Dr.EndoEP said:

Propensity score matching studies and the trial evidence: when RCTs aren't available for a specific question, propensity score-matched observational…

True, with the qualification that this is a self-selected group. The people for whom it did not work post less, and that shapes everything we think we know.

SkepticalSean said:

Bayesian meta-analysis perspective on the trial evidence: traditional frequentist meta-analyses report point estimates and confidence intervals.

That reframing is the part I needed.

Dr.EndoEP said:

Propensity score matching studies and the trial evidence: when RCTs aren't available for a specific question, propensity score-matched observational…

Filing a mild objection. Mild because I might be wrong; an objection because nobody has addressed the case that does not fit. Pushing back on the consensus forming here. Consistency is not the same as correctness, and a board that agrees with itself quickly is usually agreeing with the first person who sounded certain.

That is the short version; the long version is somebody else's post.

Adding the numbers, since they settle part of this. Worth stating the units and the reference range whenever you post a number here. A large fraction of the apparent disagreement in these threads is two people using different units and both being right.

Moderator note: leaving this open. It is being argued well and the disagreement is the useful part. No action needed from anybody.