Arrow Left Health research

Use case: How to address a research question that requires data from two databases

Doing a practical test of the Personal Health Train concept by trying to find out whether irradiation of the left breast during breast cancer treatment gives women an increased risk for heart failure.

Medical relevance

Radiation of the left breast is a treatment option for left sided breast cancer. This treatment, however, also harms the heart. Therefore, we need to find out how many irradiated patients with breast cancer undergo treatment at the cardiologist for heart failure later in life.

Description

Clinical quality registrations have improved healthcare during the last decade by collecting data and comparing information about for instance complications and mortality concerning a specific treatment. These comparisons reveal best practices hospitals that can become an example of how to realise the best care.

Patient data, however, can be scattered across several registrations, because some people are treated by different health care providers. Therefore, this project tests whether PHT is helpful when you want to combine information of two clinical quality registrations, in this case patient information from the National Heart Registry (NHR) and from the Dutch Breast Cancer Audit of the Dutch Institute for Clinical Auditing (DICA).

In order to protect privacy, this practical test uses a certified algorithm. The certified algorithm can only answer the question asked with the use of the encrypted data in the database of the DICA and the database of the NHR. The data does not leave the database and the encryption ensures that the data cannot be interpreted by data managers.

Main results

An earlier PHT test in a virtual setting has been successful. The real test will be performed in the next three months. The goal is to check whether the research question can be answered using real patient data.

Follow up

If the practical test turns out to be successful and proves the PHT concept works in practice in healthcare, DICA will start using this way of working more often. The big advantage is that DICA's data sources can then easily be linked to other data sources, thus making a lot of information easily available.

An example is the Parkinson registration. Recently it became known that pesticides could play a role in the development of Parkinson's disease. With PHT, data sources of pesticide use in the Netherlands can be examined, for instance, to check whether intensive agriculture is associated with higher incidence of Parkinson's disease. To further explore this topic many more questions can be answered, such as whether people lived in an area where certain types of pesticides have been used a lot.

Lessons learned

Nobody wants to make mistakes and get into trouble because of the new privacy legislation. That is why organisations are reluctant to give permission for the use of patient data. Trust between organisations working together in a PHT project is necessary and the involvement of the National Health Institute can help to build that trust. Many consultations are also necessary. A transparent algorithm is also important if you want to build trust. In addition, you need people with expertise and people who are willing to take the next step.

Project details

Project leader

Wouter Franke, National Health Care Institute (ZIN)

Funders and Cooperating Partners ZIN, NHR and DICA

 

Contact

Share this page…