It seems to me that the two most important goals in treating type 2 diabetes are to preserve the remaining beta cells and to reduce cardiovascular risk. And it seems to me that the main message of the UKPDS trial is that HbA1c is a useless surrogate marker for either of these. Exenatide is the first of a new class of drugs for type 2 diabetes – an incretin mimetic which has to be given by injection. This trial examined the short term effects of giving it with a thiazolidinedione, sometimes with metformin as well. Yes, it can decrease HbA1c. But the gist of the accompanying editorial is that this trial was so badly designed that it should never have been published. So why does it appear in a leading medical journal?

What counts as neutropenia? In the USA the level has been set at 1.5 x109 white cells, in the context of clinical trials of chemotherapy. But in fact 4.5% of healthy black subjects in the NHANES survey have cell counts below this, as do 0.79% of white subjects, but only 0.38% of Mexican-Americans.

A systematic review of the long-term effects of false-positive mammograms shows some interesting cultural differences: following a false alarm, women in Canada and Europe tend to have fewer mammograms, while those in the USA tend to have more. Neither show much evidence of deep psychological scarring.

In the USA, mammography is good business, and is seen as so obviously beneficial that women in their 40s are clamouring to have it, and most do. The editorial is the thing to read: it surveys the systematic review and the American College of Physicians Guideline and concludes that there is an awful lot of fudging going on for the sake of the status quo. There is no clear evidence of benefit, and some of harm. If the arguments for mammography over the age of 50 are weak, those for mammography under that age are non-existent.

Another editorial worth looking up in the paper copy if you don’t have on-line access is one on new drugs for macular degeneration. Fantastically expensive intravitreous injections of ranibizumab ($2,000 per go) have been shown to halt or even reverse the condition in an exemplary sham-controlled trial. Bevacizumab, originally developed for the treatment of metastatic colorectal cancer, has not undergone such a trial but seems to work equally well at a cost of $50 per injection. The sooner this is confirmed in a good RCT, the better. I am not getting any younger.