Francisco Bay to Palo Alto for testing. He also noticed that the phlebotomists were drawing blood from every employee twice, once with a lancet applied to the index finger and a second time the old-fashioned way with a hypodermic needle inserted in the arm. Why the need for venipunctures—the medical term for needle draws—if the Theranos finger-stick technology was fully developed and ready to be rolled out to consumers, he wondered.
Bradley’s suspicions were further aroused by the amount of time it took to get results back. His understanding had been that the tests were supposed to be quasi-instantaneous, but some Safeway employees were having to wait as long as two weeks to receive their results. And not every test was performed by Theranos itself. Even though the startup had never said anything about outsourcing some of the testing, Bradley discovered that it was farming out some tests to a big reference laboratory in Salt Lake City called ARUP.
What really set off Bradley’s alarm bells, though, was when some otherwise healthy employees started coming to him with concerns about abnormal test results. As a precaution, he sent them to get retested at a Quest or LabCorp location. Each time, the new set of tests came back normal, suggesting the Theranos results were off. Then one day, a senior Safeway executive got his PSA result back. The acronym stands for “prostate-specific antigen,” which is a protein produced by cells in the prostate gland. The higher the protein’s concentration in a man’s blood, the likelier he is to have prostate cancer. The senior Safeway executive’s result was very elevated, indicating he almost certainly had prostate cancer. But Bradley was skeptical. As he had done with the other employees, he sent his worried colleague to get retested at another lab and, lo and behold, that result came back normal too.
Bradley put together a detailed analysis of the discrepancies. Some of the differences between the Theranos values and the values from the other labs were disturbingly large. When the Theranos values did match those of the other labs, they tended to be for tests performed by ARUP.
Bradley shared his concerns with Renda and with Brad Wolfsen, the president of Safeway Health. Her faith already shaken by the delays of the past two years, Renda encouraged him to talk to Burd about them, which Bradley did. But Burd politely brushed him off, assuring the ex–army doctor that the Theranos technology had been vetted and was sound.
* * *
—
THE BLOOD SAMPLES DRAWN from Safeway employees in Pleasanton were being couriered to a one-story building with a stone façade on East Meadow Circle in Palo Alto. Theranos had temporarily set up its fledgling lab there in the spring of 2012 while it moved the rest of its growing operations from Hillview Avenue to a larger building nearby formerly occupied by Facebook.
A few months earlier, the lab had obtained a certificate attesting that it was in compliance with CLIA, the federal law that governed clinical laboratories, but such certificates weren’t difficult to obtain. Although the ultimate enforcer of CLIA was the Centers for Medicare and Medicaid Services, the federal agency delegated most routine lab inspections to states. In California, they were handled by the state department of health’s Laboratory Field Services division, which an audit had shown to be badly underfunded and struggling to fulfill its oversight responsibilities.
Had Steve Burd been allowed inside the East Meadow Circle lab, a network of rooms located in the center of the low-slung building, he would have noticed that it didn’t contain a single Theranos proprietary device. That’s because the miniLab was still under development and nowhere near ready for patient testing. What the lab did contain was more than a dozen commercial blood and body-fluid analyzers made by companies such as Chicago-based Abbott Laboratories, Germany’s Siemens, and Italy’s DiaSorin. The lab was run by an awkward pathologist named Arnold Gelb, who went by Arne (pronounced “Arnie”), and staffed with a handful of clinical laboratory scientists, or CLSs—lab technicians who are certified by the state to handle human samples. Although it made use of only commercial instruments at this juncture, there were still plenty of things that could and did go wrong.
The main problem was the lab’s dearth of experienced personnel. One of the CLSs, a fellow by the name of Kosal Lim, was so sloppy and poorly trained that one of his counterparts, Diana Dupuy, was convinced he was jeopardizing the accuracy of test results. Dupuy was from Houston and