Epic and athenahealth: Must one choose sides?
In this excerpt from his book The Digital Doctor: Hope, Hype and Harm at the Dawn of Medicine's Computer Age, Robert Wachter, MD, explores the unique characteristics of two very different EHR vendors, with very different philosophies about how technology can improve healthcare.
After spending a full day at Epic headquarters and five hours with Jonathan Bush and his top lieutenants, I felt a bit like a judge listening to two very savvy attorneys plead their case. Both Epic and athena are justifiably proud of their products and their people, and both defend their worldview with great passion and (particularly in athena's case) no shortage of hyperbole. Did I really need to choose sides?
For now, I think the answer is no – the world is plenty big enough for the two to coexist (along with a few dozen other EHR companies). Athena does a splendid job serving small office-based practices, where its cloud-based architecture and openness to a multitude of apps makes sense. Its EHR interface is simple, elegant, and standardized. Engineers in Boston can monitor the clicks of thousands of physicians and quickly change the system when they spot a problem. After the Ebola virus first appeared in Texas, athena revised its patient intake checklist to include appropriate travel questions, literally overnight.*
But building an EHR for a small office practice is a far less complex task than doing so for a huge hospital. It's a bit ironic that Bush rails against full- service hospitals (calling them "do-it-all department stores") for not focusing on a narrow core business (academic centers, he believes, should just do high- end complex care, and leave everything else to others). Yet, in trying to build a do-it-all EHR rather than sticking to the stuff his company is good at (serving the needs of office-based physicians), he is doing precisely the same thing.
In fact, although Bush talks as if he's ready to lace up his gloves and take Epic on, as of late 2014 his company didn't yet have a viable hospital electronic health record product. Speaking in March of that year, athena's CFO, Timothy Adams, noted that athena had two components of a hospital EHR: a billing system and a patient record. But, he admitted, when it came to all the modules that make up a comprehensive hospital system – the nurse rounding station, lab interfaces, bed management modules, nutrition, pharmacy, and "30 other modules out there" – well, "we don't have those." In other words, athena isn't remotely close to being able to replace the inpatient functionality of an Epic or a Cerner.** "Some day, if Jonathan is successful, he'll face some of our problems," said a nettled Dvorak, clearly tired of this particular fly buzzing around his face.
I have little confidence that a 500-bed teaching hospital – the kind that per- forms transplant surgery, cares for premature infants, and manages critically ill patients in ICUs – is going to be well served by linking to a stream of free- for-all products created by a newly unleashed universe of app developers high on Red Bull and "More Disruption, Please" pep talks. As I'll discuss in the next chapter, I desperately want to see entrepreneurs in the healthcare space, but, at least for now, I want their inventions to focus on handling simpler problems, those lacking the tight interdependencies and life-and-death consequences of the hospital.
Don Berwick, former head of Medicare and founder of the Institute for Healthcare Improvement, once described a thought experiment that captures the importance of integration. Imagine, he said, that you're trying to assemble the world's best car by using the best parts. You'd connect the engine of a Ferrari, the brakes of a Porsche, the suspension of a BMW, and the body of a Volvo. "What we'd get, of course, is nothing close to a great car," he said. "We'd get a pile of very expensive junk."
Everybody chanting "Open! Open!" points to the iPhone, but if you think about it, Apple is extraordinarily closed (or at least brutally prescriptive) about who and what plugs into its system. The company has created the standard, and nobody can just join in and start "disrupting." That is why your apps work so well. Epic is trying to do the same thing as it gradually opens up to third-party programs. That's what I want in my hospital, at least for the foreseeable future.
Does this mean that the government (or Jonathan Bush, for that matter) should just leave Epic be? Of course not. Epic and all the other vendors and providers need to be pushed toward a common standard that supports sharing, accompanied by a set of incentives and penalties to promote total connectivity. Patients deserve that. One interface that is long overdue: it should be easy for physicians to choose an office-based and relatively inexpensive EHR like athena and have it link up to their hospital's Epic system, or any of the equivalent inpatient products. Today, it is not.
Whether the winning model will be a cloud-based model like athena's or Epic's more traditional architecture is not, in the end, terribly important. The market and the state of technology will determine the optimal specs. What is important is that we don't jump too quickly onto the "open" bandwagon at the cost of the kinds of integration that, at least today, we need if we are to manage complex and vulnerable patients. Let athena be athena, and let Epic be Epic. Then, within reason, let the market decide on the better solution.
* Since the missed diagnosis of the first U.S. Ebola patient took place at an Epic facility, and the Texas hospital initially blamed its EHR for the error (a position that the hospital retracted the next day), athena's prompt revision of its intake module was accompanied, unsurprisingly, by a press release that described the company's cloud-facilitated nimbleness.
** In February, 2015, athena and John Halamka's Beth Israel Deaconess Medical Center (one of the few large hospitals still using a homegrown electronic health record) announced a deal to co-develop a cloud-based inpatient EHR. It will likely take two or three years.
See also:
