College of American Pathologists
Printable Version

  Eye the basics, not baubles, for point-of-care testing


CAP Today




January 2010
Feature Story

Anne Ford

The major gift-giving holidays may be over, but the hankering for new gadgets continues. Just ask point-of-care coordinators, some of whom continue to check their metaphorical stockings for new tests as they plead, “I’ve been a very good coordinator this year. How about, say, a POC molecular assay for respiratory infection?”

But focusing on the new—new assays, new platforms, new bells and whistles—can be detrimental to achieving excellence in your POC program, some experts say. “There’s a lot of tantalizing technology out there that looks really fun but that’s not practical or not indicated,” says Cynthia Foss Bowman, MD, medical director of clinical laboratories and director of POC testing at Long Island Jewish Medical Center, New York. “From my perspective, we should be making the bread-and-butter POC testing better. I don’t want to inhibit technical development, but I would hate to see a barrage of new tests come out without attending to the issues that we’ve got right now.”

For example? Not being aware of the limitations of each type of POC test performed in your institution. Reinventing the wheel instead of taking advantage of vendor resources. Letting clinicians dictate which tests are performed at the point of care. Not keeping abreast of controversies such as whether to use glucose meters or blood gas analyzers when implementing intensive insulin protocols.

No, not all POC coordinators make these mistakes—far from it. But issues like these are so important to the success of a POC testing program that it behooves POC coordinators to take a good, hard look at the basics and make sure they’re being done right before adding new tests just because they’re available.

Here, for your review, are some expert-recommended best practices for point-of-care testing:

  • Stick to one vendor or one type of device for each POC test. “Don’t have five different glucose meters in the hospital,” advises James Nichols, PhD, medical director of chemistry at Baystate Health, Springfield, Mass., and professor of pathology at Tufts University School of Medicine, Boston. “It’s too difficult to manage. All health systems have floating nursing staff, and if you’ve got different devices on different units, they’ve got to learn all of them, and you need to manage separate test systems, validate separate devices, have separate policies, implement a whole separate training program. It just magnifies the complexity.”

  • Standardize POC training and periodically re-evaluate staff skills. “Make sure you’ve got a written training program with solid goals and that everyone’s trained to the same standard,” Dr. Nichols says. “You don’t want to have a word-of-mouth type training program, because it’ll get diluted over time. Use a well-defined training checklist.” In addition, “you periodically need to go back and verify that the person who was trained is actually doing the test the way you think they are. Which is a problem, because the more people you have trained, the less opportunity they have to do the test.” That’s why it’s important to...

  • Minimize the number of staff performing the test “so you can maximize the frequency with which they’re doing the test,” Dr. Nichols says. “Think about signing onto a Web site—if you don’t sign on on a regular basis, do you remember all your passwords? That happens the same way with POC, except it’s more complicated. You have to remember how to troubleshoot, patient prep, how to report the test results, all of this. If you don’t do it day in, day out, you’re not going to do as good a job.”

  • Along the same lines, centralize your management “so you have one policy that is shared by all of the units doing the testing, rather than each unit developing its own policy,” says Dr. Nichols. “Definitely have the lab involved in it, because the lab has experience with preanalytic variables and with setting up a system of quality management, whereas a lot of clinical staff don’t have that background.”

  • When clinicians ask for a particular test to be made available at the point of care, require them to provide a use case and testing plan that answer questions such as: “What are you going to change in your treatment? What’s the current test’s turnaround time? How many users will be involved?” advises Brad S. Karon, MD, PhD, director of POC testing and assistant professor of laboratory medicine at the Mayo Clinic, Rochester, Minn. “Require the proponents to prove their case to somebody.”

  • Not only that, but “keep the validation and selection of the devices under the control of the lab,” says Dr. Karon. “In our program, you don’t get to tell us you want to use Device A versus Device B. You tell us that there’s a case to be made to use lactate for sepsis, and we’ll decide what lactate method is appropriate. We believe that we’re the experts in evaluating the effectiveness of methods.”

  • When negotiating the addition of POC tests, consider Dr. Bowman’s experience in the lab at LIJ. “We made an agreement with them [clinicians] that we would get our TATs so good in the lab that we would get to a point where we could assess whether they needed to do POC testing or not,” she says. “It hasn’t really stressed the lab; it’s helped the lab get more responsive and more efficient. Right now I’m turning blood gases around in two to five minutes in the lab. We have our cardiac markers down to less than an hour. I’m turning a lot of other things around in less than 30 minutes.” As a result, several units at her hospital have not requested additional POC capability.

