Zorgcowboys (healthcare cowboys)

2020

Zorgcowboys (healthcare cowboys)

Category: Open data

Country/area: Netherlands

Organisation: Pointer, Reporter Radio and Follow The Money

Organisation size: Small

Publication date: 26/06/2019

Credit: Jerry Vermanen, Dirk Mostert, Jolien de Vries, Eelke van Ark

Project description:

Pointer, Reporter Radio and Follow The Money used open data in the investigation of so-called zorgcowboys (healthcare cowboys): healthcare companies that make excessive profits at the expense of their clients. Before our investigation, publications on this subject were incidental. With our data analysis and reporting, we could structurally pinpoint what companies were suspicious and how much money is involved in this fraud.

We collaborated to make three TV broadcasts, four radio broadcasts, 38 articles and a journalistic game. We also published two datasets with suspicious companies.

Impact reached:

Thanks to our publications, we could identify almost hundred health care companies where the profits couldn’t be explained through normal practices. We used a very complicated open data set with all financial statements of healthcare companies. To give something back to the public, we wanted to give a data set with comprehensible information. We made a data set with all companies that made more than 10 percent profits: a very high and unusual profit margin in the forms of health care we researched.

The data was available for everyone, but insurance companies, municipalities and inspection services haven’t done any analysis with the data. After our publications, we heard that the Dutch health care inspection NZa used our method of analysis to do their own research from now on.

With releasing the data, we found that other media outlets could do their own findings: at least 71 publications were made by other newsrooms.

Our own investigation led to parliamentary questions, local political questions, insurers claiming money back from companies, and a few health care companies who had to stop their practices. We also won the national open data prize (Stuiveling Open Data Award) in the Netherlands for our efforts.

Techniques/technologies used:

We used open data from the ministry of Public Health, Welfare and Sport. The analysis was done in R. We analysed the financial statements of 2017, and later 2018.

Our first research focused on profit margins higher than 10 percent. The list of health care companies was 174 companies long, and we checked all reports for logical explanations for these high margins. For 97 companies the profits couldn’t be explained through legal means.

In our publications and broadcasts, we explained the systems at work in this complicated fraud. Most often, companies will write down a lot more hours than they delivered in care. But we also found a few more sophisticated forms of fraud: declaring housing costs through a shell company, paying yourself more salary through sister companies, etc.

In our second research we looked at companies who were making high profits in consecutive years. In this analysis, we found again some smart schemes. And during this phase, we collected a lot of research which we couldn’t use in our regular reporting. This is why we made a journalistic game. The game was made in Javascript.

What was the hardest part of this project?

We encountered two difficult phases in our research. The first one is related to the data. Open data in the Netherlands is mostly ‘just publish it online, no one cares’. Our data had no documentation, and even the persons responsible for the data could not explain how the data was structured. Through trial and error (and a lot of cross-referencing with the reports on paper) we could reverse-engineer how this data could be used.

The second part was finding persons on record to talk about our findings. Employees of these companies are afraid to lose their jobs, although they see a lot of neglect of clients. And these clients don’t want to get stuck with no care at all: sometimes a little bit of bad health care is even better than none.

This is why we wanted to be as open and transparent as possible: to build trust with these persons. We asked for tips constantly during our research. And after publishing our first data set, we got a lot of good tips for our follow-ups. Building trust and being clear on the way we worked was the key to our publications.

What can others learn from this project?

We proved that investigative journalism isn’t reliable on the monopoly of information. The data we used, was available for at least 4 years before we did our research. In the past, most publications were on incidents: a single company who was caught in the act. With our analysis we found a lot of companies that were under the radar.

Our research displayed how data analysis and classical reporting can enhance each other. We collaborated with our colleagues of Reporter Radio and financial experts of Follow The Money: without this collaboration, none of these newsrooms could have published anything. We were stronger than the sum of our parts.

By publishing our journalistic game, we also searched for other ways of telling our story. This is a challenge: how to make a game that is fun to play, but also does justice to our findings? We found a nice result in a game in which you can decide how to start up your own company. Do you want to be an honest entrepreneur? Of do you make high profits for your own financial well-being?