5 min read

Drive sustainable Progress: The healing power of data

Digital twin technology has the potential to revolutionize medical research and identify the right treatment for each individual patient, but progress is being held up by a lack of quality data.


In 2022 we are becoming more and more dependent on machines in our decision making — to analyze weather data to make forecasts, on apps that sift through our spending to help us budget, or through our growing reliance on the computers in our cars that help us drive safely. Particularly in the area of medical research — where they are increasingly used to make decisions on treatment options for patients with long-term conditions — when harnessed correctly, data and analytics can be powerful tools in driving forward progress outside of the limitations of our thinking power as humans. Machines can model millions of different scenarios in an environment that poses no risks to humans or animals, and can also be used to investigate scenarios that would not technically otherwise be feasible. One promising example of this is digital twin technology (also known as PBPK*) in clinical trials, an emerging area of medical research that shows huge potential in helping to accelerate the process of finding treatment options across multiple disease areas.

What is digital twin technology?

First used by the aerospace industry in the 1960s to help simulate the performance of satellites in space, digital twins are a virtual model of an object designed to accurately reflect it and how it works — and are now commonly used across most areas of industry. When it comes to medical research, digital twins are virtual reproductions of human limbs, internal systems and organs used to help simulate how they might respond to inputs that may cause a shift to the internal environment, such as a medicine or a treatment. The ‘twin’ can therefore be used to demonstrate the safety or efficacy of a particular treatment, investigate potential drug-drug interactions or refine dosing. Within the medical world, this process is referred to as an ‘in silico trial’, roughly translated, a computer simulation of a clinical study used in the development or regulatory evaluation of a medicinal product or device.

The ethical benefits of this technology are clear — medical testing that doesn’t necessitate the implication of human beings or animals and therefore cannot impact their health — but we are still years away from a full digital twin of the human body, and all of its organs and processes. Nonetheless, as this technology continues to develop, its successful application hinges on the ability of stakeholders working together to gradually increase acceptance of computer modelling in regulatory processes.

Driving progress through collaboration

It’s a space in which biomedical research company Bayer has been active for more than 20 years, developing new technologies and running these in parallel with conventional clinical trials; and using simulations to refine dosing and inform early drug development by identifying new targets. When combined with existing real-world data, the company’s in silico trials have been successfully used to identify new or optimized dosages, bio-markers and indications. A good example in the recent past of Bayer’s application of this technology came when they ran simulations with digital twins to inform dose selection for an anticoagulant medication (shown in the below graph). The predictions of the digital twin model are almost perfectly aligned with the data collected afterward in clinical studies in order to confirm the adequacy of the selected doses. In this case, patients in trials directly benefited from a minimized risk of strokes, heart attacks and thromboses, as well as any unintended side-effects.

As a partner in public-private partnerships and using its scientific and technical expertise, Bayer has also contributed to driving digital twin technology forward. A good example of this was in an EU-funded project aimed at developing a proof of concept for telemedicine and closed-loop control of insulin dosing in diabetes — a severe disease affecting an estimated 60 million Europeans. In this instance, Bayer’s virtual diabetes twins provided predictions of blood glucose levels that successfully informed insulin dosing. Another good example of the company’s influence in this sphere is the support it gives to competing biomedical research units in their adoption of this technology via Open Systems Pharmacology, with the express aim of increasing the acceptance of evidence generated via data-driven modeling and simulation.

Delivering better health care at a lower cost

In an environment of ever-growing costs for health care in Europe and beyond — driven in part by the heavy health care burden brought by ageing populations — technology holding the potential to drive down public spending is likely to be attractive to policymakers. Digital twins can be used, for instance, to enable the creation of new precision medicine options for patients, which tend to provide better outcomes and therefore are more efficient in the longer term, and therefore less costly. This would improve our ability to provide more preventative treatment options for patients with a high risk of developing certain illnesses as well. Other potential benefits include the ability to reduce the incidence of secondary health issues that may present in patients being treated for something else as well as faster treatment decisions. All of these would have a clear benefit for overall society, leading to a better quality of life for individuals and reduced public spending.

The challenges

It is clear to see that the potential for this burgeoning technology is huge, and that digital twins, if harnessed correctly and well-supported by policy, could be a game changer for the future health of European patients. But this innovation is not without its challenges.

One of the biggest hurdles that this technology faces in terms of its ability to drive progress across multiple areas of medicine is a lack of reliable access to good quality data. Without this, companies like Bayer can only get so far. The fact that health care data from electronic health records (EHRs) is not standardized is one of the greatest challenges impeding progress. Different and fragmented interpretations of data protection rules across the Continent, such as General Data Protection Regulation (GDPR), continue to present a major obstacle, as they hamper access to the data that is crucial to developing and building digital twins — and running simulations. In addition, there is a lack of technical data available; to date, there are no truly robust European databases that represent the diversity and richness of information theoretically available across the EU. A European Health Data Space (EHDS) proposal currently under discussion has the potential to improve the operating environment for digital twins in the future — and the outcome of this will be highly anticipated by medical research units across the Continent.

Another pressing issue is that the conditions and requirements for the acceptance of evidence generated via ‘in silico trials’ or computer models must be further developed and specified, while regulatory guidance should ideally be harmonized globally via the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) mechanism.

Early success as demonstrated by Bayer’s endeavors paint a positive picture of the potential of this technology. If policymakers are able to address the issues currently hampering progress in this area, it is clear that there will be numerous positive impacts on the health of people all across the EU, as well as future health care spending — a twin technology that promises to deliver twin benefits.


*Physiological based pharmacokinetic modeling and simulation (PBPK) is a type of mathematical model also known as digital twins.

Figure 3. Kamel, M. (2021). Digital Twins: From Personalised Medicine to Precision Public Health. Available at: https://www.researchgate.net/figure/Digital-twin-instances-of-the-same-person-or-patient-can-be-used-for-in-silico-testing_fig2_353513336

Figure 3. Willmann, S., Coboeken, K., Zhang, Y., Mayer, H., Ince, I., Mesic, E., Thelen, K., Kubitza, D., Lensing, A., Yang, H., Zhu, P., Muck, W., Drenth, H., Lippert, J. (2021). Population pharmacokinetic analysis of – in children and comparison to prospective physiologically-based pharmacokinetic predictions. Available at: https://pubmed.ncbi.nlm.nih.gov/34292671/

Kuepfer, L., Niederalt, C., Wendl, T., Schlender, J., Lippert, J., Block, M., Eissing, T., Teutonico, D. (2016). Applied Concepts in PBPK Modeling: How to Build a PBPK/PD Model. Available at: https://pubmed.ncbi.nlm.nih.gov/27653238/

The above column is sponsor-generated content. To learn more about our advertising solutions, click here.