On the construction of big-data in healthcare
“When you want to build something that will last, try to start with the foundation.”
I would like to quote this phrase, because I believe that it exemplifies very well one of the most common challenges faced by the healthcare ecosystem when one of its objectives is to generate data that allow decisions to be made with a degree of statistical certainty: to assert that, because data are generated, they automatically have value, may be a mistake.
Currently, an estimated 80% of medical data remains unstructured and untapped after its creation (e.g., text, image, signal, etc.)1. The two main reasons for this are the diversity of data generation sources and heterogeneity of the data, as well as the quality of the data generated.
If data are the building blocks of personalized medicine research and practice, databases are the scaffolding that ensures the integrity of medicine itself. These databases provide the structure within which data are stored and made available for future use. Therefore, it is essential to work on its design, which must respond to one or more objectives and then achieve them through a strategy.
Today, it is becoming increasingly evident how important biomedical data and the databases in which they are stored can be. However, its correct generation is key to realizing this potential. Otherwise, we will focus more on “accumulating” data than on “generating” data, a key and necessary action to convert it into evidence that will allow us to make decisions with a real-world health impact.
Sources: 1Kong H. J. (2019). Managing Unstructured Big Data in Healthcare System. Healthcare informatics research, 25(1), 1-2. https://doi.org/10.4258/hir.2019.25.1.1