Project description

This 18-month investigation revealed years of underfunding and mismanagement in New Jersey’s patchwork medical examiner system. The problems, which had been an open secret among forensic pathologists and people in state government for 40 years, had led to lost body parts, innocent people sent to jail and murders still unsolved. A massive trove of data obtained through the investigation also revealed for the first time the extent of the systemic dysfunction, showing families face months-long wait times to learn how their loved ones died, if they find out at all, because the state turns away two-thirds of the cases referred to it for review.

What makes this project innovative?

As part of the core principles of the NJ Advance Media data team, we made all of the data we used in our analysis public. We also published the python notebook showing the efforts of one of our members (Erin Petenko) to successfully replicate the results of my analysis. We also gave people as many portals into the story as possible. Knowing that not everyone has the time or desire to read a 6,000 word story on medical examiners, we also created a comic summarizing the findings, a video interactive showing details of some of our most egregious cases and a data tool where people could see how their local medical examiner's office was performing. For the main story, we built a cohesive presentation, utilizing data visualizations, audio and video that complemented the main story rather than detracted. We also employed a direct, unorthodox second-person introduction to grab readers from the get-go.

What was the impact of your project? How did you measure it?

Within 16 hours of publication, then Gov. Elect Phil Murphy joined lawmakers in calling for a complete overhaul of the medical examiner system. A bill that would rewrite the state's medical examiner act was introduced weeks later and was quickly moved out of committees in the state legislature. Lawmakers promised to pass it within 60 days of the new legislative year. The bill would dissolve the medical examiner system as it currently exists under the New Jersey Attorney General and create a semi-independent agency under the umbrella of the Department of Health. It would grant the state medical examiner oversight of the county offices, set new minimum requirements for employment at the agencies and provide pathways for families to challenge rulings — all key deficiencies highlighted by our investigation. Additionally, the bill would provide research and capital funding to the department, require autopsies to be conducted in more situations and establish a new toxicology lab exclusive to state medial examiners. It would be the most significant update to the state statutes in half a century.

Source and methodology

It took us more than six months of fighting the state to acquire a comprehensive database of all the deaths referred to the Officer of the State Medical Examiner over two decades. The state initially delayed the request repeatedly, targeting a date they expected to get a favorable court decision in a separate records case. When that court decision came in, they cited it to deny our request within hours. We eventually succeeded walking back their denial, beat back a request for several thousand dollars to complete it and, eventually, received a CD containing a database 420,000 sudden or suspicious deaths dating to 1996 for $1. Stirling spent another few months cleaning and standardizing the dataset and enlisted criminal justice reporter S.P. Sullivan to join the project. Analysis was primarily conducted using Excel and Python/PANDAS, which were used to aggregate data into offices and doctors and calculate caseload and wait times. We frequently contacted local offices to iron out questions the data presented. When completed, Erin Petenko produced a complete replication of Stirling's analysis, which we made public. We presented our findings to state medical examiners prior to publication. They were not disputed.

Technologies Used

Open RefineOpen Refine didn’t solve all of our dirty data woes, but it gave us a great headstart. Really terrific tool for identifying minor errors en masse and correcting them.Microsoft ExcelOld faithful. It was perfectly capable of handling most of the analysis we needed to do.CSVKitTerrific for slicing out portions of the data and creating new files. This not only helped us cut down the data for more detailed analysis, but carve out small sections of data for use in visualizations.Plot.lyA great tool for making customizable charts in D3. Lowers the learning curve and the product remains professional and clean.Python/PANDASUsed by data reporter Erin Petenko to replicate original analysis by Stirling. We wanted to do this in a different program to insure there weren’t any errors that were inherent to Excel or how the analysis was completed in the program.JavascriptWe used several libraries (Isotope, JQuery, Bootstrap) to make our visualizations for the project. Data was the backbone of the project, and Javascript helped it sing.AudiogramWe used NPR’s lovely audio utility to add audio to the project. Allowed us to package audio cleanly and style in line with the rest of the presentation.

Project members

S.P. SullivanChris BaxterAndre MalokCarla AstudilloDisha RaychaudhuriErin Petenko



Additional links


Click Follow to keep up with the evolution of this project:
you will receive a notification anytime the project leader updates the project page.