Groundbreaking research requires huge, rich and diverse datasets - something that cannot be found in one single location. There are several reasons why this is important.
First, while it may be convenient to draw your conclusions from one single source, taking data from just one hospital is simply not enough. New studies require bigger and bigger datasets, taken from multiple locations. This is crucial in building confidence in your results.
Second, certain conclusions simply cannot be drawn at all from just one data source. You may be researching a rare disease or performing molecular biology research on a subgroup of cancer in which case you very well may not have enough patients in one single location. You may also be tying several fields together - looking for correlations between genetics and pathology findings for example, which necessitates linking information taken from several departments. Either way, many types of research may require going beyond any single database.
Finally, data taken from one location lacks an extremely important feature - diversity. As it turns out, our race, ethnicity, socioeconomic status, geographic location, sexual orientation, gender identity and other characteristics can have a crucial effect on how we are affected by disease and how we react to treatment. For this reason, research must be inclusive and include minority and other underrepresented groups. Using a single relatively homogenous cohort simply won’t cut it and could even lead to serious lapses in medical decisions. In one example, racial and ethnic minorities who contracted COVID-19 have been found to be three times more likely to be hospitalised. To solve these types of injustices, research must use heterogeneous data taken from multiple datasets that are spread out geographically and across populations.
Health research is a real world business
Accessing multiple sources of data requires establishing multi-stakeholder partnerships that span the clinical trial ecosystem. This includes tying together information from patient groups, clinical research sites, academia, CROs, federal and state agencies and more.
But perhaps even more importantly, patients don’t live in a vacuum. They typically encounter several health providers in their lifetime, including HMOs, hospitals, and private clinics. In addition, us humans generate data that could be highly relevant to medical studies - from fitness apps and medical devices to demographic and mortality information collected by our governments.
The main takeaway is this - researchers need longitudinal perspectives on patients that are enhanced with real world data. This is the crucial next step to make a real world difference. Having a longitudinal perspective, i.e. access to the full patient journey, can be invaluable in performing deep and meaningful research that is based on a wider picture. Let’s take a simple example. Imagine someone arriving at the ER while having a heart attack. Crucial life-saving information, such as his previous stress test results or smoking habits, may be siloed in a community clinic database, out of the reach of those who need it most to make life or death decisions.
These types of incidents shed light on the importance of the complete patient journey. Longitudinal studies are picking up speed in health research, with frameworks being developed for how to perform the complex task of analysing information taken from multiple data sources.
There is in fact a growing list of studies that are proving the value of aggregating real world evidence using a patient-centred approach. Such studies are finding that consolidating data from sources such as EHRs, pharmacies, PROMs and personal digital devices can lead to much stronger evidence, accelerate drug approval and aid in assessing patient recovery and measuring intervention effectiveness.
This also highlights the importance of multi-site clinical trials that have picked up speed since the COVID-19 outbreak. Not only does collecting information from dispersed sites boost recruitment rates and save time, money and effort, it also enhances the depth and breadth of research. Integrating data from applications, wearable devices and other repositories can help shed light on some of the major uncracked medical issues plaguing the 21st century.
So how do we obtain information from a multitude of sources and make sense of it all? The traditional answer has been centralisation.
The problem with centralisation
Traditionally, researchers pulled all the data from the source or from multiple sources and indexed it in a new centralised relational database. Pooling all of your data together into a unified database may be the most intuitive thing to do, but this practice can have detrimental consequences.
First, putting all your eggs in one basket can pose a significant security and regulatory hazard. Storing a vast dataset of medical records in one place can leave sensitive information incredibly vulnerable. Healthcare breaches have been on the rise this year, increasing by 84% over the past three years. To make matters worse, data is most often aggregated through data sharing platforms that use third party servers, which account for one in three of all healthcare cyber incidents.
Second, using centralised solutions is problematic from a regulatory standpoint. When information flows to an external repository, stakeholders essentially lose control of what happens to their data. In addition, most of these third party solutions de-identify records only after they have been collected in one place and manually matched. This practice constitutes a severe violation of patient privacy.
Third, maintaining a huge central medical repository is challenging and expensive, consuming considerable storage and memory resources. In addition, indexing massive amounts of data takes a lot of time and tedious preparatory work. In fact, it can take over a year for a researcher to obtain data from just one source.
Finally, the quality of data is also compromised. The moment data is transferred to a central location it starts going stale - it loses connection to the data source, meaning researchers may be using out of data and irrelevant information.
To tackle these issues we must move past our centralised way of thinking and adopt a decentralised approach.
The Power of Decentralisation
Decentralised architecture offers a powerful solution to keep up with the demands of our world. As we have seen above, high quality research requires having access to data that is dispersed across organisations, locations and devices.
Decentralised solutions, such as Briya, are emerging on the scene, allowing researchers to access real-time data directly, from any number of sources. This is not only a more cost-effective, secure and compliant way of going about things - but it also supports research that is more potent, more accurate and more equitable. We are entering the golden age of decentralisation in healthcare, and we strongly recommend you hop on and join the ride. It’s going to be awesome.