Investigative innovator ProPublica launched a new project this week, a national database of doctors and the money they have received from pharmaceutical companies. It teamed up with the Boston Globe, Consumer Reports, NPR, the Chicago Tribune, and PBS as investigative partners. To compile the Dollars for Docs database, web developer Dan Nyugen and reporters Charles Ornstein and Tracy Weber collected the information from seven major pharmaceutical companies who have posted their payroll information online, and cross-referenced the doctors’ names with other publicly available disciplinary records. The database is searchable by state or by doctor’s name, and there’s also a list of the 384 top-earning health providers who made at least $100,000 from pharmaceutical companies (in 2009 and early 2010).
The implied story, of course, is that checks from pharmaceutical companies—usually for speaking fees at pharma-sponsored conferences, but sometimes for travel and meals as well—would inappropriately influence doctors to prescribe those companies’ drugs at a higher rate than they normally would. Indeed, there have been some hefty lawsuits through the years involving some cases where a quid-pro-quo link could be clearly drawn.
But in general, that remains a tricky thing to investigate, as individual drug prescriptions are not public information. ProPublica is quick to point out throughout the Dollars for Docs web pages that it’s perfectly legal for doctors to receive payments from pharmaceutical companies, “but it does raise ethical issues.”
What ProPublica can show, however, is that the highest earning doctors on Big Pharma’s payroll are not necessarily the most reputable in the medical field:
Drug companies say they hire the most-respected doctors in their fields for the critical task of teaching about the benefits and risks of their drugs.
But an investigation by ProPublica uncovered hundreds of doctors on company payrolls who had been accused of professional misconduct, were disciplined by state boards or lacked credentials as researchers or specialists.
(The details of some of those disciplinary cases will make you cringe.)
This project is a huge undertaking, which some have criticized for being too big, and others for being too limited (see the debate going on over at Knight Science Journalism Tracker). Regardless of the conclusions ProPublica and its partners have been able to tease out of the data thus far, the database alone is an important resource; it’s a great opportunity for reporters and editors to localize an aspect of health care reform that’s often merely abstract. For example, Judy Graham from the Chicago Tribune (one of the media partners) wrote a piece listing the twenty-five highest earning doctors in Illinois, where they work, what kinds of drugs they get paid to present on, and what they had to say when Graham asked them about it.
In a conference call on Thursday, ProPublica reporters Ornstein and Weber spoke to reporters and editors from dozens of news organizations about how to best use the database to find stories their readers would care about. Weber stressed several important caveats. First, with such a large database, there are bound to be doctors who share the same name, so it’s important to double-check before publishing anything about a particular person. For instance, Weber said that while researching disciplinary records, they found seven doctors with the same name in Manhattan alone.
Second, the database is in its beginning stages, so one can’t take the doctor payments at face value: some drug companies reported on the last eighteen months of payments, others only reported on the last six months, and only seven drug companies have thus far disclosed anything at all. So just because a doctor isn’t listed in the database, doesn’t mean that he or she hasn’t received payments from drug companies, and those doctors who are listed may have received way more than the amount listed. (By 2013, as a result of the health care bill recently passed by Congress, all U.S. pharmaceutical companies will have to disclose this information, and ProPublica will continue updating the database to reflect that information.)
It sure would be helpful to be able to search by city.
#1 Posted by Anna Haynes, CJR on Fri 22 Oct 2010 at 09:59 PM
Is there a way to export the data, for a state?
#2 Posted by Anna Haynes, CJR on Fri 22 Oct 2010 at 10:02 PM
And searching still has bugs - I used William D. Leak of Ohio as a test case, and got different results with different variants of the search.
"Leak" returned William D. Leak and W. David Leak
"Leak, William" returned just William D. Leak
"Leak, W.D." returned nothing
"Leak, W." returned just W. David Leak
FYI, it doesn't seem to matter whether you search "firstname lastname" or "lastname firstname" - and it strips any comma out.
