Research highlights – 13 May 2011

Research questions“Research highlights” is a weekly round-up of research papers appearing in the print BMJ. We start off with this week’s research questions, before providing more detail on some individual research papers and accompanying articles.

Thyroid autoantibodies, normal thyroid function, and risks to the fetus
The suggestion that women with thyroid autoantibodies are at increased risk of having miscarriages and preterm births, even when their thyroids are functioning normally, goes back 20 years. How strong are these associations, and might treatment protect the fetus? 

To answer these questions Shakila Thangaratinam and colleagues systematically reviewed and meta-analysed 31 observational studies in more than 12 000 women, plus three randomised trials of levothyroxine in pregnancy. They confirmed significant associations between the presence of thyroid autoantibodies and both miscarriage and preterm birth—and, for both outcomes, they found limited evidence from trials that treatment with levothyroxine significantly reduces the risks.

This work is a prelude to the same authors’ TABLET (Thyroid AntiBodies and LEvoThyroxine) trial, which gets underway this year. This is a randomised, placebo-controlled, double-blind trial in more than 20 British hospitals, testing the hypothesis that in euthyroid women with thyroid peroxidase antibodies, levothyroxine (50 µg orally once daily) started before conception and continued to the end of pregnancy, increases live births beyond 34 completed weeks of gestation by at least 10% compared with placebo (

Whether thyroid antibodies directly cause pregnancy loss and preterm delivery or are simply an epiphenomenon remains unanswered, note Roberto Negro and Alex Stagnaro-Green in a linked editorial. They see Thangaratinam and colleagues’ meta-analyses as a turning point in this story, and say it’s now time to do decent benchside studies of the causal mechanisms as well as the TABLET trial.

β blockers in chronic obstructive pulmonary disease
Many patients with chronic obstructive pulmonary disease (COPD) also have heart disease, and some die of heart failure. Can such patients take β blockers, or do the risks of acute airways obstruction associated with these drugs outweigh the cardiac benefits?

Philip M Short and colleagues tried to answer this question using data from several linked NHS databases in Scotland to conduct a retrospective cohort study with nearly 6000 patients aged over 50 whose COPD met the Global Initiative for Chronic Obstructive Lung Disease (GOLD) criteria. They found that adding β blockers (most of them cardioselective) to standard management of COPD was associated with lower mortality at each treatment step and a 22% overall reduction in mortality over the next four years. They minimised the impact of confounding by indication by using a Cox proportional hazard regression model and a matched propensity scoring analysis. 

Editorialists Shamsah Kazani and Elliot Israel discuss the limitations of this retrospective analysis but agree that it adds useful evidence. They call for prospective studies on this question and give some practical advice: “How should clinicians start a patient with COPD on a β blocker for cardiac indications? It is advisable to use a cardioselective β blocker and to observe the patient during the administration of the first dose. In the rare case that bronchospasm does occur, anticholinergic agents should probably be used first.”

Proton pump inhibitors  and risk of adverse cardiovascular events in aspirin treated patients with first time myocardial infarction
In this Danish study, Mette Charlot and colleagues report that treatment with proton pump inhibitors was associated with an increased risk of adverse cardiovascular events for these patients (doi:10.1136/bmj.d2690).

Inadequate reporting of research ethics review and informed consent in cluster randomised trials
In a sample of 300 trials published in 150 journals, around a quarter failed to report ethics review, found Monica Taljaard and colleagues (doi:10.1136/bmj.d2496).