On the Record

If the new administration has its way, all patient details will be converted to electronic form within five years. Jim Noga of Massachusetts General Hospital and John Glaser of Partners HealthCare tell EHM about the programs their institutions have in place to meet this goal.

“The looming challenge is to help physicians use this set of data that's now being exchanged”
-Jim Noga

According to John Glaser, CIO of Partners Healthcare, depending on the statistics you read, the adoption rate of well-developed and robust electronic health records (EHR) among US physicians is about four percent. If you loosen the definition of electronic health record so that it doesn’t include decision support, for example, then the adoption rate is about 17 percent. What does this mean for President Obama’s goal to have all patient records available in electronic format within five years?

“We would be hard pressed to bring the four percent up to 100 percent in five years,” says Glaser. “I don’t doubt that we’ll make extraordinary gains, but it’s a guess more than anything else about what percent will have adoption in a five-year period of time. It could be as low as 25 percent or it could be as high as 75 percent; but there are for example 180,000 physician groups in this country with four physicians or fewer, and it will be a challenge for us to bring them all up to full EHR within that period of time. It’s a lofty goal but it’s a good goal, and the fact that we’re unlikely to meet that shouldn’t cause us to feel that we failed.”

Jim Noga, CIO of Massachusetts General Hospital, one of the founding members of Partners HealthCare along with Brigham and Women’s hospital, agrees with this sentiment, and points out that his institution has worked hard to be ahead of the game. “From an MGH perspective, we’ve been fortunate to start early, and we’re at the 99 percent level among our physicians in terms of use of a EHR in the ambulatory environment. Where we still have some work to do is in the acute environment in terms of documentation of progress notes, nursing notes, etc., but we’ve kicked off that initiative. We will continue to focus and bring that across the line in a timely manner – we’re already targeting the 2011/2012 timeframe to complete that initiative.

“Additionally what’s a bit of a wild card still is understanding the interoperability requirements and how we might communicate with either a regional or national health network or interoperability with personal health records. That is yet to be defined but is expected as part of President Obama’s effort to computerize the nation.”

In the state of Massachusetts, by 2015 all physicians will need to be on electronic health records to maintain their licensure. Hospitals in the state will need to be on computerized provider order entry in order to maintain their hospital licensure.

Primary focus
The computerization of health records has been a big focus in the new government’s healthcare proposals – some say perhaps too much focus. Noga stresses that while EHR is important, there are other aspects that also need to be considered. “The electronic health record forms the core, but in addition to that is a whole effort around what I would call process flow initiative; for example, how do you move patients through the system effectively while not compromising quality and safety in terms of admissions and discharges?

“There’s a parallel focus besides the functionality of the EHR that is supporting the practice of medicine of the ambulatory and inpatient environment – plus supporting the research efforts, whether that’s genetics and genomics or personalized medicine – that’s extremely important for the advancement of medicine.”

Glaser agrees: “The outpatient setting and the electronic health record is the right primary focus, because that’s where most of the care occurs, and particularly most of the management of people with chronic disease. But there are also secondary focuses that are important. The hospital setting is important; the ability to do public health biosurveillance is important; the ability to see if we can accelerate clinical research of new drugs and new therapies is important.

“Across all of this, the interoperability is important, and then last but not least is the ability to reach out to patients directly through things like personal health records to help them be more actively engaged in their care. But I still think the outpatient is the most important because that’s where the vast majority of care occurs in this country.”

In December, Glaser met with the Obama team as part of the transition consultation process. Glaser describes the meetings as “effective” and voices his approval of the legislation that resulted. “The meetings were part of the administration taking hold, and they were largely to help the transition team and the President be aware of a certain variety of issues that need to be considered. And we saw a lot of those addressed in the stimulus bill that came forward. I think the stimulus bill is a nice piece of work. Like all legislation, it has flaws, but it’s a very responsive piece of legislation.”

Standards
Even if all healthcare institutions and physician offices in the country computerize their health records, it doesn’t necessarily mean they’ll be able to exchange information electronically. Individual systems are not always designed to be able to communicate with each other. How can we make this happen?

“What’s important is obviously development and adherence to standards as well as developing the core infrastructure,” Glaser explains. “Two billion of the stimulus package is geared toward some of those efforts, toward developing the office that oversees that infrastructure for developing the interoperability standard.”

