Robert Wachter, MD, noted patient safety expert and the academic leader of the hospitalist movement, has written his sixth book, The Digital Doctor: Hope, Hype, Harm at the Dawn of Medicine’s Computer Age. The tome chronicles the evolution of healthcare’s digitization and the unintended consequences of the changes. Wachter, chief of the Division of Hospital Medicine and chief of the medical service at University of California-San Francisco Medical Center, interviewed more than 100 people during the course of writing the book to track the impact of going electronic.

CIT: Did you set out to chronicle this point in time in medicine?

RW: I realized in retrospect that my practice of medicine was being transformed into a new thing, shifting from an analog to a digital business. Nobody had written about why it’s so hard and what that means. We had hoped for years that computers would fix all these problems but now that we have computers they’re clunky and we’re not looking at our patients anymore. Computers are fixing some errors but creating new ones.

You talked to a lot of big names for this book—John Halamka, Abraham Verghese, Carl Dvorak, David Brailer. How did you get them to speak so honestly?

Nobody I asked was not open to it. I was really, really pleased with how open people were and the access I was given. I’ve been around the medical block for a long time. The people I called had a sense I had some credibility. I wound up spending a day at Epic with about eight senior leaders. The president sat with just me for three hours. That reflects the idea that they wanted their point of view expressed.

I had to then come up with a narrative that’s fair but not overly cow-towing to them. There are no villains here but not everybody comes off looking spectacularly great. There is certainly a narrative out there that Epic is the embodiment of all that is wrong with health IT. I came to believe that’s mostly wrong. Epic won the game because they produced the best product at a time when there was federal support for people to buy a product. People have said I am too easy on them but there’s about a fifth of the book about an error at my organization in part because we were using an Epic system that in this circumstance didn’t serve us well.

Speaking of that error, you write that the 40-fold overdose a pediatric patient received sparked the book but early on, you write about physicians who tried to offer suggestions to improve their EHR but were told to stop harassing the IT staff.

Everyone is trying to do the best they can. If you work in IT and have a punch list of 100 fixes, of course it’s frustrating if somebody’s calling for a fix of this or that and oftentimes the answer is that the person didn’t go to the training or their computer is unplugged. In some ways, however, the job is to be open to feedback from frontline users.

I’m very proud of my own institution. This case of the overdose was brilliant in the way that it brought forth issues in IT, robots, alerts, overtrust in automation and more. But, you can imagine my first conversation with our head of risk management. This was not an easy conversation. To her credit and to the organization’s credit, they all realized that being public about this would probably save some people’s lives. But, it’s still tough to do. An outsider never would have heard about the case so we were in a unique position where people gave me extraordinary access to facts and events and also how they were thinking. They trusted me to handle it fairly.

You write about your support for some kind of national patient safety center. Do you think that kind of transparency and discussion would have a significant impact?

One of the things I discovered is that when you see an error at your own institution and say ‘how could we be that dumb?’ chances are that that same error is happening in other locations. We don’t really have a forum to talk about it. When something happens on your watch you think it’s very particular but it’s only through national or regional conversations that we realize this is a deeper and more complex issue. If everyone thinks it’s only them experiencing that problem, then the vendor may never hear about it.

Having a national clearinghouse for these things provides opportunities for centralized learning and for that learning to influence policy in ways that don’t naturally happen.

Part of the challenge is the fusion of responsibility and authority. There’s a lot of contrast with the aviation industry which has a regulator in the FAA. Boeing’s flight engineers developed ideas for cockpit design with a huge amount of input from users. They’re smart enough to know that once they’ve built it, they probably still haven’t gotten it right. They need to watch people use it in real life and then fix it and hone it until it works. In health IT, we don’t do anything that even remotely resembles that. We’re trying a little bit now and vendors are better than they used to be but they’re still very, very divorced between the product and the way this really works in the real world. Anything we can do to bridge that gap is important. A federal center where glitches get reported and disseminated is one of many maneuvers that can help bridge that gap.

We keep comparing healthcare to aviation. Why hasn’t healthcare adopted more from aviation yet?

Patient safety kind of invented that comparison by saying in the 1999 Institute of Medicine study on deaths from medical errors that it was equivalent to a jumbo jet crashing every day. I still use those examples but I’m careful about it because it is so different. To change the culture in the cockpit you’re taking two people both of similar training, language, social status and education and building in technology living in one standardized place. They have to learn once, basically. And, if weather conditions are bad, they just don’t take off. Every part of the analogy falls apart pretty quickly when you try to transfer it to a clinic or hospital.

During my time with Boeing, I saw that there are things they are able to do in the design of the cockpit that we’ll never be able to completely do. Epic can’t build a hospital but they can do a better job of bringing users into the design and having their ear open after implementation to figure out what’s working and what’s not.

It’s not just tactical and procedural. It’s actually cultural. There is this deep reverence for the experience and wisdom of the pilot because they learned it in blood. People died when the systems didn’t work well. That’s language I never heard from any of the IT vendors because they’re not connected to clinical people in a way that works and they don’t have deep respect for the wisdom they could gain. I think they need to. We can’t make the systems as good as they need to be if we don’t work together more closely than we have. I’m sure that’s on physicians’ end as well. We’ve asked vendors to do an impossible job in designing systems that meet our needs for physicians, billing, quality, risk, et cetera. We may have to fix all of that to make systems better.

So many physicians seem very unhappy. Will we eventually run out of doctors because so many leave the profession or never enter it?

Physicians have been moaning about all this for at least five to 10 years but med school applications have never been higher. I don’t want to diminish the moaning. Some of it is legitimate. It’s a very hard time to be a physician. We’re in a middle ground where the technology still is not very good yet the requirements on physicians are as if the technology is great. Measuring and improving are reasonable things to ask us to do and things we’ll be able to do much better in the future when the technology is really good.

I wish we didn’t have to go through this phase. It can be difficult but I can’t see how to get to a better place without going through this phase.

Some people want to slow things down and not have the pressure to improve quality, safety or efficiency, whether through payment changes, public reporting or accreditation changes. But, I’m pretty sure if we slowed down it wouldn’t happen. Some people say HITECH was a bad idea because the systems were not ready for prime time and we should have waited. I think without HITECH it never would have happened. The only way systems get better is if 70 percent of the market is using them rather than 10 percent. At that market share, competitors come into the market to do healthcare instead of a new version of Facebook.

If you were to write a sequel in a few years, what would you hope to be able to say?

I tried to capture the big picture of what it means to be a doctor, a patient, a vendor, a policymaker in this environment. When I asked people what do you think this looks like when the dust settles, once we agreed not to come up with a specific number of years for that, it was amazing how monolithic the answer was. All the technologies are available today. This doesn’t require any brand-new technological breakthrough. It requires a set of incentives and maybe some training that allows us to use IT that is the most productive for patients, delivery organizations and physicians. A lot of the challenges today are transitional pains. We didn’t think of this as adaptive change—the people portion and the idea of complementary innovation. Every industry that goes digital takes about 10 years to see the advantages—not because the technology is so much better but because people reengineered relationships. When that happens, I think we’re going to see something pretty magical.