You don't need to be signed in to read BMJ Blogs, but you can register here to receive updates about other BMJ products and services via our site.

Welcome to the BMJ Open blog. BMJ Open is an open access journal, dedicated to publishing medical research from all disciplines and therapeutic areas.

Find out more about the journal here.

We will be updating the blog with news about the journal, highly accessed papers, press coverage, events and matters of interest in the open access and publishing world, and anything else that catches our eye.

Higher nurse to patient ratio linked to reduced risk of inpatient death

9 Feb, 16 | by Emma Gray

Death rates 20 per cent lower in hospitals with 6 or fewer patients per registered nurse

A higher nurse to patient ratio is linked to a reduced risk of inpatient death, finds a study of staffing levels in NHS hospitals, published in the online journal BMJ Open.

In trusts where registered (professionally trained) nurses had six or fewer patients to care for, the death rate was 20 per cent lower than in those where they had more than 10.

Policies geared towards substituting registered nurses with healthcare support workers (healthcare assistants and nursing auxiliaries) should at the very least be reviewed, conclude the researchers

Previous research suggests that the nurse to patient ratio has some bearing on outcomes, but few studies have taken account of the impact of other members of medical staff.

They therefore included two measures over two years (2009-11) in their analysis: the number of beds per registered nurse, doctor, and healthcare support worker in 137 acute care trusts; and the number of patients per ward nurse, drawn from a survey of just under 3000 registered nurses in a nationally representative sample of 31 of these trusts (covering 46 hospitals and 401 wards).

They also calculated the predicted number of deaths for medical and surgical inpatients, taking account of influential factors, such as age, other underlying conditions, and number of emergency admissions during the previous 12 months.

The registered nurse headcount varied by as much as a factor of 4 between those at the top and bottom of the staffing scale. Even after taking account of all nursing staff, this variation only dropped to a threefold difference between those with the highest and lowest nurse headcounts.

Among patients admitted to medical wards, higher death rates were associated with higher numbers of occupied beds for each registered nurse and for each doctor employed by the trust.

By contrast, higher numbers of healthcare support workers were associated with higher rates of inpatient death.

When all staff groups were included in the statistical analysis of all 137 trusts, the associations remained significant only for doctors and healthcare support workers.

Among the subsidiary group of 31 trusts, death rates were 35.2/1000 medical admissions, out of a total of 1 260, 558, and 8.9/1000 surgical admissions, out of a total of 1 084 429.

But the death rate was 20% lower in those where each registered nurse cared for an average of six or fewer medical inpatients than in trusts were each registered nurse cared for 10 or more. These associations remained significant after further statistical analysis.

The results on surgical wards were similar, with higher registered nurse to patient ratios associated with a 17% lower inpatient death rate.

This is an observational study, so no firm conclusions can be drawn about cause and effect, added to which the findings on nurse to patient ratios are based on only 31 trusts.

“This [study] does not, in itself, provide a robust basis to identify safe staffing thresholds,” they say. “However, given the overall strength of evidence for an association, it does seem feasible to identify staffing levels where risk to patients is likely to be increased,” they suggest.

Economic pressures and the ageing profile of the nursing workforce internationally all point to a potential future with fewer registered nurses, they warn. But substituting them for less well trained staff may be unwise, they say.

“When determining the safety of nurse staffing on hospital wards, the level of registered nurse staffing is crucial, and there is no evidence to suggest that higher levels of healthcare support worker staffing have a role in reducing mortality rates,” they write.

“Current policies geared towards substituting [these] workers for registered nurses should be reviewed in the light of this evidence,” they conclude.

