January 2011
Feature Story

Karen Lusky

A patient with type 1 diabetes has a 450 mg/dL fingerstick glucose value when tested at a clinic but a 100 mg result when the central lab repeats the test on another specimen. An emergency department physician complains that the ED point-of-care pregnancy test results are suddenly all positive.

The Clinical and Laboratory Standards Institute’s new consensus guideline, “Quality Management Approaches to Reducing Errors at the Point of Care,” is aimed at helping labs and health care providers parse and preempt those types of problems and many more.

The blood glucose puzzle provided in the guideline contains a key clue and a couple of red herrings. The patient was taking an antihypertensive agent and undergoing peritoneal dialysis for chronic renal failure. He complained of feeling weak and short of breath for a few days before the glucose tests were done.

A hospital clinical chemist helped the affiliated clinic troubleshoot the erroneously high glucose result. And it wasn’t, as you might think, that the clinic staff failed to perform quality control on the glucose meter.

“The case example is built on the assumption that the glucose meter has a lock-out function” whereby the operator can’t proceed if QC hasn’t been performed or if QC values are unacceptable, says pathologist Valerie Ng, MD, PhD, a member of the subcommittee that wrote the guideline, which she says is geared toward hospitals but can be applied to POC testing in any setting. Lock-out is a safety feature that only some glucose meters have, notes Dr. Ng, of Alameda County Medical Center/Highland General Hospital, Oakland, Calif.

In addition, the glucose meter appeared to be in working order, the test strips were OK, and the tester was competent in performing the test.

Stymied, the clinical chemist turned to a professional listserv on which someone noted that medications can interfere with certain glucose meters.

The culprit, it turns out, was the patient’s peritoneal dialysis fluid, which is known to interfere with the meter.

“Would every place have that particular kind of case? Probably not,” says Halcyon St. Hill, EdD, MS, MT(ASCP), of Florida Gulf Coast University in Ft. Myers and a member of the subcommittee that wrote the CLSI guideline. Even so, the guideline is broadly applicable because “it provides critical steps and techniques for alleviating errors.”

As for the positive POC pregnancy tests in the ED, a sample investigative report in the guideline details a hospital laboratory’s investigation of the problem. The cause: a bad lot of reagents.

In the new CLSI guideline is a table listing the problems that can occur at every step of the testing process, from pre- to post-examination. One section focuses on patient-related variables, which proved to be the problem in the glucose testing example.

Guideline co-author Abdel-Baset Halim, PharmD, PhD, DABCC, says people tend to overlook patient-related variables that affect laboratory testing, whether it’s done at the point of care or not. The CLSI document identifies many such factors, among them nutritional status, emotional distress, physical activity, smoking habits, medications and supplements, and prior clinical or diagnostic interventions.

Emotional distress right before a test can elevate a person’s glucose, cortisol, and epinephrine, says Dr. Halim, director of clinical biomarkers, Daiichi Sankyo Pharma Development, Edison, NJ. Someone who exercised strenuously just before testing may have elevated blood creatinine kinase, lactate, and falsely elevated blood and/or protein in the urine. “Even a simple intramuscular injection can increase blood CK,” he adds.

To help ensure accurate results, “ideally, the person ordering the tests should tell patients that if they are under a lot of stress, they should relax for a certain period of time before having their blood drawn,” Dr. Halim says. The CLSI guideline recommends reporting various patient-related variables on the test report, including medications that are known interferents.

Dr. Halim calls the impact of medications and supplements on labora­tory testing overall “extremely under­estimated.” Medications containing human chorionic gonadotropin for infertility can cause a false-positive pregnancy result for POC or central lab tests. Aspirin intake exceeding 325 mg per day can lead to false-positive results for occult blood. Vitamin C in excess of 250 mg/day can cause false-negative test results for guaiac occult blood methods.

In addition, Dr. Halim says, high doses of vitamin C can lead to false results for urine glucose when using tests with Fehling and Benedict reagents. “But luckily, most POC devices these days rely on specific glucose oxidase methods.”

Over-the-counter drugs and even food additives can also skew lab test results, he warns. For example, “pseudoephedrine and ephedrine, common decongestants in cold medications, may give false-positive results for amphetamines, including the recreational drug Ecstasy [MDMA].” Ibuprofen can lead to false-positives on tests for marijuana. Well known is that poppy seeds can result in false-positives for opiates.

Specimen collection errors can also affect POC test results. And the CLSI guideline sets the record straight for those who wonder what could go wrong with a simple fingerstick blood or even a breath sample.

