Automation nudges urinalysis into spotlight

July 2007
Feature Story

Anne Paxton

Within the laboratory, urinalysis has often seemed like a hard-working but neglected stepchild that no one is quite sure where to house. But as the diagnostics industry continues to make the chemical analysis of urine more quantitative and more automated, that perception is changing.

“Urinalysis has been sort of relegated to a back room and perhaps not covered under optimum conditions,” says Sharon Ehrmeyer, PhD, professor of pathology and laboratory medicine and director of the medical technology program at the University of Wisconsin, Madison. “But there’s certainly a growing realization that a lot of valuable information comes out of a urine specimen, and it should be treated properly.”

Urinalysis instruments connected by track systems are not only producing faster, more accurate results, but also are bringing urinalysis new stature among diagnostic tests. In Hollywood terms, urinalysis may finally be ready for its close-up.

“Chemistry, hematology, and the blood bank are the big names in the laboratory,” says independent consultant Diane E. Berry, MS, CLS, MT/ SH(ASCP). “They’re always seen as the high-volume, high-dollar tests that you’ll base more life-and-death types of treatments on, and they’re more directly related to working a patient from the ER ‘to the floor or through the door.’ Urinalysis has been used as a screening tool more than for disease detection, and is usually a low priority.”

Particularly when it’s done manually, urinalysis is much more subjective than most laboratory tests. “There is a set process for reading sticks’ color change and assessing the physical properties of urines, but there is lots of variation in how it is performed,” Dr. Ehrmeyer notes. For example, in teaching her basic urinalysis course, she includes cautions to her students to follow package inserts and not to use confusing terms like magenta, chartreuse, or mauve when describing urine color.

“Most hospitals now have some type of semiautomation for urinalysis, and when those instruments were adopted, laboratory people thought it was a quantum leap from the old days when they did everything by hand,” says Berry, who formerly supervised the hematology department at Salem (Ore.) Hospital Regional Laboratory. “But in hematology and chemistry, even in the smaller laboratories, there have been much greater strides in obtaining instrumentation where you could put the specimen on and walk away, then view results and findings on a computer screen.”

“When they were able to link the semiautomated stick reader to a semiautomated microscopic analyzer, urinalysis really changed bigtime.”

Regional Medical Laboratory (RML) in Tulsa, which performs 5.3 million tests per year mostly within the state of Oklahoma, has seen great benefits from its increasing use of urinalysis automation, says Lawrence Johnson, MD, chief of hematology, flow cytometry, and urinalysis. RML uses two Sysmex UF-100s to aid in performing 135,000 urinalyses per year.

“We’ve really been trying to automate as much of our systems as we can without sacrificing quality,” he says. “Our clinicians appreciate the fact that we can turn out results so rapidly.” Automating urinalysis was the next logical step, since a fair amount of flow cytometry, hematology, and chemistry at the laboratory has already been automated.

“When we did microscopy manually, we used the Bayer Diagnostics Clinitek Atlas automated chemistry analyzer as our first step for chemistry strip analysis, and then we had several ways clients could order additional tests. For example, a routine urinalysis strip, depending on the results, would go to a manual microscopic exam, or to a microscopic exam with a reflexive culture,” Dr. Johnson says.

“However, part of the problem with the urinalysis strip analysis by itself is that it is still relatively insensitive to some urinary tract infections [UTI], and it doesn’t help distinguish if the UTI is of glomerular origin [of kidney origin] or lower down the urinary tract, which has important clinical implications. In addition, the reproducibility between even very good technologists using the manual microscopy method is not the best, and they only examine a small portion of the total sample.”

“Because of manual microscopy’s lack of reproducibility and insensitivity, and our need to increase our volume without necessarily having to add to our staff, the UF-100 was a good choice for us,” Dr. Johnson says, because it offered more reproducible results with greater sensitivity—important features in any screening test.

