What to Read Next
“What I have to do across 2,200 employees, 83 locations, 4,000 doctors and two million patients is try to take a best guess at what the future will be,” says Dr. John Halamka. Halamka is a medical doctor (his specialty is emergency medicine), but he has spent his career focused on bringing technology to the medical world.
As chief information officer at Beth Israel Deaconess Medical Center, Halamka is responsible for clinical, financial, administrative and academic information technology for one of the world’s leading hospitals. Based in Boston, Massachusetts, Beth Israel Deaconess is an academic medical center affiliated with Harvard Medical School. Its services include cardiology, obstetrics, gastrointestinal disorders and cancer care.
Halamka has been in the CIO position since 1997. Back then, he says, he personally wrote the code for the company’s health information exchange and for the foundational work of Beth Israel Deaconess’ electronic health record system (known in the industry by the acronym EHR).
Today, he doesn’t write code. “The nature of what the digital guy does today is more about influence and organization,” Halamka says. His job is to help coordinate the integration of digital technologies across the dozens of medical locations and, increasingly, right into patients’ homes. And he does it in an industry of intensely, if understandably, tight regulation.
In a conversation with Gerald C. (Jerry) Kane, an associate professor of information systems at the Carroll School of Management at Boston College and guest editor for MIT Sloan Management Review’s Digital Leadership Big Idea Initiative, Halamka explains what his five primary digital strategies are, how he deals with the challenge of finding good talent in a nonprofit world and why he found Google Glass intriguing but not ready for prime time.
You’re in health care and you’re a digitally savvy person. Set the stage for us: How is bringing digital technologies to health care different than other industries?
I have a $1.5 million fine if somebody sends a text with a patient identifier in it. I’m betting that doesn’t happen in most other industries. The constraint about data flows, audits and security, is very significant.
Data integrity is very significant. What if data is corrupted and your allergy is now missing? In another industry, if a piece of data about you, say your frequent flyer number, were missing, probably you wouldn’t die.
In our industry, there is zero tolerance for downtime. My average downtime is six minutes per year. And people say, “Were you asleep at the wheel or are you just dumb? Babies could die in those six minutes.” Try to find another industry for which six minutes per year, 24/7/365, is considered an intolerable number.
And our funding is tight — in our company, we have 1.9% of the operating budget for health care IT.
So demand is high, regulations are significant and funding is low. Those are just a few things that are different about health care.
Tell us about what Beth Israel Deaconess Medical Center is doing with respect to digital strategy. What is you digital strategy and how are you communicating that across the organization?
Sure. As we all know, health care gets better by getting bigger. So mergers and acquisitions are the theme of the day. Of course, the Affordable Care Act now pays us for performance and value, quality and outcomes, as opposed to just more volume. So that implies if we’re going to control all the means of production, if we want to ensure consistent policy and treatment guidelines and keep you out of the hospital and look at continuous wellness, rather than episodic sickness, we do need more sites of care. And specifically, patient-centered medical homes, suburban hospitals, urgent care, skilled nursing facilities, etc.
We have community hospitals, 20-, 50-, 100-bed community hospitals for which MEDITECH [software for electronic health records, or EHR] is currently used. MEDITECH is fine for that size hospital, although with every hospital we acquire, it’s a different version of MEDITECH and a different configuration, a different user experience.
So digital-strategy point number one is to take every single-instance of MEDITECH in the empire, move it to a single instance cloud-hosted web-enabled application. A little edgy. MEDITECH has agreed, as we are the first customer for this, to host it and use a subscription-based software as a service model for MEDITECH, with one single instance for all our sites.
Digital strategy number two is ambulatory care. It turns out, MEDITECH is not a great ambulatory system. So you look at the world out there and you’ve got the Epics, the GEs, you’ve got eClinicalWorks, you’ve got Athena, a few others that are not doing so well in the market. EClinicalWorks, a client-server–based application, has been our incumbent ambulatory care system, but it is very challenging in an agile world. Athena, cloud-hosted subscription model, runs on anything anywhere; much more agile. So digital strategy number two, move the entire empire ambulatory care system to Athena.
