Another Interesting Trial of an Educational Intervention – This Time Concerning Access

Young people from disadvantaged backgrounds are less likely to apply to elite universities, both in the UK and the US, than those from economically better-off backgrounds. This finding applies even after controlling for exam results prior to application – i.e. the GCSE results in England. So Sanders and co-authors from the Behavioural Insights Team and the English Department for Education did an inexpensive trial of an inexpensive intervention.[1] The outcomes were application to, and acceptance into, an elite university (defined as belonging to the Russell Group). The intervention consisted of a letter sent to students from disadvantaged backgrounds who were on track to attend an elite university given their GCSE grades. Eligible schools were randomised to control conditions or one of three interventions: to receive a letter written by a pseudonymous male student (Ben) at Bristol University on Department for Education note paper; to receive a similar letter from a female student (Rachel) at the same university; or to receive letters from both Ben and Rachel. Three hundred schools (clusters) and 11,104 students participated. It was then a simple matter to collect the outcomes from the agency that supervises the admission process (the Universities and Colleges Admissions Service, UCAS). Receipt of a letter was associated with a non-significant increase in applications, and eventual admission to, an elite university. The increase was greatest and statistically significant for students who received both letters – from 8.5% acceptance among controls, to 11.4% in the ‘double dose’ intervention group – an increase of 2.9 percentage points (or 34 percent relative risk). Certainly, these results add to growing evidence concerning aspirations in education – see recent News Blogs on keeping children back a year [2], streaming [3], and the Michelle Obama effect.[4]

— Richard Lilford, CLAHRC WM Director


  1. Sanders M, Chande R, Selley E. Encouraging People into University. London: Department for Education; 2017.
  2. Lilford RJ. Keeping a Child Back at School. NIHR CLAHRC West Midlands News Blog. 10 March 2017.
  3. Lilford RJ. Evidence-Based Education (or How Wrong the CLAHRC WM Director was). NIHR CLAHRC West Midlands News Blog. 15 July 2016.
  4. Lilford RJ. More on Education. NIHR CLAHRC West Midlands News Blog. 16 September 2016.

Length of Hospital Stay

The average length of hospital stay for patients has ‘plummeted’ over the last thirty years, from 10 days in 1983 to 5 days in 2013.[1] However, the proportion of patients discharged to a nursing facility has quadrupled over this same period.[2] So, from the point of view of the patient, the stay away from home has not changed as much as it might be inferred from an uncritical analysis of inpatient stays. So, how have home-to-home times changed? This was assessed by Barnett et al.[3] on the basis of Medicare administration claims for 82 million hospitalisations over the years 2004 to 2011 inclusive.

Yes, the mean length of hospital stay declined (from 6.3 to 5.7 days), but the mean length of stay in post-acute care facilities increased from 4.8 to 6 days. Total home-to-home time increased from 11.1 to 11.7 days. This is not necessarily a bad thing, but it must be taken into account in assessing costs and benefits of care. The risk of iatrogenic harm and costs are lower in nursing facilities than hospitals. However, the article cited here does not consider the possibility that these risks and costs are not lower for the group of people in nursing facilities who would otherwise be cared for in hospital.

— Richard Lilford, CLAHRC WM Director


  1. Centers for Medicare and Medicaid Services. CMS program statistics: 2013 Medicare Utilization Section. 2017.
  2. Tian W (AHRQ). An All-Payer View of Hospital Discharge to Postacute Care, 2013. Rockville, MD: Agency for Healthcare Research and Quality; 2016.
  3. Barnett ML, Grabowski DC, Mehrotra A. Home-to-Home Time – Measuring What Matters to Patients and Payers. N Engl J Med. 2017; 377: 4-6.

Update on Scientists Being Held Accountable for Impact of Research

I recently wrote a news blog on the dangers of researchers being advocates for their own work. Readers may be interested in an article from an authoritative source that I chanced upon recently, published in BMC Medicine (Whitty JM. What Makes an Academic Paper Useful for Health Policy? BMC Med. 2015; 10: 301).

