Research highlights – 18 March 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.

End of a short story?

Although growth hormone therapy can increase growth rate, controversy continues over its use in children with idiopathic short stature, partly because this is a heterogeneous condition that includes normal variants of growth. Trials have also tended to be too small, short-term, and inconsistent to ascertain the benefits for adult height. Because of this uncertainty, the European Agency for the Evaluation of Medicinal Products has not yet approved use of growth hormone for this indication.

Annalisa Deodati and Stefano Cianfarani systematically reviewed all randomised and non-randomised trials of growth hormone therapy in children with idiopathic short stature, defined as height >2 SD score below the mean. Of 19 studies included, only three were randomised controlled trials carried out to adulthood. The limited evidence indicated that growth hormone therapy does increase adult height, but the increase was small compared with those achieved in other conditions and it varied greatly between individuals.

Reviewers thought the study provided “the ending of the story,” since funding for further trials is unlikely. The authors suggest that debate should now shift to whether the gains seen here—around 4 cm—are of true value.

In an editorial, Kerstin Albertsson-Wikland emphasises that the response to growth hormone treatment can vary substantially between individuals and also at different stages of growth. Therefore, she says, it would be safer and more efficient to tailor doses to the individual, rather using a standard dose based on the child’s size—the approach that is currently recommended by drug agencies and that was used in this study.

Increased mortality with osteoarthritis
Eveline Nüesch and colleagues report that patients with osteoarthritis are at higher risk of premature death compared with the general population and that most of the excess is due to deaths from cardiovascular disease.  In their cohort study, 1163 patients aged 35 years or over with symptomatic and radiographically confirmed osteoarthritis of the knee or hip showed a 55% excess in all cause mortality. Major risk factors included a history of cancer, diabetes, or cardiovascular disease, but the most striking association was with walking disability. In essence, “the more severe the restriction in walking ability the more likely a person was to die early,” and most of the excess deaths associated with walking problems had cardiovascular causes. The authors recommend that management of patients with osteoarthritis and walking disability should focus on treatment of cardiovascular risk factors and increasing physical activity.
In their editorial, Cyrus Cooper and Nigel Arden concur that evaluation and amelioration of cardiovascular risk factors (which already routinely occurs with rheumatoid arthritis) could usefully be extended to management of osteoarthritis. They also discuss the idea that osteoarthritis should be seen as a manifestation of biological ageing (as osteoporosis and sarcopenia already are), potentially with new approaches to prevention and treatment.

Reporting genetic risk: time to get a GRIPS
Genetic risk prediction studies usually analyse large numbers of genetic variations (from tens to tens of thousands) using statistical algorithms and models of variable complexity. All too often they are poorly reported, despite existing guidance on reporting observational study designs (STROBE statement), genetic association studies (STREGA), prediction models (REMARK), and test evaluation (REMARK and STARD). 

So an international group of risk prediction researchers, epidemiologists, geneticists, methodologists, statisticians, and journal editors has developed the GRIPS (Genetic RIsk Prediction Studies) statement to help authors report this type of work fully (Research Methods and Reporting).

As well as covering the building, validation, and limitations of the statistical models, the 25 point GRIPS checklist asks for the down to earth stuff that often gets missed, such as “Report the numbers of individuals at each stage of the study. Give reasons for non-participation at each stage. Report the number of participants not genotyped and reasons why they were not genotyped” and “Discuss the generalisability and, if pertinent, the healthcare relevance of the study results.”

The supporting  “explanation and elaboration” documentation for GRIPS is published in PLoS Medicine, and the statement is being copublished in nine journals, to ensure that it reaches both specialists and generalists.

Prevention of pain on injection of propofol
Leena Jalota and colleagues aimed to determine the most efficacious approach for preventing the pain caused by injection of propofol to induce anaesthesia. Their systematic review and meta-analysis suggests it’s use of the antecubital vein, or pretreatment using lidocaine in conjunction with venous occlusion for the hand vein, but other useful options are discussed.

Effect of statins on atrial fibrillation
Kazem Rahimi and colleagues pulled together evidence from both published and unpublished trials in a meta-analysis. Their results failed to back up the suggested beneficial effect of statins on atrial fibrillation shown in published shorter term studies.