Digital strategy number three is expanding caremanagement. In a world where your funding is dependent on outcomes and wellness, you need something beyond an EHR. An EHR is fine for a single doctor to do analytics, but it is not enough for population health or care management. Two years ago, we built a care-management medical record that today receives thousands of transactions a day from disparate EHRs across the empire. That provides a single, consolidated view to care managers, non-physician folks, who are looking for gaps in care and for deviations from protocols and guidelines. It can be used for initiating visiting nurses in the home, telemedicine, teleservices visits, scheduling appointments, looking at activities of daily living and social support. These are the sorts of things you need to do to survive in a risk-based world.
So far, by the way, using that care-management medical record, Beth Israel Deaconess care organization has become the number-one ACO [accountable care organization, tying reimbursements to reductions in the total cost of care] in New England and the number-three ACO in the entire United States. That’s because we are preventing redundancy and waste, and we’re focusing on keeping you well in your home, as opposed to seeing you for more ICU stays or emergency visits.
Digital strategy number four is engaging patients and families. They are seeing the notes written about them. They are contributing to the records, structured and unstructured data. Their devices in the home are connected, like blood pressure cuffs and something like Apple Watch one day. The idea is that you need data about weight and glucometer values and various vital signs if you’re going to look for interval change between office visits and make interventions. The patients and families articulate their goals for the visits, and the care team then shows them progress against those goals. We don’t want to give more care, we want to give right care. Giving patients and families on BYOD [bring your own device] devices the ability to see digital dashboards of their communication preferences, care plan, goals and progress certainly helps.
And then digital strategy number five is to make sure we comply with all the various federal regulations, whether that’s ICD-10, meaningful use stage one, two or three, HIPAA rule, ACA, etc. Of course, we have done things like leverage the state’s health care information exchange to its greatest extent to achieve the various interoperability requirements.
So that’s, in brief, the digital strategy, and that is probably communicated to every manager and every supervisor. I go to what are called leadership meetings to explain it. It’s well known among the senior management team. Published extensively on my blog. Broadly communicated in every stakeholder meeting I have with other CIOs across both our community and the country. As well, at Beth Israel Deaconess, we have engaged governance. With all the board members we reemphasize the digital strategy we’ve had in FY15 and what we’re executing in FY16.
You have a very clearly articulated digital strategy. What have been the most significant challenges you’ve dealt with as you’ve sought to implement the various aspects of the strategy?
Well, strategy number one. Do you think that the CEO or the senior executive business owners come up to you and say, “Do you know what we really need? We need an omnibus, continuous care-management platform for population health and analytics that will help us understand variations in cost and care, so we can maximize quality, safety and efficiency?” No. The senior executives say, “I don’t know what IT we need. I’m not really sure.”
So how do you get a set of crisp requirements and specifications that are going to be foundational to an IT project? Or do you try to get together, bottom up, a bunch of people and skate where the puck is going to be? You guess.
In effect, what I have to do across 2,200 employees, 83 locations, 4,000 doctors and two million patients, is try to take a best guess at what the future will be. That’s sort of strategy issue number one. Not top-down command and control.
Number two, budgets. In this world of declining reimbursement, it’s not as if I have vast amounts of cash to go try a bunch of things. I have to be extraordinarily focused in my efforts, and be cheap. The entire cost of the MEDITECH cloud ends up being not particularly different from the costs that IT and the community are incurring today. In effect, we’re doing innovation and doing projects in a totally cost-neutral way.
Challenge three, compliance and regulatory burden. I have to share more data with more people for more reasons, but never spill a byte. Because if I do, I have a $1.5 million HIPAA omnibus rule fine and the attorney general crawling down my back.
