Research News

All you need to read in the other general journals

BMJ 2012; 345 doi: (Published 04 July 2012) Cite this as: BMJ 2012;345:e4517

Hydroxyethyl starch linked to excess deaths from acute severe sepsis

More safety concerns about the colloid hydroxyethyl starch (HES) have emerged from a randomised trial of adults with severe sepsis. Those given HES 130/0.4 for volume expansion were significantly more likely to die than controls given Ringer’s acetate solution (51% (201/398) v 43% (172/400); relative risk 1.17, 95% CI 1.01 to 1.36). They were also more likely to need renal replacement treatment during 90 days of follow-up (22% (87/398) v 16% (65/400); 1.35 1.01 to 1.80).

The authors recruited patients from 26 intensive care units in Scandinavian hospitals and allowed local doctors to make all treatment decisions except choice and maximum volume of resuscitation fluid. They tested a hydroxyethyl starch solution with a lower molecular weight than traditional starch solutions to find out if the different formulation would be safer. It wasn’t. The colloid caused an excess of deaths in both per protocol and intention to treat analyses, with an estimated number needed to harm of just 13. Adults resuscitated with HES 130/0.4 needed more blood products than did controls. Both groups received about the same volumes of resuscitation fluid overall. The trial was double blind and independently funded.

Results were broadly in line with the results of smaller trials testing older formulations of HES. The increased mortality took about a month to emerge, and the authors suggest that toxic effects of HES on kidney, liver, and bone marrow were responsible for the late extra deaths. Participants had a mean age of 66 years, and 84% were in shock when randomised. The lungs and abdomen were the most common sources of sepsis.

Assisted reproduction works well for young women

OpenUrlCrossRefPubMedWeb of Science

Women and couples embarking on a course of in vitro fertilisation want some idea of their eventual chances of a live baby if they keep trying. An analysis of national outcomes data from the US suggests that the answer for women under 31 lies somewhere between 63.3% and 74.6%. These are more or less conservative estimates for the cumulative chance of a live birth after up to three cycles of treatment using autologous oocytes. Cumulative live birth rates fell steadily with age, to between 6.6% and 11.3% for women of 43 or over. Births continued to accumulate, but more slowly, after more than three cycles of treatment. Women using donor eggs were more successful, with cumulative live birth rates between 60% and 80% for all age groups, over three cycles.

The authors analysed data from 246 740 women who started treatment between 2004 and 2008 at clinics across the US. More than half (57%) had a live birth. Among women using their own fresh eggs, transfer of two embryos was associated with higher cumulative birth rates than transfer of one. Transfer of embryos aged 5 or 6 days looked more successful than transfer of embryos aged 2 or 3 days. Infertility caused by male factors, endometriosis, tubal problems, or polycystic ovary syndrome (PCOS) was easier to overcome than infertility caused by diminished ovarian reserve or uterine factors.

Under the most favourable conditions, the chances of a live baby after assisted reproduction come close to the chances reported for women trying to conceive naturally, say the authors. This was especially so when there were no social or economic barriers to assisted reproduction.

Surgery sooner rather than later for selected adults with severe infective endocarditis

OpenUrlCrossRefPubMedWeb of Science

Emergency valve surgery (replacement or repair) is already recommended for people with severe infective endocarditis and heart failure. Early surgery may also be best for people without heart failure who have large vegetations and valve dysfunction, say researchers, after their small trial showed that this strategy helped prevent systemic emboli without increasing mortality.

All the participants had severe left sided endocarditis and large vegetations. Most had mitral or aortic regurgitation, and almost half had had a systemic embolus. None of the 37 patients who had early surgery had further emboli, compared with eight of the 39 patients treated initially with antibiotics (0% v 21%; P=0.005). One patient in each group died within six weeks.

The two groups had comparable mortality at six months, but the combination of death, embolic events, recurrence of infective endocarditis, or hospital admission for new heart failure was significantly more common after conventional treatment (28% v 3%; hazard ratio for early surgery 0.08, 95% CI 0.01 to 0.65). The researchers treated their control group according to American Heart Association guidelines, which allow surgery for recurrent emboli or persistent valve vegetations. Thirty controls needed surgery after initial medical treatment and 27 had surgery during their first hospital admission.

These findings should help clarify a grey area of practice, says a linked editorial (p 2519). But selecting the right patients for early surgery could still be a challenge for non-experts. The editorial’s authors recommend rapid referral to a specialist centre for all adults with left sided infective endocarditis complicated by valve dysfunction, large vegetations, or invasive disease beyond the cusps or leaflets.

“Non-inferior” linagliptin could be safer than sulphonylureas

Boehringer Ingelheim is currently evaluating its new antidiabetic agent, linagliptin, as a second line treatment for type 2 diabetes in adults who are inadequately controlled by metformin alone. In a head to head trial, linagliptin was slightly less effective over two years than the sulphonylurea, glimepiride, but it still passed the manufacturer’s non-inferiority test. Mean glycated haemoglobin (HbA1c) concentration dropped 0.16 of a percentage point in the group taking metformin plus linagliptin and 0.36 of a percentage point in the group taking metformin plus glimepiride (mean difference 0.20 percentage points; 97.5% CI 0.09 to 0.30; P=0.0004). Both groups started the trial with a mean HbA1c of 7.7% (60 mmol/L).

