Partners HealthCare CIO: Inside Our Clinical Transformation

As CIO of Partners Healthcare, James Noga leads an IT organization poised to replace a generation's worth of hospital software with an industry-standard electronic medical record and revenue management system -- and that's only the beginning.

Figure 1: James Noga

In May 2012, Partners announced what the press release called "a historic commitment to, and investment in a health information technology infrastructure of the future," with software from Epic Systems at its core. In its annual report for 2012, the healthcare non-profit talked up the promise of this new enterprise system, dubbed Partners eCare, calling it "a new, fully-integrated system that will provide a single, complete and up-to-date electronic record for all Partners patients and providers" that when fully implemented "will enable a seamless flow of clinical and administrative information, reduce duplication of services, and avoid unnecessary tests -- while assuring quality, safety, and access."

However, as 2013 comes to a close, the eCare project is just getting started on a five-year rollout plan, with a first phase of deployment scheduled for 2014, and clinical systems initially untouched. The initial implementations at Massachusetts General Hospital and Newton-Wellesley Hospital will be limited to revenue cycle management -- the billing and collections side of a hospital's enterprise system -- while Brigham and Women's Hospital will be the first to field both revenue management and clinical software.

Noga said this software overhaul is important, but not really the point. "What I say is, Epic is one of the enabling tools of Partners eCare, but eCare is really about clinical transformation," he said.

Still, this will be a major technological change for the largest healthcare organization in Massachusetts -- actually, the largest private employer in the state -- an organization with a rich heritage of IT innovation. Partners was built around two Harvard-affiliated teaching hospitals, Massachusetts General Hospital and Brigham and Women’s Hospital, the largest private hospital recipients of National Institutes of Health funding in the nation. From that core, Partners expanded to encompass hospitals and other healthcare facilities across the state.

Massachusetts General in particular pioneered the development of information systems that are commonly used across the healthcare industry today -- even if they may sometimes be derided as "legacy." Over the past 10 years, Partners says it has invested more than $2 billion in improving healthcare IT at its facilities, moving all doctors to electronic medical records and implementing online medication orders at its major hospitals and most other facilities.

Noga went to work at Massachusetts General in 1990 as Director of Clinical Applications, and in 1997 was promoted to CIO for the hospital. He became Partners' corporate CIO in 2011, succeeding John Glaser, who left to become CEO of Health Services at Siemens Healthcare.

Noga spoke with InformationWeek from Partners headquarters in the Prudential Tower in downtown Boston.

Epic just the beginning
When Noga speaks of clinical transformation being more important than any one computer system, what he is really talking about is coping with change in the healthcare industry and all the pressures it is under from regulators and insurance companies. Partners is a pioneer Accountable Care Organization -- one of only 32 organizations chosen by the federal government to explore how to do business under new ground rules where payment is based on the quality of care delivered and the actual health of patients rather than fees for each service delivered. That's a framework being pushed hard by the Centers for Medicare and Medicaid Services and it is destined to be imitated by other payers under the influence of the Affordable Care Act and other regulations.

Those new ground rules are translating into new systems requirements.

Previously, Partners had a custom-built enterprise electronic medical records (EMR) system for ambulatory care -- that is, walk-in patients -- at all its facilities. On the other hand, different Partners hospitals had their own systems for tracking patients who had been admitted for inpatient care. Epic will replace both.

Partners has long had a strategy of feeding data from all its clinical systems into a common repository for analysis. "That really served us well for many years," Noga said. "But there were just limitations to that approach once we started trying to really bend the curve on medical expenses. We realized we needed more seamless integration." To implement a "population health" approach to maximizing the wellness of all patients, admitted or outpatient, across all facilities, Partners recognized "we need to have that one record, one patient view" of all care delivered, he said.

Yet it's a mistake to focus too much on the effort and expense of implementing an enterprise hospital system, whether from Epic or a competitor like Cerner, because the software is only the beginning of what's required.

