Skye Gould/Tech Insider
Faster, cheaper, better.
An innovation that accomplishes those three things has the potential to disrupt an industry, but such innovations are rare.
Theranos, a company founded by Stanford sophomore Elizabeth Holmes in the fall of 2003 (she dropped out a few months later) has generated a lot of buzz for developing a revolutionary approach to the blood test. Theranos' innovation could theoretically upend the branch of medicine that provides the data used in roughly 70% of medical decisions.
But the more buzz Theranos gets, the more questions people have asked.
On Oct. 15, The Wall Street Journal (WSJ) published an investigative report that claims the company does not use its own "revolutionary" blood testing technology, called Edison, for the majority of lab testing, largely because of concerns about accuracy. Instead, the Journal's sources claim Theranos relies primarily on standard blood-testing technologies.
WSJ writer John Carreyrou spoke to four former employees, none of whom are named in his story, including one who "accused the company of failing to report test results that raised questions about the precision" of their tests, which could be "a violation of federal rules for laboratories," the Journal article said. Theranos' lawyer acknowledged to the WSJ that proprietary technology was not used in all of the company's tests, but in a later statement called doubts about Edison's accuracy "baseless" and "erroneous," claiming that the quoted sources "were never in a position to understand Theranos' technology."
This is not the first time that experts and insiders have raised questions about the secretive technology behind what has become the hottest and most highly valued health startup in the world.
When Holmes landed on the Forbes billionaire list in 2014, Theranos had raised $400 million, and the company had a valuation of $9 billion. Now Theranos is valued at $10 billion, and Holmes owns about half of the company.
Holmes assembled a board of directors for Theranos that Fortune magazine described in 2014 as "what may be, in terms of public service, the most illustrious board in U.S. corporate history." Among those board members are a retired Navy admiral, a retired Marine Corps general, former US Sens. Sam Nunn and Bill Frist, former CDC director William Foege, former Wells Fargo CEO Dick Kovacevich, and two former secretaries of state, George Shultz and Henry Kissinger.
The caliber of the board suggests that Theranos must have developed a transformative innovation, but other than Frist, who has not practiced medicine in many years, only Foege is a medical professional. The technical details about Theranos' seemingly revolutionary tests are hard to come by, and the company is known for its secrecy about its founder's invention.
Skye Gould/Tech InsiderThere's one fundamental question, which is partly unanswerable without Theranos revealing information that it wants to keep confidential: How, exactly, does what Holmes invented work?
"It's impossible to comment on how good this is going to be — it may be wonderful [or] it may bomb, but I really can't be more definitive because there's nothing to really look at, to read, to react to," Dr. David Koch, president of the American Association for Clinical Chemistry and a professor at Emory University, told Tech Insider in April — before Thursday's WSJ report surfaced and before the company received Food and Drug Administration (FDA) approval for one of its tests in July.
One thing we do know is that people who have seen the company's data are convinced there's something promising about Theranos.
Holmes has teamed up with Walgreens to put Theranos labs inside its pharmacies (Theranos is in 41 of more than 8,000 Walgreens so far), and it has also set up at least four direct-to-consumer "wellness centers." Theranos has also conducted tests for GlaxoSmithKline and Pfizer and several hospital systems; it recently established one partnership with Cleveland Clinic and another with Capital BlueCross in Pennsylvania. The company also earns revenue from the work it does for the US military, though Holmes has previously declined to elaborate on that, beyond saying that it's an important area with the potential to save lives. Each lab is certified, and it can operate officially in a clinical capacity.
But Theranos hasn't published peer-reviewed studies comparing its tests to traditional ones, and the company hasn't allowed independent experts to publicly assess its labs, citing the need to protect its intellectual property. It did submit data from its herpes blood test to the FDA (which approved it), but a former Theranos employee cited by the Journal said that, at the end of 2014, the company only used its technology for 15 blood tests out of at least 225. Theranos now offers more than 240 tests but told the Journal that they would not clarify how many use Edison (though they said it was more than 15).
Holmes tweeted that the WSJ story was "full of falsehoods" and claimed that Theranos offered to show the media outlet the Edison technology, but The Journal declined.
