Medical Device Link.

Originally Published IVD Technology January/February 2003

In Person

While laboratory instrumentation has come a long way in recent years, there is still room for improvement.

Sue Stern is chief technologist in the chemistry section of the University of Michigan Health System's pathology department (Ann Arbor, MI). She can be reached at scstern@med.umich.edu.

Developing novel laboratory instruments and producing efficient upgrades that satisfy the needs of laboratorians and provide affordable solutions can be challenging to IVD manufacturers. This is especially true in a time of rapid technological advancements and economic instability.

To find out what laboratorians are looking for in instrumentation improvement and development, IVD Technology editor Richard Park spoke with Sue Stern, chief technologist in the chemistry section of the University of Michigan Health System's pathology department (Ann Arbor, MI). In this interview, Stern talks about which instrumentation components are most important to laboratorians, what factors most affect purchasing decisions, and what future improvements in laboratory equipment capabilities would be most useful to end-users. IVD Technology: What factors guide your selection of which instruments to purchase for your laboratory?

Sue Stern: We look at how the instrument fits into the laboratory, both in terms of utility and physical fit. Our general selection criteria are what assays the instrument offers and whether it does everything we need it to do. We look at ease of use for the technologist, difficulty of training a person to use it, whether it has laboratory information systems (LIS) capability, and whether we can integrate it into the laboratory quickly. The cost of the instrument or the cost per test also influences our decision, depending on how we are going to finance the purchase. A more recent consideration is whether this piece of equipment can eventually be modified to be added to a line of automated instruments, because laboratories are moving toward total automation.

Do you consider whether there will be future upgrades of an instrument?

Yes, as well as what is in development. If it is a newer piece of instrumentation, it probably won't have all the menu items we need. So, we look at the manufacturer's timetable for integrating assays, and at how much effort they are putting into further developing the piece of equipment.

Many new instruments are more modular than in the past. Being able to combine instruments often minimizes the need for purchasing upgrades. If we need an instrument that does basic chemistries and one that does immunoassays, like our therapeutic drugs or thyroids, we can hook two modules together to accomplish that. That way, we cut down on workstations and keep our instruments similar, which simplifies training.

Crucial Features and Capabilities

How important are issues of connectivity and the ability to set up a laboratory information system (LIS) for managing information both within your laboratory and among other departments within the university?

We would not be able to function without a laboratory information system. We are deathly afraid of our LIS going down. We would not even consider an instrument that could not interface with a computer and allow us to disseminate information.

Which features and capabilities offered on laboratory instruments are the most important to laboratories and laboratorians?

In terms of selecting a main high-volume instrument for our laboratory, we would not consider anything that didn't have primary-tube sampling, meaning it could sample right out of the collection vacutainer flood tube in the laboratory, thereby reducing aliquoting mistakes.

Primary-tube sampling reduces turnaround time. If the sample comes to us centrifuged, we can stick it right on the instrument. Also, by avoiding contact with the serum or whole blood, safety is increased by lowering the risk of infection. For our more esoteric type-testing of smaller samples on which we don't want to spend the money to send out, we may have to use a small analyzer that doesn't do primary-tube sampling. But we definitely would not use this type of analysis for a high-volume test.

We look for quality control [QC] of precise sampling, or a sample integrity check. In other words, the instrument should have the ability to detect clots in the sample, and also have a mechanism for determining that it doesn't have enough sample to run the assay. This is important for ensuring accurate results. It gives the technologist peace of mind by alleviating the need to determine whether every abnormal result is valid, and saves time that would be lost in reworking the sample.

Another important capability is reagent detection. Although it isn't as critical, it is helpful if an instrument can detect how much reagent is on board, and most of them do that now. Having refrigeration on board is important as well, so we can leave our reagents on 24 hours a day, even when we aren't using the equipment.

It has also become more important for instrumentation to be able to automatically dilute a specimen if it produces high values. There's variability in how technicians make dilutions, so it's convenient for an instrument to be able to determine if the concentration is too high, and if so, to be able to send the sample back around for dilution.

We also consider the ease of performing maintenance and how much time it will take. We do not like our instruments to be down for very long, because we are open 24 hours a day, seven days a week. We need to plan when we can take our instruments down, and if something's going to be down for 4 hours, we will need another instrument to back it up.

Do you look for a machine with a backup system to minimize downtime, or are backup systems built into the laboratory?

We do often look for a backup system, or create a backup plan, depending on what test we are doing. We have multiples of most instruments, because our general chemistry analyzers here do stat work for the ER and OR, so if one goes down, we have an exact backup to use.

We don't have an exact backup for some instruments that do more esoteric low-volume testing that isn't stat, though we have duplicates of most things in our laboratory.

What other features and capabilities of laboratory instruments would help laboratorians run better tests?

We could save time by further automating the intruments. For example, we had a device that uncaps a container, takes it to the instrument, and runs it. If the instrument could pierce the vacutainer tube and take out the serum or whole blood, we could save ourselves one more step. We want to be protected from touching serum as much as possible, so providing a stopper through which the machine could directly draw the sample would be helpful.

Incorporating calibrators and controls into the instrument would allow us to load it less frequently and improve productivity. Whenever a test needed to be calibrated or controlled, the instrument would automatically do it. Instruments can already tell us when we need to calibrate them, but we still have to put in the fluids and the QC material.

Nuts and Bolts

Do laboratorians prefer certain subsystems or components over others?

Laboratorians prefer sampling systems that can accommodate any size of collection tube. Years ago, there was only one kind of collection tube. Now we can use short tubes, tall tubes, fat tubes, and skinny tubes.