"If I was just doing meat-and-potatoes HIT, I could do it -- I don't want to say inexpensively, but probably for a third of the expense," Noga said. "If you really do it correctly and take on the clinical transformation -- that's really where the expense lies. It's in the people. It's not in the hardware, and it's not in the software."

To maintain that focus on the ultimate clinical goals, Noga works closely with Chief Health Information Officer O'Neil Britton, an MD who is the executive leader of the eCare program. "It's really important to have a clinical leader of that program," Noga said, adding that this is just a healthcare-specific version of the rule all CIOs should follow to always have a clear customer with a stake in the success of any big IT initiative.

"I don't see Partners eCare as an IT initiative at all," he added. "We're just providing enabling technology."

Yet the first round of enabling technology will be enabling collections, not clinicians. Noga said the revenue cycle part of the implementation isn't necessarily easier to tackle, but it will be faster, and getting it done first will allow Partners to establish a stable foundation to build on as clinical components are added. "It means we'll have a base to work with," he said. The alternative, "going big bang, across the board," would mean deploying more interlocking software components, making it harder to track down any problems that might arise.

Also, by implementing Epic's revenue module, the Massachusetts General IT team will avoid the necessity of rewriting the billing software in use there for compatibility with the new ICD-10 diagnosis and billing codes, Noga said.

Besides, as important as the clinical components may be, "you've heard the adage no margin no mission -- we do need to make sure we're still able to get cash in the door," he said.

Once the clinical software starts to roll out, it will be implemented "in one fell swoop" at each hospital, for both ambulatory and inpatient use.

Because the transition to Epic will take time, Partners continues to refine its homegrown ambulatory care EMR, which in recent years has been updated to compete with what today are more commonly called electronic health records (EHR) systems, reflecting an expanded vision of how these systems should operate and interoperate. In the near-term, Partners' efforts at keeping up with federal Meaningful Use standards for electronic records are based on certification of that homegrown system.

I asked about the concerns I shared about these systems in a column on Why Doctors Hate EHR Software, but Noga thinks Partners doctors are past that point.

"The first six months can be very burdensome on a doctor, before they start seeing the benefit," Noga says, but once physicians have made the transition "the majority will tell you they provide better patient care through use of the record." It helps to provide niceties like mobile access to records, allowing doctors to look up patient records from wherever they may be. That's something Partners already offers for its custom ambulatory records system -- and those are usually the records doctors most need mobile access to -- but the Epic version will be an improvement, offering a more complete record of hospital and outpatient visits, he said.

Where there is dissatisfaction, it's likely as a result of hospitals and practices chasing Meaningful Use incentives in a way that led them to move too quickly, with too little planning, Noga says. "Most really good implementations plan for a drop in productivity first 6 months," he says, meaning they make accommodations for the likelihood that a doctor learning to work a different way might wind up seeing fewer patients in a day.

"Once they've run the gantlet and gone through a stabilization period, no one would say 'I'm going to give back the electronic health record,' " he says.

Standardizing, while still innovating
Part of the transition Partners is going through is striking a new balance between innovation and standardization. It created its own healthcare IT systems at a time when there were none available that met its requirements. Now, it finds itself paying a price for having non-standard systems. At the same time, investment required for keeping custom EHR systems current no longer makes sense, he says.

"At some point, we realized we're supporting these transactional systems that are available in the market today," he says. "What we want to focus our energy on is: what aren't the healthcare IT vendors doing?" Epic was chosen as a system his staff felt would provide the right base functionality, while still providing web services interfaces and other means of innovating beyond that base.

"They also use Cache as the underlying database, which is what we've used here -- of course, the origins of Cache go back to Mass General and MUMPS," he added. That is, the data management and middleware product marketed as InterSystems Cache is derived from what was originally known as the Massachusetts General Hospital Utility Multi-Programming System when it was created in the late 1960s for managing hospital data. Although it's not a standard relational database management system, that's just fine with Noga because Cache is "a very scalable database that can do transforms into relational" when necessary, responding to queries in the RSDBMS Structured Query Language.