Even with Theranos’ proprietary technology in hand, however, the lack of peer-reviewed studies — the kind of data scientists traditionally use to assess a development in their field — has generated skepticism from experts outside the company.
"They completely bypassed the traditional process of peer review or publishing in peer-reviewed journals or having peer labs evaluate their product," Dr. Jerry Yeo, a professor and director of Clinical Chemistry Laboratories at the University of Chicago, told Tech Insider in April.
They completely bypassed the traditional process of peer review.
Yeo says it's normal for companies when they launch a new medical product to publish their results and allow experts to analyze their tests, especially as they become major players in the field.
"Why haven't they shown us that information, why haven't they been willing to publish it, and why haven't they shown comparisons with existing technology?" Yeo said.
Theranos does show detailed charts on its website, which portray a close correlation between its tests and more traditional blood-testing technologies, but the data used to generate those charts is not publicly available. The specific tests flagged as problematic by the WSJ report — for vitamin D, prostate cancer, and thyroid hormones — are not among the charts the company has made public.
The company says it wants to protect the details of its unique product from its competitors, just as any company would protect an innovation. It also says it wants the FDA to evaluate its product to show that it works. The recent FDA approval for its herpes test was the first public sign the company is making headway on this goal, though the company and the FDA declined to comment to the WSJ on the status of the more than 130 other tests it has submitted to the agency.
A representative for Theranos provided this statement to Tech Insider in April:
The FDA, which is the ultimate arbiter of safety and efficacy — of high quality tests that are proven to work — is the gold standard, and Theranos wants its tests to be the best and safest for its patients. We have called for an unprecedented level of review with the FDA, something we are not obligated to do. We've been submitting all of our tests to them and are committed to continuing to do so.
Tech Insider wanted to see what we could figure out about the science behind Theranos, even though it's currently impossible to say what the company's technology is or how well it works without published data. However, we can study what Holmes has said about Theranos and look at the latest developments in related scientific fields. We spoke to experts in the fields of clinical pathology and laboratory medicine, biomedical engineering, and healthcare investing, all to see if Theranos really could have an approach to blood testing that's faster, cheaper, and better.
Our investigation raised as many new questions as answers.
Clues about a secretive technology
Many of the headlines about Theranos talk about how Elizabeth Holmes reinvented the blood test:
- Wired: "This Woman Invented a Way to Run 30 Lab Tests on Only One Drop of Blood"
- The New Yorker: "Blood, Simpler: One woman's drive to upend medical testing"
- Business Insider: "This Woman's Revolutionary Idea Made Her A Billionaire — And Could Change Medicine "
But the details of Theranos' technology aren't what Holmes usually focuses on when she talks about the company.
The details of Theranos' technology aren't what Holmes usually focuses on.
When Holmes talks about transforming medicine, she talks about the idea that her system could eventually make it possible for people to more easily access data about their health by going and getting a cheap blood test on their own. The promise is that, just as easily as diabetics can check their glucose levels, anyone could skip going to a doctor to get their blood cell counts, cholesterol levels, potassium levels, and other readings.
We don't know that having access to all that information is helpful yet, but many think that if we get better at interpreting the data, it could have huge effects on our health and help us catch disease sooner.
Still, whether or not easier access to blood tests could transform our health, such conversations don't tell us anything about how Theranos' technology works. Most of the time, Holmes says little about it. The New Yorker's Ken Auletta described her explanation of what Theranos does as "comically vague."
She told him:
"A chemistry is performed so that a chemical reaction occurs and generates a signal from the chemical interaction with the sample, which is translated into a result, which is then reviewed by certified laboratory personnel." She added that, thanks to "miniaturization and automation, we are able to handle these tiny samples."
More useful is something Holmes said in an interview with Fortune in June 2014.
As Roger Parloff wrote then:
Precisely how Theranos accomplishes all these amazing feats is a trade secret. Holmes will only say — and this is more than she has ever said before — that her company uses "the same fundamental chemical methods" as existing labs do. Its advances relate to "optimizing the chemistry" and "leveraging software" to permit those conventional methods to work with tiny sample volumes.
