Waiting for p<0.05

In this submission, PatientsLikeMe considers the risks, benefits, and opportunity for collecting and analyzing data for ALS patients who are taking part in clinical trials. Our intent is to encourage discussion and debate about the best way to address the complex medical, ethical, and scientfic issues surrounding the collection of such data.


3.11.2012 Update:

(1) Replaced 4-month NP001 results with 6-month results in Tables 1 and 2, and Figure 1.

(2) Removed the statement that ALS Untangled reports have found "no hint of efficacy"; they have actually recommended possible followup study for 2 out of 18 treatments.

(3) In the acknowledgements, corrected the statement of which experimental treatments Rob Tison had reported.

These updates do not affect the conclusions.


26.10.2012 Update: After receiving feedback from study investigators who passed on the concerns of patients, as well as one of the sponsors of an ongoing trial, we are updating to the paper. While we stand by the methods and the data shown previously, we accept that given that between 50-67% of trial patients were on study drug while between 33-50% of patients were on placebo, the results as we originally presented them may be open to misinterpretation. Therefore we have clarified Table 1 and Figure 1 to be more easily interpretable. This does not change our conclusions.

Table 1 was modified to present results of two different scenarios, based on different assumptions about for the calculation of the treatment effect size. (Please see the manuscript for details.)

We also added Table 2 to more completely present the side effect analysis, as well as providing more detail on the interpretation of these results for the different trials.

Finally, we modifed Figure 1 to reflect the most likely estimate of the treatment effect and its distribution, under the assumption that the balance between treatment and placebo trials arms was the same for PatientsLikeMe participants as it was in the clinical trials as a whole.