...We Welcome You To The Resverlogix HUB withIn The AGORACOM COMMUNITY!

Free
Message: Re: The scariest thing for a clinical trial......

"My question is: does the medical science community have a general consensus on what is an acceptable and unacceptable P value, applied to any and all trials and target groups?"

p<0.05 is the generic threshhold for statistical significance. However, often the threshhold p-value is lower in clinical trials due to adjustments for multiple comparisons and accounting for any efficacy interim analyses. For certain, a primary outcome measure with a resulting p-value greater than 0.05 will be interpreted as a negative result. The actual threshold p-value for BETonMACE may or may not be below 0.05, depending upon their statistical analysis plan.

"Or are there allowances for less-than-perfect P values for certain types of trials and target groups?"

The primary outcome measure must meet the minimum threshhold p-value (see above) to be considered a positive result (statistically significant). From my understanding, secondary outcome measures (i.e. HDL-C, apo-AI, fasting glucose, fasting insulin, HbA1c) can only be considered positive if the primary outcome measure is also statistically significant. Otherwise, they are only exploratory. I do not know for certain if sub-group analyses (i.e. rosuvastatin vs. atorvastatin) in BETonMACE are valid if the primary outcome measure for the entire trial population was not statistically significant.

"BetOnMace targets very sick individuals who's very poor "end-stage" health is likely the result of very complex, diverse and entangled health issues.  Would any "slack be cut" by those watching in response to news of less-than-perfect P values given the obviously very complex and severe (deathly) variables in both the dosed and the placebo groups."

No

"Also, how would you describe our trials' P values recorded to date?"

This is too difficult to summarize concisely. ASSERT, SUSTAIN and ASSURE have all have different primary outcome measures (apo-AI, HDL-C, percent atheroma volume reduction, respectively). Post-hoc analyses of these trials led to the realization that patients with diabetes and/or high baseline hsCRP and/or basline low-HDL and/or also treated with rosuvastatin are better responders for plaque reduction and/or apo-AI/HDL-C modulation and/or 5-point MACE reduction. BETonMACE has been designed to take advantage of this information learned from the post-hoc analyses to choose the best population to assess apabetalone's MACE reducing effects in BETonMACE.

BDAZ

Share
New Message
Please login to post a reply