Research highlights – 17 and 24 September 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.

Poor physical capability linked with mortality in older people

Whether someone can manage the physical tasks of everyday living can reflect on their general health. Rachel Cooper and colleagues undertook a systematic review and meta-analysis to investigate whether this ability, known as physical capability, has any effect on mortality among people in the community.

Overall, the authors found that people who performed less well in four objective tests of physical capability—grip strength, walking speed, chair rising, and standing balance times—were at higher risk of all cause mortality. The association of walking speed, chair rising, and standing balance with mortality only applied to adults aged more than 70; however, the association of grip strength with mortality was also found in populations under 60.

This research received substantial media coverage, although many outlets took the main message to be that a strong handshake increases longevity. BBC News, for example, ran the headline “Firm handshake link to long life.”

Grip strength, however, is more than just the strength of a person’s handshake and is rather a technical, physiological characteristic that is measured with a machine.

Supervision of methadone therapy

Methadone maintenance treatment can substantially reduce mortality among people dependent on opioids such as heroin. But misuse, unsanctioned coadministration, and poor compliance can all lead to death from overdose. In fact, during the 1990s, methadone was implicated in as many drug related deaths in the United Kingdom as was heroin. In the mid to late 1990s, prescribing practice in the UK was changed with the aim of reducing deaths from overdose; daily dispensing and supervision of methadone dosing were introduced or expanded.

John Strang and colleagues looked at publicly available mortality data from England and Scotland between 1993 and 2008 to assess the effect of these changes. They saw a striking reduction in deaths caused by methadone overdose in both Scotland and England over that period, with the timing of the improvements related to the introduction of widespread supervision of methadone dosing.

The authors believe that this improvement in safety might previously have gone undetected because methadone prescribing also increased substantially over the same period. Here they have controlled for that confounding by using an index of the number of deaths related to methadone per million defined daily doses. Given the continuing political controversy over opiate substitution, tools like this index that help to accurately assess the value of policy changes are welcome.

Venous thromboembolism and antipsychotic drugs

The BMJ has published several studies arising from the QRESEARCH database, which holds the anonymised primary care clinical records of more than 11 million people registered over the past 16 years with more than 500 UK general practices.

The latest, from Chris Parker and colleagues, is a nested case-control study showing that prescribing of antipsychotic drugs (for indications including nausea and vertigo as well as mental health disorders) is associated with significantly increased relative risks of venous thromboembolism, particularly among new users and those prescribed atypical antipsychotic drugs.

The relative risks look alarming. But the authors rightly report the absolute risks too, which show four extra cases annually per 10000 among patients of all ages and 10 extra cases per 10000 among the over 65s.

Research online: For this and other new research articles see

Techniques to combine data in mixed methods studies

Mixed methods studies that collect both qualitative and quantitative data—for example, patient views on a new intervention as well as information on its efficacy—are becoming more common in health research, but integrating these two types of data is tricky.

Alicia O’Cathain and colleagues cover in a Research Methods and Reporting paper three ways to combine qualitative and quantitative data in mixed methods studies.

The techniques they recommend are triangulation—comparing the findings from both approaches to find common themes and important differences—”following a thread”—following a question or theme from one approach across to the other approach—and creating a mixed methods matrix—using a table where rows contain cases that have both qualitative and quantitative data and columns display different data collected on each case.

By combining qualitative and quantitative data effectively, mixed methods research should yield a “whole greater than the sum of the parts.”

 Screening for prostate cancer

Screening based on prostate specific antigen (PSA) has led to increased rates of diagnosis of prostate cancer, but whether this translates into more lives saved has been in doubt. Mia Djulbegovic and colleagues aimed to clarify the matter with a systematic review which identified six randomised controlled trials for meta-analysis.

Their results confirmed that screening increased the probability of being diagnosed with prostate cancer—although the quality of evidence was low—but they showed no effect on overall or disease specific mortality, based on moderate quality evidence. Little information was available about the potential harms of screening.

The authors conclude that we don’t have enough good evidence to justify routine population screening at present.

As editorialist Gerald Andriole Jr points, out another concern is the substantial human and economic cost associated with screening—largely down to overdiagnosis.

Andrew Vickers and colleagues investigated an approach that might help to limit “overscreening” by identifying men who are most at risk. Their case-control study looked at the relation between concentration of PSA at age 60 and subsequent diagnosis of clinically relevant prostate cancer in an unscreened population of over 1000 Swedish men. The results indicated that this measure could predict lifetime risk of metastasis and death from prostate cancer. Since the great majority of deaths from prostate cancer were in men who had PSA concentrations in the top quarter at age 60, the authors suggest that screening should focus on this high risk subgroup. These results require further validation, and in the meantime Andriole advises clinicians to individualise their approach to screening on the basis of factors such as age and family history.

Screening for Barrett’s oesophagus—a new approach

The usual way to detect Barrett’s oesophagus is by using white light gastroscopy and biopsy, but this approach is invasive and expensive. Sudarshan Kadri and colleagues’ prospective cohort study suggests that a new device called the Cytosponge, when coupled with a single immunomarker trefoil factor 3, might be a promising new way to screen for Barrett’s oesophagus.

The Cytosponge is an ingestible gelatine capsule that contains a compressed mesh attached to a string. The capsule and string are swallowed with water, where the capsule dissolves in the stomach. The mesh then expands and is withdrawn by pulling on the string, thus obtaining a specimen from the oesophagus that can be examined in the laboratory.

The Press Association reported this story with the headline “‘Sponge’ could help prevent cancer.” As Peter Bampton points out in his linked editorial, Barrett’s oesophagus increases the risk of oesophageal cancer by 30-40-fold. By detecting Barrett’s oesophagus early, the Cytosponge could help make sure these patients receive appropriate treatment and don’t “convert” to adenocarcinoma. But he warns that this approach does not fulfil the criteria for a population screening programme.

Nearly all (99%) of the patients in this study of diagnostic accuracy were able to swallow the Cytosponge, and most (82%) reported low levels of anxiety before and after the test, indicating that this approach is acceptable to patients. Compared with gastroscopy, the sensitivity and specificity of the test for segments of 1 cm or more were 73.3% and 93.8%, respectively, whereas these values were 90.0% and 93.5% for segments of 2 cm or more.

Elsewhere, Janusz Jankowski and colleagues discuss the natural history and diagnosis of Barrett’s oesophagus in a clinical review. They also cover treatments to prevent progression of Barrett’s oesophagus to adenocarcinoma and highlight new NICE guidelines that recommend clinicians consider offering endoscopic ablative therapy as an alternative to oesophagectomy for people with high grade dysplasia and intramucosal cancer.

 Podcast: Barrett’s oesophagus

Rebecca Fitzgerald, an author of this week’s paper on testing for Barrett’s oesophagus, talks to Duncan Jarvies about how improvements in treatment for the disease have spurred on the search for population screening methods, and about how the suitability of the Cytosponge for such a role is being evaluated.