In other interviews she makes it sound more like she's improving and streamlining processes while incorporating new but maybe not unheard-of technology.
The WSJ report questions how much Theranos relies on their new technology and how many of their tests are run on traditional machines.
Regardless, what Holmes has said is, essentially, that she's come up with a way to optimize blood tests using chemistry, automation, software, and the latest in biomedical engineering.
An early academic interest might be key
Screenshot/Theranos.comAs an ambitious undergraduate (she worked in a Ph.D. lab her freshman year), Holmes expressed a lot of interest in what Auletta describes as "lab-on-a-chip technology, which allows multiple measurements to be taken from tiny amounts of liquid on a single microchip."
This type of technology, microfluidics, has the potential to transform biomedical research by allowing people to glean meaningful data from minuscule amounts of liquid — in this case, blood.
The basic definition of microfluidics is "fluid flow in a channel that has a dimension of less than one millimeter," according to Ben Moga, president and cofounder of a company called Tasso Inc. His company is working on a device that collects blood samples from people at home who attach the collection vessel to their skin with an adhesive so it can draw a clean sample. The ease-of-use factor, lack of big needles, and tiny amounts of blood involved mean the company definitely has something in common with Theranos — Tasso uses microfluidics to work with minimal quantities of blood.
It wouldn't be possible to revolutionize the blood test and get away from big needles if you couldn't use these tiny quantities of liquid.
Holmes' first patent, filed before she started Theranos in 2003 and approved in November 2007, was for a technique to do multiple tests on a drop of blood. The term "microfluidics" appears in nine of the 31 patents that appear when searching the US Patent Office for "Theranos," and she's the co-inventor on more than 270 global patent applications.
Holmes may have the most ambitious ideas about how to use microfluidics technology, but she's not the only one working in that field.
Others have developed quick blood tests using microfluidics before, including an IBM lab in Switzerland. Experts said that not only were these tests becoming more used in general, but that this sounds like exactly what Holmes and Theranos are doing, as it's what she studied and what would make her blood tests possible.
In other words, the core technology behind the Theranos blood test might not be mysterious or even particularly new. But there are still some crucial missing pieces, especially since Theranos' version sounds more powerful than anything else out there.
"The technology is emerging," said the University of Chicago's Yeo, and "some version of it is already out there." But his question is whether Theranos could have really developed a way to run as many tests as it offers (more than 240 so far) on the large scale needed at hospitals and in major labs without relying on already existing machines to automate processing. The WSJ story suggests that they do, in fact, lean on existing technology for most of their tests.
Experts we spoke with said it seemed that running more than 20 or 30 tests from a drop of blood would be an impressive feat. Theranos is the first company to start using blood tests that primarily rely on microfluidics, and it says it gets unparalleled results.
"We can perform hundreds of tests, from standard to sophisticated, from a pinprick and tiny sample of blood, and we have performed more than 70 tests from a single tiny sample," a representative for the company told Tech Insider in April.
Moga's company, Tasso, spun out of leading microfluidics expert Dave Beebe's lab at the University of Wisconsin. More than a decade ago Beebe wrote that "microfluidics has the potential to significantly change the way modern biology is performed."
Last year, in a review published in the journal Nature, Beebe and coauthors wrote that hematology, the study of blood, was one of the leading areas of use for microfluidic technology, though "a 'killer application' that propels microfluidics into the mainstream has yet to emerge." Still, he wrote in that review that using this technology it was possible to use "microliters of blood from a finger prick" instead of "milliliters of blood from a [vein]" and process those samples in a few minutes instead of an hour.
This is cutting-edge technology, but it's not exclusive to Theranos.
This is cutting-edge technology, but it's not exclusive to Theranos.
Others are working in the same area, even if no one else has launched a similar commercial effort. If Theranos is the company that has finally come up with Beebe's "killer application" for microfluidics, that may explain its reluctance to show the patented details that make its technology unique, even though that rubs scientists the wrong way.
If, as Yeo said, a version of the technology is "already out there," Theranos' innovation may be a lab that implements a new form of that technology in an elegant and efficient way, putting it into action on a commercial scale for the first time in history.
