The one thing I did notice about the control group was the average age was less than the treated group (37 for the control group and 51 for the treated group). The sample size was small; treated group was 20 and the control group was 16 and they did have 6 dropouts. The average number of days between the onset of symptoms and the inclusion in the study was essentially the same for the treated group and the control group. They did present the numbers of patients who were asymptomatic, had upper respiratory symptom, and had lower respiratory symptoms for both the treated and control groups. There were more patients with upper respiratory symptoms than lower respiratory symptoms for both groups; about 60% of both groups had upper respiratory symptoms. I suspect that means that most of the patients in both groups were not very sick (yet?). They did do a statistical test for significant of the results. This was clearly a preliminary study and the authors did make suggestions for further studies.
Quote:
Originally Posted by Mikeod
What stood out to was the selection bias. If you choose only those with mild cases of the virus and don’t include a control group similarly selected, you have no way of knowing if the recovery was due to the treatment or not. You could have treated them with Perrier and touted that as a potential cure. Bad science is worse than no science.
|