Ann Intern Med 5 June 2007 Vol 146

Another negative trial of homocysteine-lowering therapy: this is a substudy of HOPE-2 and shows that oral supplements of folic acid, pyridoxine and cobalamin lower HCy but do not reduce venous thromboembolism.

Dementia is associated with a mean survival of about three years, but this survival study did not go beyond 180 days, and examined the association between death and the use of antipsychotic medication in these patients. Sure enough, the association is positive and may be higher in those given “atypical” antipsychotics. But this was no watertight study; medication was often given for a month or less and it’s far from clear where, if at all, the relation was causal.

“Physiotherapy” for most of us doctors is a black box, and we want to keep it that way for fear that we might have nothing otherwise to offer our patients with subacute back pain. This study from Australia and New Zealand attempted to prise open the black lid and see what’s inside the box: it randomised patients to physiotherapist-directed exercise or sham exercise sessions plus physiotherapist-directed advice or sham advice. The canny Ozzies and Kiwis weren’t supposed to know the difference. And actually there wasn’t much at 6 weeks and none at all at 12 months, in terms of outcome.

Here’s a Japanese meta-analysis of a new marker for rheumatoid arthritis – anti-cyclic citrullinated peptide antibodies. These antibodies may even be causal in the disease process. The editorial shows that their predictive value still varies with the clinical context, as is the case with all diagnostic tests. “Bayes clears the haze”, it claims, but even a high risk patient with a clear clinical history and positive anti-CCP antibodies has no more than a 67% risk of having rheumatoid: quite a diagnostic sea-mist, I would say.