Why health information policymakers should read “Moneyball”


Recently, I happened to pick up and read a chapter in the middle of Michael Lewis’ great book Moneyball, which I enjoyed even more than the film, although I liked the film as well.

In recounting the history of collection and use of baseball statistical data, Lewis described what I consider to be a most interesting parallel between the history of baseball statistical data and the history of healthcare data. In both cases, the statistics that should have been freely available to the broader public or even to the individuals whose data was being collected were withheld by those who collected the data. In both cases, the hoarding of the data was detrimental to the broader public interest. In the case of baseball data, a populist movement created a solution that was forced on the data hoarders. With respect to healthcare, we have not yet gotten to that point.

The relevant history of baseball statistical data

Michael Lewis wrote about Bill James, a fanatical amateur baseball fan who wanted to collect baseball statistical data and to implement different ways of analyzing and using it. Just as those of us who try to get our patient data from doctors and hospitals hit proverbial brick walls, James hit a brick wall called the Elias Sports Bureau, the official collector of Major League Baseball statistics.

Oddly enough, James was not trying to threaten Major League Baseball, but to provide it with analytical tools to improve the quality of the game for everyone, especially the fans. However, the idea that someone outside the baseball fraternity of managers, coaches, scouts, general managers and player agents should have unrestricted access to baseball statistical data collected from the official scorers of each game was deeply threatening. James did not give up. He found a way to collect the data outside of the official scoring system.

James realized that in every Major League baseball stadium there are season ticket holders who not only attend every game, but do their own scoresheets. He created a data collection system called Project Scoresheet, and built his own database of what happened at every Major League during the seasons with which he operated Project Scoresheet.

He then published an annual collection of data-driven analytical essays in books called Baseball Abstracts, which, I might note, are both out of print and impossible to buy on Amazon.com’s secondary book market. I have the 1985-1988 Baseball Abstracts and have no intention of ever selling them because they are as insightful today as they were when I first read them.

As Lewis recounts, eventually, the combination of Rotisserie League marketing opportunities (“Rotisserie leagues are groupings of people who “select” players and track their performance over a season, with the winners being those whose players perform the best), the salary negotiations between agents and teams, the financial disincentives for overpaying for players, and the voracious appetite for statistical data from consumers paying increasingly greater amounts for season and individual game tickets liberated the data that had been hoarded by Elias.

Today, Yahoo, mlb.com, ESPN and other web sites make it far easier for someone who wants to collect data to do so. Moreover, Major League Baseball is installing cameras in every stadium that will enable it to collect and analyze data that is not able to be captured through the official scoring system. For example, one of the challenges in evaluating players since the beginning of baseball history is how to determine the quality of defensive play. So many variables enter into the defensive skills of catching, fielding and throwing that official scorers could not capture them, primarily because they were particularly difficult to capture from the naked eye.

For example, a great defensive player has to master positioning, the best path to catching the ball, the best fielding position, the decision on where to throw the ball, how to make an accurate throw (preferably on the fly), and how to coordinate with the other players involved in a play. Traditional official scoring only captured whether the individual made a play that resulted in an out, or made some error in failing to catch a ball or in throwing it inaccurately. The whole subject of inquiry relative to whether the player could have gotten to the ball was incapable of being addressed by traditional scoring. The new cameras will address this question, as well as providing better data on a player’s offensive skills. I do not believe this whole line of inquiry would have happened without the explosive growth in statistical analysis that Bill James’ work triggered.

We must go through the same process in collecting health and healthcare data and in making intelligent use of it.

The equivalent of the Elias Sports Bureau and Major League Baseball relative to health information on individuals and population are the healthcare providers (e.g. doctors and hospitals) and insurance carriers and other payers (e.g. health insurance plan administrators). They withhold data, but, worse than that, they make no use of it for intelligent individual and population health management.

Dossia, the company I run, is in the business of liberating health and healthcare data from those who hoard it. However, it is also in the business of finding the best ways to use individual and population level data to make people healthier and to make them better consumers of healthcare when they need to access the healthcare system. Let’s compare the state of play in the vital healthcare data we need to what the baseball statistical system would be like (or, for that matter, any sports statistical system) if it operated like the healthcare information system does today:

  • Imagine that each player could get access only to his statistics on a portal provided by his team, but that data were not available to him on anything other than on that portal. In effect, he could view the data, but not download it or easily share it with anyone else, including his agent. That is the way our healthcare data gets communicated to us today by the vast majority of providers and health plan administrators.
  • Imagine that there was no system for collecting population-level statistics available to the public, but only to the Elias Sports Bureau, and imagine further that the Elias Sports Bureau was one of multiple statisticians and that it only collected data on one of the six divisions of Major League Baseball. Imagine that, to collect overall statistical data, we had to go to six organizations, not just the Elias Sports Bureau. That is the way our healthcare system data is fragmented and siloed. No entity has all the relevant population level data to address major trends, including potential flu or infectious disease pandemics.
  • Imagine that the great games of the past that we enjoy seeing on ESPN Classic or MLB Network were wiped out after seven years and all records of those games disappeared. Our healthcare system does that with diagnostic imaging data stored in many healthcare systems if the patient does not demand access to the records.
  • Imagine if Major League Baseball captured all this wonderful data from its newly-installed on-field cameras, but it was not used for anything or communicated to anyone other than the players shown on camera. That is what the healthcare system does with daily activity and biometric data that is captured outside a particular clinician’s system. If you or I are tracking our sleep, exercise patterns, or nutritional decisions, not only do the clinical electronic medical record systems not capture this data, they block it from being imported into the system.

When are elected officials and regulators going to do their jobs in forcing the data hoarders to serve the interests of the people they are supposed to be serving: patients and consumers?

Too many of us accept this idiocy I have described above as the natural order of things. We clearly need more people like Bill James to embarrass and harass those who collect and hoard our health and healthcare data today.

The object of the hoarded or excluded data is not simply to enable doctors to do a better job treating us and to be more financially successful, although that is one of its core goals. The most important goal of this data is to get us engaged and committed to a successful effort to maximize our health and wellbeing.

We should not settle for anything less!