Major Changes Needed in Sharing and Applying Clinical Data, Says MIT Researcher

Leo Anthony Celi, a principal research scientist at the MIT Laboratory for Computational Physiology (LCP), and at the Institute for Medical Engineering and Science (IMES), emphasizes the need for significant changes in how medical researchers gather, share, and apply clinical data. Celi, who is also a practicing intensive care unit (ICU) physician, believes that open access to clinical data, multidisciplinary collaborations, and focusing on the needs of diverse populations are key to advancing medical knowledge and improving patient outcomes.

Open Access to Clinical Data

Celi advocates for making all types of clinical data openly available with proper privacy safeguards. He believes that openly sharing data can lead to better understanding of health and disease, and foster collaborations between hospitals, universities, and industry partners.

Focusing on Diverse Populations

Celi highlights the importance of addressing the varying needs of populations across different countries. He suggests empowering experts in these regions to drive advances in treatment and research. To achieve this, Celi and his team organize “datathons” in different countries, where local professionals and students collaborate with MIT students and faculty to analyze health data and generate innovative solutions.

Fighting Bias in Medical Data

Machine learning and advanced data science techniques have revealed pervasive bias in medical data, resulting in suboptimal outcomes for certain groups. Celi gives the example of pulse oximeters, which have been found to overestimate oxygen levels in people of color, leading to inadequate treatment. He warns that as medical data expands, computers can easily learn sensitive attributes and perpetuate biased recommendations. Addressing these biases is crucial to ensure equitable healthcare.

Opportunities for Industry Collaboration

Celi emphasizes the need for collaboration between pharmaceutical companies, vendors of electronic health records, and medical device manufacturers to better understand societal needs and design more inclusive health products. He suggests that corporations share the results of their clinical trials and participate in datathons to witness the transformative potential of data analysis by diverse teams.