Research highlights – 5 November 2010

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.

Colorectal cancer and decision aids
As discussed elsewhere in this weeks’ BMJ, shared decision making for patients is a major part of the UK government’s plans for the NHS, in which evidence based decision aids (it’s hoped) will play a part. In a randomised controlled trial Sian Smith and colleagues investigated a tool to assist informed choice about screening for colorectal cancer.

The decision aid included illustrations and a supplementary video and was aimed at adults with low educational attainment, with the trial population recruited from disadvantaged areas in Australia. The results showed that the tool improved participants’ ability to make an informed choice about faecal occult blood testing, leading the authors to recommend its integration into primary care.

However, the study also found reduced uptake of screening among participants who used the decision aid. Since screening is known to reduce mortality from bowl cancer, says editorialist Hilary L Bekker the effect of this decision aid is in conflict with the objectives of screening programmes. She proposes the aim should be “informed uptake rather than informed decision making.”

In another of this week’s papers, R J C Steele and colleagues report that sending repeat invitations increased the uptake of screening for colorectal cancer, and that incidence screening detected important neoplastic disease. Meanwhile, Helene Kirkegaard and colleagues found that a substantial amount of mortality from colon cancer might be attributable to a failure to adhere to lifestyle recommendations.

Treating ankle sprains
Received wisdom these days seems to be that exercises are a more beneficial form of treatment for ankle injury than immobilisation. Indeed, a randomised controlled trial published in the BMJ earlier this year showed that early therapeutic exercise produced a greater improvement in ankle function in the first two weeks after injury than conventional treatment with protection, rest, ice, compression, and elevation.

On the other hand, a systematic review from the Netherlands published in this week’s issue is less equivocal, concluding that adding supervised exercises to conventional treatment for patients with an acute lateral ligament ankle sprain only modestly improves recovery and time to return to activity compared with conventional treatment alone.

The authors do point out that they were able to identify only 11 eligible studies for their review, however, most of which had a high risk of bias or were lacking power.

Perinatal mortality in the Netherlands
The second Dutch study in this week’s issue looks at perinatal mortality and severe morbidity in births that take place in primary care compared with those that occur in secondary care. The obstetric care system in the Netherlands is based on risk selection at two levels: pregnant women and women in labour are divided into a low risk group, who are supervised during birth by a midwife at home or in a midwifery unit, and a high risk group, who are supervised by an obstetrician in a hospital.

Despite a high standard of medical care, the Netherlands has one of the highest perinatal mortality rates in Europe (10.0 per thousand in 2004), so the authors of this study tried to establish whether the two tier obstetric care system is a contributing factor.

They found that infants of low risk pregnant women who started labour in primary care had a higher risk of delivery related perinatal death than did infants of high risk pregnant women who started labour in secondary care. The highest mortality was among infants of women who were referred from primary care to secondary care during labour because of an apparent complication.

The authors suggest that the difference in outcome could be because some high risk women were inappropriately booked for labour care under a midwife and, therefore, the system for risk assessment in the Netherlands may be inadequate.

Research online: For these and other new research articles see

PROMS in cardiovascular trials
Patient reported outcomes help to assess the impact of interventions from the patients’ perspective, but how relevant are such outcomes to contemporary cardiovascular trials, and do investigators use them appropriately? Kazem Rahimi and colleagues did a systematic review and analysed 413 trials to address the question. Sixty five trials used at least one measure of patient reported outcomes; in most of these studies, such information would have been important for clinical decision making. The authors judged patient reported outcomes to be of little value for a large number of (mostly explanatory) cardiovascular trials—but of 174 trials in which such outcomes would have been important for clinical decision making, 122 did not report them (doi:10.1136/bmj.c5707).