A podcast for this manuscript is available to listen to here: https://soundcloud.com/bmjpodcasts/bmjopen-podcast-alison-walker-peter-griffiths-30-11-2015-davidd-mixdown/s-7Hqki?in=bmjpodcasts/sets/bmj-open-podcast

The NHS Health Check in England, bereavement by suicide, and public involvement in medical research: Most read articles in January

5 Feb, 16 | by Emma Gray

NHS Health Checks, the risks of bereavement by suicide, and public involvement in medical research

File:USMC-100209-M-1998T-001.jpg

The Top 10 Most Read list for January is almost completely comprised of new entries, with the exception of the ever popular review by Kristensen et al on the effects of statins. In the top spot this month is an evaluation of the first four years of the NHS Health Check in England, by Robson et al. At number two we have a cross-sectional study seeking to test the hypothesis that young adults bereaved by suicide have an increased risk of suicidal ideation and suicide attempt compared to their peers bereaved by other sudden deaths. Parsons et al enter the list at number five with a paper asking what pharmaceutical industry professionals believe about public involvement in medicines research and development, and Joost et al enter at number eight with a paper on persistent spatial clusters of high body mass index.

Rank Author(s) Title
1 Robson et al. The NHS Health Check in England: an evaluation of the first 4 years
2 Pitman et al. Bereavement by suicide as a risk factor for suicide attempt: a cross-sectional national UK-wide study of 3432 young bereaved adults
3 Kristensen et al. The effect of statins on average survival in randomised trials, an analysis of end point postponement
4 Carey et al. Systems science and systems thinking for public health: a systematic review of the field
5 Parsons et al. What do pharmaceutical industry professionals in Europe believe about involving patients and the public in research and development of medicines? A qualitative interview study
6 Friis et al. Gaps in understanding health and engagement with healthcare providers across common long-term conditions: a population health survey of health literacy in 29 473 Danish citizens
7 Ovseiko et al. Markers of achievement for assessing and monitoring gender equity in translational research organisations: a rationale and study protocol
8 Joost et al. Persistent spatial clusters of high body mass index in a Swiss urban population as revealed by the 5-year GeoCoLaus longitudinal study
9 Parsons et al. ‘Please don’t call me Mister’: patient preferences of how they are addressed and their knowledge of their treating medical team in an Australian hospital
10 Greenhalgh et al. Virtual online consultations: advantages and limitations (VOCAL) study

Most read figures are based on pdf downloads and full text views. Abstract views are excluded.

Requesting clinical trial protocols

19 Jan, 16 | by Fay Pearson


We have recently made some changes to our submission system. If we receive a clinical trial, we will now ask that authors include the original trial protocol. Editor-in-chief, Dr Trish Groves, explains why:

Transparent reporting of research has always been at the heart of BMJ Open. From the very start the journal has published all research with open access, open peer review, and sometimes – when authors opt for sharing via the Dryad repository – open data.

Openness is specially important for clinical trials, and that’s why we are now requiring that trialists submit the study protocol to BMJ Open along with their paper:

  • This should be the version of the protocol that gained ethics approval and was used to run the trial
  • If the protocol was amended between getting ethics approval and starting the trial, the authors must say why in their cover letter
  • The submitted protocol should be labelled with a version number and should include a history of substantive amendments, giving the dates when amended
  • If the trial protocol has been published in an open access journal, and if that published version includes all the information and dates as listed above, then authors can simply provide BMJ Open with the reference and link to the published protocol
  • If BMJ Open publishes the submitted trial, the protocol will be posted alongside as part of the pre-publication history

Why is BMJ Open now requiring protocols? For clinical trials (and protocols for clinical trials) BMJ Open has always required prospective registration, and we ask authors to give the trial’s registration ID in the paper’s abstract. This allows editors, peer reviewers, and ultimately readers to see a summary of the planned design of the trial and to compare it with the paper that reports the trial’s results. And dates in the registration record allow us to check that the trial was, indeed, registered prospectively – before any patients were enrolled. Isn’t this enough? Not quite. Trial registration provides a lot of useful information about a planned study, but registries have space only for a minimum of key information.

You can’t interpret the results of a clinical trial without fully understanding how the study was planned, designed, and conducted. Yet it’s all too common for methods to be reported selectively, partly, or not at all. And it has become clear that registration does not always prevent this, particularly when it comes to outcome switching.  Too often, prespecified outcomes are not reported, and others are added and reported in analyses that may not be valid. By making trial protocols available when trial results are reported in BMJ Open, we can all help to optimise the evidence base for treatments and other interventions.

