company B had to offer in the hematology area, those different instruments could be linked up and be track compatible?
I see just a slight indication in that direction. Today, if you choose a particular track from the several that are available, you will be able to attach analyzers to that track. There will be a selection of analyzers from that supplier, or through that supplier, that will be attachable.
What doesn’t exist is the variety. A lab can’t go to one track manufacturer and expect to find it has seven different types of chemistry analyzer that attach to its track. That obviously doesn’t happen. More likely, one type of chemistry analyzer is attached to it, or one type of hematology analyzer.
Having IVD manufacturers make their instruments universally attachable to tracks is perhaps an attractive idea, but I think the reality is that attaching analyzers to tracks is often not quite that straightforward. There may often be competition issues involved in allowing one diagnostics company to have access to what’s needed to make its analyzer attachable to a track. I think the seller of the track, who perhaps already has its own analyzer that attaches to it would see that—and does see that—as something that’s competitive, and wouldn’t want to encourage it.
So, it’s not necessarily the case that an IVD company that offers an analyzer that isn’t track compatible can do anything about it. That company may not easily be able to attach its analyzer to the track, because information it needs may not be available. Also, the track company may have some issues with the idea. It may feel that an attachment that it hasn’t designed and built to enable an analyzer to go onto the track is something that might be detrimental to the track.
Those are some of the issues that have stopped the development of universal interfaces for tracks. Obviously, robotic interfaces between a track and an analyzer can be built, but those tend to be custom, and custom things aren’t always the most desirable type of arrangement in the laboratory.
So then, some companies have responded to the market demand and have enabled track compatibility while others haven’t?
If an IVD company is already selling tracks, then it usually has as near to a total solution as possible in terms of having various types of analyzer that will attach to the track. So, in that sense, when a lab buys the track, it does have a choice of analyzers.
But a company selling analyzers that is not a track company has a situation that’s a little more difficult. Building or designing in the compatibility may or may not be a possibility. It depends on how open and welcoming the track company is, especially if the track company is hoping to sell total systems and not just the track.
For labs like yours that are totally automated, that limits the market selection.
Yes. I think most clinical labs that are going the track route are selecting from the analyzers that’ll attach to the track that are offered by the IVD company that makes the track. It’s just not possible to have completely open choice in these matters. It’s not practicable anymore.
Instrument Features and Capabilities
Lab instruments over the years have offered more and more features, and they are capable of running many different types of tests. Which of their features and capabilities are most important to labs and laboratorians now?
There’s such a large list of issues to think about when you’re looking to purchase an analyzer, that ranking them is sometimes difficult. The cost of the analyzer, along with the cost of operating it, obviously is important, but perhaps not always the most important thing. Throughput is important. Fast analyzers clearly are very desirable. Ease of use, I think, is becoming a big issue.
As the medical technologist and clinical chemist workforces shrink, labs are looking for instruments that are easier to use and easier to train people to use. Also, I think they are looking for broad menus. That way we can have fewer different types of analyzers. I’m sure that’s a trend.
A number of companies have produced hybrid analyzers where, perhaps, a chemistry analyzer and an immunoassay analyzer are put together into what is essentially one instrument. They have done that to try to simplify instrument operation and bring together more tests so that there are fewer “benches” in the laboratory. That diversity of menu is becoming increasingly attractive to laboratories.
Things on the analyzers that streamline their general operation—indices on chemistry analyzers so that you can automatically find specimens that are hemolyzed or icteric or lipemic, the capability of the instrument to also dilute so that you don’t have to do it—are important. Stable calibration is another important issue. Also, some of the information technology (IT) aspects of analyzers—bidirectional interfaces, obviously.
But some other things are starting to look rather attractive as well, such as the modem-based diagnostics some companies are offering so that the lab’s analyzer is being watched constantly and failures are detected early. Things like that I think are of interest to people in laboratories, especially when they’re faced with maintaining a 24-hour operation where machine downtime can become a really important issue. All of these electronic and IT-based solutions to help keep analyzers running effectively and efficiently with minimum breakdowns and downtimes are important to labs today.
Would you say that IVD manufacturers, like the clinical labs, are trying to respond to demands coming from whatever hospitals and clinics? Have they done a good job of recognizing the trends and responding to them appropriately in terms of the technologies and devices they’ve introduced?
