CAP TODAY Pathology/Laboratory Medicine/Laboratory Management
Pathology in a big data world
October 2022—When I was in pathology training back in the ’90s, physicians carried around an index card for each patient, with all of the information we needed to know about them easily covered in that small space.
Today, the practice of medicine—and specifically the practice of pathology—looks very different in the era of big data. Of course, we still have to fill our traditional roles: making the correct diagnosis for individual patients and ensuring the integrity of laboratory results. But increasingly large data sets inform the diagnosis in individual cases and, at the same time, individual cases become data points in large data sets that inform the health of populations.
Beginning in the 2000s with the value-based care movement and accelerating with the rise of high-parameter tests, we find ourselves having to be data scientists as much as physicians. We are being asked to incorporate data-heavy tests and pipelines, some of which require clinical decision support algorithms that demand a certain fluency with more sophisticated software. We find ourselves in the new position of considering population health in addition to patient health, an element that can involve predictive analytics and data mining.
For those of us who started our pathology careers in the days of the patient index card, this has been a really big change. In oncology, for example, we used to rely solely on morphological and immunophenotypic subtyping to help guide treatment decisions. That is no longer enough. Pathologists today have to be familiar with the genetic signatures of tumors and the various therapies each signature might imply to help their clinician colleagues in the treatment selection process.
Dr. Volk
This new paradigm has an impact both on individual patients and entire patient populations. On the patient level, we can tailor treatment guidance better than ever thanks to the tremendous amounts of very specific data we gather on the way tumors will behave, or the way diseases may progress. On the population level, the evolution of electronic health records (and their somewhat improved interoperability) has given us the opportunity to look across our patient community to gain new insights.