The system works much the same way that a flow cytometer combined with features of a hematology analyzer might work. “Urinary formed elements are stained with fluorescent dyes specific for nucleic acids and membranes. The laser measures size and complexity based on fluorescence staining and light scatter characteristics. It combines this information with impedance measurements in order to detect white cells, red cells, casts, epithelial cells, bacteria, yeast, sperm, and it offers quantitation of most formed elements.”

RML’s chief reasons for using the Sysmex system are that each unit can perform up to 100 tests per hour, and it produces reliable results without loss of sensitivity. “We used the culture as the gold standard in order to check the results against a manual microscopic exam and the automated microscopic exam provided by the UF-100, and from that we derived the cutoffs as to when to reflex to a culture, because the clinicians were concerned, especially the infectious disease doctors, that we were over-culturing urine samples, and they were correct.” He estimates that before the Sysmex UF-100 came in, at least 25 percent of the total urine cultures done were unnecessary. “They would be negative or flag a bacterium that was a contaminant.”

RML was also able to determine cutoff values where clinicians would not necessarily treat or culture the patient even if the white count was at a level that, on the face of it, might appear to be elevated. “In the old way of reporting, the clinicians would say that 2+ white cells, as determined by manual microscopy, is the cutoff for normal. But that is a meaningless value to me as you cannot reliably reproduce it even between very good microscopists. Our cutoff is now a quantitative value of 61 white blood cells per mL,which required educating our clinicians about what this quantitative value means versus the old method of 0, 1+, 2+, etc. At that value or more, the specimen would be flagged and indicate that it is worth culturing those patients.”

However, Dr. Johnson notes, cutoffs vary quite a bit among laboratories. “I’ve seen some as low as 25 cells per mL up to more than 111 as a cutoff. They’ll also use bacterial counts as an indicator to reflex to a culture, which I think is a very poor indicator.” There are significant differences, he notes, in baseline bacterial counts between inpatients and outpatients, as in some cases samples from outlying areas may sit around in suboptimal conditions, without refrigeration, leading to high bacterial counts, making bacterial counts highly variable without good predictive value.”

“We didn’t see the same problem when we just narrowed the cultures to reflex based upon nitrate positivity, the presence of yeast, and elevations in the urine white cell count. When we did that, we had very reproducible criteria for whether a culture was indicated.”

While manual methods of urinalysis are very much technologist-dependent, he says, “there is no way you could have one tech manually examine 200 samples an hour and do a good job. We’ve been able to get one technologist to handle a markedly increased number of samples without slowing down his work or compromising the quality of the testing.”

Clearly, those kinds of savings can bring a quick return on investment. “There are tests where we accept that we’ll lose money because of the benefit to the patient and the client. In this case, we were lucky in that we were able to bring the instrument in, provide even better care, and not lose money. In addition, it was very cost-effective because we didn’t have to add another technologist.”

After Salem Hospital Regional Laboratory installed its iQ200 automated urinalysis system (consisting of the iQ200 Automated Urine Microscopy Analyzer and the Arkray Aution Max AX-4280, made by Iris Diagnostics), it found that moving from semiautomation brought huge and unexpected improvements to its order-to-result turnaround time. (Related article: Unexpected findings)

But the laboratory also reaps savings from having results in digital form as well as from improved accuracy, Berry says. “Normally when you are doing a manual microscopy, the images would not be saved. If the microscopic cellular elements are allowed to sit at room temperature, they disintegrate as chemical changes take place.”

“If they’re refrigerated, it slows down, but if you save them you hope the same things you saw that were abnormal will still be there to identify. Storage becomes a huge issue in laboratories doing 100 to 300 urines a day. Where do you put all of this, and for how long do you keep it?”

With their Iris system, “if a dialysis doctor or renal specialist came down and said, ‘I don’t believe those were calcium oxalate crystals,’ we are not only able to show the chemical analysis but also saved digital images of what was seen in the urine at the time of testing. Images are archived and serve as a great training tool.”