— Richard Lilford, CLAHRC WM Director

Reducing the Global Burden of Diagnostic Errors

A recent issue of the BMJ Quality and Safety carried an interesting review on the global burden of diagnostic errors in primary care.[1] The review looked at the most common symptoms and conditions affected by such errors; summarised the current interventions; and suggested what could be done next to reduce the global burden of diagnostic errors. The authors found that:

  • Typically there are multiple ‘root causes’ for errors, including both cognitive errors, such as failing to synthesise evidence, and system flaws, such as communication issues.
  • The most common categories of harmful diagnostic errors are infections, cardiovascular disease, cancer, and diseases in children.
  • Very few interventions to reduce errors have been tested empirically.
  • In order to reduce errors successfully there is likely to be a need for multiple interventions.

They go on to propose eight themes for interventions to measure and reduce diagnostic error:

  1. Improving diagnostic reasoning.
  2. Encouraging government policies that support primary care.
  3. Improving information technology.
  4. Involving patients.
  5. Improving access to diagnostic tests.
  6. Developing methods to identify and learn from diagnostic errors.
  7. Optimising diagnostic strategies in primary care.
  8. Providing systematic feedback to clinicians about their diagnoses.

The authors call on the World Health Organization to bring together concerned bodies to address the many challenges that are common across all countries and the opportunities that can be taken to reduce diagnostic error. CLAHRC WM collaborators are working on a more detailed classification system for the theoretical basis for diagnostic error.

— Peter Chilton, Research Fellow


  1. Singh H, Schiff GD, Graber ML, Onakpoya I, Thompson MJ. The global burden of diagnostic errors in primary care. BMJ Qual Saf. 2017; 26: 484-94.

Mindfulness for Low Back Pain

Lower back pain is fast becoming a major public health problem. Perhaps because of our increasingly sedentary life style it affects around 75% of the population at some point during their lives. However, there is no optimum clinical treatment. In light of this, many people turn to complementary therapies. A recent systematic review by Anheyer and colleagues [1] looked at the effectiveness of such a therapy, mindfulness-based interventions. Mindfulness-based stress reduction programmes (MBSR), and mindfulness-based cognitive therapy (MBCT) (see main article) usually involve activities such as meditation, yoga, and focusing attention on different parts of the body. The authors identified seven RCTs involving 864 patients, and found that MBSR was associated with statistically significant short-term improvements in pain, compared to standard care, though these weren’t sustained in the long term, and could not be deemed clinically meaningful. However, there were no significant differences when compared to active comparators, such as health education programmes. These were only a limited number of RCTs and there is still a need for long-term RCTs that compare MBSR against active treatments.

— Peter Chilton, Research Fellow


  1. Anheyer D, Haller H, Barth J, et al. Mindfulness-based stress reduction for treating low back pain. A systematic review and meta-analysis. Ann Intern Med. 2017; 166: 799-807.

Cognitive Behavioural Therapy vs. Mindfulness Therapy

It is known that mindfulness therapy is effective in improving depression and, in many circumstances, in improving chronic pain (see later in News Blog). What is not so clear is whether it is better than the more standard therapy of cognitive behavioural therapy (CBT).

Cognitive behavioural therapy aims to abolish or reduce painful and harmful thoughts. Mindfulness therapy on the other hand does not seek to extirpate the depressing thoughts, but rather to help the person disassociate themselves from the harmful consequences of these thoughts. It often involves an element of meditation.

We have found three recent studies which compare CBT and mindfulness therapy head-to-head for depression.[1-3] In all three RCTs the two therapies were a dead heat. In short, both methods seem equally effective and certainly they are both better than nothing. But does this mean that they are equal; that the choice does not matter one way or the other?

In this article I argue that the fact that the two therapies all equally effective in improving mood, does not mean that they are equivalent. This is because they are designed to have different effects – abolition of harmful thoughts in one case, learning to live with them in the other. So it is reasonable to ask which one would prefer, abolishing the painful thoughts or simply learning not to be affected by them.

Philosophically, the argument behind CBT is that thoughts, at least at a certain level, are a kind of behaviour. They are a behaviour in the sense that they can be changed under conscious control. Mindfulness therapy does not attempt to ‘over-write’ thoughts. This means that the two therapies, in so far as they achieve their objectives, are not philosophically equivalent. Moreover, there are arguments in favour of removing the harmful thoughts, even if this does not result in any greater improvement in mood than the counter-factural. Consider a man whose wife is annoyed by certain movements that he is unable to control. It is surely much better, both from her point of view and from the point of view of the husband, that these painful thoughts should be removed altogether, rather than just tolerated. Alternatively, consider a person who is chronically distressed by a recurring memory of the painful death of a parent. Again, it is surely better that this person trains himself to think of another aspect of the parent’s life whenever the troubling thoughts recur, than to simply continue to remember the death, but not get upset by it.

So, I think that CBT is philosophically preferable to mindfulness therapy, even if it is no more effective in improving mood. From a philosophical point of view, it is important to develop a high rectitude way of thinking. When negative or morally questionable thoughts pop into the brain, as they do from time to time, these should be suppressed. A racist thought, for example, should be replaced with thoughts of higher rectitude. It is the purpose of the examined life to be able to control negative or bigoted thoughts and supplant them with more positive thoughts under conscious control. From this philosophical perspective CBT can be seen as an extension of the human ability to supplant negative or reprehensible thoughts with ones that are more positive or of higher rectitude. I choose CBT over mindfulness; for all that they might be equally effective in elevating mood, psychiatric treatments have implications that go beyond purely clinical outcomes – since they affect the mind there is always a philosophical dimension.

— Richard Lilford, CLAHRC WM Director


  1. Manicavasagar V, Perich T, Parker G. Cognitive Predicators of Change in Cognitive Behaviour Therapy and Mindfulness-Based Cognitive Therapy for Depression. Behav Cogn Psychother. 2012; 40: 227-32.
  2. Omidi A, Mohammadkhani P, Mohammadi A, Zargar F. Comparing Mindfulness Based Cognitive Therapy and Traditional Cognitive Behavior Therapy With Treatments as Usual on Reduction of Major Depressive Symptoms. Iran Red Crescent Med J. 2013; 15(2): 142-6.
  3. Sundquist J, Lilja A, Palmér K, et al. Mindfulness group therapy in primary care patients with depression, anxiety and stress and adjustment disorders: randomised controlled trial. Br J Psychiatry. 2015; 206(2): 128-35.

‘Information is not knowledge’: Communication of Scientific Evidence and how it can help us make the right decisions

Every one of us is required to make many decisions: from small decisions, such as what shoes to wear with an outfit or whether to have a second slice of cake; to larger decisions, such as whether to apply for a new job or what school to send our children to. For decisions where the outcome can have a large impact we don’t want to play a game of ‘blind man’s buff’ and make a decision at random. We do our utmost to ensure that whatever decision we arrive at, it is the right one. We go through a process of getting hold of information from a variety of sources we trust and processing that knowledge to help us make up our minds. And in this digital age, we have access to more information than ever before.

When it comes to our health, we are often invited to be involved in making shared decisions about our own care as patients. Because it’s our health that’s at stake, this can bring pressures of not only making a decision but also making the right decision. Arriving at a wrong decision can have significant consequences, such as over- or under-medication or missing out from advances in medicine. But how do we know how to make those decisions and where do we get our information from? Before we start taking a new course of medication, for example, how can we find out if the drugs are safe and effective, and how can we find out the risks as well as the benefits?