"I always smile when people say Cache is an antiquated database -- if you look into the history SQL and others are equally as antiquated," Noga said. "To me, it's how you use them -- and what you use them for."

Having confidence in the ability to extend Epic was important because the research institutions that are part of Partners are often ahead of the curve in the ways they use new kinds of healthcare data. "In areas like genetics and genomics, we're doing some leading edge work, so maybe there we'll build alongside Epic -- and maybe someday Epic catches up," he says.

Partners plans to take advantage of Epic's MyChart patient portal, but will probably embed components of that software in its own website rather than accepting the Epic personal health record system wholesale, Noga says. Partners already offers a patient portal that's fairly successful -- close to 700,000 patients who've signed up and actively use it -- and "we think there will be things in addition to Epic functionality we'll want to provide," he says.

Overall, he was impressed by the catalog of extensible web services Epic provides, which is more extensive than what Partners was able to create internally.

Meanwhile, one of his concerns is IT's role in "allowing innovation to continue to thrive." Given the research mission of the university hospitals in the Partners system, much of that innovation originates outside of IT, but IT still has a role to play in supporting it. For example, he is working on refining Partners strategy for making use of big data technologies.

"There's a lot of opportunity there, when you look at phenotyping, genetics, image data, and biologic data repositories -- all of that is going to significantly advance our ability to advance medicine through research," Noga says. "Within the organization, we do have a cadre of what I call true data scientists -- but we'll have to grow that even more. It's true that the analytics you do for operational purposes or financial analysis is different than the data science you need to mine data from big data approaches." Fortunately, Partners has the opportunity to tap local resources like MIT for help in some cases, he says.

The centralization of patient data stemming from the Epic implementation should also have some benefits for research, Noga says. "In other words, how do we notify a primary care physician that they have a patient eligible for clinical trial? Epic will allow us to do that much more easily than we can today."

Accountable care
At an even more fundamental level, Partners needs to make better use of data to manage patient health -- not just out of altruism but because reimbursement will be increasingly based on demonstrated success at keeping entire populations of patients more health, on average. That means organizations like Partners are going to be taking on more risk, and the best way to understand and minimize that risk is with better data.

"This goes back to why we're going to Partners eCare and clinical transformation," Noga says. "Being a pioneer ACO has really allowed us to be a learning organization," Noga says, as has participation in CMS demonstration projects. "That's helping us architect what Partners eCare will be in the end."

For example, just by implementing a system for alerting primary physicians when a patient shows up at the emergency room, Partners found that it was often able to avoid the need for the patient to be admitted to the hospital and also deliver a better outcome for the patient. Similar alerts and notifications could pay off in many other areas, and Partners is working to expand the system. Still, so far, but only "a sliver of the patients" are notifications and only a fraction of the potential is being met, he says.

What's important to understand is that you never achieve the perfect system in one step, Noga says. "You bring up a system, you stabilize it, and then you learn, and then you have to optimize it. The mistake I've seen is people make is they think when you cross the finish line, the race is won. In fact, the race has only begun. That's when you have a baseline system in place, then you really need to focus on analytics and asking, am I seeing an improved process?

"I think we're going to see a shift in IT workers," he adds, "from writing functional specs and thinking about transactions to being process improvement specialists. They'll be operations-type folks who will help with workflows -- with IT providing the enabling technology."

David F. Carr is the Editor of Information Healthcare and a contributor on social business, as well as the author of Social Collaboration For Dummies. Follow him on Twitter @davidfcarr or Google+.

Though the online exchange of medical records is central to the government's Meaningful Use program, the effort to make such transactions routine has just begun. Also in the Barriers to Health Information Exchange issue of InformationWeek Healthcare: why cloud startups favor Direct Protocol as a simpler alternative to centralized HIEs. (Free registration required.)