Elizabeth shot an annoyed look at the bespectacled officer in army fatigues as he enumerated the various regulations he thought the approach she'd described fell afoul of. This was not what she wanted to hear. Shoemaker and the little military delegation he was leading had been invited to Palo Alto on this morning in November 2011 to bless Theranos's plans to deploy its devices in the Afghan war theater, not to raise objections about its regulatory strategy.

Lieutenant Colonel David Shoemaker had been politely listening to the confident young woman seated at the head of the conference table explain how her company intended to operate when, fifteen minutes in, he couldn't hold his tongue anymore. "Your regulatory structure is not going to fly," he said, interrupting her.

The idea of using Theranos devices on the battlefield had germinated the previous August when Elizabeth had met James Mattis, head of the U. S. Central Command, at the Marines' Memorial Club in San Francisco. Elizabeth's impromptu pitch about how her novel way of testing blood from just a finger prick could help diagnose and treat wounded soldiers faster, and potentially save lives, had found a receptive audience in the four-star general.

Jim "Mad Dog" Mattis was fiercely protective of his troops, which made him one of the most popular commanders in the U. S. military. The hard-charging general was open to pursuing any technology that might keep his men safer as they fought the Taliban in the interminable, atrocity-marred war in Afghanistan. After meeting Elizabeth, he'd asked subordinates at CENTCOM to set up a live field test of the Theranos device.

Under military rules, such requests had to be routed through the army's medical department at Fort Detrick in Maryland, where they usually landed on the desk of Lt. Col. Shoemaker. As deputy director for the Division of Regulated Activities and Compliance, Shoemaker's job was to ensure that the army abided by all laws and regulations when it experimented with medical devices.

Shoemaker wasn't your average military bureaucrat. He had a Ph. D. in microbiology and had spent years doing medical research on vaccines for meningitis and tularemia, a dangerous bacterium found in cottontail rabbits that was weaponized by the United States and the Soviet Union during the Cold War. He'd also been the first army officer to complete a one-year fellowship at the Food and Drug Administration, making him the army's resident expert on FDA regulations.

With his genial smile and his southern Ohio drawl, Shoemaker had a calm, self-effacing manner about him, but he could be direct with people when he needed to be. Theranos's strategy, which envisioned bypassing the FDA altogether, was a nonstarter, he warned Elizabeth, especially if she planned to roll out her devices nationwide by the following spring, as she had asserted to him. There was no way the agency would allow her to do that without going through its review process, he told her.

Elizabeth disagreed forcefully, citing advice Theranos had received from its lawyers. She was so defensive and obstinate that Shoemaker quickly realized that prolonging the argument would be a waste of time. She clearly didn't want to hear anything that contradicted her point of view. As he looked around the table, he noted that she had brought no regulatory affairs expert to the meeting. He suspected the company didn't even employ one. If he was right about that, it was an incredibly naïve way of operating. Health care was the most highly regulated industry in the country and for good reason: the lives of patients were at stake.

In his eighteen-year career in the army, Shoemaker had come across a lot of people who seemed to think the military was exempt from civilian regulations and free to conduct medical research as it pleased. That was simply not the case, though this wasn't to say it hadn't happened in the past. The Pentagon tested mustard gas on American soldiers during World War II and Agent Orange on prisoners in the 1960s. But the days of unsupervised, freewheeling medical experimentation by the military were long gone.

During the Serbian conflict in the 1990s, for instance, the Pentagon made sure to get the FDA's assent before offering troops deployed in the Balkans an experimental vaccine against tick-borne encephalitis. And only soldiers who wished to receive the vaccine were given it. Similarly, the army worked closely with the agency to make an investigational vaccine against botulinum toxin available to soldiers in Iraq in 2003. At the time, concerns were high that Saddam Hussein had stockpiled the lethal biological agent, and the promising vaccine, which had been developed by researchers at Fort Detrick, hadn't yet been approved by the FDA.

