Medical lab test company Theranos offered a rebuttal to the critical article about the firm published in the Wall Street Journal last week, but most analysts and experts were not impressed.
The “rebuttal” was notably short on details or proof, and Theranos continues to claim any kind of details about its new Edison small-sample blood testing technology, including the number of tests that are currently performed with Edison, are “trade secrets”, which just adds to the growing cloud of uncertainty and potential fraud surrounding Theranos and it ultra-ambitious founder Elizabeth Holmes.
Problems at Theranos as noted in recent WSJ article
The damning article about Theranos in the WSJ last week was largely based on interviews with four former employees of the company. One key point was that Theranos “does the vast majority of its tests with traditional machines bought from companies like Siemens AG.” While there’s nothing illegal or even unethical about that, it is certainly not a “disruptive business model” using a new technology to drive down lab test prices.
Even though the company will not directly confirm it, multiple sources confirm that as of December 2014, only somewhere between 15 and 20 of the more than 200 tests it offers are performed using the new technology for which it is famous: taking a few drops of blood by and then analyzing that small sample in a compact Edison device.
According to the WSJ sources, another 60 of Theranos’ tests were being performed by finger stick, but were then diluted into greater volume samples so that they could be read by third-party lab test machines. Diagnostic test experts say saying that dilution is not a preferred lab practice, and can produce less reliable results.
The remaining tests on the menu were performed by the traditional venipuncture method, merely using a smaller than ordinary needle which draws a less than usual sized sample, which is then analyzed by conventional third party blood sample diagnostic devices.
The WSJ sources also allege that in the first few months of 2014, company president Sunny Balwani instructed employees to engage in questionable behavior regarding reporting of proficiency tests that the firm must pass to keep its lab certifications. Based in part on internal emails, the WSJ claims that Balwani told Theranos employees to test the regulatory proficiency test sample on a conventional, third-party-manufactured machine, even though Theranos was routinely performing the same diagnostic tests with its Edison system. It seems the third-party machines produced more accurate results than Edison.
The reliability of Theranos’s tests is also called into question by some of the sources, but no direct proof of the claim was offered. In its effort to become a full service medical lab test provider, Theranos is seeking FDA approval for close to 130 of its proprietary tests, and has had one test approved by the FDA.
Keep in mind that while an FDA approval does in effect prove the basic validity of Holmes’ platform, we need Edison approved for at least a half dozen tests before we can really say Theranos’ technology is validated.
Theranos weak response to criticisms
When you get down to details, however, Theranos’ rebuttal falls woefully short, not refuting several specific allegations and rebuffing requests for details or records as not possible because they are “trade secrets.”
Evercore disagrees and opined in a note last week:
To level set, Theranos has officially replaced Epic Systems as the “private company we try to learn as much as we can about since they are as secretive as possible,” so we view this article as important to dive into relative to its impact on the clinical lab/diagnostic market. After what we would describe as an incredibly effective PR campaign, this report pulls back the veil on some of the challenges faced by this ultra-secretive company. The question now, however, is what does this actually mean? We have believed that Theranos, if it were a risk to the traditional testing market, represented more of a long-term tail risk rather than something that could impact LabCorp and Quest Diagnostics over the near-to-medium-term. We have also noted that while the company appears to have a meaningful amount of capital (both financial and intellectual), the level of secrecy given the pitch CEO Elizabeth Holmes has made to the market has created a number of questions we are curious about (Why so secretive if she is trying to change the world? Why isn’t this expanding faster in Walgreens if it is SO revolutionary?). All-in, though, we have been more skeptical than anything about this business and the opportunity to disrupt the market, particularly based on the fairly limited amount of knowledge we had on this company, and based on this article, our skepticism becomes even higher.
Raymond James disagrees, stating:
While the claims in the Wall Street Journal and other media sources are not completely verified, we find Theranos’ lack of willingness to disclose details or data as perhaps giving some credence to the reports. The article, which cites unnamed former Theranos employees, claims that at the end of 2014 only 15 of the 205 tests offered by the company were run on the proprietary Edison machines. The remaining 190 were run on traditional lab machines. Of these, roughly 60 were allegedly run using the smaller sample collection methods and a special dilution method to obtain adequate sample volumes, something with which the article’s “lab experts” take issue. The article also questioned accuracy, citing an Arizona woman who saw six test results all outside the normal range according to samples run by Theranos. Hospital results two days later placed each within normal ranges. Theranos and doctors cited in the article disagree here as well, with the company claiming normal variations and the doctor claiming some of the metrics should have steady values over such short time periods. Theranos has remained unwilling to share details in responses to most of the questions raised on the grounds of “trade secrets.” That said, Holmes has stated that over 1,000 pages of documentation were sent to the Wall Street Journal (which has stood by its reporting) prior to publication, and even offered to bring an Edison to the newspaper’s office to demonstrate the system. The company admits some tests are run on traditional lab equipment and refutes the claims, but without any numbers. Given the simplicity of stating “X% of our tests are typically run on Edison machines” for example, we find this lack of disclosure as somewhat validating some claims. To be sure, an answer is unlikely in any near-term scenario, but with this being the first negative public commentary, we do view the turmoil as perhaps lifting some pressure from public players LabCorp and Quest.