“The next set of challenges will be the fact that even though you might be able to interoperate, that doesn’t mean you will interoperate,” says Noga. “There’s a lot of work we have to do with the providers to help them understand why they should do this and what the care gain and the practice will be in sharing data. They do some of that today in the paper world, but obviously the country’s envisioning a lot more.

“There are still privacy issues to wrestle with, which are complicated. For example, there is still a lot of diversity of privacy laws across the states. So if you’re in a small geography like New England with Vermont and New Hampshire and Connecticut and Rhode Island nearby, differences in state law can confuse people about what they can and can’t do.”

Noga also points out that there is another set of looming challenges that some of the research will go to. “An example would be physicians who in the future will be getting all this data from all other providers and fundamentally becomes overwhelmed by the amount of data. They might say, ‘There’s too much here; I don’t have time to go through it,’ and so a lot of the value of interoperability will be wasted because there are hundreds of notes.

“They need to be able to pick out the four or five pieces that are the most important to them. We have incentives and workflow to deal with, and privacy to deal with, and the looming challenges to help physicians maximally use this set of data that’s now being exchanged.’

Size does matter
Most large hospitals and healthcare institutions have already made strides in the direction of EHR. However, a large percentage of the healthcare community is made up of offices of four physicians or fewer, and it can be more difficult to persuade these smaller entities of the need to take their records online.

“It is more difficult with the smaller operations,” says Glaser. “Although clearly the economic stimulus language provides money to them. It says if you’re doing this by 2011 or some of the years after, you can make $40,000 or $50,000, which is a lot of money relative to the expenses. That will clearly help them move in the right direction.

“The other challenge they now face is they say, ‘I’m ready to do this, and I’d like to – who will help me? Who will help me pick the right type of record? Who will help me figure out where to put all the terminals? Who will help me do the workflow stuff?’ All the challenging work that goes on with an implementation.

“If you work at Mass General, Jim’s got a team that can help you do that, but if you’re out in the community, there is no team. One of the things that is in the legislation is the creation of regional extension centers, which would fund organizations to help those physicians with an effective implementation. Right now, not many of those organizations exist, though there are some. We have some work to do to put these organizations in place so physicians have someone to turn to.

“It’s also a real challenge communicating with that many physician practices to make sure they understand what they have to do to get these funds and what they have to do in order to implement the technology effectively. Many state and national medical societies and state governments are trying to sort through how best to reach everybody.”

Identification required
Another challenge facing the introduction of a nationwide EHR system is the fact that currently there is no single patient identifier in the US. If we computerize everyone’s records, how will we tell apart people with the same name?

“I would add that to the list of challenges we have,” says Glaser. “You don’t have to go too far outside of Massachusetts to have a problem with Mary Smith. If you’re in the Southeast Los Angeles/Texas area, there are a lot of Hispanic names that are consistent, and if you’re in California many Asian names are quite common. The country has been loathe to assign an identifier for privacy reasons, and there’s some thought that we could use Social Security numbers, but then there is a large immigrant population that doesn’t have a Social Security number.

“I don’t know how we’re going to address this issue. There’s an area of computer science research that is being done to help identify people using matching algorithms. At Partners we have a team of people who look at situations in which the algorithm does a partial match; for example, that this Mary Smith could be the same as that, the ages are two years apart. We have to have people who go look at records and determine whether they really are or are not the same Mary Smith. We’ve no idea how you do that on a national basis.

“Right now, we don’t have a single identifier, which will hinder us at some level, and we will have to rely on human beings to make the match, with all the peril that that introduces. I don’t know whether this problem will become so crushing that we will decide that in fact we will issue a number, but right now it is politically problematic.”

Moving around
Patients are increasingly mobile and often attend different institutions for different aspects of their care. As Noga explains, institutions like MGH are using a number of technologies to support this increasing mobility. “One is the product we call Patient Gateway that is available at all Partners entities. It’s a patient portal into a practice where you can communicate with physicians and where you can request prescription refills, and where visit letters are returned to you.