Shoemaker told Elizabeth she would need to get something in writing from the FDA supporting her position if she wanted him to greenlight the use of her machines on army personnel. Her face conveyed deep displeasure. She resumed her presentation but gave Shoemaker the cold shoulder for the rest of the day.

What was valid for vaccines was also valid for medical devices. If Theranos wanted to try out its blood-testing machines on troops in Afghanistan, Shoemaker felt certain that it would need to put together an IRB-approved study protocol. But since Elizabeth had been so adamant and he was also getting second-guessed by CENTCOM, he decided to bring in Jeremiah Kelly, an army lawyer who'd previously worked at the FDA. He scheduled another meeting with Elizabeth so Kelly could hear from her directly and provide a second opinion. They agreed to meet at 3:30 p. m. on December 9, 2011, at the Washington, D. C., offices of Theranos's law firm, Zuckerman Spaeder.

In both instances, the army consulted an institutional review board, or IRB -- a committee within the military that monitors medical research to ensure that it is conducted safely and ethically. If the IRB deems that a proposed study doesn't pose significant risks, the FDA will usually allow it to go forward, provided it's carried out under a strict protocol the committee has reviewed and approved.

Elizabeth came to the meeting alone with a single-page document outlining the same regulatory approach Shoemaker had heard her present a few weeks before in Palo Alto. He had to give it to her: the structure she laid out was creative. One might even call it sneaky.

The document explained that Theranos's devices were merely remote sample-processing units. The real work of blood analysis would take place in the company's lab in Palo Alto, where computers would analyze the data the devices transmitted to it and qualified laboratory personnel would review and interpret the results. Hence only the Palo Alto lab needed to be certified. The devices themselves were akin to "dumb" fax machines and exempt from regulatory oversight.

There was a second wrinkle Shoemaker found equally hard to swallow: Theranos maintained that the blood tests its devices performed were laboratory-developed tests and therefore beyond the FDA's purview.

The Theranos position then was that a CLIA certificate for its Palo Alto lab was sufficient for it to deploy and use its devices anywhere. This was a clever theory, but Shoemaker didn't buy it. And neither did Kelly. The Theranos devices were more than just dumb fax machines. They were blood analyzers and, like all other blood analyzers on the market, they would eventually need to be reviewed and approved by the FDA. Until then, Theranos would need to consult with an institutional review board and come up with a study protocol that the agency could live with. It was a process that typically took six to nine months.

Elizabeth continued to disagree despite the army lawyer's presence. Her body language wasn't as hostile as it had been in Palo Alto and she was more willing to engage in a discussion, but they remained at an impasse. What was strange was that no one from Zuckerman Spaeder was there with her in the room. Shoemaker had expected her to show up accompanied by several of the firm's partners, but there she was on her own. She continued to invoke the firm's legal advice, but no one from the firm was present to attest to it.

The meeting ended with Shoemaker reiterating that he would need to see something in writing from the FDA backing up Theranos's regulatory stance before he signed off on any experiment in Afghanistan. Elizabeth agreed to get such a letter. She acted as if it was a formality. Shoemaker very much doubted so, but at least things were now clear: the ball was in Theranos's court.

Shoemaker didn't hear anything more about the matter until the late spring of 2012, when he started receiving queries from CENTCOM again. He couldn't help but be annoyed. Not only had Theranos failed to produce the letter he'd asked for, the company had gone completely silent since he and Kelly had traveled to Washington to meet Elizabeth in December.

With the approval of his boss, he decided to make contact with the FDA himself. On the morning of June 14, 2012, he sent an email to Sally Hojvat, the head of the agency's microbiology devices division. The two had worked together during Shoemaker's FDA fellowship in 2003 and had just run into each other at a conference the previous week. Shoemaker described to Hojvat the Theranos situation and, calling the company's regulatory approach "quite novel," requested the agency's guidance on it. Although he didn't intend his email as anything more than an informal request for advice, it set off a sequence of events that would have made him think twice about sending it if he could have foreseen them.