In a world with change, the health care environment is basically undergoing a radical redesign as we move from fee-for-service to value-based purchasing. Every stakeholder wants their own thing, maximizing each individual silo. Governance is a real challenge. Saying no, making sure the institution agrees on what we’ll do and what we’ll not do, and trying to keep people satisfied when they’re told no. That’s kind of hard.
And then finally, a strange thing. Most for-profit companies will pay out generous bonuses, give you stock options, everybody has a Tesla Charger in their parking space. But in nonprofit organizations, I have very little to offer employees to recruit and retain them. I am competing with social media dot-coms and EHR vendors that pay much better than I do. So recruiting and retaining talent while I’m executing a digital strategy is hard.
On that last point, how do you recruit and retain digital talent in this highly competitive environment? And what to do you look for?
The first is to find people who are mission driven. Do you have any idea what the CIO of HCA [Hospital Corporation of America, a U.S. for-profit operator of health care facilities] makes per year?
He makes $3.7 million. That would be slightly greater than 10 times my salary. That’s in Nashville. So people who are working in nonprofit health care are not doing it for fame or fortune. They are doing it because of a sense of mission. And so you find people who really feel just good about their making a difference.
Number two, you make the organization seem so innovative and cutting edge that people want to work there to be part of the next new thing. Beth Israel Deaconess was the pilot site for Google Glass, the pilot site for Apple HealthKit, we have the Apple Watch.
And number three, we create a family environment. In 20 years, I have never had a resignation of one of my direct reports. You create a collegial place to work, where we never have blame or negativity, and where we examine the process that failed, not the person who failed. People feel very supported in that environment. So it’s a combination of mission and excitement and family that seems to be successful.
What have been the biggest outcomes? What are the benefits that are being realized?
Our goal is zero harm. It is very difficult in a paper-based world to reduce harm, but we haven’t had a handwritten order in 12 years. Every single med is dispensed with a triple check, with barcoded meds and barcoded delivery and the pharmacist oversight. It’s all digital. So you reduce harms — that’s certainly a significant outcome.
We have 80,000 monthly users of our shared medical record, where patients and families are contributing and looking and reading and engaging and securely messaging their clinicians. That digital workflow has led to significant patient satisfaction and patient retention.
We’re able to provide layers of decision support and analytics that are probably greater than most institutions because we have the data collection instruments and the warehouses, data normalization and health care information exchange. We’re able to manage risk because of this digital footprint.
If you can look into your crystal ball, what are the biggest changes facing digital and health care coming in the next three to five years, or whatever timeframe you can project into?
How about a month from now? We all are seeing this mass migration to mobile. Eighty percent of the website accesses at Beth Israel Deaconess are mobile based. The desktop is dead, the laptop is dying. Instead it’s tablets, it’s mobile phones. For us, ensuring that our patients and our doctors have the tools they need to do the tasks they want in a mobile environment — while securing it — is kind of an interesting challenge.
I mean, the number of security issues that we’ve had over the last year — state-sponsored cyber terrorists, hacktivists and organized crime. Embracing BYOD and mobile in a world filled with an Internet that’s a swamp, that’s a challenge.
What do you get excited about as far as how digital is changing health care? Do you geek out on anything? Do you say, “This is really cool, what we’re going to be able to do”?
I don’t geek out on anything. I’ve got a lot of equanimity. Not high highs or low lows.
But from personal experience — my father died two years ago and my father-in-law four months ago — what you realize is that there are gaps in our digital world we need to fill. How do we ensure that physician orders or medical orders for life-sustaining treatment, end-of-care preferences, goals, are transmitted across the continuum? Are they stored on your phone and then shared with a caregiver? Are they on a registry where they can be accessed by anyone? There are still a lot of advance directives, health care proxies and physician orders for treatment moving around on paper. It’s very disconnected.
I think it’s increasingly important that we do get information from the home of all kinds, which means teleservices, telemedicine, care in the home. I believe that Minute Clinic and those kinds of urgent cares will have an increasingly important role. Walgreens is beginning an infusion program, for example. So when my wife was diagnosed with breast cancer, she was driving in on the Mass Pike hours every week, when, in fact, there’s a Minute Clinic 50 feet from our farm that now could conceivably do her chemotherapy infusions. That’s lower cost, with higher patient satisfaction. It requires that the digital record be sent through the health care information exchange, to share it with all the caregivers.
