Rice research questions blood drop test accuracy
Study suggests samples from pricking a finger can vary greatly
With advances in rapid, point-of-care testing, doctors can test for many diseases and irregularities by taking just a drop of blood. But new research from Rice University suggests that the contents of blood drops taken from the same patient can vary greatly.
“There is in fact a significant variation in some of the parameters of a complete blood count, when you go from one drop to the next drop to the next drop,” said Rebecca Richards-Kortum, a Rice professor of bioengineering and the lead researcher of the study.
Although the research was just a pilot study, the findings raise questions about the accuracy of a range of tests performed using blood obtained by pricking a finger rather than a sample taken from a vein.
Students in RichardsKortum’s lab have been developing low-cost tests for anemia, platelet and white blood cell counts. They noticed wide variation in some of the benchmark tests performed with hospital-grade blood analyzers.
They decided to test individual drops of blood to measure variation.
They took six consecutive, 20-microliter drops of blood through fingerpricks from 11 donors, and 10 consecutive, 10-microliter droplets from seven additional donors. They then drew a tube of blood from each donor’s vein using a needle.
“Some of the differences were surprising,” said Meaghan Bond, the researcher in Richards-Kortum’s lab who first noticed the discrepancies. “In some donors, the hemoglobin concentration changed by more than two grams per deciliter in the span of two successive drops of blood.”
The normal range for a hemoglobin test is between 13.5 to 17.5 grams per deciliter for men, and 12 to 15.5 grams for women, so a 2-gram swing could be significant.
Averaging results
Bond used the example of the quick test used to rule out anemia before donating blood.
“You might be accepted to donate if they were to measure the first drop and rejected from donating if they were to measure the third drop,” she said.
The researchers found that averaging the results of the droplet tests could provide results on par with venous draws, but it took six to nine drops of blood to achieve consistent results.
The study only looked at hemoglobin, white blood cells and platelets, so Bond couldn’t speculate on whether similar discrepancies could be found in other types of tests. Dr. Thomas Wheeler, chair of pathology at Baylor College of Medicine, said the significance of those differences depends on the type and purpose of the test.
“For certain things, it doesn’t make a difference, like a heel stick on a baby, where you’re screening for some genetic disease,” he said. “If you’re using it to determine whether somebody needs a blood transfusion … you want to make sure it’s spot-on accurate.”
It’s also unclear whether the findings have any implications for the millions of Americans with diabetes who test their blood sugar with fingerprick tests.
“It certainly raises the question that there could be variability between early drops versus later drops,” said Dr. James Nichols, a pathology professor at Vanderbilt University School of Medicine. While point-of-care testing relies on increasingly lower amounts of blood, other issues have arisen with tests using blood from fingerpricks.
Recently the Food and Drug Administration and the Centers for Medicare & Medicaid Services advised doctors not to use glucometers, devices that measure blood sugar using a drop of blood, with critically ill patients.
‘Just because it’s red’
Those patients may not have as wide a margin for error, and the glucose meters may not provide results accurate enough to guide medical decisions for such patients.
“I don’t think we have a good concept of what really happens during a fingerstick. We’ve always assumed that it’s just capillary blood that we’re getting, but we know that it’s mixed with tissue and juices and other cellular components,” Nichols said. “Just because it’s red, it doesn’t mean it’s the same blood sample.”