Big Data in Epidemiology

Too Big to Fail?

The authors respond:

We thank Dr. Chiolero¹ for his interest in our commentary.² We agree that observational data must be used with caution, and lots of data must be used with lots of caution. The concluding paragraph of the commentary echoes Dr. Chiolero’s point: we need to avoid the illusion that big data is a panacea for all the threats to the validity of observational studies.

Sengwee Toh

Richard Platt

Department of Population Medicine, Harvard Medical School and Harvard Pilgrim Health Care Institute, Boston, MA [email protected]