So laboratorians appreciate systems that have modified or grown with how samples are drawn because now we can load a system with a tube of any size, and not have to make adjustments. Instead of using one rack for a 12 x 75 tube and another for a 13 x 100 tube, there are now systems with one rack that will hold any tube, no matter how big or small it is.

Laboratorians also appreciate systems that do not require a large sample. We had an old piece of equipment that used to require 250 µl of serum to perform a troponin, which is quite a bit of serum these days. The system we've used for about two years takes only 10 µl. This improvement is wonderful because when we were screening at the ER with the old system, we always needed more sample. Now we can do creatine kinase MB isozyme and a troponin on a third of what we used to.

They prefer to have refrigeration for reagents on board because most reagents need to be refrigerated to keep their stability. With refrigeration, we don't need to load and unload reagents that shouldn't be left out at room temperature when the system is not in use. Also, they won't have to waste reagents by disposing of them when they have been left at room temperature for too long.

They would never get an instrument that could not read a bar code.

In terms of what subsystem is preferred, I do not think that technogists have a preference for how things are read, whether it is an absorbence or an immunoassay like chemiluminescence, they just want them to be read accurately. However, clinical chemists do prefer certain methologies or detection systems over others.

When laboratorians look for a new product or an update, they will often have a list of things that they want changed on an instrument. Because no instrument is perfect, we have to weigh what features are offered on which equipment from each vendor and determine what we can live with and what we can live without—pick our poison. After we've received the instrument, we try to get the manufacturer to upgrade it to where we want it.

What trends have recently emerged among subsystems and components in laboratory instruments?

The biggest advancement I have seen in terms of subsystems and components is the ability to combine instruments. Instead of having to interface with two or three instrument computers, we can hook two instruments up to one data manager, allowing us to do things on each instrument, but manage both through one system, which increases our throughput. It's all about getting more tests and more samples through in a shorter amount of time.

Instruments are becoming more modular. For instance, a manufacturer could produce a general chemistry module, and then develop an immunoassay module. If we do a lot of general chemistries, we can hook two chemistry modules together and have the sample system run down both modules, placing our throughput as quickly as if we only had one module. Then, if we do immunoassays, we can either attach another module, or leave it freestanding.

In addition, further automating the instrumentation enables us to develop our own track for pushing samples through so we don't have to separate a sample for general chemistry and more-specialized chemistry tests. We'll have one tube, put it on, and the instrument will do all of our tests, without having to either split the sample or draw two tubes to get our throughput.

Technological Advancement

What advancements in laboratory instrumentation technology during the past couple of years have had the biggest effect on how laboratorians conduct their tests?

There has been growth in bringing more esoteric-type testing to an automated platform and helping smaller laboratories do more than they could in the past. For example, automated radioactive assays take much less time than before they were automated. And although chemiluminescence has been around for a long time, it really seems to have caught on and spurred the development of automated tests in the last five years.

What further improvements would laboratorians like to see made to laboratory instruments?

Incorporating enough reagents into the instrument so that we don't need to stop and start would save time.

Another improvement would be having the whole dilution on board, so we can put a sample on, let the instrument process it, and have a complete result by the time we look at it. Anything that can improve throughput and give us a complete result, minimizing the need to move or handle the sample, will be important. It would be perfect if the instrument could store the sample on board, and if it requires a dilution, remove it and run it through the dilution.

Costs—anything that cuts costs would be an improvement. Making instruments less expensive, without sacrificing our product or the assay, is important. However, cost is one of my lower priorities. I don't worry about it as much as I should.

How important is cost?

If we don't choose the cheapest option, we certainly have to justify our choice. We tend not to buy our instruments, because they become outdated so fast. We tend to do either reagent leased or cost per test. We'll have a contract for five or seven years stating that we'll buy a certain amount of reagents or perform a specific number of tests. It is a leasing-type mechanism. A company may offer us a price of $1 to do creatines. In which case, we are buying reagents, consumables, controls, and calibrators for a zero amount, so for a 100-kit test of creatines we pay $100.

But in terms of cost, greater precision and accuracy can justify a higher cost for an instrument.

We also look at repeat rates. We get a feel for how many dilutions we'll have to run because a dilution is really another test.

We might get a price of $1 per test on one instrument, but on another we may buy reagents and components separately. The reagent might appear to be cheaper, but when we add in all our consumables, calibrators, and controls, it may end up being more expensive. When we do a cost analysis, we try to include every consumable and the percentage of repeats we expect to see on the precision. We can justify purchasing a piece of equipment by itemizing what we are actually paying for.

Future Directions

What advice would you give to IVD manufacturers that are beginning to design and manufacture their next generation of laboratory instruments?

Detection systems need to be able to utilize small sample size without sacrificing accuracy and precision. Manufacturers should provide simple, logical software that is useful, but not elaborate. It needs to be user-friendly. As a person on the bench, I do not want to have to click and open 100 screens to see one thing.

In developing a new instrument, manufacturers should look at what assays clinicians are focusing on now, but they should not build an instrument that can do everything. I think modular is the way to go, because we don't want all our eggs in one basket. If we have components or modules, when one requires service or maintenance, we can remain functional.

Having the capability to do our general chemistries, our immunoassays, and everything else we need to do, and being able to mix and match to fit our workload and workflow, is preferable to purchasing one megainstrument that can do everything.

I think manufacturers have been envisioning these megainstruments, but that is not the way to go. If we want to stat the task, we do not want it to spend 10 minutes on an instrument going through several modules, we just want to do our stat potassium in 3 minutes and be done with it.

Sue Stern is chief technologist in the chemistry section of the University of Michigan Health System's pathology department (Ann Arbor, MI). She can be reached at scstern@med.umich.edu.

Copyright ©2003 IVD Technology