The urinalysis analyzers could have an even larger impact as they start to include other body fluids on their menus. “Body fluids are extremely technologist-intensive,” Berry says. “The Iris is programmed and has FDA approval to test cerebrospinal fluid, semen, and other body fluids like bronchial lavage through a module of the iQ200.” Iris Diagnostics says urine samples and other body fluids may be run at the same time without operator intervention.

Berry adds that the consumables to perform body fluid testing manually are not that expensive, “but the cytospin disposable funnels that you use can be costly, and Iris may eliminate that cost.”

Several fluid types can be run on the newest high-end hematology analyzers, and have pending FDA approval, Berry notes. “But some viscous fluids like synovial fluid cannot be run through a hematology analyzer because the openings are too small and it clogs the system. You don’t want to put them through the cell counter and jeopardize its functionality.”

“As a supervisor and technologist, I’d be more comfortable putting body fluid through a machine like the Iris than I would be compromising the quality functioning of a hematology analyzer.”

Installing any automated system is a good occasion to ask why things are done a certain way, she says. “When we reported out bacteria prior to the Iris, we agonized over grades of 1+, 2+, 3+, 4+, because in some institutions the difference between a 1+ and 2+ meant an automatic culture. So some staff would try to err on the side of caution and artificially inflate the bias on the count.”

“They’d call it a 2+ because the patient was a woman and the diagnosis code on the computer said ‘burning on urination,’ and they didn’t want to miss anything. So we said, Let’s not do grading; how about just ‘present’ or ‘not seen’? That’s what we ended up doing, and that’s added to the streamlining of the processes.”

But to ease the adjustment for technologists, she advises a “stepwise” approach to installing an automated system. “The first month we went live, we were doing only clear urines on the Iris. We wanted people to get used to looking at regular images of uncomplicated specimens and get used to the software and Iris menus. In the second month, when all the urines were going through the Iris, whether they were clear or turbid or green, people were much more confident, much more ready to handle it.”

Now the technologists are automation advocates. “Once the system was up, about four months after the live date, we had something break. They fixed it in 24 hours, but I’ve never had so many techs come and complain to me, in my six years there, that they had missed something.”

Berry maintains that it’s not necessary to be running a huge volume of urinalysis testing to make the move to automation. Even small hospital laboratories can do so. “I’m consulting now at a hospital that’s 100 beds and is using automation even though they may do only 30 to 40 urines a day,” she says.

With 350 beds, Providence Hospital in Anchorage isn’t that small; it’s probably the largest laboratory in Alaska. But in 2002 it was still performing more than 120 urinalyses a day by largely manual methods, says core laboratory manager Jim E. Atwell, MPA, MT(ASCP)H.

Though the laboratory had a Bayer Diagnostics Clinitek strip reader, the time spent pouring out samples, dipping, capping, centrifuging, and adding stains was a problem. “One urine including spin time would easily take at least 15 minutes. So on evening and night shifts, it was just eating our lunch. We could never keep up,” Atwell says. The laboratory was only meeting its 45-minute turnaround time goal about 25 percent of the time.

Choosing an Iris iQ200 system because it allowed “true walkaway” operation, the Providence Hospital laboratory was able to improve its turnaround time significantly. It now has 92 percent of urines reported in 30 minutes from the time the laboratory receives them.

It took very little time to recoup the cost of investing in the system, Atwell says. “I got my return on investment in about eight months. It was quite a surprise. When we were doing a cost analysis, we actually found we’re using less consumables than before, and I was able to relocate one FTE technologist to another area of the laboratory.”

On top of reducing turnaround time, the automated system standardizes the process of reporting out microscopies and makes it easier to respond to clinicians’ concerns. “If there is a question on sediments, with the manual test you’ve already poured it down the sink. With the Iris system it stores those images so you can go back and share them with the physician.” That makes it a great teaching tool as well, he adds.

