Don’t look away now

This week we have updates on several medical stories that seem to have been running for ages. But they’re still very much worth your attention.

For more than a decade we’ve been publishing news stories reporting that one or other jurisdiction has legalised cannabinoids for one or other medical indication. But we haven’t been very curious about whether they actually worked. At our request Michael Farrell, Rachelle Buchbinder, and Wayne Hall pored over the literature. They came away distinctly underwhelmed (doi:10.1136/bmj.g2737).

Nick Freemantle and Greta Rait couldn’t even manage ambivalence when they read the DAMASCENE study of the relation between the number of discrepancies in trial reports and the effect size of bone marrow stem cells in heart disease. In summary, the study found that the dodgier the trial, the larger the effect size (doi:10.1136/bmj.g2688). Trials with no errors found no benefits. The trial reports included in the DAMASCENE study are evidence of “shambolic and poorly conducted research that reflects poorly on every part of the research process,” conclude Freemantle and Rait in their editorial (doi:10.1136/bmj.g2750). They attribute some of the problem to the lack of scrutiny received by non-commercial trials in comparison with that received by trials contributing to market authorisation.

Another candidate for the rubbish bin may be therapeutic cooling after out of hospital cardiac arrest. Stephen Bernard (doi:10.1136/bmj.g2735) dates the enthusiasm for this practice to two papers published in 2002. But earlier this year a huge trial involving hospitals in Europe and Australia found no benefit. This trial “should change practice immediately,” writes Bernard. “Attention should now turn to rigorous investigation of other aspects of care after cardiac arrest that have been proposed for this patient group.” How many of them will survive such scrutiny?

Sadly, it’s impossible to be emphatic about every contentious medical question. One of the most vexed of the day is whether vitamin D deficiency is the root of many evils. Good research supporting this widespread belief is elusive. A few weeks ago we published a meta-analysis reporting that, overall, vitamin D supplements had no effect on mortality, though subgroup analysis indicated that vitamin D3 reduced mortality (doi:10.1136/bmj.g1903). In their letter to The BMJ Mark Bolland and colleagues deny the true believers even this modest consolation (doi:10.1136/bmj.g2860). When four ineligible trials are excluded from analysis the benefit disappears.

Electronic cigarettes continue to exercise us—and editorialists at large. Do e-cigarettes represent strategies for harm minimisation (good) or a gateway to the real thing (bad)? The Financial Times is in no doubt. As “the aim of health regulation must always be to reduce harm,” e-cigarettes “should therefore be embraced as part of the solution to the growth of cancer worldwide.” Regulators should “follow the science,” which “provides overwhelming evidence that electronic cigarettes are a benefit” (http://on.ft.com/1tZn015).

On the other hand the Lancet worries about the potential for the “increased uptake of conventional cigarettes, renormalisation of cigarette smoking, and increases in nicotine addiction.” It notes the “alarming expansion of the e-cigarettes market and active targeting of youth” and wants other countries to follow the European Union’s suit and urgently regulate e-cigarettes (doi:10.1016/S0140-6736(14)60701-7).

In such illustrious company it may seem quaint to quote anything as old fashioned as data. Zosia Kmietowicz’s news story provides data from a survey commissioned by the UK antismoking charity Action on Smoking and Health (doi:10.1136/bmj.g2987). It found that the use of e-cigarettes has tripled in the past two years in the UK, with 2.1 million people now regularly using them. Yet “use of e-cigarettes among people who had previously never smoked remains negligible, with only about 1% of the respondents trying them and virtually none continuing to use them.”

On this issue at least, the jury seems very much still out.