For example, a breath sample for alcohol testing contaminated with residual alcohol in the mouth after recent consumption of an alcoholic beverage can significantly distort the breath test result because the actual quantity measured in breath is extremely small, Dr. Halim says. A 15-minute waiting period after consumption can mitigate this contamination. A small incision can lead to expedited blood clotting before the specimen reaches the POC device, due to slow blood flow or the increased release of tissue factor when the finger is squeezed. Both factors, or even one of them, can contribute to false reduction in INR readings, he says.

A partially clotted sample may produce a falsely low platelet count. To prevent that problem, “use a lancet with the proper size at the proper application pressure,” the guideline advises. And, it says, “collect drops as fast as possible.”

Post-examination variables that can derail patient care involve failing to identify and act on abnormal test results, including critical values. The guideline says, “Some devices have symbols indicating a critical value and a mandatory comment code for actions taken before proceeding to the next patient.”

Based on the literature, using a single laboratory result as a basis for diagnosis is not a good idea, Dr. Halim stresses. “If blood glucose is 450 and urine glucose is also positive, then this is a good indication the patient has a problem. But if blood glucose is 450 and urine is negative for glucose, then that’s suspicious. If blood glucose is 450 and A1c is 7.5 or 8, then, yes, that makes sense.”

The CLSI guideline also reviews the pros and cons of POC testing. The testing does provide a shorter turnaround time and sidesteps transport, labeling, and centrifugation issues that can affect central lab testing accuracy, Dr. Halim points out. And it spares ICU patients from iatrogenic blood loss associated with repeated venous sampling for central laboratory testing.

Also, most POC devices are CLIA-waived, he notes, which makes simple-to-operate devices available for use in emergencies and even by patients without medical supervision. “On the other hand, it can also be a disadvantage for the end-user” because the testing is not subject to the same quality control and proficiency testing required for nonwaived testing.

While the CLSI guideline is designed for worldwide use, Dr. Ng says, CLIA-waived tests in the U.S. have a single regulatory requirement to follow the manufacturer’s test instructions, which may include QC and PT. However, “most [tests] do not require PT, and only a subset requires external liquid QC.” (For an update from CMS’ Judy Yost on oversight of waived laboratories, see “More education for waived labs,” above.)

If a hospital laboratory is affiliated with a POC setting, Dr. Ng says, the lab will usually require it to conduct external liquid QC and PT. Whether that’s true in freestanding physician laboratories depends on their grasp of laboratory quality issues. “They may be blissfully ignorant,” she says.

The CLSI guideline does not specifically recommend using external liquid quality control. Dr. Halim, who advocates using that form of QC, acknowledges that it does add to the already higher price tag for POC testing. “It would cost about $10 to get two levels of liquid QC for INR [point-of-care] testing to run just a few samples a day.” By contrast, in the central lab, the liquid QC for INR might cost only “pennies per day.”

Dr. Ng notes, however, that the extra cost pales in comparison to the cost of inaccurate POC test results that lead to erroneous treatment decisions.

Some POC users may also be falsely reassured that a device’s built-in QC covers all aspects of the testing system when, in reality, many times it simply tests a device’s electronic system, Dr. Ng points out.

And, Dr. Halim warns, a device with a low battery or slight electronic failure may continue to provide inaccurate results without producing a device error message. “In such cases, it’s difficult to detect the first incident, and errors will accumulate until someone raises concerns about an [unexpected] result.”

Dr. Ng requires external liquid QC for a nonwaived blood gas instrument at her institution because the device’s built-in QC cannot determine whether the user knows how to correctly place the tiny blood specimen in the sample port. And clots and bubbles in the sample can alter the test results.

In Dr. Ng’s experience, POC devices perform fairly reliably well, with most errors caused by the personnel who perform the testing. And that gets back to the basics: reading the instructions, training, and demonstrating competency, she says. At her institution, she has overheard nurses ask each other how many drops of urine they use for the POC pregnancy testing. The correct answer should be three. Yet “one will say, ‘I use six,’ and another will say, ‘I fill up the well.’”

Even urine dipsticks can provide false results if users don’t know they have to read the various patches on the strip based on designated time frames, says Lou Ann Wyer, MS, MT(ASCP), of Sentara Healthcare in Norfolk, Va., and CLSI subcommittee chairholder.

The new POC testing guideline makes it perfectly clear that what some may view as relatively simple in the POC testing arena often isn’t simple at all. If so, then Wyer feels the mission has been accomplished.

“We want people to get the best out of the testing,” she says.

Related Links

• More education for waived labs