Successfully pulling this off would be huge, and experts say it's plausible. But in a February article in Journal of the American Medical Association, John Ioannidis, a widely respected expert in scientific accountability, questioned how scientists could fairly evaluate a closely guarded biomedical innovation.
"How can the validity of the claims made be assessed, if the evidence is not within reach of other scientists to evaluate and scrutinize?" Ioannidis wrote.
The issue, as Yeo said, is that medical technology has the potential to save or harm someone's life — and for that reason, doctors recommending a Theranos test want to be absolutely certain that it works.
Does it work?
For Theranos to take lab-testing business from the two major lab companies used by doctors, Quest and LabCorp, and to partner to run tests for more hospitals, pharmaceutical companies, and government agencies, it has to show it can get it right at every stage, from the preparation before it pricks someone's finger to the accurate processing of hundreds of thousands of samples.
The WSJ account cites former employees who allege that Theranos' tests aren't as accurate as they say they are, and at least one of those employees alleged that the company manipulated results in tests meant show the proficiency of their new Edison machines.
These accounts were roundly denied by Theranos and Holmes, with the company issuing a statement saying that facts it presented "to prove the accuracy and reliability of its tests" were ignored by the WSJ and that the anonymous sources in its story were "disgruntled."
Theranos' labs have been evaluated and deemed acceptable by drug companies, hospitals that work with them, Walgreens, and other entities. There's even a small study coauthored by Holmes and published in an online journal that compares Theranos test results with traditional tests favorably, though it looks only at results for one characteristic of blood among six people. The charts on Theranos's website indicate favorable comparisons between its Edison tests and at least 24 traditional tests, though again, the source of that data is not public. (Tech Insider has requested the 1,000-or-so pages of data reportedly offered to the WSJ, but we did not hear from Theranos by the time this article was published.)
Yet none of these evaluations are a substitute for the extensive and thorough publicly available peer-reviewed studies that researchers want to see when faced with something bold and new.
Theranos is obviously not doing things in the usual way that pure research works; it's also selling a service, which it says it wants to protect. This might explain why it says that FDA approval is a more important way of showing its tests work. And a nontraditional way of doing things may be guaranteed to draw skepticism from researchers who want to see the traditional scientific processes followed, even if that new method works.
A nontraditional way of doing things may be guaranteed to draw skepticism.
Still, since labs are the biggest source of objective data that most doctors have to base decisions on, they need to trust new tests before deciding they're going to start using them. It could be a matter of life or death, which is why people in the medical profession want to see studies that show how accurate Theranos tests are compared to traditional ones.
"Scientists by nature are skeptical people, but we are driven by data," says Yeo. "If data shows your test is fast and accurate and cheap, more power to you." He adds: "Maybe they are really good, and if they are, I wish they would subject themselves to some peer review."
Even if the tests work, which experts we spoke to think is perfectly plausible, Theranos still has to execute them well enough to take over existing business.
"Every lab test, no matter how innovative, has limitations," says Yeo. Even the finger prick presents challenges.
The finger prick doesn't open a vein and let blood flow out; instead, it draws capillary blood from tiny vessels just below the skin.
Yeo explained that you need to prepare someone's finger for this process to make sure blood is flowing, but if you end up squeezing the finger too hard or have to prick it too many times, that may damage and break down cells — and this cellular damage could skew various test results.
Moga, of Tasso Inc., says there are ways to expect and compensate for this damage. Still, he says he understands that the medical community wants to see that any new company can do a new test accurately on a large scale.
The WSJ report claims a former senior employee told them Theranos was diluting fingerprick blood samples in order to test them on traditional machines. This can work but "introduces more potential for error," according to what blood pathologist Timothy R. Hamill from the University of California, San Francisco told WSJ writer John Carreyrou.
The public statement from Theranos says the "sources relied on in the article today were never in a position to understand Theranos’ technology and know nothing about the processes currently employed by the company."
The company acknowledges it needs to execute these finger-stick tests perfectly for their tests to work. Holmes told Auletta: "We have data that show you can get a perfect correlation between a finger stick and a venipuncture for every test that we run." And another of its patents is for a finger warmer, which may help prepare the finger for the blood draw. Still, it's an area where, if the phlebotomist operating the machine is improperly trained, there could be problems no matter how small the amount of blood drawn.
