Most of us don’t have a twin – or do we? Traditional twins account for about 3% of live births in the United States,1 but each of us has a digital twin. These virtual copies of us don’t look identical like physical twins do, but they do have a lot in common. Using a digital version of patients, researchers gain unique insights into current and future health risks. Combining any number of digital twins, health leaders can see the bigger picture, allowing them to focus on improving the overall health and wellbeing of a specific community or population.
Why go to the trouble of creating health digital twins with the same risk factors for diabetes, heart disease, mental health concerns or more? Because this strategy gets around some limitations that plague conventional research – including high logistic costs, recruiting sufficient patient populations, and following patients over long periods to see significant differences in outcomes. Creating health digital twins for research can also help address ethical concerns around withholding treatment for control groups.
In contrast, health digital twin technology – particularly if powered by artificial intelligence or buoyed by a large language model – can leverage health, social and environmental data to pinpoint individual and community risks more precisely. Researchers can use digital twins to see patterns in the data that might otherwise be missed, based on factors such as age, co-existing conditions or social determinants of health.
One definition reads: “health digital twins are defined as virtual representations [‘digital twin’] of patients [‘physical twin’] that are generated from multimodal patient data, population data, and real-time updates on patient and environmental variables.”2
This means that using digital twin data from decentralized records, biological information and mobile sensors can measure and predict response to medication, behavior change and environmental factors. The management and treatment of obesity and its comorbidities, especially for at-risk populations, is one example of how an individual might benefit from research done on their digital twin. Health digital twin research can not only be used for precision medicine but also for predicting health outcomes as they relate to equitable access to care.
On a population level, one aim is to improve the overall health of a community or group at risk. For example, the National Institutes of Health awarded Cleveland Clinic and Metro Health a $3.14 million grant to accelerate health disparity research using digital twin technology.3 Organizations like Patient-Centered Outcomes Research Institute (PCORI) have announced future funding opportunities that can assist with the expansion of digital twin research as it relates to the improved use of artificial intelligence (AI) and machine learning (ML) in clinical research.4 Health digital twin research can help fill the gaps in scientific evidence by capturing more diverse populations, contexts and lived experiences of individuals with varying health related social needs.
Creating robust datasets that combine social, economic and neighborhood factors in addition to medical and EHR data through innovative solutions like digital twin technology can provide data not as sensitive to historical bias. Measuring and predicting health risks based on an individual’s or community’s inherent health status can be precise while controlling for bias.
The recent executive order from President Joe Biden on AI confirms the need to ensure AI doesn’t perpetuate disparities with algorithmic bias or discrimination. As such, the quality and accuracy of data being used in health digital twin research is paramount. There is an opportunity for every study to be a health equity study by combining comprehensive, individual-level real world data like social drivers of health along with clinical data. When doing multi-site data collaborations or using disparate data sources there can be challenges around combining datasets with confidence. Because not all data linking is created equal, it is critical to have a trusted data partner who can accurately link individuals through referential data matching technology.
Sources
Please fill out the form below and we'll be in touch shortly, or call us for immediate assistance at 1-866-396-7703