Long-awaited Theranos data shows irregularities that could affect care

Company, which still has not published data, criticized new independent study.

Beth Mole – Mar 29, 2016 8:02 pm | 47

Founder and CEO of Theranos, Elizabeth Holmes at TEDMED 2014. Credit: https://www.youtube.com/watch?v=dBvzKp0AERE

Finger-prick blood test results from Theranos, Inc., the beleaguered medical startup once valued at $9 billion, had more irregularities than those from the top two clinical laboratory testing services, LabCorp and Quest Diagnostics, according to a new study.

Though Theranos’ results mostly fell in line with those from the other labs, researchers say that some of the discrepancies could lead to additional testing and medical costs or incorrectly sway medical decisions.

The results, published in The Journal of Clinical Investigation, are the first published, peer-reviewed data on the company’s finger-prick blood tests.

Theranos made waves and drew big donors last year with claims that its new technology could run a multitude of medical tests with just a few drops of blood—dramatically cutting costs and eliminating the need for needles and arm punctures. But buzz around the company quieted after a Wall Street Journal story dug up questions about the tests’ accuracy. Since then, public records show that the Food and Drug Administration has repeatedly flagged problems at the company, including one that posed “immediate jeopardy to patient health and safety.”

Theranos’ CEO, Elizabeth Holmes, said last October that the company would release data that would validate the technology, but it has yet to do so.

In the new study, researchers at Icahn School of Medicine at Mount Sinai had 60 healthy adults have their blood drawn and tested by the three services all on the same day. The blood draws were done last July, before concerns about Theranos arose.

Each blood sample was used in 22 tests, looking at things such as cholesterol and blood cell count. Though much of the three labs’ results fell in line, Theranos stood out with a few problems.

Theranos had slightly more measurements out of a normal range—12.2 percent—compared with Quest and LabCorp—7.5 percent and 8.3 percent, respectively.

The authors also found the Theranos’ measurements on total cholesterol tended to be nearly 10 percent lower on average. This bias in the data may alter clinical decisions, the authors note.

In a letter (PDF) signed by Theranos laboratory directors, the company called the study “flawed and inaccurate,” citing a list of complaints about the study’s design and the authors’ conduct. In particular, Theranos noted that the study did not try to determine which of the three company’s tests results were actually accurate.

The company also noted that the study did not follow proper protocol for collecting the finger-pricked blood samples, invalidating the results.

Theranos also claimed that the authors lied to the media in their claim that they tried to contact Theranos prior to publishing the data. The company said it has no record of any contact.

Still, some experts were not swayed by the company’s arguments. “This represents just another effort for a failing company to cover up bad press for a technology that is not ready for prime time,” Frederick L. Kiechle, medical director for clinical pathology at Memorial Healthcare System in Hollywood, Florida, told The New York Times.

Journal of Clinical Investigation, 2015. DOI: 10.1172/JCI86318 (About DOIs).