A health data disconnect between clinicians and data scientists is wasting precious medical research and healthcare resources, hampering innovation and resulting in poorer outcomes than would otherwise be achievable.

That’s the conclusion of researchers from the Massachusetts Institute of Technology’s Critical Data group, an affiliation of research labs at MIT focused specifically on data that has a critical impact on human health.

As the use of IT and data grows within healthcare, the researchers suggest that data science should be included in the core curriculum during medical school and residency training.

Despite the digitization of healthcare and abundance of health data from disparate sources including EHRs, mobile devices and wearables, they say the fundamental quality, safety and cost challenges of providing care have not been resolved and that better use of clinical data has the potential to address these issues.

According to Leo Anthony Celi, MD, head of the MIT Critical Data group, the problem is that a lot of the data exists in silos and is not integrated. In particular, he believes that the idea of data sharing is still foreign in healthcare because of stubborn cultural barriers that continue to stand in the way of science and progress.

“If we are to learn as a healthcare system, there has to be more data integration and harmonization,” contends Celi, who specifically calls out the health data divide between clinicians, the domain experts and the technical experts, such as data scientists.

“There is a persistent gap between the clinicians required to understand the clinical relevance of the data and the data scientists who are critical to extracting useable information from the increasing amount of healthcare data that are being generated,” wrote Celi and his MIT colleagues in a recent viewpoint article published in the Journal of Medical Internet Research.

Leo Anthony Celi

However, the MIT researchers contend that the health data divide can be narrowed by creating a culture of collaboration between clinicians and data scientists, exemplified by events such as datathons, as well as reforming medical education, rethinking academic incentives and providing funding opportunities.

In particular, they believe the education of medical trainees in data analysis methods and data scientist trainees in the domain issues of clinical practice and data “would be a primer for future collaboration between the two groups later in their careers,” states the article.

When it comes to health IT, the researchers charge that medicine has “clumsily entered” the digital age.

“Vast and costly electronic medical records systems have been implemented largely without careful and planned consideration for their impact on the entire healthcare system, including education, practice, workflows, and research,” they write. “Education and practice systems have not taken this new digitized world into full account, and consequences include students who are unprepared for their digital futures, very unhappy physicians stuck behind computer screens selecting seemingly endless items in reams of dropdown lists, and the unconscious loss of many opportunities for improvements in practice and research.”

As a result, the researchers say: “It is time, even if a bit tardy and somewhat less than proactive, to acknowledge and address this transition of medicine from paper to computer, from opinion and experience to evidence, and from memory to search engines.”

They argue that an introduction to the use of digital health records for research may provide a foundation to be able to contribute to knowledge discovery regardless of the career path medical students and residents eventually choose.

Also See: New AMA textbook schools students on EHRs, emerging medical practice

Matthieu Komorowski, MD, a visiting scholar at MIT who works with Celi and co-authored the JMIR article, makes the case for inclusion of data science in the core curriculum in medical school and during residency training.

“It’s not common at all to receive any formal training in these subjects. Even in biostatistics, you have to take optional courses to learn about statistics—we’re not even talking about machine learning and advanced computing,” says Komorowski.

Celi acknowledges that the system-wide changes he and his colleagues have recommended will not be easy to achieve in healthcare and that the path forward is daunting. “There will be resistance. There are people who are benefiting from the current health data divide and will want to maintain the status quo,” he concludes. Nonetheless, Celi remains optimistic.

Register or login for access to this item and much more

All Health Data Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access