Yes, I think they have been responsive to what has been a changing situation in the clinical laboratory. Obviously, there are more things that could be done. More simplification would be an advantage. I mean, operating analyzers is a fairly complex matter, but you can find other types of instruments that are also complex yet require a very low amount of training.
The one I always like to cite in comparison is the photocopier, which has cartons of reagents inside it. One is rarely trained to use one of those machines. It’s the sort of thing you can walk up to and, by inspection of the control panel, make it do what you want it to do. Clinical analyzers obviously are considerably more complex, but perhaps their manufacturers could learn a little from the photocopier type of technology, especially the user interface.
And simplifying the user interface—making machines more intelligent, easier to use, more intuitive—I think represents an opportunity waiting to be explored and exploited. That hasn’t been worked on as much as it should have been. Just look, for example, at the telephone, how it has evolved and how its user interface has evolved—especially into things like the iPhone.
User interfaces generally are becoming simpler and simpler, and perhaps some of that technology could come over into the analyzer world and be beneficial to the lab. Obviously, we have touch screens on analyzers. We have modem links. There are a lot of things we do have, but perhaps there are even more that we could have.
I think this is especially important in light of the fact that the pool of employees is either shrinking or static. It’s not increasing. I think many labs around the country have difficulty finding new staff. Perhaps one of the ways to improve matters is to make work in the lab in some areas a little easier. This is why I think improving the interface could be beneficial.
What do you foresee as other features and capabilities of lab instruments that might help labs run better tests and run them faster and more efficiently?
Well, one thing to do, perhaps, is to look at what technologies are emerging that might be helpful to the clinical lab. Two areas that come to mind are microfluidics and nanotechnology. Innovations might come from those technologies that could change and improve the way we do things, making lab operations easier and simpler.
And there have been some developments in those areas that are promising. In microfluidics, devices that facilitate sample processing might have an application in the central lab. Obviously, a lot of microfluidics research and development has been directed toward the lab. I’m referring to the lab-on-a-chip idea, where it is intended that this microlab would be used in the field or at the point of care rather than be part of a central lab analyzer. Perhaps there is scope for some aspects of microfluidics to be adapted to large central laboratory instruments.
It’s difficult to say exactly what nanotechnology is going to offer. But one thing that it is offering already is different end points and detection systems, which lead to greater sensitivity. There are a number of assay systems that are using nanoparticles of various sorts to try to improve detectability and reduce detection limits.
Some of that might come into the lab and be helpful for some of the analytes for which we really need ultrasensitivity. It remains to be seen how that will play out in the central lab setting. A lot of these developments are focused, I believe, on smaller instruments for smaller labs, or for point-of-care testing, rather than being something that’s going to change the way in which we do central laboratory testing.
And that brings up another aspect of this that many people are extremely enthusiastic about: point-of-care, or POC, testing. The most-enthusiastic voices predict a future in which there’s a lot of point-of-care testing.
If that turns out to be the case, one wonders what effect that would have on central laboratory testing, especially for a laboratory that’s doing a lot of outpatient testing. If POC testing becomes more prevalent, then perhaps the outpatient testing that labs like ours do now might be taken up by it, and our volumes would be reduced.
It’s very hard to see how that will all play out. There’s obviously a tremendous amount of optimism about POC testing, but it hasn’t gotten to the level yet where it has an appreciable impact on central laboratories’ test volumes.
POC Testing
Do the big reference laboratories and hospital labs—the major labs—feel concern about the growing trend toward POC testing, that it may take away some of their business? Or do they have a different attitude toward it?
I can give our viewpoint. Because we’re a hospital laboratory, we actually run the point-of-care for the hospital; so we have a very extensive POC testing service here. We don’t feel threatened by the growth of point-of-care. If that’s the way things evolve, that’s the way they will.
And when people want to do point-of-care testing, we don’t think in terms of losing a lot of testing from the lab, because it’s usually things that are not done in very, very high volumes. With POC testing, there’s usually also going to be some follow-up in the central lab as well.
So it’s not as if we’d lose all the testing. We’d lose some of it—cardiac testing in emergency rooms, and things like that. That obviously takes away some of the testing, but we’ll still probably end up testing those patients in the central lab once the initial POC test has been performed.
