HIStalk Interviews Paul Brient, CEO, PatientKeeper

Paul Brient is CEO of PatientKeeper of Waltham, MA.

Tell me about yourself and the company.

I’ve been CEO of PatientKeeper for almost 14 years. Our company is focused on automating physicians, primarily in an inpatient setting. We offer an overlay solution that allows doctors to automate their entire days, regardless of the back-end system that they are working on in their hospital.

Given the data entry that’s expected of physicians, is it possible to make usability better?

Certainly usability has come to the forefront as we have gotten past the adoption question and people are using it. But now the question is, can people use it in a way that saves them time? Clicks and keystrokes are the enemy of saving time. Lack of intuitiveness is as well. If you have to puzzle over a screen and figure out what is being asked of me, or how do I find that order that I’m looking for, those things all kill productivity.

Clearly we think it’s possible to create systems that save physicians time, but it requires a very thoughtful set of work. Not only on software design, but also on, what are we going to ask the physician to do? 

Obviously in our current healthcare environment, there are a lot different people in different organizations that have very legitimate things they would like physicians to do. Unfortunately, without some sort of filter or prioritization of them, you end up with all of them being thrust on the doctors. That just kills their productivity.

How do you go beyond the technical definition of usability to design software that physicians will at least tolerate and maybe even enjoy using?

In healthcare, that is a particularly challenging question. If you go back to the days of Hewlett-Packard, they were engineers building software or systems for engineers. They had this next-bench idea, where literally they would be building a tool for an engineer at the next workbench at Hewlett-Packard. They had this great environment for design.

In the healthcare world, that’s just not practical. You can’t just go sit in a hospital and have doctors write software while they are taking care of patients. That would be a bad thing for lots of reasons.

We think the best approach is get as close to that as you can, though, which is to have full contact with practicing providers to get feedback on what the real world is in healthcare delivery. Not a theoretical world, a theorized world, or a world they way we would like it to be. The actual world of all the crazy data patterns and situations that occur.

Then, get experienced designers who have usability training who understand how to build good software. If you don’t expose them to the chaotic and complicated world that physicians face every day, they just can’t build software that works for them. It’s really hard. It’s a difficult challenge to get access to that environment and then also to digest it in a way that makes sense.

The handful of significant inpatient EHR vendors are running decades-old code. Are they challenged to meet customer demands without rebuilding their products from the ground up?

Cerner Millennium — which I think is the most modern of the systems — was released before the millennium, in 1997. They certainly all have some legacy aspects to them in terms of technology. They weren’t built yesterday. You couldn’t have built them yesterday, because it takes a long time to build these systems. They’re big and complicated and they have many, many elements to them.

But I do think that some of the vendors — with the move towards interoperability and some of the standards that are being proposed, the FHIR concept if not the standard — pressure is starting to get applied that will allow these systems to become more open and allow innovation to occur that hasn’t before. Even a system as old as Meditech Magic can be made very open. It’s not a technological limitation, it’s a philosophical limitation. The push towards interoperability is helping to get the philosophy aligned more where we would like the technology to go.

When we talked three years ago, you said that healthcare is the only area left where it’s OK to have a monolithic, closed system that doesn’t support interoperability or an ecosystem. Where do you see that going?

Certainly in the last three years it has improved a lot. The FHIR standard has come out. At HIMSS, we saw Cerner demonstrating applications running against Millennium and moving across and running those same applications against Epic or even PatientKeeper, since we support it as well.

That’s a big change. That’s awesome. But it’s not yet sufficient. Even if you make the software interoperable, the data underneath in many hospitals isn’t yet. It’s not LOINC encoded and all that stuff like it would be if you started from scratch. But they did their implementations 30 years ago as well.

There’s still a lot of work to do as an industry. It’s a little bit chicken-and-egg. The more we open stuff, the more people can innovate and invent and other vendors can create cool applications that motivate people to want to exercise interoperability. That says, we’ll make more interoperability. It becomes a virtuous cycle. Without that pull, it’s just theoretical, “Hey, you should be interoperable and make some new APIs available” and no one really uses them. That isn’t going to drive it.

I think we’re starting to see that cycle start a little bit. You see a variety of organizations — like xG health, for example — taking some products that Geisinger has written for in-house and trying to bring them out to the market. It’s starting. It will be really cool to see that happens over the next three or four years.

How will that impact your business? PatientKeeper has been connected to these systems for more than a decade and new entrants will then have the bar lowered to do the same.

We had to spend a tremendous amount of money building all these integrations, but we would just as soon not have to build them. We built them so that we could build the software that we expose to physicians and that they use.

We embrace it. We’ve implemented the FHIR standards on both ends of our application. Somebody can run FHIR on top of us. We can run using FHIR on top of something that is FHIR enabled.

We think openness is philosophically the way to go. That means if someone finds a better application than we have, well then, shame on us. Our job is to have the best applications, and if we don’t, then someone should buy one that is different from ours and have it work with ours that they do think are best.