Editors’ Picks 2015: Part Seven

10 Jan, 16 | by Ed Sucksmith

Polypharmacy and potentially inappropriate prescription of medicines

P7

This week we have been looking at some of our staff editors’ favourite articles from 2015. These papers have covered a variety of important issues in medical research, from the transparency of clinical trials to the publication of negative findings. Our final editors’ pick takes us to Ireland, where Professor Tom Fahey and colleagues from the Royal College of Surgeons in Dublin looked at pharmacy claims data to investigate long-term trends in prescribing of medicines between 1997 and 2012. Their findings included the positive news that prescribing quality had improved across the time period. However, there was a substantial increase in the number of people taking multiple regular medicines (known as ‘polypharmacy’), and this significantly predicted potentially inappropriate prescription of medicines. The authors suggested that patients should be made more aware of the trade-off between taking more medicines to prevent disease and the potential for harm from potentially inappropriate prescribing of drugs, which is associated with taking more medicines.

Editors’ Picks 2015: Part Six

9 Jan, 16 | by Ed Sucksmith

The 2D:4D digit ratio and autism risk: is there an association?

P6

In December we surveyed our staff editors to tell us about their favourite articles from 2015. Our penultimate editors’ pick investigates the relationship between autism risk and a putative marker of testosterone exposure: the 2D: 4D digit ratio.

In spite of large investments in autism research, the underlying causes of autism diagnoses remain largely unknown, although twin and family studies suggest that autism is highly heritable. Perplexingly, autism is more frequently diagnosed in males, with male-to-female ratio estimates ranging from 2:1 to 9:1. A popular theory put forward to explain the sex differences behind autism diagnoses is the ‘Extreme Male Brain theory’, which proposes that people with autism are exhibiting an exaggerated profile of the male cognitive profile. A proposed physiological mechanism underlying this theory is the effect of prenatal sex steroids (such as testosterone) on the developing brain. The index to ring finger ratio (2D:4D digit ratio) is believed to be a proxy for fetal testosterone exposure; males have a lower 2D:4D digit ratio on average than females.

In this study by Anna Guyatt and colleagues, the 2D:4D digit ratio was examined in 6015 children from a British birth cohort known as The Avon Longitudinal Study of Parents and Children. The 2D:4D digit ratio was then tested for association with autistic traits in the general population and with autism diagnoses. Analysis of the data did not indicate a significant association between 2D:4D digit ratio and autism diagnoses or high autistic traits in the general population for either sex. Thus, the data did not support the extreme male brain theory of autism.

The study is just one of many published in BMJ Open in 2015 reporting negative findings. We believe that a study should be published irrespective of the direction of the findings providing that the study meets our publication criteria. Publishing negative findings has the added benefit of combating the damaging effects of dissemination bias in medical research (see Editors’ Picks 2015: Part Three)

Editors’ Picks 2015: Part Five

8 Jan, 16 | by Ed Sucksmith

How well reported are interventions in systematic reviews?

P5

In December we surveyed our staff editors to tell us about their favourite articles from 2015. Part 5 focuses on another important editorial issue: the quality of reporting in systematic reviews.

Clinicians and healthcare professionals often rely on systematic reviews and meta-analyses for information on how effective interventions are, as they do not have the time to read through the literature on each individual trial that is published. But how useful are systematic reviews for clinicians and are the interventions reported in sufficient detail for clinicians to implement them in clinical practice? Interventions should also be reported in sufficient detail so that other researchers could independently replicate the interventions in future studies.

In our fifth editors’ pick, Tammy Hoffman and colleagues selected a random sample of systematic reviews of non-pharmacological interventions for stroke, and examined the completeness of reporting of the interventions in the reviews. Using an established reporting instrument called the Template for Intervention Description and Replication (TIDieR) checklist, the authors found that a number of checklist items were incompletely reported in the reviews, including: (1) modifications to the intervention; (2) a description of intervention adherence/ fidelity; (3) a description of the materials used in the intervention; (4) a description of the procedures used in the intervention and; (5) if there was a plan for the intervention to be personalised or adapted during the study. The incomplete descriptions of the intervention materials and procedures were highlighted as crucial missing elements that have previously been reported as most frequently missing in publications of individual randomized trials.