Hojvat forwarded his query to five of her colleagues, including Alberto Gutierrez, the director of the FDA's Office of In Vitro Diagnostics and Radiological Health. Gutierrez, who had a Ph. D. in chemistry from Princeton, happened to have spent a not insignificant portion of his twenty-year career at the agency pondering the question of laboratory-developed tests.

That changed in the 1990s when laboratories started to make more complex tests for mass use, including genetic tests. By the FDA's own reckoning, scores of flawed and unreliable tests had since been marketed for conditions ranging from whooping cough and Lyme disease to various types of cancers, resulting in untold harm to patients. There was a growing consensus within the agency that it needed to start policing this part of the lab business, and the biggest proponent of that view was Gutierrez. When he saw the email Hojvat forwarded to him from Shoemaker, Gutierrez shook his head in disbelief. The approach it described was exactly the type of regulatory end run around the FDA that he wanted to put a stop to.

The FDA had long considered it within its power to regulate LDTs, as laboratory-developed tests were known. However, in practice, it had not done so because back in 1976, when the Federal Food, Drug, and Cosmetic Act was amended to expand the agency's authority from drugs to medical devices, LDTs weren't common. They were only made by local laboratories occasionally when an unusual medical case required it.

After some back-and-forth, Gutierrez, Yost, and Keller all reached the same conclusion: the Theranos model didn't comply with federal regulations. Yost and Keller decided it wouldn't hurt to send someone to Palo Alto to see what exactly this company that none of them had heard of before was up to and to correct its misconceptions.

The job fell to Gary Yamamoto, a veteran field inspector in CMS's regional office in San Francisco. Two months later, on August 13, 2012, Yamamoto arrived unannounced at Theranos's offices in Palo Alto. By then, the company had completed its move to the old Facebook building located at 1601 South California Avenue, less than a mile from its former home on Hillview Avenue.

Gutierrez's view that it was the FDA, not the Centers for Medicare and Medicaid Services, that should regulate LDTs did not mean he didn't get along with his colleagues at CMS. To the contrary, they had a good working relationship and often communicated across agency lines to try to bridge the regulatory gap spawned by outdated statutes. Gutierrez forwarded the Shoemaker email to Judith Yost and Penny Keller, two members of CMS's lab-oversight division, adding a note at the top:

How about this one!!! Would CMS consider this an LDT? I have a hard time seeing that we would exercise enforcement discretion on this one.

Alberto

Sunny and Elizabeth ushered Yamamoto into a conference room. When he explained that his agency had received a complaint about Theranos and that he was there to look into it, he was surprised to learn that they knew where and who it came from. Someone had apparently tipped them off about Shoemaker's email to the FDA. Elizabeth was not pleased, a sentiment made clear by the scowl on her face. She and Sunny professed not to know what Shoemaker had been talking about in his email. Yes, Elizabeth had met with the army officer, but she had never told him Theranos intended to deploy its blood-testing machines far and wide under the cover of a single CLIA certificate.

Why then had Theranos applied for a CLIA certificate? Yamamoto asked. Sunny responded that the company wanted to learn about how labs worked and what better way to do that than to operate one itself? Yamamoto found that answer fishy and borderline nonsensical. He asked to see their lab.

They couldn't deny him access the way they had to Kevin Hunter. This was the representative of a federal regulatory agency, not some private lab consultant they could thumb their noses at. So Sunny reluctantly took the inspector to a room on the second floor of the new building. After Dupuy's firing, Theranos had moved the lab there from its temporary East Meadow Circle location.

