The lead medical informaticists in hospitals, medical groups, and health systems are juggling an unprecedented number of "menu" items on their to-do lists these days. Which are the most strategically important? And do the CMIOs in those positions have what it takes to deliver the goods? Healthcare Informatics asked a virtual roundtable of industry-leading CMIOs for their perspectives.

CMIOs in hospitals, medical groups, and health systems are facing a complex stew of challenges right now, from fulfilling the requirements for meaningful use under the Health Information Technology for Economic and Clinical Health (HITECH) Act, to helping their organizations lay the foundations for accountable care, value-based purchasing, and readmissions reduction under the Affordable Care Act (ACA).

What are the top items on their strategic “to-do lists” right now? To find out, Healthcare Informatics’ Editor-in-Chief Mark Hagland spoke recently with five industry-leading CMIOs, and wove their responses together into a virtual roundtable. These five industry leaders offered quite varied responses, and diverse perspectives, on the challenges ahead. But all are absolutely agreed on the centrality of the role of the CMIO as someone whom his or her organization will need to turn to in order to help lead forward change.

Our panel:

Ferdinand Velasco, M.D.
Chief Health Information Officer
Texas Health Resources
Arlington, Texas

G. Daniel Martich, M.D.
CMIO, University of Pittsburgh Medical Center (UPMC) Health System
Pittsburgh, Pennsylvania

Christopher Longhurst, M.D.
CMIO, Lucile Packard Children's Hospital, Stanford University
Palo Alto, California

Brian Patty, M.D.
CMIO, HealthEast Health System
St. Paul, Minnesota

Colin Banas, M.D.
CMIO, VCU Health System
Richmond, Virginia

Below are excerpts from his discussions with these visionary leaders.

WHAT ARE THE MOST PRESSING CHALLENGES?

Healthcare Informatics: Gentlemen, when you look at the huge to-do list facing CMIOs across the U.S. these days—a to-do list driven very strongly by policy and regulatory mandates and initiatives right now—what do you see as the top few strategic challenges on your lists?

Colin Banas, M.D., CMIO, VCU Health System: On my end, the two biggest issues right now are interoperability and the exchange of data and information with the right person, at the right time, and in the right format. Creating true interoperability and data exchange are probably the single-biggest impact we could have, and that matters to me more than anything else. The problem is that this gets all balled up with ICD-10 and meaningful use Stage 2, and now the former PQRS [Physician Quality Reporting System] program, the outcomes measures. It’s no longer going to be good enough to construct spreadsheets on this; you’re going to have to totally automate the process. Now, obviously, that’s important, and we all want to get there, but in the sea of computing priorities, that isn’t nearly as important as interoperability, both on its own and as part of Stage 2.

G. Daniel Martich, M.D., CMIO, University of Pittsburgh Medical Center (UPMC) Health System: The first issue for me is around clinical documentation. You know very well the copy-paste dialogue that’s going on among CMIOs now; and at the grassroots level, it’s painful at times. So we’re stuck in this land of, how do we capture what we need for billing purposes and for the ICD-10 transition so that we can continue to code appropriately; and code medical-legally appropriately; and yet make the process easy for clinicians? The second for me is around leveraging the power of analytics. At UPMC, we’re starting down the path on this, and we continue to live the dream of our organization’s $100 million investment in data analytics, but under the dream layer, the real work is a few layers below that, in terms of NLP [natural language processing] and applying that to quality outcomes research and other important purposes.

Chris Longhurst wrote that article in the [Nov. 2, 2011] NEJM [New England Journal of Medicine] about mining the data. The reality is, every one of us would like to be able to go in there and say, gosh, I’ve got this 14-year-old girl with lupus and nephritis and blood clots, and I’d like to scan the EHR to support my diagnostics in her case; but it’s much harder in practice, so the entire set of issues around deriving information from the record would be number two on my list. Number three has to do with how you derive information from the EHR and clinical information systems to support academic research, and to support the emergence of personalized medicine.

Brian Patty, M.D., CMIO, HealthEast Health System: We’re in the middle of switching EHR vendors, so that’s first and foremost on my list right now. The other ones for me are continuing to hit on meaningful use, though we’ve pretty much gotten it down now, so it’s just making adjustments; and the ICD-10 transition. ICD-10 is a big, looming issue. And the other is making sure our EHR is serving us to meet the needs of healthcare reform, especially the ACO [accountable care organization] market, and the continuum of care. Within the area of initial initiatives around where we can reduce cost, one of the biggest areas is around unnecessary readmissions. In the current world, we get paid for those; in the future world, those will hurt us. We’re looking at strategies to reduce unneeded readmissions, and that ties into the ACO strategy.

TACKLING READMISSIONS

HCI: Dr. Patty, your organization is a Medicare Shared Savings Program ACO. How hard has work on readmissions and ACO development been so far?

Patty: Readmissions are tricky. We obviously look at the top chronic diseases, like CHF [congestive heart failure], heart disease in general, and total joints [total joint replacement surgeries]. Diabetes plays into all of those, of course, as a readmissions risk factor. We’ve got a risk assessment tool, and once we understand the levels of risks for different disease states, we’re building that knowledge into the clinical decision support, whether through order sets or care plans, or nursing documentation, or to-do lists—really trying to nail things down that cause readmissions. A big part of it is medications, making sure the patients understand their medications; and obviously, making sure they have follow-up visits after their discharge.