The authors concluded that better reporting of interventions were needed in systematic reviews if they were going to be effectively used and interpreted by readers. Recommendations are provided for authors of systematic reviews to use the TIDieR checklist as a guide for reporting interventions, and to provide an ‘intervention options table’ that summarises the pros and cons of all usable and feasible interventions reported in the review.

Editors’ Picks 2015: Part Four

7 Jan, 16 | by Ed Sucksmith

Long-term health effects of exercise for the elderly: study protocol

Aerobics_P4

Each day this week we’ve been looking at a published paper from 2015 that sparked the interest of one of our staff editors. Our fourth pick is a study protocol for a randomised controlled trial of exercise training in an elderly population.

BMJ Open supports the publication of study protocols to improve transparency in the ‘research to publication’ process and to protect the wider community against numerous damaging research practices, such as publication bias and HARKing.

In this study protocol, Dorthe Stensvold and colleagues from Norway, Australia and the USA aim to evaluate the effects of 5 years of exercise training on mortality in an elderly population. Whilst previous epidemiological studies have indicated that physical activity is associated with reduced risk of premature death and that changes in fitness predict mortality, it is unclear whether the level of physical activity is the direct cause of favourable health effects rather than other factors associated with a person’s state of health.

The proposed study in just under 1600 participants is considered to be the first randomised controlled trial to assess the effect of exercise training on mortality and morbidity in an elderly population. The protocol currently has 6 citations in Google Scholar, suggesting that the article has been a useful addition to the scholarly literature since it was published in February 2015.

If you are interested in learning more about study protocols then check out BMJ’s recently launched e-learning programme: Research to Publication. The programme includes a free taster module on how to write and publish a study protocol. If you successfully complete the module, then you will receive formal recognition with a certificate from The BMJ’s Editor-in-Chief and UCSF’s associate dean. If you provide this certificate when you submit your study protocol to BMJ Open, then you will receive a 75% reduction on the open access publication fee if your article is accepted.

Antidepressants and risk of mania, medically unexplained symptoms and the association between alcohol and mortality in Alzheimer’s: Most read articles in December

6 Jan, 16 | by Emma Gray

Antidepressants, medically unexplained symptoms, and alcohol consumption in Alzheimer’s patients

File:Glass of white wine.jpg

December saw a lot of new papers entering the Top 10 Most Read, with half of them being from the current issue. Patel et al‘s research paper, complete with video abstract and press release, questions whether antidepressants increase the risk of mania and bipolar disorder in people with depression in the top spot this month. Miller et al and Kristensen et al maintain their places at numbers two and three, with papers on clinical trial registration and the effect of statins on survival. Our fourth most read article examines junior doctors’ knowledge and experiences of managing patients with medically unexplained symptoms via a qualitative study using in-depth interviews. Berntsen et al enter the list at number seven, with a prospective cohort study investigating the association between alcohol consumption and mortality in patients with Alzheimer’s disease, and Jonas et al write about surgery and the placebo response.

Rank Author(s) Title
1 Patel et al. Do antidepressants increase the risk of mania and bipolar disorder in people with depression? A retrospective electronic case register cohort study
2 Miller et al. Clinical trial registration, reporting, publication and FDAAA compliance: a cross-sectional analysis and ranking of new drugs approved by the FDA in 2012
3 Kristensen et al. The effect of statins on average survival in randomised trials, an analysis of end point postponement
4 Yon et al. Junior doctors’ experiences of managing patients with medically unexplained symptoms: a qualitative study
5 Martin-Misener et al. Cost-effectiveness of nurse practitioners in primary and specialised ambulatory care: systematic review
6 Bourne et al. The impact of complaints procedures on the welfare, health and clinical practise of 7926 doctors in the UK: a cross-sectional survey
7 Berntsen et al. Alcohol consumption and mortality in patients with mild Alzheimer’s disease: a prospective cohort study
8 Jonas et al. To what extent are surgery and invasive procedures effective beyond a placebo response? A systematic review with meta-analysis of randomised, sham controlled trials
9 Boaz et al. Does the engagement of clinicians and organisations in research improve healthcare performance: a three-stage review
10 Hysing et al. Sleep and use of electronic devices in adolescence: results from a large population-based study

Most read figures are based on pdf downloads and full text views. Abstract views are excluded.

