1. Provide the title of your story or series and the names of the journalists involved.
"Revealed: Pfizer's Payments To Censured Doctors" by Peter Aldhous, Jim Giles and Brad Stenger,
2. List date(s) this work was published or aired.
First published online: April 22, 2010; Abridged version published in print: May 1, 2010,
3. Provide a brief synopsis of the story or stories, including any significant findings.
They are billed as "healthcare professionals who spend years building expertise in their fields." Using materials grounded in science, they educate their peers in the risks and benefits of drugs. This is how Pfizer, the pharmaceuticals giant, describes the experts it hires to lead forums in which doctors are lectured on the use of its products. Yet we found that some of Pfizer's experts have been disciplined by state medical boards for deficiencies in patient care, while others have been reprimanded by the US Food and Drug Administration for how they conducted drug research trials. Our findings added to a growing controversy surrounding the pharmaceutical industry's efforts to market drugs by influencing patterns of prescribing.
4. Explain types of documents, data or Internet resources used. Were FOI or public records act requests required? How did this affect the work?
As part of a $2.3-billion deal with the federal government to settle charges of illegal drug marketing, Pfizer was obliged to release details of its payments to physicians. Initial data covering the second half of 2009 were released on a searchable website on March 31, 2010. Stenger wrote a Perl script to scrape the data in its entirety from this website. Aldhous had already obtained medical license and disciplinary data through public records requests to the state medical boards of California, New York and Texas. He was able to download data from Florida, giving comprehensive data on physician discipline from the four largest U.S. states. Aldhous also collated data on Food and Drug Administration warning letters to clinical investigators. Aldhous then loaded all of the data into Microsoft Access, and ran queries to match Pfizer payment data to disciplinary records using first and last name. Aldhous then went through all of the matches thrown up by Access, using physician license look-up sites maintained by each of the four states California, New York (another link ), Texas and Florida, as well as other online reference sources such as www.healthgrades.com to verify in each case whether or not a match represented the same individual. Where the matches were genuine, Aldhous downloaded records of the doctors' discipline from the state physician license lookup websites and warning letters from the FDA website. Finally, Aldhous and Giles conducted interviews with the disciplined doctors and other sources to complete the story.
5. Explain types of human sources used.
Doctors whose disciplinary records we investigated, Pfizer officials, patient advocates and academic researchers studying doctors' prescribing habits and conflicts of interest in medicine.
6. Results (if any).
While New Scientist was not able to devote the resources to continue this line of investigation, we are delighted that ProPublica rose to this challenge with its subsequent Dollars for Docs series. The opening story in that series extended our strategy of relating drug company payments to physician disciplinary data to six other companies that have disclosed payment data.
7. Follow-up (if any). Have you run a correction or clarification on the report or has anyone come forward to challenge its accuracy? If so, please explain.
There have been no challenges to the accuracy of our reporting.
8. Advice to other journalists planning a similar story or project.
Put in public records requests for potentially useful data, such as state medical board records on licensed physicians, as a matter of course. Look for opportunities to use these data in your reporting, and respond quickly when these opportunities arise. Familiarize yourself with techniques of computer-assisted reporting, so that you are comfortable searching, summarizing and matching across large data sets.