“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.

• What clinical characteristics could help to rule out subarachnoid haemorrhage in people with acute severe headache?
• What is the effect of vitamin E supplementation on stroke?
• What is the cost effectiveness of one-off screening for chronic kidney disease?
• Are patient reported outcomes relevant and appropriately used in cardiovascular trials?

Spotting subarachnoid haemorrhage
How do you know when a headache is a sign of something more serious, like intracranial bleeding? Identifying subarachnoid haemorrhage in patients who present to the emergency department can be tricky (particularly in patients who are “neurologically intact”) and involve expensive medical imaging and invasive procedures including dural and lumbar puncture.

Jeffrey Perry and colleagues looked at nearly 2000 patients presenting with severe headache at six university teaching hospitals in Canada to see whether any clinical characteristics might predict a diagnosis of subarachnoid haemorrhage (p 000). They identified seven variables strongly and reliably associated with the condition: age =40, witnessed loss of consciousness, complaint of neck pain or stiffness, onset with exertion, arrival by ambulance, vomiting, and raised diastolic ( =100 mm Hg) or systolic (=160 mm Hg) blood pressure.
The authors of this study are now carrying out prospective validation of three clinical decision making rules formulated on the basis of their findings. “Since 4% of all emergency department visits are for headache and everyone is worried about missing subarachnoid haemorrhage, this is a potentially important paper,” says Elizabeth Loder, clinical epidemiology editor at the BMJ and chief of the Division of Headache and Pain in the Department of Neurology at the Brigham and Women’s/Faulkner Hospitals in Boston.

Supplements and stroke
Up to 12.7% of adults in the United States take vitamin E supplements in the hope of staving off cardiovascular disease, but Markus Schürks and colleagues have found that these supplements increase the risk of haemorrhagic stroke by more than 20% (p 000). Interestingly, their meta-analysis of almost 120000 patients also found that supplementation reduced the risk of ischaemic stroke by 10%.
The authors stress that the absolute risks are small: vitamin E could cause one additional haemorrhagic stroke for every 1250 people taking the supplement but would prevent one ischaemic stroke per 476 people. However, given that haemorrhagic stroke is associated with worse outcomes and the risk reduction for ischaemic stroke is modest, they caution against indiscriminate widespread use of vitamin E supplements and recommend other strategies to prevent ischaemic stroke, such as blood pressure and cholesterol lowering drugs and having a healthy lifestyle.

Peter Coleman, deputy director of research at the Stroke Association, told BBC news: “This is a very interesting study that shows that the risk of haemorrhagic stroke can be slightly increased by high levels of orally taken vitamin E, although what is a high level has not clearly been ascertained. More research is required to discover the mechanism of action and the level at which vitamin E can become harmful.”

Population screening for chronic kidney disease
The <a href="www.nice.org.uk/nicemedia/live/12069/42119/42119.pdf"2008 NICE guideline on chronic kidney disease (CKD) says: “Offer people testing if they have any of the following risk factors: diabetes, hypertension, cardiovascular disease (ischaemic heart disease, chronic heart failure, peripheral vascular disease and cerebral vascular disease), structural renal tract disease, renal calculi or prostatic hypertrophy, multisystem diseases with potential kidney involvement, family history of stage 5 CKD or hereditary kidney disease, or opportunistic detection of haematuria or proteinuria” and recommends annual testing with estimated glomerular filtration rate (eGFR) for all of these groups.

But is there also a case for population based screening? It seems not, as 16-21 people would have to be screened to detect one case of disease and this would not be cost effective. But studies so far have been based on screening with urinalysis for proteinuria and macroalbuminuria, and now Braden Manns and colleagues have brought the evidence up to date by modelling the cost effectiveness of screening in primary care with a one off test for eGFR (p 000). They found that, in a cohort of 100000 people, screening and subsequent treatment with angiotensin blockade would cut the number of people developing end stage renal disease over their lifetime from 675 to 657, with an unfavourable incremental cost per QALY of more than £62 000. For people with diabetes, however, routine screening looks promising and has a cost per QALY similar to other publicly funded interventions.

Research online: For these and other new research articles see www.bmj.com/research

Detecting endobronchial intubation

Endotracheal intubation is routinely performed by clinicians with different levels of experience, but misplacement of an endotracheal tube in a mainstem bronchus can lead to serious complications. Christian Sitzwohl and colleagues did a randomised trial to determine which bedside method of detecting inadvertent endobronchial intubation in adults was most sensitive and specific. Clinicians were randomly assigned to perform bilateral auscultation of the chest (the currently recommended method); observation and palpation of symmetrical chest movements; estimation of the tube position by the insertion depth; or all three. When using auscultation, doctors with limited experience missed over half of endobronchial intubations, and even experienced anaesthetists were often unable to detect the problem. When tube insertion depth was used, sensitivity was 85% in first year residents and 90% in experienced anaesthetists; optimal tube depth was 20 cm for women and 22 cm for men. The highest sensitivity and specificity was achieved by combining all three methods.