Journalists and doctors who have tried the test called it painless, but a look at Yelp reviews of one Theranos location, at a Walgreens in Palo Alto, shows that at least one person claims their finger was bruised. In an editorial about Theranos in the journal Clinical Chemistry and Laboratory Medicine earlier this year, Eleftherios P. Diamandis, the head of clinical biochemistry at Mount Sinai Hospital in Toronto, argued that fingerprick tests are not necessarily any more painful than needle draws — though there still might be an appeal for someone afraid of needles.
Other Yelp reviewers said they still had to have their tests done with traditional blood draws (something Theranos is equipped to do, so this could have been a test that required that), and yet another wrote that their results — which signaled alarming medical problems — turned out to be wrong. Yelp reviews aren't verified, aren't guaranteed to be accurate, and are often written by the people who think they've had the most negative experience. There's also always the potential that operator error caused a specific problem. But they are an indication that the system may still have kinks to work out.
Auletta reported in The New Yorker that in 2016 Theranos plans on doing 1 million tests, while Quest will do 600 million. If the company is going to grow and be a real option for clinical use, it needs to show its system can work accurately and handle every step from the individual finger prick to the processing of hundreds of thousands of samples every day.
Skye Gould/Tech Insider
What's yet to be proven
Many think Theranos will be forced to reveal more data about exactly how its product works as it expands. Holding information close to the vest has worked as it has gained big partnerships, Ambar Bhattacharyya, a vice president at Bessemer Venture Partners, who has handled investments in a number of other healthcare companies (they are not investors in Theranos), told Tech Insider in April. But he thinks it'll need to open up more to the scientific community to "win the trust of folks that would be deciding whether to refer [patients] to Theranos versus LabCorp or Quest."
The new allegations made by the WSJ may increase this pressure.
Bhattacharyya said that from a business perspective, winning the medical community's trust is the most important thing, but it's not the only thing.
Theranos is already offering tests at prices that are a fraction of what hospitals charge.
Theranos already offers tests at prices cheaper than what Quest and LabCorp charge — and a fraction of what hospitals charge. And unlike its competitors, Theranos lists its prices clearly on its website. Yet to conduct tests all over the country the company will need to build labs and machines to help process all those samples. It will need to train staff to collect samples and operate every stage of a complex pipeline.
Theranos needs to show it's a profitable, sustainable enterprise that will be able to scale its business as it grows, and "that's a huge risk" with a lot for them to prove and challenges to overcome, Bhattacharyya said.
There's also the question of whether regulatory hurdles could slow the company down. As Theranos has noted, it wants all its hundreds of tests approved by the FDA, even though it isn't required to get that approval right now.
The company has received approval for one test and is certified to operate lab testing facilities, but the FDA's assessment of the company's other tests — more than 240 are listed on Theranos' website and they say they've submitted over 130 for approval so far — is still pending. The company has also received FDA approval to run tests outside of traditional labs.
A company that designs tests that it uses only in its own labs, as Theranos does, doesn't need to get them approved, but the FDA is considering changing its policy. Theranos supports that change and, by having already submitted tests to the FDA, it has gone further than its competitors to gain approval. (While it wasn't originally clear that the FDA would consider those tests without changing policy, one approved test plus a statement from the administration indicates that they are indeed evaluating those tests.) But all that legwork doesn't mean the regulatory process couldn't slow Theranos down — it could yield the validation it seeks or end up throwing a wrench into its plans.
Bhattacharyya said he wouldn't be surprised if competing labs, feeling threatened by this ascendant newcomer, try to support legislation that would impede Theranos' growth by adding additional regulatory questions for new tests. Quest has already said it doesn't believe tests that rely on capillary blood can be accurate.
Theranos' board, stacked with Washington insiders, "knows DC as well as anyone else," Bhattacharyya said, which would probably help in navigating that sort of fight.