  • Set up your test ordering system so that it guides clinicians toward ordering the most appropriate test, Dr. Nichols says. “As laboratorians, we really need to help the physicians in their ordering pathways. Physicians don’t have the training or experience to know what the differences in cost are between a POC test and a lab test. They just think ‘convenience’ or ‘faster,’ without thinking, ‘This is a more costly test; maybe I shouldn’t use it in this population.’” A well-set-up ordering system can also help clinicians understand the limitations of POC tests and evaluate whether they are willing to accept those limitations for the sake of a speedy result.

  • Train staffers who perform this type of testing not to rely blindly on the POC device’s results. “If it has a digital readout, they believe it,” Dr. Bowman says. “On one device—and I know this device has a variance of less than 10 percent—[staff members performing a POC test got] one value of 100 percent and another of 300 percent. They averaged the result and got 200. I said, ‘That’s just too wide a swing. I know that shouldn’t swing more than 10 percent.’ They said, ‘Well, it came up on the screen.’ There just seems to be this faith in a digital response.” It’s for similar reasons that Eberhard Fiebig, MD, director of the clinical laboratory at San Francisco General Hospital and professor of laboratory medicine at the University of California, San Francisco, advises nurses and physicians: “You cannot rely blindly on a glucose meter reading. You always have to take it into context. If a patient looks hypoglycemic, regardless of what the meter shows, you don’t ignore that patient’s symptoms or plot a treatment based on that one glucose strip reading. Always confirm with a sample to the lab, at the minimum.”

  • Understand the limitations of your POC devices. Of particular concern at the moment, Dr. Nichols says, is the question: “Is a glucose meter accurate enough to manage intensive insulin protocols, or do you need a more precise device like a blood gas analyzer?” That’s a question that’s best answered by examining your own data, Dr. Karon says. “Ask yourself: What’s your target now? How low are you trying to go? What’s the level of accuracy you’re getting?” He suggests bearing in mind, too, that “it’s very controversial, the role that device accuracy actually plays in bad outcomes with tight glycemic control. It’s a lot more complicated than just device accuracy. Somebody’s got to measure the glucose, and you can do that wrong. Someone’s got to interpret the result, and you can do that wrong. I would hate to see everyone everywhere abandon their glucose meters and delay treatment. You can get hypoglycemia from delayed results as well as inaccurate results.”

  • Another area of concern: Maltose interference. “If a patient is taking a medication that has maltose in it, and they do a glucose fingerstick on that patient, on some glucose meters the maltose can make the glucose levels appear artificially high, so they get a false-positive,” Dr. Nichols explains. “The patient may get higher levels of insulin, and that can send them into hypoglycemia. That’s actually happened. Because of that, you have to understand the limitations of the meter you’re using, and whether maltose is used in the medications in your institution.” You can also do as Dr. Fiebig is doing, that is, collaborate with the pharmacy to require that whenever hospitalized patients receive a medication that’s known to cause potential maltose interference, providers are warned not to use glucose meters on those patients.

  • Be aware of forthcoming POC-related resources, such as two CLSI documents that are in development. Dr. Nichols explains: “Two guidelines are in the works that involve glucose meters and criteria for comparison of glucose meter results with central laboratory methods. While both documents currently adopt the ISO criteria of ±20 percent, there is pressure from U.S. government regulatory agencies to increase the stringency of this criteria.”
  • One reason to be aware of these forthcoming CLSI guidelines is to provide feedback to CLSI by commenting on proposed drafts of the guidelines before their final approval. In Dr. Bowman’s opinion, the new CLSI guidelines may put POC testing programs in a difficult position if the glucose meters they are using cannot meet the proposed error limits. “Programs will either accept wider error ranges or they will run critical care glucose testing in the central laboratory or on blood gas analyzers,” she says, adding: “That could be very disruptive and a logistical challenge, to say the least. We really need to assure the accuracy of the glucometers we use.”

    Point-of-care coordinators should be aware, too, that the CAP’s POC Testing Committee is in the process of creating a different POC-related resource, she adds. “We’re deeply into writing what we’re calling a tool kit for pathologists and lab directors,” she says. “We’re hoping to summarize all the elements of POC testing—how to approach it, the different elements you have to consider—and we’re hoping it’ll be a basic medical director’s primer or resource on POC testing. We’re hoping to have a rollout by this time next year.”

    Steven J. Melnick, MD, PhD, chief of the Department of Pathology and Clinical Laboratories at Miami Children’s Hospital, sums up what may be the best reason for adhering to these and other POC-related best practices. “If we had to go back and do in the lab the amount of stat testing that has been taken over by POC testing, we’d be in deep trouble,” he says. “POC allows us to be far more effective in the remaining stat testing that needs to be performed that can’t be done on a POC platform. The more POC testing we do, the more we decompress the lab.”

    Anne Ford is a writer in Chicago.