Skip to content

University Gazette

The University of North Carolina at Chapel Hill

Researchers uncover reason for false-positive marijuana tests in babies

What would you think if, immediately after birth, a baby tested positive for marijuana? Well, you might think the baby’s mother used drugs. But what if the mother didn’t? What if the test was wrong?

False positives happen, and they’ve stumped lab chemists and nursery departments across the country. But now researchers at Carolina have figured out what’s going on.

Nurses collect urine from a newborn for a drug test if doctors determine that the mother hasn’t had adequate prenatal care. If the test for THC, the psychoactive chemical in marijuana, comes back positive, then doctors have a conversation with the mother. A single positive drug-test result doesn’t mean the nursery will call child protective services, but it is one of several determining factors, so getting that test right is important.

In July 2011, nurses at UNC Hospitals’ Newborn Nursery found that some babies’ urine samples came back positive for THC even though tests of their meconium – a newborn’s first bowel movement – came back negative.

Catherine Hammett-Stabler, executive director of UNC’s Core Laboratory at the William W. McLendon Clinical Laboratories, and Carl Seashore, director of the Newborn Nursery, asked the nurses to continue collecting urine samples from newborns. If any came back positive, the nurses were immediately to request that the doctor order a confirmation test using the more sophisticated mass spectrometry to confirm or correct the initial finding.

A week later, nurses had compiled a number of positive results using the standard test, but mass spectrometry showed that they were all negative. This was odd, Hammett-Stabler saod, because the hospital never had a problem with the standard test when screening other patients for THC.

“The mass spectrometry test always matches,” she said. “So this bothered me. It made me think something was happening in the nursery that was different.”

UNC social worker Lizzy Burch organized a meeting with nurses, clinical staff and doctors. The team walked through the results, and then nurses described every step that takes place in the nursery between childbirth and sample collection. It turned out that nurses collected urine samples in different ways. Some used cotton balls or gauze to dab samples from diapers. Others turned diapers inside out, which caused urine to interact with colored dye from the diaper. Maybe that dye was made from hemp, Hammett-Stabler thought. Other nurses used collection devices to get urine directly from babies. Some nurses cleaned babies with towelettes; others gave babies baths with soap.

Resident Daniel Duncan and chemist Steve Cotten collected all these materials and ran experiments to determine whether any would trigger a false-positive result when mixed with urine. All those tests came back negative except for one item – baby soap.

To confirm the finding, Hammett-Stabler bought as many kinds of baby wash as she could find at local stores and ran them through the same tests. Sure enough, nearly all the baby soaps on the market produced some false positives. She then sent her samples to Duke and Rex Hospitals, which use slightly different immunoassays to test urine for THC. Both hospitals got false-positive responses as well.

“We realized this problem wasn’t unique to the assay we use,” said Hammett-Stabler, who is quick to point out that these soaps don’t contain THC. Chemical components in the soap – cocamidoprophyl betaine, polyquaternium 11, PEG 80 and sorbitan laurate – cause the false-positive results. Hammett-Stabler’s team tested these chemicals individually and found that they seem to cause false positives independent of each other.

“Exactly how the false positive happens – what the mechanism might be – we don’t know,” she said. Structurally, the chemicals in question don’t look like marijuana metabolites, but for some reason the standard test sometimes thinks they are.

As a result of the collaboration between the nursery and laboratory staffs, testing protocol has changed at UNC and other hospitals around the country. Before, the laboratory had to receive permission from a doctor to order the confirmatory mass spectrometry test. In the time it would take for the request to be processed, many urine samples would have been discarded. Now, when there’s a positive THC result, the laboratory is allowed to automatically order the mass spectrometry test.

Since Hammett-Stabler and Seashore’s research paper was published in the “Journal of Clinical Biochemistry” in June, Hammett-Stabler has been contacted by colleagues across the country who had been chasing the false-positive THC problem but hadn’t been able to figure out what was going on.

“We could not have done this without the nurses and the hospital staff,” Hammett-Stabler said. “It’s been a tremendous team effort.”