What Yamamoto found in the room didn't impress him but didn't raise any big concerns either: it was a small space with a couple of people in white lab coats and a smattering of commercial diagnostic instruments that were sitting idle. It looked like any other lab. No sign of any special or novel blood-testing technology. When he pointed this out, Sunny said the Theranos devices were still under development and the company had no plans to deploy them without FDA clearance -- flatly contradicting what Elizabeth had told Shoemaker on not one but two occasions. Yamamoto wasn't sure what to believe. Why would the army officer have made all that stuff up?

There were no clear violations he could point to about the way Theranos was currently operating, however, so he let Sunny off with a long lecture about lab regulations. He made sure to emphasize that the scenario Shoemaker had described in his email to Sally Hojvat -- experimental blood analyzers operated remotely from one CLIA-certified mother base -- was out of the question. If Theranos intended to eventually roll its devices out to other locations, those places would need CLIA certificates too. Either that or, better yet, the devices themselves would need to be approved by the FDA.

We are taking swift action to correct these misleading statements. I would very much appreciate your help in getting this information corrected with the regulatory agencies -- LTC Shoemaker communicated to the FDA that he was giving them "a heads-up" about "what Theranos was up to" and provided the agency with incorrect information that makes us appear to be in violation of the law. Since this misinformation came from within DoD, it will be invaluable if this information is formally corrected by the right people in DoD. Thank you for your thoughts and as always for your time.

Elizabeth wasn't one to sit still and quietly take it when she felt her company was under attack. In a blistering email to General Mattis, she hit back against the person who had dared to put hurdles in her way. Shoemaker, she wrote, had communicated "blatantly false information" to the FDA and CMS about Theranos. She went on to heap several paragraphs of scorn on the lieutenant colonel and listed seven inaccurate statements he had allegedly made to the agencies "compiled with assistance from our counsel." Her email closed with a request:

When he read Elizabeth's email a few hours later, Mattis was furious. He forwarded it to Colonel Erin Edgar, CENTCOM's command surgeon and the aide-de-camp he'd put in charge of making the Theranos field test happen, with a note that conveyed his anger:

The CMS inspector's surprise visit had put Elizabeth on the warpath. In a phone call to Colonel Edgar, she threatened to sue Shoemaker. Edgar relayed her threat to his Fort Detrick colleague, along with news of the inspection. He also forwarded to Shoemaker Elizabeth's email to Mattis and Mattis's reaction to it.

Elizabeth.

With my best regards,

Erin: Who is LTC Shoemaker and what is going on here?… I have tried to get this device tested in theater asap, legally and ethically, and I need to know did this visit happen as related below and how do we overcome this new obstacle… Bottom line, I need ground truth for the accuracy of the statements below. If I need to see LTC Shoemaker and LTC Mann so they can explain how I am pushing for something unethical or illegal please set up a time for them to meet with me in Tampa when I return to the states [sic] (I am going to be delayed in theater past my original return date). Thanks, M

Shoemaker also felt genuinely bad that his actions had led to an inspection of the company. He was well placed to know how unpleasant such visits could be: his previous assignment had been at the Army Medical Research Institute of Infectious Diseases, where he'd taken over as director of biosurety, the department responsible for securing biothreat agents used in army research, two weeks before Bruce Ivins killed himself in July 2008. The suicide had led to the disclosure that Ivins, an institute researcher, was the likely perpetrator of the 2001 anthrax attacks and to an avalanche of inspections from an alphabet soup of government agencies that had continued unabated for two years. Shoemaker had been the officer on the receiving end of every single one of them.

When he read the email string, Shoemaker blanched. Mattis was one of the most powerful and fearsome people in the military. The blunt-spoken general had once famously told Marines stationed in Iraq, "Be polite, be professional, but have a plan to kill everybody you meet." This was not a guy you wanted to get on the wrong side of if you were a lower-ranking army officer.