“I happen to be a patient at a practice that is using Patient Gateway, and I find it very valuable in terms of the asynchronous communications, being able to the day after my visit see my lab results and have my physician push his summary of the visit and follow-up items to me. That’s one way we’re reaching out. We’re not complete with that rollout, we’re in the early stages.

“In the home setting, we have an initiative called Connected Health, which is aimed at helping patients manage either chronic diseases or things like wound management from home. They’re not having to come into a hospital to see the physician, instead they can be monitored and managed remotely. We’re not yet looking at things like the Google Personal Health Record, though we’re watching it carefully.

“We’ve also done some things for our providers, in terms of them being able to access electronic health records just about anywhere where they can get an internet browser in a secure way, and we’ve also done some things such as pushing out summary data to smart phones over a cellular network. Those are the types of things we are trying to do to extend healthcare to the mobile population.”

Staying secure
With electronic systems, there can be concerns about data security. MGH and Partners both have initiatives in place to address these concerns. “Our approach to audits has become more and more robust in terms of being able to audit down to the transaction level,” explains Noga.

“On a national level as we look at interoperability, which is something that is going to have to be addressed from a security perspective. We can easily do that within the Partners network; as things become externalized, obviously systems will have to be in place so that people can’t make a request for electronic health information on an individual that would result in a breach. There are still some more things that need to go into that as the interoperability standards are developed.”

Partners has begun experimental work looking at algorithms similar to those used by credit card companies to detect suspicious patterns of purchasing. These kinds of algorithms allow you to see whether there are anomalous patterns of access within the systems and to the degree that the algorithms suggest a pattern that’s unusual, to then ask the privacy people to take a look at that and do further diligence if necessary.

MGH offers both to our patients and staff at MGH – many of whom are patients –the ability to review the audit records of who has looked at their records. Noga says this is important knowledge for people accessing the records, so that they know the capability exists and hopefully it helps them act more responsibly.

Getting personal
The increasing computerization of health records and the growth of personalized medicine have happened more ore less in tandem. How has the first affected the second?

“We’re spending some time looking at how these issues affect each other,” says Glaser. “One is looking at in the era to come, if we store either all or large portions of your genome along with what they call the proteome, which is the protein expression of the genes, what do we have to do in terms of the data structures that we have underneath the electronic record? Also looking at the decision support to the degree that physicians’ decisions are guided by knowledge of genetic makeup, what types of decision supports are needed, and the other piece – which is I hadn’t really thought of until it was pointed out to me – is that we are physicians.

“We know how to present to our cholesterol results or operative notes, and they can look at them and know what to do based on that. We don’t really know yet how to present 1000 genetic test results so that a physician at a glance can say, ‘Oh, my goodness, this is what we need to do.’ So we’ve done some initial work to understand how the EHR is altered by the personalized medicine revolution, and that’s Part A.”

“Part B is looking at some of the stuff that Jim had mentioned around the patient portal and progressively incorporating family history into that. That is obviously a terrific indicator of whether you’re at great risk of cancer or all kinds of other diseases. Even if we’re not quite sure of your genetic makeup, if there are hot zones in your family then we have a greater risk than not.

“The third part, or Part C, is we have a lot of investigators here at Mass General. We have a large research enterprise and a lot of individuals looking at the genomic basis behind disease or treatment variability. For example, people who are studying bariatric surgery in the weight loss clinic at the Mass General note that if you’re obese and they do the surgery, on average you lose one-third of your weight following the surgery, but sometimes you lose no weight.

‘Sometimes you lose so much weight that you need nutritional supplements for the rest of your life. Why does the same surgery have such different outcomes? Is there a genetic underpinning there? What we are doing with them is seeing to what degree can we leverage the existing EHR data such that research like that can be done much faster and for a lot less money, and it appears that you can.

“It’s still very exploratory, the genetic and EHR family history research, but nonetheless, we’re getting something we think will be quite common in the medical care delivery in about three to five years. It’s already common in some areas; they’re just some areas of cancer and for kids who are born with genetic anomalies, but it is yet to be broadly part of medicine.”

It’s clear there are many challenges still to be overcome on the route to computerization of patient health records, but with institutions like MGH and Partners HealthCare leading the way, others should soon follow.

Jim Noga is CIO at Massachusetts General Hospital. John Glaser is CIO at Partners HealthCare.