There's a lack of Real-World Data for rare diseases, they say?

Yet, patients with rare conditions are (eventually) diagnosed & treated. How do we reconcile these 2 facts?

To start off, the large majority of healthcare encounters are being recorded for clinical decision-making purposes that are not necessarily aligned with the clinical research intent. Not that these are two objectives are competing or clashing in any way; however, the way the data is organized does matter. Traditional clinical decision-making "allows" for useful data to be buried in free-text fields and documents. Data dispersion is often an unintended side-product of historical EHR systems.

Then, data privacy concerns are definitely amplified in rare diseases. Risk of identification increases exponentially with 2 things: it's directly proportional to data depth and indirectly proportional to disease prevalence. When in doubt, governments & healthcare providers prefer to trade innovation for safety - which is only fair. Primum non nocere.

But seemingly marginal changes to the way the data is collected, organized and maintained can make a huge change for those looking to use clinical data to inform important decisions - decisions that, all added up, can have a significant impact on patients. Can we use technology to accurately abstract information in unstructured fields and transpose it to structured datasets? Can we act upstream and actually reduce the unstructured : structured data ratio?

While many capable people are currently trying to crack this nut from different angles, I at least further reassure myself with the following: I've always been able to find the clinical data I was looking for. Now actually gaining access to it... well, that's an entirely different story - one that pushed me into the pathway of strategic partnerships.