9 Comments

Summary:

The average doctor has likely never heard of Fred Trotter, but he has some provocative ideas about using physician data to change how healthcare gets delivered.

Fred Trotter image

Fred Trotter calls himself a “hackivist” — but not the kind that assists fugitives or spills national security secrets online.

His mission isn’t to expose highly classified documents — it’s to turn hard-to-reach, poorly organized or overlooked public data into meaningful maps that the littler guys in healthcare can use to improve quality and lower costs.

Over the years, Trotter has built a reputation for advocating open-source software in healthcare and finding interesting applications for seemingly mundane datasets. But his most recent round of internet fame is the result of a successful crowdfunding project called DocGraph that used Medicare data to show how physicians team up to treat patients.

That kind of information could not only give patients a sense of which doctors are most respected by their peers, it could help hospitals understand the referral patterns associated with poor care coordination, and provide health IT startups with a map of the most plugged-in doctors in each city.

Todd Park, the chief technology officer in Obama administration, was previously the CTO for the Department of Health and Human Services.

Todd Park, the chief technology officer in Obama administration, was previously the CTO for the Department of Health and Human Services.

The idea for DocGraph goes back to a conference a couple of years ago, when Trotter ran into fellow health data geek Todd Park, the co-founder of electronic health records company athenahealth and now CTO in the Obama Administration. Trotter was just starting to learn about the kinds of data that could be made available from a Freedom of Information Act request, and he asked if he could FOIA his way to data showing when two physicians billed Medicare for the same patient around the same time.

When Park said it should work, he decided to give it a shot: “I said, ‘Why the hell not?”‘

Launching a ‘not only for profit”

It took more than year of back and forth with various FOIA officers in different states – not to mention an 11th-hour assist from Park, who had assumed his powerful role at the White House by then – but Trotter eventually got his data.

DocGraph, in turn, has led to the creation of his two new businesses. In the months since his first crowdfunding campaign, Trotter has created DocGraph Journal, a site that intends to uncover “opaque” health data — data that’s technically open to anyone but difficult to dig out of bureaucratic public agencies — and release it as open datasets for a small fee to startups, researchers and others in health IT.

He also recently launched Not Only Dev, a “not only for profit” startup that helps Accountable Care Organizations (ACOs) make management decisions. (ACOs are groups of hospitals, doctors and other healthcare providers that are financially incentivized to cut costs and improve outcomes instead of billing for more medical procedures.)

“We’re the first open-source data vendor in healthcare,” Trotter said. “Inside the hallowed halls of big pharma and insurance companies, these patterns have been known for years. What’s new is for this information to be open and for patients and researchers to get convenient and inexpensive access to this information.”

Data scientist Ryan Weald visualized the geographic connections between physicians using DocGraph data.

Data scientist Ryan Weald visualized the geographic connections between physicians using DocGraph data.

Mapping the relationships in healthcare

Much like Facebook’s social graph maps the connections between friends, Trotter’s DocGraph includes 50 million connections showing the overlapping relationships between doctors. “It’s a map of how healthcare in this country is delivered,” he said. Beyond highlighting how doctors are teaming up to deliver care, he believes it provides the blueprint for a comprehensive ranking system that could give patients transparency on their doctors’ history, outcomes and assessment by peers.

Even though services across the web already offer doctor rankings – from U.S. News to Healthgrades and HealthTap – (unless they use Trotter’s data) they mostly rely on patient feedback and surveys, not organic patterns and objective data.

Getting into the family business

In a sense, Trotter is putting a new spin on the family business. Decades ago, his grandfather started a business that provided doctors with technology to manage their practices, but Trotter began his career working in the more lucrative field of internet security. After his mother died of ovarian cancer, his aunt and uncle invited him to join the company, and he said he decided to move into an industry with more meaning.

Interestingly, even though he technically does data science, he resists the label “data scientist,” arguing that more formally trained data scientists have more tricks up their sleeves.  His big advantage, he suggests, is in his intimate understanding of who the health IT players are and how the data flows between them — and how he can manipulate that information.

Despite his growing reputation in health IT, the average doctor has likely never heard his name. But his work could well help change how they do business.

What does big data mean for doctors?

Consumer Reports and other groups provide detailed reports on the hospitals that carry out the best surgeries or offer the best care. But Trotter wants to amass as much data as possible to drill down to the doctor level – to understand the different approaches doctors are taking and the kinds of outcomes they produce.

