Problems in interpretation of the Surgisphere results even if they weren’t fraudulent

Colleagues on the ORCHID trial committee and I wrote this response to the Surgisphere hydroxychloroquine results. In the event, the papers were discovered to be based on fraudulent data, so Lancet ended up not publishing the response. However, our arguments were relevant even if the results had not been fraudulent. Given the importance of what we term “trajectory bias,” I thought it reasonable to post the paper here for future reference.


Banner image is of Mount Mkinwartsveri (Kazbek), with the Church of St. Mary foreground left, image © Samuel Brown 2000