Meanwhile, on the outpatient clinic side, what we find is that poorly managed diabetics get readmitted, while well-managed ones do well. But about 20 to 30 percent of them inadvertently don’t have diabetes on their problem list. Their physician has been treating for diabetes for many years, but somehow hasn’t put diabetes into their problem list, so the items on the checklist aren’t being managed aggressively; they’re not being treated as aggressively. They tend to fall through the cracks more often, and so just making sure we’re managing our diabetics effectively is a big first step.

HCI: Coming from an organization that is very far ahead in leveraging analytics to improve care quality and support clinicians, Dr. Longhurst, where do you think CMIOs and their teams are right now in that important area?

Longhurst: The challenge with analytics is that it really needs to follow a thoughtful, well-executed EHR implementation, right? Everything out in the field is further behind in clinical analytics than it is in EHR implementation, by definition, but hopefully, it’s developing in parallel with the EHR. You take a place like UPMC, where they’re far beyond their initial implementations, they’re farther ahead on analytics as well; I think you can pretty much guess where an organization will be on analytics, based on where they are on EHR right now.

INTEROPERABILITY AND HEALTH INFORMATION EXCHANGE

HCI: Dr. Banas, as you and your colleagues at VCU work your way forward with creating and improving interoperability and health information exchange [HIE], what are you finding to be the most challenging elements in those areas?

Banas: To me, the government is doing a pretty decent job in setting standards and specifying norms on this, but for us, it’s very expensive and difficult to jump into a health information exchange; and because there aren’t a lot of other players out there for us to connect with, I’m not sure that we’re getting our bang for the buck. We’re essentially being asked to leap out there and do this. In a world of restricted budgets, what happens is that when only 60 doctors are on our physician portal right now, for the c-suite, that doesn’t exactly resonate yet. This would be a way for VCU to differentiate ourselves: we could say, look at us, if you come to us as a patient, we’ll have your record and information everywhere to send to people; but it’s hard to convince the c-suite of this.

HCI: That’s part of the challenge, convincing them, correct?

Banas: Indeed. Strategically, it seems like a no-brainer and the right thing to do, but operationally, it might be a luxury we can’t afford at the moment.

HCI: Dr. Velasco, would you also agree that the exchange of data is a strategic imperative?

Velasco: Yes, I would gather that under the headings of meaningful use and population health as a key strategic imperative. Real, meaningful health information exchange, and quality measurement, are both very central to Stage 2 meaningful use.

HCI: When it comes to working to develop more advanced versions of clinical decision support, where do you perceive CMIOs and their organizations to be right now in their progress?

Velasco: Well, it’s about time for that [advanced clinical decision support]. Many organizations that are at the HIMSS Analytics stages 5 and 6 have very advanced clinical decision support systems. I’m not worried about the requirements for clinical decision support in Stage 2, because you’d better be doing good clinical decision support in Stage 2; it was rather under-emphasized in Stage 1. The other one that’s very important in Stage 2 that wasn’t represented as well in Stage 1 was meaningful patient engagement. That will be necessary.

HCI: Patient engagement sounds soft, but it will actually be necessary to engage in true population health, correct?

Velasco: You’re absolutely right. When you get into the details of what will be involved in population health, yes, the patient absolutely needs to play a role in that.

HCI: And it will hit physicians at the point of care and at the workflow level, right? So CMIOs will be under pressure to do what’s needed there, right?
Velasco: Yes, because population health and accountable care are really driving providers to revisit their operational models and performance; and CMIOs will be central and critical to that transition, because their having led implementations for the last several years, gives them the perspective needed.

ARE CMIOs UP TO THE TASK?

HCI: When you consider your colleagues nationwide, Dr. Martich, how close would you say the match is right now between CMIOs’ level of preparation and the level of challenges facing them ahead in the next few years?

Martich: It’s probably a bell-shaped curve, and probably most of the bell is distributed where you might think it would be; and some of it has to do with age, with experience, with clinical specialty background, with level of informatics development. I think that there are plenty of people who understand both the cultural and technological aspects of this. But where CMIOs often fall astray is where they fall to one or the other end of the spectrum. They might get too technological and forget the human element of what the practicing physicians actually have to do and accomplish every day in their workflow, or vice-versa.

HCI: Dr. Velasco, where are you on the spectrum of optimism versus pessimism, when it comes to CMIOs being able to master the challenges facing them in the next few years, and being able to successfully lead their organizations forward towards the new healthcare?

Velasco: I think it depends on the organization the CMIO is in. I’m in an organization where I have the luxury of being very optimistic; that’s also my nature. And there are a lot of challenges, but I think the challenge is worthwhile, because the value, the reward, is even greater. There are CMIOs in situations perhaps less ideal than mine, and they may be challenged even to get to meaningful use Stage 1. And if they’re challenged with infrastructure problems and downtime, they will be challenged to move ahead. So it really depends on the context.