Hypoglycaemia and weight gain are two of the main drawbacks associated with sulphonylureas. Linagliptin caused significantly less of both, and severe hypoglycaemia was virtually absent from the linagliptin group. A significant reduction in major cardiovascular events was more of a surprise (2% (12/776) v 3% (26/775); relative risk 0.46, 0.23 to 0.91). The authors and linked comment (doi:10.1016/S0140-6736(12)60859-9) agree that this result needs confirming—the trial wasn’t big enough to rule out the play of chance.

Linagliptin is one of a handful of new antidiabetic drugs that inhibit the enzyme dipeptidylpeptidase-4 (DPP-4) and increase circulating concentrations of glucagon-like peptide-1. All are jostling for position among other second line agents, says the comment. Lack of weight gain and a better safety profile than sulphonylureas are important strengths, but we don’t yet know if the long term benefits are big enough to justify the extra cost.

Early intravenous aspirin does not help adults with ischaemic stroke

The addition of intravenous aspirin to thrombolysis with alteplase failed to improve neurological outcomes for adults with acute ischaemic stroke in a recent trial. The extra aspirin, given within 90 minutes of alteplase, caused significantly more intracranial bleeds than alteplase alone early (4.3% (14/322) v 1.6% (5/320); P=0.04), and a data monitoring board decided to stop the trial. Participants given intravenous aspirin had more serious adverse events than controls and slightly more died within three months, although the difference wasn’t significant (11.2% (36/322) v 9.7%(31/322)). All participants started oral aspirin 24 hours after thrombolysis, as recommended by international guidelines.

The authors describe their trial as pragmatic and low budget. They couldn’t afford a placebo, so patients and doctors knew which treatment group they were in. Doctors might have looked harder for intracranial bleeding in patients given intravenous aspirin, and we should interpret these results with care, they write. Perhaps, says a linked comment (doi:10.1016/S0140-6736(12)61043-5), but, overall, the findings are secure enough and support current best practice. Eligible adults with acute ischaemic stroke need oral aspirin a day after thrombolysis not intravenous aspirin given earlier.

Future attempts to improve cerebral reperfusion after stroke should start with more sophisticated imaging to select patients, says the comment. We need to see blood vessels and assess occlusion directly, preferably with multimodal computed tomography or magnetic resonance imaging.

MRSA bacteraemia falls significantly across the US

OpenUrlCrossRefPubMedWeb of Science

More than nine million men, women, and children of all ages are entitled to medical care at facilities run by the US Department of Defense. A detailed epidemiological study of Staphylococcus aureus infections in this population reported a significant drop in the annual incidence of meticillin resistant bacteraemia between 2005 and 2010. Resistant bacteraemia acquired in the community fell in parallel with resistant bacteraemia acquired in hospital (from 1.7 to 1.2 per 100 000 and from 0.7 to 0.4 per 100 000). The incidence of meticillin sensitive bacteraemia fell significantly during the same period.

Trends in skin and soft tissue infections were less clear cut, although the proportion of community acquired infections caused by meticillin resistant S aureus (MRSA) fell from 62% in 2006 to an all time low of 52% in 2010 (P<0.001).

The very young and the very old had the highest incidence of staphylococcal bacteraemia acquired in the community. Young men, including those on active service, had the highest incidence of staphylococcal skin and soft tissue infections acquired in the community.

Although this population is healthier and more affluent than the general US population, it is diverse enough to signal that the epidemiology of staphylococcal infections might be changing nationally, say the authors. Further studies will need to explore the factors driving these changes.

Herpes zoster vaccination for adults with immune mediated diseases?

OpenUrlPubMedWeb of Science

Live vaccines against herpes zoster are contraindicated in adults taking biological treatments for immune mediated diseases, such as rheumatoid arthritis. Experts worry that a live vaccine might trigger the infection it was designed to prevent, in this case shingles. A recent analysis of Medicare data from the US could find no evidence of a link between vaccination and shingles in older adults treated for various immune mediated diseases, and the authors say a blanket ban should be revisited.

The 463 541 participants in the cohort had psoriasis, rheumatoid arthritis, inflammatory bowel disease, or ankylosing spondylitis. The crude incidence of herpes zoster was 7.8 cases per 1000 person years among the 7780 who were vaccinated and 11.6 cases per 1000 person years among the rest. Just 663 adults were treated with biological treatments at the time of vaccination. None developed herpes zoster within 42 days. Vaccination was associated with a significantly lower risk of shingles over two years of follow-up in adjusted analyses of the whole cohort.

No safety signals emerged from this admittedly imperfect dataset, say the authors. A randomised trial is now justified.


Cite this as: BMJ 2012;345:e4517

View Abstract

Sign in

Log in through your institution

Free trial

Register for a free trial to to receive unlimited access to all content on for 14 days.
Sign up for a free trial