When he confronted Colonel Edgar with that information, Edgar sheepishly admitted that he had been the one who'd shared his email to Sally Hojvat with Elizabeth in what he described as an oversight. He apologized and invited Shoemaker to come to CENTCOM headquarters in Tampa, Florida, the following week to walk Mattis through the regulatory issues. Shoemaker was nervous about meeting the general face-to-face, but he accepted the invitation. He reached out to Alberto Gutierrez to see if he could join him on the trip, figuring his opinion would carry more weight if it was supported by someone high up at the FDA. Although it was on short notice, Gutierrez agreed to come along.

With Colonel Edgar's encouragement, he tried to defuse the situation by emailing CMS officials that he had never meant to imply that Theranos had already implemented the regulatory strategy he'd described, merely that it was considering it. He also expressed surprise that the agency had told Theranos he was the one who requested the inspection. The response he got brought another surprise: CMS had told Theranos no such thing; the company already had a copy of his correspondence with the FDA when the inspector arrived.

Shoemaker had gone over everything again with Gutierrez and felt confident he was on solid ground. He spoke first, giving a brief overview of the issues raised by an in-theater test of the Theranos technology. Gutierrez took over from there and told the general his army colleague was correct in his interpretation of the law: the Theranos device was very much subject to regulation by the FDA. And since the agency hadn't yet reviewed and approved it for commercial use, it could only be tested on human subjects under strict conditions set by an institutional review board. One of those conditions was that the test subjects give their informed consent -- something that was notoriously hard to obtain in a war zone.

At 3:00 p. m. Sharp on August 23, 2012, Colonel Edgar escorted the two men into Mattis's office on MacDill Air Force Base in Tampa. The sixty-one-year-old general was an intimidating figure in person: muscular and broad shouldered, with dark circles under his eyes that suggested a man who didn't bother much with sleep. His office was decorated with the mementos of a long military career. Amid the flags, plaques, and coins, Shoemaker's eyes rested briefly on a set of magnificent swords displayed in a glass cabinet. As they sat down in a wood-paneled conference room off to one side of the office, Mattis cut to the chase: "Guys, I've been trying to get this thing deployed for a year now. What's going on?"

Mattis was reluctant to give up. He wanted to know if they could suggest a way forward. As he'd put it to Elizabeth in an email a few months earlier, he was convinced her invention would be "a game-changer" for his men. Gutierrez and Shoemaker proposed a solution: a "limited objective experiment" using leftover de-identified blood samples from soldiers. It would obviate the need to obtain informed consent and it was the only type of study that could be put together as quickly as Mattis seemed to want to proceed. They agreed to pursue that course of action. Fifteen minutes after they'd walked in, Shoemaker and Gutierrez shook Mattis's hand and walked out. Shoemaker was immensely relieved. All in all, Mattis had been gruff but reasonable and a workable compromise had been reached.

The limited experiment agreed upon fell short of the more ambitious live field trial Mattis had had in mind. Theranos's blood tests would not be used to inform the treatment of wounded soldiers. They would only be performed on leftover samples after the fact to see if their results matched the army's regular testing methods. But it was something. Earlier in his career, Shoemaker had spent five years overseeing the development of diagnostic tests for biological threat agents and he would have given his left arm to get access to anonymized samples from service members in theater. The data generated from such testing could be very useful in supporting applications to the FDA.

Yet, over the ensuing months, Theranos inexplicably failed to take advantage of the opportunity it was given. When General Mattis retired from the military in March 2013, the study using leftover de-identified samples hadn't begun. When Colonel Edgar took on a new assignment as commander of the Army Medical Research Institute of Infectious Diseases a few months later, it still hadn't started. Theranos just couldn't seem to get its act together.

In July 2013, Lieutenant Colonel Shoemaker retired from the army. At his farewell ceremony, his Fort Detrick colleagues presented him with a "certificate of survival" for having the courage to stand up to Mattis in person and emerging from the encounter alive. They also gave him a T-shirt with the question, "What do you do after surviving a briefing with a 4 star?" written on the front. The answer could be found on the back: "Retire and sail off into the sunset."