#3 Posted by Anna Haynes, CJR on Fri 22 Oct 2010 at 10:10 PM
ProPublica should do some fact checking themselves before blindly accepting the data provided by the drug companies. When incorrect, such claims will damage a physician's reputation and harm his/her practice. This is what happened to me when ProPublica made the false claim that I received $90,000 from Pfizer. In fact, the money went to fund a trial at a not-for-profit hospital. This trial was conducted by a pharmacist who had a name similar to mine. It took hours of phone calls, emails, and legal billing before ProPublica grudgingly removed my name from their database.
#4 Posted by MD Baker, CJR on Fri 22 Oct 2010 at 11:12 PM
Anna and MD Baker, thanks for these comments. You both speak to the flaws and limitations inherent in this database, which the ProPublica reporters were the first to acknowledge in last week's conference call. They explained that they decided not to use city info because different drug companies sometimes wrote cities differently, and the inconsistencies could cause search problems (the example they used was St. Louis, St Louis, Saint Louis). I know it sounds weird, but I think they just wanted to be as inclusive as possible, and some shortcuts were necessary in putting together that much data.
But I shouldn't speak for them - I'm sure the ProPublica folks would be able to answer your questions much better than I can. For instance, I don't know the answer to the question about exporting the data, sorry, but I bet if you called them they could walk you through it.
As for MD Baker's problem, that's very unfortunate. I think they've heard about a number of problems like that, as they stressed over and over again to reporters that they should check and double-check all doctors' names before making accusations in the press. This database is merely a resource, a starting point for much more thorough reporting. One could argue that it's irresponsible to put that information out there in the first place (as MD Baker would argue, surely), but keep in mind that ProPublica is only re-publishing raw data that is already publicly available on the drug companies' websites, and compiling it into a more searchable database. Unlike many of their other investigations, they didn't actually "uncover" anything here; it was a feat of data compilation.
#5 Posted by Lauren Kirchner, CJR on Mon 25 Oct 2010 at 10:29 AM
Lauren, I contend that ProPublica was doing more than simply "re-publishing raw data."
Here's what I mean: On Pfizer's website, I was not listed as the payee or recipient of $90,000 from Pfizer. It was made clear that the non-profit hospital to which I am affiliated received that money. But on ProPublica's website/database, I was listed as the recipient of the funds. And, to make their point, ProPublica put the number in large, green print. A patient who searched my name on ProPubica's website would think that I had pocketed $90K from a big drug company.
Of course, as I mentioned in my previous post, I was not the recipient of any funds from Pfizer nor did the hospital receive funds in my name. It was a case of mistaken identity. A pharmacist with a name almost identical to mine was responsible for the study. Clearly, ProPublica tweaked the data for its own purposes. They wanted to catch the bastards, to bring to light the unseemly relationship between doctors and Big Pharma. But I am innocent and was unfairly sullied by their investigative report. Oddly enough, Pfizer's database, while incorrect, was less damaging to my reputation.
So, to clear my name I had to fight a strong, journalistic entity. To me, that's ironic given ProPublica's mission to shine "a light on exploitation of the weak by the strong." Perhaps they don't know their own strength.
~MD Baker
#6 Posted by MD Baker, CJR on Mon 25 Oct 2010 at 10:27 PM
I see what you mean, MD. Why do you think they were reluctant to take your name down? Did they not believe you at first?
#7 Posted by Lauren Kirchner, CJR on Thu 28 Oct 2010 at 09:42 AM
Yes, they did not believe me. They wanted proof from Pfizer. Meanwhile, Pfizer told me that it would take twenty days for them to to complete their investigation into the matter. So I was very frustrated. I felt as if I was a small, powerless individual caught between a huge journalistic giant on one side and a strong pharmaceutical giant on the other.
In the end, the hospital with which I am affiliated called ProPublica and managed to convince them to remove my name from the database.
Pfizer has since sent me an email with an apology. It took less than twenty days.
#8 Posted by MD Baker, CJR on Thu 28 Oct 2010 at 12:41 PM
This is Dan Nguyen from ProPublica, the reporter who compiled the data from the varied reports, including Pfizer's. As the intent of this project was to create an easily searchable database, I would be the person, second to Dr. Baker, most frustrated by this kind of discrepancy. However, the entity responsible for putting his name online is Pfizer. I copied his name, among thousands, from what Pfizer reports in their official disclosure mandated by their settlement of fraud allegations.