Editors’ Picks 2015: Part Three

6 Jan, 16 | by Ed Sucksmith

Combating Dissemination Bias in Clinical Research: Recommendations

Iceberg_P3

Tip of the iceberg: Dissemination bias includes publication bias, where published papers that are exposed to readers are biased in favour of particular characteristics (e.g. positive findings). Papers with other characteristics (e.g. negative findings) are not published and, as a result, remain out of view from readers.

 

 

 

 

 

 

 

 

This week we have been highlighting some of our favourite articles from 2015, which were selected by our staff editors in December. Today’s pick looks at what can be done to combat dissemination bias in clinical research.

Dissemination bias is a problem that extends beyond clinical trials to include all clinical research. It is estimated that 50% of research findings are not published at all or only present partial information, which means healthcare professionals and other stakeholders can’t make informed decisions. The focus of our third editors’ pick, by Joerg Meerpohl and colleagues from the OPEN project consortium, was to address how this problem could be resolved. The research group drafted a comprehensive set of evidence-informed recommendations for combating dissemination bias that was discussed by a variety of external experts and stakeholders. These recommendations included the need to raise awareness among stakeholders, e.g., patients participating in clinical trials who may need to be informed that their participation only contributes to scientific knowledge and progress if the results are published or at least publicly available. Other important recommendations included the need to promote trial registration and posting of results and to support initiatives that help trials to be easily searched and identified across registries.

A large number of more specific recommendations were proposed that targeted particular stakeholder groups. For journal editors/ publishers, these included (1) the consideration of research papers for publication regardless of the direction of the findings; (2) making trial registration a compulsory pre-requisite for publication; (3) checking manuscripts next to study protocols to identify selective reporting and; (4) checking for redundant publication of results.

Combating dissemination bias in medical research is an important issue for BMJ Open and its editors. We welcome these proposals and further studies addressing this vital issue.

Editors’ Picks 2015: Part Two

5 Jan, 16 | by Ed Sucksmith

Clinical Trial Transparency: disclosures for new drugs below ethical and legal standards

Transparency_P2

In December we surveyed our staff editors, who oversee and manage the peer review process, to tell us about their favourite articles from 2015. Part 2 looks at clinical trial transparency for new drugs. 

The ability of medical and healthcare professionals to select the most effective and appropriate treatments for patients depends on an unbiased and scientifically sound evidence base. Clinical trials, particularly randomized controlled trials, are the most rigorous methods for investigating the effectiveness of drugs and other treatments. If clinical trials are not registered or only selectively disseminated then the medical evidence is distorted and clinicians are unable to make informed decisions about which treatments are best for patients. As a result patients may be harmed by taking inferior treatments and participants in clinical trials are not sufficiently protected because they may be unknowingly recruited into a failed experiment that has already been conducted previously but has not been publicly disseminated.

In our second editors’ pick, Jennifer Miller and colleagues examined the levels of clinical trial registration, reporting and publication rates for new drugs approved by the Food and Drug Administration (FDA) in 2012. Out of a total of 318 clinical trials reviewed, 57% were registered per drug on average. 56% of trials were published whilst 65% of trials were either published or had reported results. The authors concluded that disclosures for new drugs frequently fell below the legal requirements and did not meet the ethical standard that results of clinical research should be publicly accessible.

Along these lines, BMJ Open supports editorial policies and procedures that improve clinical trial transparency. These include: (1) our commitment to making all papers accessible to the public; (2) compulsory prospective registration of all clinical trials submitted to the journal in a WHO or ICJME approved registry; (3) the publication of clinical trial protocols and; (4) the publication of negative findings. BMJ is also a founding member of the AllTrials campaign, an initiative specifically set up to improve clinical trial transparency by calling for all past and present clinical trials to be registered and their results reported.