“I’d have to say that a system to automate one of the most common tests ordered is a definite advantage. Most of my issues with urines have disappeared,” Atwell says. Now he is free to focus on the remaining bottleneck in conducting urinalysis: getting the samples to the laboratory in a more timely fashion. “We have a pneumatic tube system that comes from the ED. But it’s not a direct line, so since it’s shared with pharmacy and admissions, it may sit there before it’s sent to the lab. We’re still struggling with that.”

The Sansum Clinic’s decision to install an automated urinalysis system was driven, in part, by the laboratory’s location in the resort town of Santa Barbara, Calif.

“About a year ago, we were seeing a volume of about 100 UAs a day. It required one clinical laboratory scientist to be on the urine bench 10 hours a day, manually testing, reviewing results, and performing microscopy on those specimens that met the review criteria,” says Lisa Heinrichsen, CLS(NCA), hematology and urinalysis supervisor at the clinic, which is a nonprofit medical foundation.

The clinic has had a job opening in the lab for the past year, and with the price of a house in Santa Barbara averaging $1 million, Heinrichsen says, they were not getting any response to their job announcements. She herself commutes 75 miles to the clinic every day. “The lab was taking on more outreach, resulting in more outpatient work being sent our way, so we looked at automation as a way to deal with increasing volume,” she says.

The Sansum Clinic chose an Advia Urinalysis WorkCell, a marriage of the Bayer Diagnostics (now Siemens) Clinitek Atlas urine chemistry analyzer and the Sysmex America UF-100 urine cell analyzer, because it was similar technology to some of the automated CBC machines.

“It looked like it would give us the chance to reduce the number of manual reviews we perform, so we chose to go in that direction,” Heinrichsen says. “I liked the idea of the Sysmex UF-100 linked to the Atlas using an Instrument Manager software. This would allow us to program rules for reflexing the Sysmex UF-100 testing based on the results that were obtained from Atlas.”

The Advia Urinalysis WorkCell does take up a larger space. “Before, we had just a tiny instrument that took up about two square feet of space. The AUW, which takes up about eight feet of counter just for the automated portion, actually doubled the amount of space we were using for urines. Space is still needed to get the specimens together, put them in a rack, and then place the racks on the instrument.”

But she estimates there has been a nearly 50 percent improvement in efficiency in performing urinalyses. “In the past, I would need the equivalent of one full-time clinical laboratory scientist on the bench only doing urines for 10 straight hours. Now, that same technologist can multitask and be utilized in other areas of the department.”

Until now, says RML’s Dr. Johnson, “there have been a lot of efforts to standardize the way urinalysis is performed, but they have been very expensive. It’s difficult to make it both reproducible and cost-effective.” But automation is making standardization much more feasible.

Urinalysis automation could have even broader implications for the patient population, he believes. One of the most exciting applications, which is not happening yet but is theoretically possible, is the diagnosis of other diseases using the automated urinalysis instruments.

“We use flow cytometry in hematology to diagnose hematopoietic tumors, lymphomas, and leukemias, and we use a variety of antibodies, and depending on the combinations, form, size, and other characteristics, we can infer certain diagnoses. In theory, the same would be true using a flow-based urinalysis instrument, for example, to diagnose transitional cell carcinoma. That’s not been done yet, but that would be one way of screening a large number of patients, depending on how you set it up, to look for underlying cancers.”

In the short term, for laboratories making the jump from semiautomation to true automation, Heinrichsen has practical advice.

“It is a new way of approaching urines,” she says. “I think it’s really important if someone is interested in doing urine automation that they make sure to go to a site that’s actually using the instrument. You need to really see the instrument work and talk to the laboratory scientists, and maybe take time to review the data that’s coming off and see how it would work for your site—because there are always challenges when you’re evolving to a new technology.”

Related Links

• What's new
• Unexpected findings