DJA nominee of the day: Methadone and the Politics of Pain
The nominees for the 2012 edition of the Data Journalism Awards (DJA) were announced on 27 April at the International Journalism Festival in Perugia, Italy. In this series of posts we are featuring the 57 nominated projects one by one in order to tell the story behind each project. Every day we are showcasing a different project from the six categories of the competition.
Graphics and charts featured on the main page of the Seattle Times' project
Methadone and the Politics of Pain is a three-part series published in December 2011 in The Seattle Times. It has been nominated in the DJA category Data-Driven Investigation (local/regional). The project took ten months and a team of five to complete. The major findings are articulated in three points:
Part 1: “State pushes drug that saves money, costs lives”
Since 2003, at least 2,173 people in Washington have fatally overdosed on methadone, a narcotic painkiller that is both cheap and unpredictable. Washington steers people with state-subsidized health care – most notably, Medicaid patients – toward the drug in order to save money. In terms of the drug’s toll, the poor have been hit the hardest. Medicaid recipients account for about 8% of Washington’s adult population but 48% of the methadone deaths. Year after year, a committee of state-appointed medical experts has sanctioned methadone’s use, but the committee’s deliberations have been skewed by misinformation from state officials who have a financial stake in methadone’s selection as a preferred drug.
Part 2: “New law leaves patients in pain”
Washington’s new pain-management law – lauded in some circles as a possible model for other states to emulate – has become a case study in unintended consequences. Passed with the intention of curbing overdose deaths, the law has driven so many medical providers from the field of pain management that many patients now struggle to find help for their chronic pain. This story chronicles the struggles of Charles Passantino, a patient with a crippling liver disease who went months without his modest prescription of oxycodone while his wife worked frantically to get him help – calling doctor after doctor, writing to the governor, pleading with state officials.
Part 3: “In pain clinic’s wake: doubts, chaos, deaths”
A pain clinic in southwestern Washington managed to stay open for several years while prescribing potent painkillers in often extraordinary amounts. At least six people prescribed narcotics at the clinic overdosed and died. The state’s anemic response to the more than 100 complaints that rolled in against the clinic’s staff highlighted the medical community’s bitter debate about pain management and the murky landscape that regulators confront when dealing with allegations of overprescribing.
Visualisation is key: plotting the demographics of methadone-related deaths
The data team behind the project includes Ken Armstrong (staff reporter), Michael Berens (staff reporter), James Neff (editor), Justin Mayo (data specialist) and Danny Gawlowski (multimedia producer). When asked about the project's aims the team wrote that: "Among other things, we wanted to show how an arcane area of policy can have such a dramatic impact on a community. When covering the politics of health care, the media tend to gather in high-profile settings – legislators engaged in floor debate, the governor at a microphone. But often it’s the obscure panel convening in some hotel conference room that warrants greater scrutiny."
What were the tools and data used in this investigation?
Through requests filed under the state’s Public Records Act, the project team was able to obtain three databases from the Washington State Department of Health that proved invaluable in their reporting. These databases were:
1) The state’s death-certificate database, containing dozens of fields – for example, cause of death (primary and contributing), along with the deceased’s age, address and occupation – that break down tens of thousands of deaths each year. This database was used to analyze fatalities going back to 1993. "We isolated accidental overdoses tied to methadone and used the information in other fields to identify and enumerate key co-factors, such as the presence of anti-anxiety medications or illicit substances."
2) The “literals” database, containing text notes left by doctors, medical examiners and coroners involving each death. These notes were analysed to obtain additional details about many deaths linked to methadone.
3) The Comprehensive Hospital Abstract Reporting System (CHARS), containing data on hundreds of thousands of hospitalizations each year, including diagnoses, procedures performed, cost of stay and the economic status of each patient.
One of the follow-up stories in this investigation and an overview of items published in the series
The project also drew upon federal data. "We analyzed drug-consumption figures compiled by the U.S. Drug Enforcement Administration in its ARCOS (Automation of Reports and Consolidated Orders System) database, downloading information available online and filing a federal public-records request for missing portions of the data," explained the project team in the DJA application form. "Lastly, we took poverty and income data from the U.S. Census Bureau and layered it in with our analysis of the state’s methadone-associated deaths. From this overlay we were able to spot geographical and economic disparities, and create multiple interactive graphics and charts." Further data collection was also carried out through public-disclosure requests, which returned thousands of investigative records from the state Department of Health. "We collected court records and discovery documents pertaining to lawsuits filed against a pain clinic in Vancouver, Wash. We also interviewed more than 100 people, including families of victims; state officials and agency physicians; private doctors; law enforcement officials; attorneys; patient advocates; pain management experts; and epidemiologists."
The tools used for data analysis were Access, Excel and ArcGIS mapping software. For display and explanatory purposes – with the video and interactive graphics – the team used Google Fusion Tables, Tableau, Final Cut Pro and Adobe After Effects. DocumentCloud was also used for posting and highlighting key records.
A good presentation includes a mix of formats
The project ran as a three-part series with follow-up stories. The online presentation included a short video that highlighted the data that formed the project’s spine; source documents, a timeline, and four interactive graphics. The most ambitious of the interactive graphics plotted the addresses of those people who had died in Washington from an accidental overdose linked to methadone. Using color coding, the graphic broke the state into five income brackets, allowing viewers to see how there were more deaths in poorer communities.
The data team used multimedia to explain the topic to the audience. Video: Methadone's toll
"To take the story beyond points on a map, we profiled a Medicaid patient, Angeline Burrell, who was prescribed methadone while being tapered off a more expensive painkiller. In medical notes, a doctor who doubled Burrell’s dosage said she warned Burrell that mixing methadone with her other medications could 'cause her to stop breathing.' Two days later, Burrell overdosed and died. Once we isolated deaths linked to methadone, we used our data in all kinds of ways to make the story more personal and compelling."
The project’s revelations had an immediate impact on public policy. The state, at the first chance it had, responded by issuing an emergency public-health advisory to more than 1,000 pharmacists and to about 17,000 licensed health-care professionals, warning of methadone’s risks. In January the state went further, sponsoring training programs in which physicians are instructed to treat methadone as a drug of last resort instead of a first-line agent.
Advice for aspiring data journalists
"Prepare for fierce opposition from state officials and physicians who may try to sidestep your questions. Follow the money and the science. Study research reports with a close eye; many are flawed and incomplete, often based on unreliable estimates or surveys. Anchor the story to state records and public data (via public record requests); it’s hard for the government to refute its own records."
The Data Journalism Awards is a Global Editors Network initiative supported by Google and organized in collaboration with the European Journalism Centre. Please visit the Data Journalism Awards website for the full list of nominees.