A SARS-CoV-2 specimen labeling mix-up that resulted in an infected individual erroneously receiving a negative result highlights the very real consequences of errors that can occur during the pre-analytical phase of laboratory testing.

In February, the University of California San Diego (UCSD) Medical Center collected samples from four individuals who were in quarantine after returning from China and sent the specimens to the Centers for Disease Control and Prevention (CDC) for testing. Some confusion in differences between the two organizations’ labeling systems resulted in miscommunication about the test results, with CDC incorrectly advising that all four patients tested negative for the virus, according to news reports.

As the four individuals were being transported from the UC San Diego Health hospital back to Marine Corps Air Station Miramar to wait out a 14-day government-mandated quarantine, word came that one of the individuals had in fact tested positive. That individual was then returned to the hospital and placed in isolation.

The episode underscores a finding that is becoming more relevant as clinical laboratories have rushed to offer testing for the new coronavirus and to come to terms with stressed supply chains and unusual sample collection procedures: that improving communication and collaboration with those outside labs remains the best strategy to ensure accuracy and protect patients.

A Reality Check on Labeling Standards

Clinical laboratories have been working for decades to overcome the problem of specimen labeling errors. Researchers estimate that more than 160,000 adverse patient events occur each year in the United States because of patient specimen identification errors involving clinical laboratories, while 11% of all transfusion deaths occur as a result of phlebotomists not properly identifying patients or mislabeling tubes of blood, according to an article in The Journal of Applied Laboratory Medicine (2017;2:244-58).

The Clinical and Laboratory Standards Institute (CLSI) in 2011 developed a standard to reduce the unacceptably high incidence of mislabeled specimens in clinical laboratories. The standard, “AUTO12–Specimen Labels: Content and Location, Fonts, and Label Orientation,” specifies locations and formats for the required human-readable elements that must appear on the label for each clinical laboratory specimen. The standard also specifies rules for truncation for long patient names, the location and size of the bar code on each label, a list of the most commonly used variable elements that can appear on specimen labels, and the required orientation of labels on specimen tubes. That standard is still in effect.

The Joint Commission in 2014 acknowledged the issue of specimen identification errors and released two National Safety Goals to address this problem. The first goal called for healthcare providers to use two patient-specific identifiers, such as name and date of birth, to ensure each patient receives the correct medication or treatment. The second goal was to make sure the correct patient gets the correct blood when they get a transfusion.

But while standards are in place nationwide for specimen labeling and handling, the incidence of patient identification errors—including mislabeled and misidentified specimens—remains much too high, say experts. The best data on errors in U.S. laboratories comes from three separate College of American Pathologists (CAP) Q-Probe studies, in which the reported rates of mislabeled specimens were 0.39/1000 in 120 institutions (2006), 0.92/1000 in 147 clinical labs (2008), and 1.12% of blood bank specimens in 122 clinical labs (2010) (CLN 2014;4:12–13).

Over the past decade, laboratories have tried many interventions deemed to be effective in reducing specimen error rates. But how effective are they really? A 2016 study concluded that computer-generated identification systems and interdisciplinary cooperation can significantly reduce patient identification errors (PLoS One 2016;11:e0160821).

For the article in The Journal of Applied Laboratory Medicine, researchers conducted a review of published studies to determine which interventions resulted in substantial decreases in specimen labeling errors. Specifically, they evaluated the effectiveness of four categories of laboratory practices to reduce the incidence of specimen labeling errors involving blood or any other type of patient specimen. The four categories were: improved communication and collaboration between laboratory and healthcare professionals, education and training of healthcare staff responsible for specimen collection, audit and feedback of labeling errors and real-time event reporting, and implementation of new technology.

The authors concluded that improved communication and collaboration between laboratory and other healthcare professionals in the form of multidisciplinary teams was the top recommendation for decreasing specimen identification errors. While the other evaluated practices also led to a decrease in specimen labeling errors, the investigators couldn’t make a recommendation in favor or against the effectiveness of those practices because of insufficient data.

Paramjit Sandhu, MD, an epidemiologist at CDC in Atlanta and author of the study, told CLN that despite CLSI standards and improvements in bar-coding, the rate of errors probably is even higher than what is known. “In general, published error rates are usually underreported because of inadequate detection methods and reluctance to publish or otherwise share errors,” he said. “Accreditation organizations like The Joint Commission and CAP continually reinforce patient safety goals and standards. However, despite those efforts, labeling errors remain one of the leading types of preanalytical errors associated with ancillary services.”

