Free and accessible data for research is threatened by politics in modern era
Scientists and the freedom of research which are enshrined in the German constitution, in the modern climate are increasingly under threat.
Politically driven attacks against the science behind climate change have left their mark on society. Since 2016, demonstrations under the name “March of Science” have been taking place around the world. Physicist Sascha Vogel of the Frankfurt Institute for Advanced Studies said in an interview with Deutschlandfunk Kultur that science is definitely under threat. “Brazil is a nice example, some subjects are being abolished there now. Hungary is a nice example, Turkey definitely. And that is the reason why we are demonstrating with the ‘March for Science`.”
Science must be independent of state and church. But with the increasing complexity of research processes due to a growing number of actors in a globalizing world, other factors are coming to the fore that threaten free research. There is a growing scepticism towards science among the population of Western democratic industrialized countries. To combat this, diverse and innovative research must be promoted, especially in the economic sector.
The freedom of research is not only endangered by external political threats, such as populists and science sceptics. Internally, it is endangered by misguided developments in the scientific community. Commercial interests get the upper hand and determine the research process. The effort for the press release seems to be more significant than the research itself. For example, a research project at Heidelberg University Hospital was marketed as a “milestone in blood cancer diagnostics”. In fact, it lacked a scientific basis and the advertising campaign contained some questionable claims. The case became known as the Heiscreen affair, triggered by the same-named company. In addition, medical innovations that are not ready for the market are launched to boost the stock price.
Publication of data depends on commercial success
The use of data is regulated by the DSGVO (engl.: Privacy policy). This regulation affects not only corporations like Google and Co. but also all SMEs, associations, doctors and schools, etc. According to this, the collection of data is only permissible according to Section 28 (1) No. 1 BDSG, if it is necessary or required for the execution of a contract. The creation of data categories for patient data is also prohibited – health data is considered to be particularly sensitive data. Additional security measures are intended to specifically protect this data from third-party access. In 2020, the German Federal Ministry of Education and Research published a text touting Germany’s leading position in the field of AI research, which is to be secured and expanded. “The focus and degree of novelty of the projects must be on intelligent approaches in which the AI that is applied leads to a clear added value compared to established processes. At the same time, the protection of personal data and privacy must be ensured […]” ( Announcement – BMBF ) For the realization of data-driven applications, one cannot rely exclusively on existing and highly restricted data sets. But where should usable data sets come from?
Commercial interests of pharmaceutical companies and other players determine the research process not only in the commercialization but also in the publication of data. Peter Sawicki, the former head of the Institute for Quality and Efficiency in Health Care, puts it succinctly: “So you’re dependent on the data you get in the process, and for most drugs, that’s the pharmaceutical industry, which generates these data and then also publishes them or, what’s much worse, just doesn’t publish them.”
According to experts, half of all data from medical research is misappropriated and therefore not accessible. Often, the pharmaceutical industry is behind it. For example, researched advantages of a drug are published, whereas studies with disadvantages are not published. Such studies remain trade secrets of the corporations and are thus kept away from the public. The right to do so is subject to the fact that the pharmaceutical industry commissions the studies. Consequently, research institutes are under financial pressure and want to fulfil the pharmaceutical industry’s orders first. In this context, favouritism studies arise, which are not lies, but biased. How can we circumvent dependency structures and make data more accessible?
Researcher Peter Sawicki suggests that public authorities should issue the research orders, not drug manufacturers. Thus, research institutes are independent of the funding agencies and the data could not be kept under lock and key. However, the basic funding that universities receive from the federal states has long been insufficient. Financial assistance from third parties is indispensable for maintaining research and teaching – especially in cost-intensive medical research. Heidelberg University Hospital, for example, raised 34.6 million euros from private donors in 2017, compared with only half that amount 11 years earlier.
German law to “protect” free science?
Reliable and free processing of data must be given in research. Medical research should be able to develop independently of commercial interests. In politics, it is forbidden to act politically for the benefit of those whom one’s own party supports with financial means – why not in medicine?
There is a lack of evidence that hospitals pass on sensitive data to corporations. But rather, access to this data for research purposes should be given to SMEs, who could then modify the data to suit customers or patients. Hospitals are very sensitive about disclosing their transactions. In 2008, the University of Cologne concluded a cooperation agreement with Bayer Healthcare. To date, the contracts between the University of Cologne and Bayer are not public. Jurisprudence in Germany “protects” the freedom of science – even against public interest and SMEs.
We at Kipoly want to do it right. With the increasing number of science sceptics and success-oriented sponsors, we want to advance scientific work in such a way that efficiency and effectiveness can be increased. We ensure the security of the datasets by applying Differential Privacy and Federated Learning. Here, the data is modified in such a way that, unlike anonymized data, no conclusions can be drawn about the original data. Nevertheless, a statistical analysis is possible when using Differential Privacy. The implication of Federated Learning allows AI applications to still be able to classify different data collections. In intelligently processing and comparing datasets through AI applications, the performance of research institutes benefits significantly. This approach allows us to be able to develop new models without sharing sensitive clinical data. The availability of data from studies and medical examinations can drive the progress of AI applications in medicine. In the end, all of society should be able to benefit.
Reach out to Reza Esfahanian per Email with questions or ideas for the next steps esfahanian@kipoly.com
LET’S TALK MORE
+49 511 4753950
hello@kipoly.com
OUR VALUES
Kipoly is transforming the healthcare industry by offering sustainable solutions such as EHR, EMR and artificial intelligence-based CDSS. We assist health organisations to leverage their data and optimise their clinical operations. High integrability and intuitive understandability are what we define ourselves with.