That’s the way innovation works. That’s the way it works in the tech world. That creates a great ecosystem, an ecosystem that has all ships rising because it puts competitive pressure on everybody. I’m a huge fan, philosophically. I think it can do nothing but good things for us and for other vendors like us.

You just added imaging appropriate use criteria to your product. Are you seeing more interest in having point-of-care systems offer guidance, reminders, or other features that keep providers on the best practices track?

Hopefully it’s the tip of the iceberg. I believe the reason that we as a country spent $40-plus billion getting doctors onto electronic systems isn’t so that we can just get rid of paper, although that was nice. It’s so that we can take this next step of improving healthcare and making the computer an essential tool for physicians.

The analogy I like to use is if you go to most doctors today and say, "Would you write this order on paper instead of putting it into the computer?" Depending on what kind of computer they have, they might gladly say, "Yes, please give me that paper. I can’t wait to write it on paper." If we do our job right as informaticists and as healthcare IT providers, the answer to that should be, “No. I would never write it on paper, because that’s dangerous. I get so much good information and so much help from the computer to do my job that I would never consider practicing without the computer.”

We’re not there yet. PatientKeeper isn’t there. I don’t think anyone is there. But that is the ultimate test. Imaging criteria is one small step. As we start to deploy more advanced techniques, with all the big data analytics techniques, we’ll have computers that know everything about that patient that is all codified. 

The computers aren’t really helping the doctors that much. In some cases, the computer asks the doctor questions the computer knows about. Did you give aspirin to this patient? Well, yes, because I put the aspirin order in the system — why are you asking me? It’s even worse.

The next four, five, six years is going to be that renaissance, helping the physicians with what they do in a way that works for them. Interoperability is such a key to that because it’s going to require the entrepreneurial horsepower of an industry. It’s not going to be one company that solves that problem.

We’re seeing early steps in using little data, where instead of waiting years for big clinical studies to be completed, doctors are getting immediate data analysis from their own systems, such as, “If I have 10 patients in my database who are somewhat like this one, how many of them benefited from this treatment option I’m considering?” Is that concept ripe for development?

I am so excited about that concept. If you think about clinical trials the way they have existed to date, we have a molecule or we have a procedure or a hypothesis. We go out and recruit people, we do all kinds of stuff, and we see whether it works or not.

But every day, there are millions of clinical trials being done. Patients are seeing providers. Things are happening. Outcomes are happening. If we can learn from all of that, even in the smaller cohort, that here are patients like you and and let’s observe how they work. Here are different protocols.

Our parent company HCA has been doing clinical research essentially by just observing different practice patterns across their hospitals. They have done groundbreaking research around sepsis prevention and what things worked and what things didn’t work around preventing infection. Just by observing that there are three or four different ways people do this in terms of washing hands, prophylactic antibiotics, et cetera. They figured out which ones work better without a clinical trial — just by observing the data they have.

That is the future. It might even change the clinical trials industry. At some point you still have to come up with new molecules, but when you start getting into these practices and procedures and off-label use, there is a lot we can learn.

I haven’t heard much about the HCA acquisition since it was first announced. What has changed since?

Certainly the goal of the acquisition was to have exactly what you just described happen, which is business as usual for PatientKeeper from a customer perspective and from an organization perspective. I’m pleased to report that we have achieved that goal. We’re a year and a half in to the acquisition. I’ve talked to some of our customers and they didn’t even know we were acquired. That’s awesome.

The big thing that has changed, which our customers will start to notice over time, is that we’ve made some very big investments in our R&D organization and our hosting center operations. We now have a world-class hosting operation. We had a pretty good one before, but we have a much better one now.

That’s really the big change that we have made. We’ve accelerated R&D efforts and accelerated a variety of projects that we had on the back burner. We’re in the pipeline that we’ve now pulled forward. We haven’t gotten those out to the market yet, so if you are a customer of ours, you haven’t seen the benefits of that. But in the next six to 12 months, you’ll start to see those things hitting the release cycle.

Otherwise, it is just business as usual for us. We’re deploying our advanced clinical software throughout the HCA hospitals and having a great time continuing to go against our original vision.

Do you have any final thoughts?

We’re at the beginning of a new era in healthcare IT. Up until now, it’s been, get rid of paper, get stuff automated. We’ve mostly done that. I wouldn’t say we’re complete, but that phase is coming to an end, where you’re taking processes that have never been automated and automating them.

Now it really is about that next generation. If you think of the evolution of the Internet, we now have concepts like Facebook and EBay that were not possible on paper. They are new concepts. What we’re going to find is a whole new set of innovation in healthcare IT around concepts that were not possible until everybody is electronic. As a company, we’re excited to participate in that. We’re excited to see the ecosystem and the healthcare IT industry itself blossom as that occurs.