Rigor in Quality Improvement Studies and the Role of Time-Series Methodologies

To the Editor In response to the editorial by Grady et al¹ published in a recent issue of JAMA Internal Medicine, most quality improvement activities in medicine are not research, which is defined as “a systematic investigation, including research development, testing and evaluation, designed to develop or contribute to generalizable knowledge.”^2(p94) Because of expectations for generalizability, study designs that test hypotheses, maximize external validity, and test multiple contexts, are expected. Conversely, most quality improvement, though such activities may be systematic and data driven, is not designed to be generalizable, prompting recommendations exempting quality improvement from usual human subjects research review.²