Toon Segers

CPO at Roseman Labs

Published on: 23 February 2021

Secure data COLLABORTION for HEALTH consortia

Care providers or health insurers can rapidly deploy secure analytics, without the traditional burdens of (leaky) pseudonymization or trusting the data to a (new) third party.

Data collaboration in the healthcare industry is extremely valuable. It helps to increase the effectiveness of treatments and special medicines, to increase the efficiency of care, to share capacity and to provide the right care in the right place.

However, health care information is very sensitive, in many cases not standardized, and not easy to access. Roseman Labs is the first Dutch company that provides a solution that enables collaboration on sensitive data by keeping data encrypted at-use. Based on Secure Multi-Party Computation (MPC), Roseman Labs offers a solution that enables organizations to perform analyses on joint data with unprecedented privacy guarantees. 

How does Multi-Party Computation Work?

The Secure Multi-Party Computation (MPC) technique allows multiple parties to perform calculations on a joint data set, without parties exchanging or combining data sets. The data input from one party remains hidden from the other parties, and the result of the analysis is only known to pre-specified parties. With MPC one can calculate with multiple data sources as if it were one data source, without the data being brought to a central place. This makes it possible, for example, to combine data from multiple sources in an analysis, without entrusting this data to a “trusted party”. From a privacy point of view, this is highly beneficial.

Example: Biologicals in rheumatism care

One example of how MPC can be applied within the healthcare industry, is that of biologicals in rheumatism care. Biologicals (medicine developed with special biological techniques) are an important new development in rheumatic care. However, not enough clinical test information is available to indicate the effectiveness of these biologicals. Effect studies of actual treatments are desirable, provided that populations are large enough. With our solution, patient populations can be safely combined to scale the study of treatment effectiveness without centralizing the data, preserving privacy and accelerating time to insights.

Example: Therapy for wounded feet

Another example is that of wounded food therapy. More than 40% of patients treated for a wounded foot will recur later in life. However, a podiatrist who is aware that a wounded foot has been treated, can effectively contribute to prevention. To share this signal between second line (hospital) and first line (podiatrist), the patient’s consent is required. However, sharing patient information in advance is often disproportionate and unwanted by patients, who do not want to give consent for access to their data when purposes are not predefined. MPC allows privacy-preserving computations on patient information, while the patients data is inaccessible to any single party. This way, alert signals can be generated “in the blind”, and data will then only be shared between 1st and 2nd line care if relevant for the patient.

In both cases, the Roseman Labs solution ensures maximally protected patient privacy. Also, the data processing follows the spirit of modern privacy guidelines such as data minimization, proportionality, purpose limitation, privacy by design, etc.

To conclude

Care providers or health insurers can rapidly deploy the Roseman Labs solution, without the traditional burdens of (leaky) pseudonymization or trusting the data to a (new) third party. Patient data remains at the source, and only pre-agreed statistics are shared. We believe this solution could revolutionize how care providers collaborate, at a critical time when care providers are looking for novel ways to increase effectiveness through collaboration and partnership.

Image Credit: Hush Naidoo on Unsplash