Q&A: Butler Health CIO Thomas McGill, MD

'Pure technology people seem to think, "If I buy the right stuff and implement it, my problems will be solved."'
By Skip Snow
07:14 AM
Doc with stethoscope and medical icons

Thomas McGill, MD, vice president of quality and safety and chief information officer at Pennsylvania's Butler Health System, describes the system as "Community Hospital USA."

Located about 30 miles west of Pittsburgh, Butler is a single-hospital health system with 14,000 annual discharges, 50,000 emergency department visits and 500,000 outpatient encounters a year. It services a population of 200,000 with 300 docs on medical staff and revenue of approximately $275 million.

McGill spoke with Healthcare IT News about the role of a physican CIO at this time of change, as healthcare shifts away from volume and toward value-based care. He offered the following bits of perspective.

[See also: Q&A: Geisinger CIO Frank Richards ]

MDs make good CIOs
I'm an infectious disease doctor by training. I came to Butler Health System in 1992. I was the founder of the quality department at Butler; infection control naturally developed into quality. Then I was functionally the CMIO because that's how you get data for quality work. The CIO retired. They asked me to be interim. After six months, they asked me to stick with it. I said, "Yes." That was four years ago this month.

I'm still a newbie in terms of the technology. You have to trust your technology people. You have a little more uncertainty, say, than if you came more from the tech side. On the other hand, I know very clearly what's important for the business and the operations. If I were to say what's more important I would say that business acumen trumps technology chops right now, especially with the transformation of healthcare.

I practice 50 days a year. I do one weekend a month, and I cover some vacations for my infectious disease practice partner. After a weekend at the hospital, I can talk to my [IT] team about what is hard and what is easy. I ask them to improve the workflow. It's all there – getting the workflow and the information more easily accessible, making it easier. EMR is quite customizable. It's quite an art form to get it so that it's there but not distracting. A lot of solutions, when we first lay them out, they are distracting.

[See also: Q&A: Kaiser Permanente CIO Dick Daniels]

Everyone has a certain amount of cognitive bandwidth, and the distractions or stress in life, the fatigue, can narrow your bandwidth. You get tunnel vision. You're focused on the immediate problem, which could be trying to get that system to work, and outside of that tunnel is what's really important, which may be something with the patient that isn't exactly obvious, but if you had time to contemplate it, you recognize it and you're actually doing a better job.

Butler Health System is managing transitions of care
The way I think about population health management is, No. 1, coordination of care. And what that means is on an individual basis, when you ask providers, they know what to do, but getting all the different pieces that are not part of the same exact system working in the interest of the patient – I consider that phase one of population health, so that's where this messaging comes in. But right now, we're not able to integrate it directly into their record, but we can send it to them in some secure format where they can see it as a human readable or ingested in some other way.

We pass information forward down the line more effectively than the continuity of care documents office. It could be to home health or it could be long-term care or a patient's doctor. The CCD is a good start. But there's a lot of important stuff that's not in there. We're starting to send additional messages outside of the CCD to important downstream providers that they say they need or we think they may need in order to enhance the care downstream. We're often sending them to humans, because we can't integrate it into their record at this point. But it seems to be key for preventing readmissions, for instance.

Diagnose early and often
We're changing our workflows asking providers to identify a provisional diagnosis for the problem early. Once there's a provisional problem, then there's actually a good medical practice that already has a pretty good conceptual framework about it. That can then help you not forget things and prioritize the order of things. The system can present other information. You have to be aware of that could be related to that. You, of course, have to program all this.

Let's pick a common problem: heart failure. Let's say a patient has heart failure. So, there are certain medications that are good for heart failure, certain labs you want to follow, and even on the diagnosis and coding side, for proper payment and burden of disease.

Butler integrates first, buys second, builds last
We make buy decisions all the time when looking at a problem and conversely a solution: How long would it take? How much does it cost, or do we have to buy it? That's what it really comes down to, to really pushing the vendors into telling you what they can really do. There seems to be quite a bit of vaporware out there. If I get this, what am I going to have to do to make this effective? They fail to tell you that you have to figure it out for yourself. But we figure it out before we buy it.

Pure technology people seem to think, "If I buy the right stuff and implement it, my problems will be solved." But, really, technology is a tool – so you have to know how to use it. When it comes to information technology, I focus on having people asking the right questions. I notice that when I'm at conferences, there does seem to be that belief that if I just buy the right stuff, my problems will be over.