So again, I don’t think labs necessarily feel threatened by that. It’s just part of the development. Things may slowly move toward point-of-care in areas where a genuine need for doing that can be shown.
What about other labs that don’t work closely with point-of-care technicians within a hospital organization? Will a switchover to more and more POC testing force those labs to rethink the model they’ve set up?
If that were the case, then probably they would. My personal experience of community and physician’s offices and places like that—certainly in the primary-care practices I know—is that they tend not to do very much POC testing at all. The question is, who’s going to pay for it and who’s going to reimburse them for it?
So, at the moment I don’t think that’s a strong force. I see a continuing mix of the central laboratory and some POC, with perhaps slow growth in POC. To really push the growth of POC would require some sort of external factor that would drive people in that direction. At the moment, the menu of tests is somewhat restricted.
As for what people really want, look at all the types of POC tests that are available now in the pharmacy. And if consumers wanted to expand the menu of tests they could do on themselves or their relatives, there are a lot of options where they essentially buy a collection device and send the sample off to a distant lab.
I don’t know how well those sorts of labs are doing, but they might make an interesting place to look in order to try to assess how popular point-of-care in its broadest sense is becoming. Beyond that are the companies that are offering the very large health profiles and the ones offering genome testing. They probably represent the far end of the spectrum in terms of self-testing ordered from the home.
Technology Advancements
Which advancements in lab instrumentation over the past couple of years have had the biggest effect on how laboratorians conduct tests?
Well, that depends on what “a couple of years” means. Certainly, from our perspective here in this laboratory, the thing that has changed the lab the most has been introducing total automation. I know that total automation is not a particularly new technology; we introduced it eight or nine years ago, so that’s more than a couple of years. But that has probably had the biggest effect on what we do in this laboratory because of its impact on so many of the specialties here. It brought high-volume chemistry and hematology and coagulation and toxicology and endocrinology together in one setting.
Not only is it a change in the way in which we do testing, but it’s also something of a change in the culture. Specialties that each have their own separate labs gave up some of their testing into the central facility.
Do you think more and more labs will adopt total automation as a part of the way they test?
I think more and more labs will adopt automation. Many, many labs have opted for just the front end of automation, dealing with the specimen processing, rather than building total systems and having, after the specimen processing stage, a track with analyzers on it.
But at least automation of that part of the processing is something that I’m sure will continue to grow. I don’t know what the numbers are in terms of systems installed in this country, but I think it’s fairly large if all of the labs that have got the front-end automation for specimen processing are counted.
What sort of factors are driving the expansion of test menus, and what kind of new tests do you think clinical labs will be asking to see included on lab analyzers’ test menus?
Progress is obviously being made on cancer markers, so more cancer markers that were effective would be welcomed by the clinical testing community.
Then there is another issue that labs are going to be dealing with: the increase in popularity—especially in the research arena—of microarray-type devices, either protein microarrays or nucleic acid microarrays. Those are slowly emerging in the clinical laboratories. Already, a number of FDA-approved tests are based either on planar microarrays, where you have reagents on a flat surface in the form of a small chip, or on liquid microarrays, which involves a collection of beads with characteristic signatures, each bead designed to test a particular analyte.
Those sorts of tests that are just appearing are going to create challenges for the laboratory, especially in quality control. It’s an open question what impact they will have.
There’s already one IVD manufacturer that has a chip-based system for immunoassays. How that will influence the way in which we do immunology testing remains to be seen. But that’s part of a larger trend to build analytical systems that are based on an array format of one sort or another.
How do you think the continuing growth of the molecular diagnostics market will affect the clinical laboratory, and also IVD companies in terms of how they fit molecular diagnostics into their instrumentation and automation offerings?
Currently, of course, molecular diagnostics is a separate specialty in most institutions. But the idea of having molecular tests along with other tests on the same platform is an idea that is not new. It is one that might perhaps become increasingly interesting, however.
Whether the appropriate technology for this exists, and whether it would be the right thing to do, again is open to discussion. But it might be an interesting thing to try to do, especially having immunoassay and nucleic assay technology on the same platform, with perhaps a common end point to make the two things compatible.