So those sorts of things — more connectivity, reaching into the home, more patient centricity — those are exciting.
Where do you find the time to stay ahead of what’s next, or what you could be doing, given new technologies coming on the market? How do you remain innovative?
It’s important to have an individual contributor pushing innovation. Because otherwise, innovation will get lost in the tyranny of the urgent.
Do you think that applies beyond health care as well?
Oh, I do. Do you think it’s IBM that makes the big innovations? Of course it doesn’t. It’s the two or three guys or gals who come together and innovate and drive some new disruptive solution home. Not large, lumbering companies. So whether it’s a person or a small group, a skunkworks, it’s that sort of thing.
Tell us about your Google Glass project in delivering health care. Did it work?
I was asked to pilot Google Glass in health care. What we decided was the use case would be the emergency department.
The plan was this: We lock the devices in a safe. A physician comes in, unlocks the safe, takes out a device, puts it on. Logs in by looking at a QR code that is unique to that individual physician. So a physician is now in the department, on shift and logged into the glass. As they walk into any room, they look at the door. On every door is a QR code. The registration system assigns patients to a room. Therefore, the physician knows which patient is in the room they’re walking in. They can see the problem list, medication list, allergies, laboratory results, care plans and other things in Google Glass as they are talking to the patient.
So we went live with it and deployed it. It was, I think, four pairs of Glass across eight doctors, who used it for several months. And then we threw it away. Because it worked flawlessly from our perspective, but the Google Glass was a horrifically engineered device. Used a TI processor that overheated. We had one pair melt. Had a one-and-a-half-hour battery life. Wi-Fi didn’t roam. It didn’t have Android updates. It had security problems.
So it was just a prototype that was not ready for prime time. Of course, we think wearable computing is really great and important, and we will be happy to use a device in the future that actually works.
So you’re not using that at all right now?
We can’t, because it can’t hold a charge for more than an hour and a half and they melt.
Wow. Okay. What about Apple Watch, you mentioned that one.
Yes. The idea was this: If you have a patient with, for example, multiple medications, that patient may not understand what medications to take when. So what if, on the Apple Watch, you gently tap them using the optics. You show them a picture of the pill to take. After they take it, they tap the watch and you have a patient-generated electronic medication administration record showing compliance with medications.
And? How did that experiment go?
So I am a white male, and I have very little body hair, and it turns out that the watch works fabulous on me. But if the watch is put onto a dark-skinned person, especially one with a lot of body hair, the sensors don’t work. Apple in its release scaled back a lot of its health-related telemetry goals, and focused more on fitness — how much did I run, that sort of thing. We may get there, but at the moment the design of the sensors just don’t work for everybody.
We’ll see. It’s, again, kind of an expensive thing, and it requires that you have your phone. So I think the jury is out as to what adoption will be in health care.
So that’s two examples, the Apple Watch and the Google Glass, that didn’t live up to the expectations you had initially. How do you approach new technologies given these kinds of hiccups? Does it dissuade you?
Failure is a valid outcome. What I mean by that is, we’ve learned that wearable computing is great, but Google Glass is not. And therefore, let’s just wait until the technology is ready. Or with Apple Watch, we may discover the patients love it, and therefore it is a platform. Or not, and that’s okay. We just don’t know.
My traditional final question, is there anything I should have asked that I didn’t?
Just recognize that as you talk to IT leaders, it is just really hard to be an IT leader in the current environment when you are being asked to change the wings on a 747 while it’s flying. Total security, total reliability and stability, with complete innovation at the same time. And so I sort of wonder, as we go forward to the future, will people want to take this role? Or maybe the role has to be recast and the tasks divided across multiple individuals so they can deal with the pace of change and the stress.