Post

1) 🇬🇧 The results of the big STIMULATE-ICP study are in. It tested three repurposed drugs in hundreds of Long Covid patients: - colchicine: an anti-inflammatory - famotidine and loratadine: antihistamines - rivaroxaban: an anticoagulant Sadly, all three had poor results.

2) The graph below shows the main findings. Unfortunately, there wasn't a placebo control group only a group that received no drug, only usual care. The primary outcome was the Fatigue Assessment Scale (FAS, range: 10-50), with higher scores indicating worse fatigue.

3) The authors indicated that 3 points on the FAS can be seen as the minimally clinically important difference. For all three treatments, the benefit compared to no treatment was lower than 3 points, so they didn't produce a significant benefit.

4) The negative result was clearest for the anticoagulant rivaroxaban (difference: –1·06 [95% confidence interval: –2·47 to 0·35]; p-value= 0·139). This finding questions whether blood clotting is a relevant factor in Long Covid patients.

5) The paper is more optimistic about colchicine and antihistamines, but the results are pretty clear. The confidence interval is fully below 3 points. colchicine (–1·49 points [95% CI –2·92 to –0·06]; p=0·041) famotidine + loratadine (–1·48 [95% CI –2·88 to –0·08]; p=0·038)

6) So I think these treatments also had largely null results. This could be because LC is very heterogeneous. On the other hand, with around 200 patients per treatment arm, the sample size was quite large, so it could have picked up on relevant subgroups.

7) Lastly, the authors highlight that all groups improved over the course of the trial, even the one that received no drug. They think this "shows the importance of access to and evaluation of specialist care". That's a rather questionable interpretation.

8) The change over time more likely reflects natural prognosis and regression to the mean than a positive effect of the specialist care being offered. I find it quite unprofessional to interpret this as a treatment effect without any control group to compare it to.