That was something suggested at one stage for a particular company’s analyzer; one of its features would be that it could do both of those things. I’m not sure whether anyone is making that sort of claim for a future analyzer now. But that would be an interesting development, I think.
Various projections are showing that molecular diagnostics will take a bigger part of the testing market eventually. Won’t IVD companies, and also clinical labs, have to address that and implement molecular diagnostics in their analyzers?
Well, if this testing area grows, then I think it’s inevitable that automation solutions will have to be developed for high-volume laboratories. You only have to look back at the growth of immunoassay testing, which started off with test kits that were manual, then evolved into systems consisting of components that were put together to produce some level of automation, and then moved finally to total automation on platforms that are all but indistinguishable from the chemistry systems that preceded them.
Perhaps molecular testing will follow the same route. It’s not there yet, but that might be the logical progression if testing numbers rise to the levels we see for immunoassays.
Changing Workforce
Less-qualified laboratorians making up a larger share of a shrinking laboratory workforce on the one hand, and, on the other, interest among IVD companies in developing tests on which they can get a clear waiver so as to broaden their market—both of these trends feed into the issue of instrument ease of use. Are these factors that IVD companies should take into consideration when developing their lab instruments and analyzers?
Yes, especially if it leads to greater reliability and instruments that are less prone to operator error. These things are desirable for improving the quality of analytical data. And some of that will come with greater ease of use.
I don’t think anyone particularly wants to keep analyzers complex. People in the lab certainly would be in favor of making analyzers easier to use and simpler to use, more intuitive. I don’t see that as a bad thing. And it will ultimately help with the workforce issues.
Are laboratorians generally aware of how the various subsystems and components that make up a complex lab instrument work? Should IVD manufacturers explain these things to them, or does it not matter?
Usually there’s going to be a number of people within the lab who have a fairly good knowledge of how an analyzer works in terms of components and subcomponents. And the primary operators of analyzers usually have been trained by the IVD manufacturer during the postpurchase phase of analyzer integration into the lab. Those operators usually have had the most intensive training in, and exposure to, the analyzers, and would probably know the most. But I think most people working in labs do have a reasonable understanding of how analyzers work.
Now, the nitty-gritty level of how things happen, the mechanics and the electronics, is the level about which some people in labs don’t have a great deal of knowledge. Nor are they motivated to acquire it. But I know they have a general understanding even of that, the extent of it varying among members of the staff.
With hospitals and clinics constantly demanding more and more tests, are increasing numbers of labs going to have to turn to automation solutions engineered to allow less-trained and less-qualified laboratorians to do the tests required?
Ultimately, making things simpler will perhaps provide the lab with more options when it comes to the workforce. We may end up at that point. There’s obviously a lot to be said for having highly trained people running clinical analyzers—people who’ve had a particular level of education and training—and the idea of using people who are less trained is something that not everyone embraces. But we may eventually have to adapt in order to get by with a workforce that has mixed levels of qualifications, which we do already to some extent.
Labs’ Wish List
Besides microfluidics and nanotechnology, what further improvements in lab instruments would laboratorians and clinical lab managers like to see?
You have to look at all the things that are inconvenient in a clinical analyzer: the time required to get it going so it’s ready to start testing, difficulty of interfacing with lab information systems, certainly unreliability. Reliability and ease of serviceability are two very important issues.
The amount of waste that analyzers produce is an issue. The need for water and drains on analyzers is an issue. Some IVD companies offer technology solutions that avoid these problems, but they are few and far between.
Finally, if an IVD manufacturer just beginning to design its next generation of lab instruments came to you for advice, what would you say?
I think I’d probably tell them that one of the most important things they could do in terms of designing for the future would be to work on the instrument interface. The operator interface would be an important thing to simplify and make easier to use.
And beyond that, the test menu. A limited test menu is a major problem. The manufacturer is not going to do very well if its instrument doesn’t have an adequate menu. That to some extent goes without saying.
An analyzer that was automatically track compatible with all available tracks would also be something interesting. Then, no matter which track the lab had, this instrument would be an option for it. That’s easier said than done, of course, because that would require negotiating in advance with various track companies to make sure that they would allow the laboratory to make the instrument compatible with their track.
