Jul 18 - Journal Club F/U

Thanks to Dr Anderson for being the guest discussant at today's journal club.  Below are links to the articles reviewed:

Continuation of Low-Dose Aspirin Therapy in Peptic Ulcer Bleeding from Annals of Internal Medicine, Jan 5, 2010.  Though small in size, this study offered compelling evidence that patients taking low dose aspirin for secondary prevention of cardiovascular events should be continued on aspirin even when they are treated for GI bleeding.  The clinical event rate with aspirin withdrawal was quite high, even in the brief 30 day follow up period.

The major methodological issue I had with this trial was the use of a "non-inferiority" design.  I do not think it was the appropriate way to look at this clinical question, as there was no well-established clinically proven baseline to compare it to.  A randomized controlled trial would have sufficed.  Also - beware of non-inferiority trials in general.  In up to 12% of publications reviewed by one group it was found that conclusions drawn by the authors were incorrect and missed by the editors. Another concern in general with non-inferiority trials is that they may lead to "biocreep" where drugs of decreasing effectiveness are deemed non-inferior to a prior drug that was non-inferior to a prior drug that actually was superior to placebo.  After a few generations of non-inferiority the actual effectiveness compared to placebo may actually be nil.

The second article: Rifaximin Therapy for Patients with Irritable Bowel Syndrome without Constipation from the Jan 6, 2011 NEJM brough up several interesting points.  One was the difference between statistical and clinical significance.  In this trial, most people that took rifaximin experienced no relief.  Additionally, 1/3 of patients experienced relief with placebo.  The overall proportion of patients who's improvement could be attributed to an expensive medication was quite small.  At >$900 per treatment course it would be best to start with some cheap innocuous meds like amitriptyline which improved nearly 70% of patients (compared to 30% placebo) and costs pennies a day.

It also brought to mind another recent NEJM (Jul 15, 2011) trial that shed quite a bit of light on the placebo effect in general, especially for clinical conditions with subjective clinical endpoints.  In this trial of asthma therapy they compared albuterol to three other treatments -  placebo inhaler, pretend quackery (AKA sham accupuncture), and no treatment.  Interestingly, the symptomatic improvement was equal with real or fake albuterol as well as fake accupuncture.  All three of those arms experienced better relief of symptoms compared with no treatment.  Tellingly - the only intervention that improved the objective outcome of an increase in FEV1 was real albuterol.  The only issue I had with their methods is that the phrase "sham accupuncture" is a tautology.