Read on Twitter
Incredibly proud to be part of this amazing @OHDSI collaboration: LEGEND. This is next generation, #bigdata observational research. The pioneers are @suchard_group @nicolepratt_ Ryan You @davidbmadigan Hripcsak. https://academic.oup.com/jamia/advance-article/doi/10.1093/jamia/ocaa103/5895561?guestAccessKey=064eff9f-5ae1-4702-9b2c-4fdcab861148 #openscience @AMIAinformatics
And check out these principles of the Large-scale Evidence Generation and Evaluation across a Network of Databases (LEGEND) initiative. This is the foundation for the #bigdata, next generation observational research. https://academic.oup.com/jamia/advance-article/doi/10.1093/jamia/ocaa103/5895561?guestAccessKey=064eff9f-5ae1-4702-9b2c-4fdcab861148
Transparency, pre-specification, sharing and more… see how a typical study in LEGEND adheres to the principles. https://academic.oup.com/jamia/advance-article/doi/10.1093/jamia/ocaa103/5895561?guestAccessKey=064eff9f-5ae1-4702-9b2c-4fdcab861148
Another innovation, let interested readers explore the results with an interactive viewer so they can fully understand the study. Not just a static paper, but a dynamic interface. https://academic.oup.com/jamia/advance-article/doi/10.1093/jamia/ocaa103/5895561?guestAccessKey=064eff9f-5ae1-4702-9b2c-4fdcab861148
"LEGEND embodies a new approach to generating evidence from health-care data that is designed to overcome weaknesses in the current process of answering and publishing (or not) 1 question at a time.” https://academic.oup.com/jamia/advance-article/doi/10.1093/jamia/ocaa103/5895561?guestAccessKey=064eff9f-5ae1-4702-9b2c-4fdcab861148
"A key guiding principle of the LEGEND approach is transparency. Prior to any analysis, the prespecified LEGEND study protocol and full analytic code are made available in open-source format ( https://github.com/OHDSI/Legend ).” https://academic.oup.com/jamia/advance-article/doi/10.1093/jamia/ocaa103/5895561?guestAccessKey=064eff9f-5ae1-4702-9b2c-4fdcab861148
"following LEGEND principles that address bias, P hacking, & publication bias, LEGEND seeks to augment existing knowledge by generating reliable evidence from existing health-care data, answering questions simultaneously using transparent, reproducible, and systematic approach."
Here is our companion piece: "Large-scale evidence generation and evaluation across a network of databases (LEGEND): assessing validity using hypertension as a case study” https://academic.oup.com/jamia/advance-article/doi/10.1093/jamia/ocaa124/5895564?guestAccessKey=3f845cae-005c-47bd-9772-ab2d587dec45
