Good research starts with data protection

“The interesting thing is that a lot of our field is becoming more digitalised. In the project ‘Sensing Potential in the food supply chain’, we look at how we can use sensors to estimate food intake. An example is using a colour depth and near infrared camera that uses artificial intelligence to try to estimate portion sizes and thus predict quantities of ingredients and carbohydrates, proteins and fats.

“For clinical nutritional research that is covered by the WMO (medical research involving human subjects act), we adhere to the Central Committee on Research Involving Human Subjects guidelines. Then a research proposal must be submitted that is then assessed for its ethical aspects. In doing so, we also comply with the AVG legislation by, for example, ensuring test participants are properly informed about the collection and processing of both personal data and research data, and requiring consent from them.

“As a researcher, I make sure that the data we collect from participants is only processed using systems and applications that are in the ApprovedApps Tool. The privacy-sensitive data that we use can include names, telephone numbers, addresses and IBAN numbers; we need this for payment purposes. Some participants are aware of potential privacy issues while others put their whole lives in an email, so to speak. In these cases, I tell them that they can bring the reimbursement form to us, for example, and we then enter that into the system. That paper then goes straight into the blue container for test participant information and is destroyed.

“I have noticed that there has been a greater focus on privacy since 2018 when the AVG legislation was introduced. We are all much more aware of it. This also applies to what you write in your emails; no personal data, but a participant number. We also take the retention period into account. Participants’ data should not be kept longer than necessary. This depends on the applicable retention period for the type of information.

In addition to personal data, we also have retention periods for research data. For clinical nutrition research, this is 15 years; for consumer research, we follow the WUR guideline of ten years. One major improvement in recent years is that we have extended these guidelines to digital folders that we keep on the secure digital WUR drive.”

“We have protocols on who should have access to this data. These individuals also have to sign a declaration that they will not leak any confidential data. Not everyone has the same rights and access; most employees use data tables in which the names of fishing companies and vessels are anonymised using a code. In the analyses, catch data is also aggregated to prevent it from being traced back to specific companies.”

“Even as an administrator, I cannot access everything. The most important databases are managed at the highest level by FB-IT, the facilities company of the IT services within Wageningen University and Research. We use databases they host. The advantage of this is that they take care of the databases’ security and provide technical updates. We also use managed servers where other applications are installed, and these need to be managed too. FB-IT largely takes care of security, taking control of standard updates and management of the firewall and user accounts. We are responsible for the management of the applications we install on these servers, which can create risks if they are not updated in time. I try to avoid those risks as much as possible by using software that is technically maintained by FB-IT.

“Things can always go wrong with privacy-sensitive information. We have reported two potential data breaches. These were acted on quickly, according to procedure. The last thing you should do in these instances is to freeze or sit around. There is a whole page on the intranet with instructions on what steps you need to take. It is like traffic. You can be as careful as you like, but accidents can happen. You have to do your best to prevent them, but if things go wrong, you have to deal with the situation as quickly as possible.”

“Every ship has the name of its home port as well as a number: the Urk 1 or the Scheveningen 65. An important choice is which employees have access to the actual name of the ship. Others only see its name in anonymised form: the Urk 1 is called ship 47, for example. You can still see that the ship is fishing for herring or blue whiting, but you cannot trace which ship it is.

“The datasets are currently stored on the well-secured WUR servers and will remain there. Our hard drives on the PCs and laptops are encrypted with bitlockers, so unauthorised figures cannot access them easily. But you don’t want this to become a potentiality, so our policy is to keep everything on the network drives. Only a select group has access to these files. Everyone has to provide a statement in writing to access them. I am the first to do this every year.

“Not everyone has access to everything; many employees can only access the data in its aggregated form or only a part of it. If we make maps or products based on the data, then I check with a colleague whether there is adequate assurance of anonymity in what we make public. If it is not, then we do not publish it.

“If we do not handle the data we have been given with care, there can be huge consequences. The loss of trust — something especially important with fishers. The government may also decide to not supply us with any more data sources, or cease to provide data in its current format: raw data that enables us to do innovative research. So there are implications for research trajectories as well as for the knowledge we develop about fisheries.