Is it possible to get to a 0% error rate? According to Sandhu, a 0% error rate should be a goal, though he noted that specimen mislabeling can occur during multiple continually changing steps of the preanalytic phase of the total testing process. Errors can occur at the time of specimen collection, from an incorrect bar code read, or due to labeling mix-ups right before or after sample collection and during laboratory accessioning of a specimen.

Unique Challenges for Point-of-Care Testing

Point-of-care testing (POCT) is mostly automated and uses bar-coded patient identification and bar code scanners with a test device at or close to a patient, but that doesn’t mean these systems are perfect. Typically, device manufacturers build in tools that detect certain preanalytic errors, said Brenda Suh-Lailam, PhD, DABCC, FADLM, director of clinical chemistry and point-of-care testing at the Ann and Robert H. Lurie Children’s Hospital of Chicago.

“When preanalytic errors do happen, the system should render the test invalid,” she explained. “The healthcare provider would need to recollect the sample and repeat the test.”

Automation, bar-coding, and interfaces with electronic medical records (EMR) have helped reduce specimen collection labeling errors, but the potential for error in manual POC tests still exists, she added, especially if there is not an interface between the POCT device and the EMR. In addition, errors can occur in other parts of the preanalytic process, such as specimen collection and reagent storage.

While the goal may be to get to 0% labeling and identification errors, the human component means the possibility of an error probably will never be eliminated, said Suh-Lailam. “It’s not foolproof,” she said. “If you scan the wrong thing, you’ll get the wrong information.”

Ultimately, the extent to which a laboratory minimizes errors in labeling and specimen identification comes down to how closely the lab follows CLIA regulations and the CLSI standards, the amount of automation the lab employs, and the policies and procedures the lab has in place for its preanalytic processes.

“Because of the potential patient adverse consequences associated with mislabeled specimens such as transfusion-related death, medication errors, and misdiagnosis, every labeling error should be treated seriously,” said Sandhu.

Using Data to Drive Better Practices

If the standards have not changed, but the errors remain—what is the next step beyond improving communication to minimize errors in specimen labeling and handling? Some laboratories organize their efforts in this area according to the Six Sigma approach. David Rogers, senior operations director for support services at ARUP Laboratories in Salt Lake City, explained ARUP’s success with using Six Sigma to ensure that specimens are labeled, transported, and handled properly. ARUP also invests in keeping staff focused on detecting errors.

“We look for opportunities to improve each process,” he said. “First, we identify the steps that go into a process, such as selecting the correct patient and printing the associated label, and then we establish checks for each step of the labeling process. We have built in detection processes on the front end and secondary detections within the individual testing labs on the back end.”

When ARUP receives a specimen, the processing staff use unique identifiers, such as a client’s accession number or a medical record number (MRN), to query the system to ensure the correct label is printed. The system used in specimen processing defaults to query unique identifiers in order to eliminate risks associated with using patient names. Other processes involved require processing staff to electronically confirm orders to received specimens. Before a specimen is sent on for testing, at least four identifiers are double-checked during the labeling process. The most common are patient name, client accession or other container ID, MRN, collection date and time, and order. The labeled tube is then routed through an automated delivery system to the proper lab.

“We have a program that incentivizes our employees to find and report any labeling discrepancies,” said Rogers. “We find it motivates our staff to check labels very carefully and, more importantly, report issues to facilitate optimal tracking and trending of the individual processes moving forward.”

For tests that require complex collections and/or multiple specimens, such as adrenal function testing, additional secondary checks are required. For example, when two or more specimens are processed for these tests, a notification report goes to lab review staff and to support services. These individuals must certify and document electronically that they have reviewed the specimens and they are labeled correctly.

“In addition to implementing quality checks during the individual processing and labeling steps, we target additional processes and reviews where they are needed the most and where they will be most effective,” explained Rogers. “That’s what our data does for us. By engaging all staff in these efforts, and through effective tracking and trending, we have been able to pinpoint scenarios that inherently present higher risks, and we have identified staff whose specific task it is to secondarily review the associated specimens to ensure they are labeled correctly. Any lab can do this—it’s all about effectively using the available data, not the size of the lab.”

Kimberly Scott is a freelance writer who lives in Lewes, Delaware. +Email: [email protected]