As Bhattacharyya pointed out, there's plenty of room for Theranos to expand. It has thousands of Walgreens locations it can open and thousands more potential hospitals and research projects it could become involved with, provided Theranos and Holmes find a way to handle the hurdles of clinical acceptance, regulatory red tape, and rapid growth.
Steve Jennings/Getty Images for TechCrunch
The future of medicine
Modern medicine is still a young science. We've figured out some drugs that usually work, but as we learn more about the human body and our genetic code — the things we have in common but also the things that make us unique — we may come up with a new sort of medicine, tailored for each person.
This concept, sometimes called personalized medicine, could transform the way we understand and treat disease.
While most people associate that idea with genome sequencing, any measure of a person's health data could help develop personalized medicine for them. Analyzing someone's blood test could theoretically show changes in certain characteristics that indicate an illness or a bad response to certain types of drugs.
Researchers are talking about tracking blood tests to measure the appearance of cancer-related DNA molecules. Traditional lab companies like Quest have also started talking about using blood tests for a personalized medicine approach to the early detection of diseases.
Theranos may be able to do all of those things.
"We believe every individual has a basic human right to access actionable healthcare information when they need it the most — at a time when they have an opportunity to change outcomes and live better lives. Individual engagement with lab testing will improve early detection and prevention of disease," a Theranos representative told Tech Insider in April.
Not everyone is convinced that people should order their own medical tests, and some think that doing so could create chaos. But many believe it will be central to the future — Theranos' competitors are now pivoting in that direction — and the idea that a person could order tests to collect their own medical data is core to Theranos' mission.
A company representative put it this way:
Right now, our healthcare model is set up so that insurance will only cover a test (and therefore a physician will only order a test) when a patient becomes symptomatic for a given condition. Based on the concept that the test is then 'needed,' testing only once someone is already sick results in costly outcomes and limits individuals from fully taking ownership of their own health, in a world in which so much of our healthcare costs and outcomes are driven by lifestyle related decisions. Our work is to change that.
Even though we may not be ready yet to process the vast amount of information that would come from more frequent blood tests, there are many people who think this could be a life-saving innovation. If that turns out to be true, Theranos could be the first company helping people get there.
And that's the key. By positioning itself where it has, Theranos may have taken the latest existing technology and tweaked and refined it into a unified system that's more streamlined and more elegant than any existing system, creating a disruptive lab business. It may have positioned itself into a place that's anticipated the future, setting itself up to be a leader in the personalized medicine field for years to come.
But that position relies at least in part on the belief that they really are doing things "faster, cheaper, and better."
In his June editorial, Diamandis argued that "most of the company's claims are exaggerated."
"History teaches us that there are numerous examples of seemingly disruptive technologies in healthcare (and especially in diagnostics) ... which later collapsed, due to their inability to deliver the promised goods," he wrote, noting that the merits and quality of the Theranos technology in particular were difficult to judge.
Theranos' story reminds a lot of people of Apple.
So far, Theranos' story reminds a lot of people of Apple. The company has a brilliant leader with a penchant for wearing the same color turtleneck every day. That leader theoretically came up with an elegant system that can leverage the most advanced technology and sell it to the world, though Holmes lacks the mercurial nature of Steve Jobs. Theranos controls every aspect of its technology, just like Apple.
There's a reason Holmes has generated the buzz that she has.
But here's where we also have to remember the key difference between healthcare and consumer technologies.
"I'm all for innovation and for revolution — that's what we need," says Yeo. "But it's very difficult to judge a lab that's very secretive … and unlike the Apple products, this is a product that affects people's lives. If an iPhone is faulty, we go buy something else." But if a blood test is wrong, it could endanger someone's life.
Holmes is aware of the questions people have. "Every time you create something new, there should be questions," she told CBS News. But these latest questions seem to have upset the founder. "Unclear why anonymous anecdotes outweigh scalable statistics," she tweeted today, with a link to the company's response.
None of these doubts necessarily mean that Theranos' tests don't work, though the latest allegations add further support to the idea that the technology still has a long way to go.
A health innovation has the potential to change people's lives and be transformative in a way that a phone never could, and that's precisely why, as excited as they are, scientists remain skeptical.
Lauren F. Friedman contributed to this story.