I was not involved in making the multiple calls to the various parties involved, but from my vantage point, Dr. Baker's complaint was handled with the upmost priority. I know that as soon as it was confirmed to us that Pfizer's database was incorrect, I immediately made the change to our database.
There were other doctors who had questions about their inclusion in the database. But, as in Dr. Baker's case, these weren't cases in which our list deviated from the companies. It was either the companies' words or the doctors', and we, as is the case for any outside party, can't make an a priori judgment on who is correct.
Discrepancies between what physicians and companies reported are what spurred transparency in this area in the first place. U.S. Sen. Charles Grassley (R-Iowa), who led the Sunshine provision efforts, found doctors whose reports were far more incomplete than the companies'.
http://www.nytimes.com/2008/10/04/health/policy/04drug.html
So without having direct access to company accounting records, outside observers who aren't inclined to take a side can't make an automatic judgment on who is in err.
Pfizer's database (and thus, Dr. Baker's entry in Pfizer's listing) has been up since April, and was a result of Pfizer's $2.3 billion settlement of illegal marketing claims in Sept. 2009. I don't expect Dr. Baker to owe us any gratitude for our part in revealing flaws and inaccuracies with Pfizer's database six months later. But his case only illustrates the unintended consequences of companies making this process less than transparent (i.e. withholding identifiers more specific than first/last name, and the subject area of the service. And of course, making the information impossible to aggregate without programming expertise), as it creates the potential for misunderstandings.
For us to make a publicly available database and then, as Dr. Baker alleges, to "tweak" it, would be counter to the purpose of the project, which is to make accessible what drug companies have declined to do beyond a nominal implementation.
#9 Posted by Dan Nguyen, CJR on Fri 3 Dec 2010 at 01:22 AM
Mr. Nguyen misses my point regarding the tweaking of data. Yes, Pfizer's data base was incorrect. Pfizer incorrectly listed me as the principal investigator for a Pfizer-funded study. But at least Pfizer made it clear that I had not personally pocketed $90,000. Rather, a non-profit hospital received the money to fund clinical research. But on ProPublica's website, I was listed as the direct recipient of $90,000 from Pfizer. So, regardless of the goal of the project, ProPublica took Pfizer's flawed data and made it even more damaging to me and my practice.
I find ludicrous the implication that I might be grateful to ProPublica or to Mr. Nguyen for what he did. . While I have no philosophical quarrel with ProPublica's crusade against big pharma, I find it unconscionable to slander my name in the process. By the way, as far as I could tell, ProPublica made little effort to correct the problem. Any phone calls and emails to Pfizer or to the department of research at the hospital involved were made by me. Yes, many emails were exchanged between me and officials at ProPublica but their tone was more obstructionist than helpful.
#10 Posted by MD Baker, CJR on Sat 4 Dec 2010 at 03:23 PM
Dr. Baker, you don't owe me any gratitude, that was not my implication. As someone who is interested in accurate disclosure, I'm grateful to anyone who helps resolve discrepancies in these records.
To clarify, the listing you're referring to was in the widget view of our main database, which allowed users from other sites to access the search function and see a list of names and payments. The results directly linked to the corresponding records on our main site, which display the company payment descriptions verbatim, including payee and purpose of the payment. Research records were included in our database because they were part of Pfizer's disclosure requirements. But the primary focus of the stories was on physician payments for speaking and consulting, not sponsored research.
Regarding the issue of Pfizer's record listing an incorrect Dr. Baker, like I said, there was no interest in delaying the appropriate fix. But Pfizer ultimately has the payroll records. You said yourself Pfizer gave you a 20-day estimate to investigate the issue. From my vantage point, upon Pfizer's error being confirmed, making the change to our database was not a "It can wait until after lunch" matter, nor did we wait for Pfizer to reflect the change on their official database.
I don't fault you for venting, if this situation was frustrating to me, it was obviously more frustrating to you. I want to respond and clear up what I can, but you're free to feel however you want in the end. I'll continue to assert that I've never felt your concerns were considered trivial, Nor do I think that any of the parties you feel were arrayed against you in this, including Pfizer, gain from purveying an inaccurate record.
#11 Posted by Dan Nguyen, CJR on Mon 6 Dec 2010 at 01:15 AM