It’s the rare healthcare system that, like Kaiser Permanent or the Department of Veterans Affairs, is able to take all kinds of physicians and deliver healthcare that works. Most others offer error-prone care, while some offer top-quality care but only by assembling all-star teams of doctors, he said. With the right data, Trotter explained, hospitals could scale quality healthcare by identifying doctors’ strengths and matching them with the most complementary peers.

moneyballThe end of ‘doctor knows best’

Instead of taking the New York Yankees’ approach and picking the best players, “we can play ‘Moneyball’ in healthcare,” he said. “We can create big data systems that create teams of doctors where not every physician is brilliant at everything, but the team as a whole does very, very well.”

Of course, doctors may not like their big data-driven future. As Trotter sees it, after being optimized “Moneyball-style,” they’ll be enhanced with digital tools, and then charged with merely monitoring those tools. Eventually, they’ll start to be phased out.

“It will take years to fully replace them, but I think it is going to be like the cowboy. That is a profession that still exists but with far fewer people doing very different things than they did before. We are reaching the end of the era of the notion that ‘doctors know best,’” he said. “I warn every doctor I know that if they want to have a job in 20 years, they had better learn to program.”

  1. Thank you for making me sound so smart!

    -FT

    Share
  2. HappyHealth.me Friday, August 9, 2013

    The internet is helping to inform and empower patients and healthcare consumers in all sorts of ways, which is why we created HappyHealth to help consumers store their health data, manage it, and auto-fill those annoying doctor’s office registration forms. Check us out, and let your health data start empowering you today!

    Share
  3. HappyHealth.me Friday, August 9, 2013

    The internet is helping to revolutionize healthcare in all sorts of ways. Big data and complex health analytics will help make patients healthier, make physicians more productive, and help insurers reduce costs.

    We created HappyHealth to help consumers store their health data, manage it, and auto-fill those annoying doctor’s office registration forms. Check us out, and let your health data start empowering you today!

    Share
  4. Great post on a valuable cause. Thanks @kheussner for covering & @fredtrotter for offering great data for companies like BetterDoctor.

    Share
  5. “It will take years to fully replace them, but I think it is going to be like the cowboy. That is a profession that still exists but with far fewer people doing very different things than they did before. We are reaching the end of the era of the notion that ‘doctors know best,’” he said. “I warn every doctor I know that if they want to have a job in 20 years, they had better learn to program.”

    These comments show an ignorance to the complexity of human health.

    The ‘data’ can never show the intricacy of an ‘outcome’, nor assess the complexity of patient/physician/healthcare interaction and ones health.

    Share
    1. I think your point would be better served if you used the word “intimacy” rather than “complexity”.

      Note that I did not say that “nurses are going to be out of a job” I do believe that healthcare required intimacy, but I do not go to the doctor to get “intimacy”. I go to get the right answer.

      Eventually, computers will be much smart than people and better able to handle any kind of complexity. Believe me, I do not at all underestimate the level of the complexity involved in medicine. I suspect that medicine will be one of the problems that holds out close to, or even slightly beyond, the singularity.

      I guess you could say that I believe that you are ignorant of the “simplicity of complexity”. At least that makes for a catchy retort!!

      Share
  6. “I warn every doctor I know that if they want to have a job in 20 years, they had better learn to program.”

    Not too sure about that. Well integrated health systems like Kaiser are on primary care physician hiring sprees.

    We have a pretty darn good algorithm for determining the need for antibiotics in strep throat, but inappropriate antibiotic use in viral URIs is rampant. Data (clinical studies) drove the creation of a clinical support tool for strep, and people are still getting appropriate care. And this is for the most common of conditions. Hmm.. I wonder why that is.

    Share
    1. Actually, meant to say “people are still getting inappropriate care”.

      Share
  7. The sentiment that docs will be replaced is intentionally provocative and directionally correct but far from certain–and certainly further than 20 years off. Yes, nurses add intimacy to the equation but their licensure doesn’t allow them to take ultimate responsibility for the patient under most circumstances. And that is where futurist technologists like Mr. Trotter tend to slip up–they forget that technology is always ahead of culture and regulation. I believe that it will be a long time before people want to get treated by an algorithm. Algorithms have hard time setting a fracture, taking out a gallbladder, or dealing with a mourning family. I should add, btw, that I am a physician that stopped practicing full time over 10 years ago to enter the world of health IT/data analytics, so I am on his side. And I am a pariah at my med school reunions.

    This said, I think his work is fantastic and keep at it. This work on transparency is essential and tremendously worthwhile.

    Share

Comments have been disabled for this post