The Academy of Medical Sciences produced a report, ‘Enhancing the use of scientific evidence to judge the potential benefits and harms of medicine’,[1] which examines what changes would be necessary to help patients make better-informed decisions about taking medication. It is often the case that there is robust scientific evidence that can be useful in helping patients and clinicians make the right choices. However, this information can be difficult to find, hard to understand, and cast adrift in a sea of poor-quality or misleading information. With so much information available, some of it conflicting – is it any surprise that in a Medical Information Survey, almost two-thirds of British adults would trust experiences of friends and family compared to data from clinical trials, which only 37% of British adults would trust?[2]

The report offers recommendations on how scientific evidence can be made available to enable people to weigh up the pros and cons of new medications and arrive at a decision they are comfortable with. These recommendations include: using NHS Choices as a ‘go to’ hub of clear, up-to-date information about medications, with information about benefits and risks that is easy to understand; improving the design, layout and content of patient information leaflets; giving patients longer appointment times so they can have more detailed discussions about medications with their GP; and a traffic-light system to be used by the media to endorse the reliability of scientific evidence.

This is all good news for anyone having to decide whether to start taking a new drug. I would welcome the facility of going to a well-designed website with clear information about the risks and benefits of taking particular drugs rather than my current approach of asking friends and family (most of whom aren’t medically trained), searching online, and reading drug information leaflets that primarily present long lists of side-effects.

Surely this call for clear, accessible information about scientific evidence is just as relevant to all areas of medical research, including applied health. Patients and the public have a right to know how scientific evidence underpinning important decisions in care is generated and to be able to understand that information. Not only do patients and the public also make decisions about aspects of their care, such as whether to give birth at home or in hospital, or whether to take a day off work to attend a health check, but they should also be able to find and understand evidence that explains why care is delivered in a particular way, such as why many GPs now use a telephone triage system before booking in-person appointments. Researchers, clinicians, patients and communicators of research all have a part to play.

In CLAHRC West Midlands, we’re trying to ‘do our bit’. We aim to make accessible a sound body of scientific knowledge through different information channels and our efforts include:

  • Involving patients and the public to write lay summaries of our research projects on our website so people can find out about the research we do.
  • Communication of research evidence in accessible formats, such as CLAHRC BITEs, which are reviewed by our Public Advisors.
  • Method Matters, a series aimed to give members of the public a better understanding of concepts in Applied Health Research.

The recommendations from the Academy of Medical Sciences can provide a useful starting point for further discussions on how we can communicate effectively in applied health research and ensure that scientific evidence, rather than media hype or incomplete or incorrect information, is the basis for decision-making.

— Magdalena Skrybant, CLAHRC WM PPIE Lead


  1. The Academy of Medical Sciences. Enhancing the use of scientific evidence to judge the potential benefits and harms of medicine. London: Academy of Medical Sciences; 2017.
  2. The Academy of Medical Sciences. Academy of Medical Sciences: Medical Information Survey. London: Academy of Medical Sciences; 2016

A Drug Treatment for Autism

Autism affects 1-2% of children. These children may have problems with social interaction, adhere to strict routines, have repetitive behaviours, restricted interests, poor self-care, and/or heightened sensory experiences. A very wide array of genetic mutations and environmental exposures interact to produce the phenotype. It is a neurological disease and one theory, the “cell danger hypothesis”, holds that certain neurological pathways are prone to become over-activated and respond as though they were under ‘threat’. Purines released from mitochondria leech through the cell membrane where they play a role in activating microglia and affecting synaptic remodelling – a topic covered in other News Blogs.[1][2] A drug called suramin inhibits the action of purines such as ATP. It is used in high doses to control trypanosomiasis (sleeping sickness). It is toxic at high dose, but might it be effective at a lower dose for autism? A very small, double-blind trial has been carried out in which five matched pairs of autistic children were randomised to a single intravenous dose of suramin or saline.[3] Metabolic pathways were affected as expected, and the treatment was associated with improvement on a standard score two days after the infusion. It is early days, but it is just possible that we are entering a period where autism will be added to the growing list of neuro/psychiatric disorders that can be mitigated by pharmacological therapy based on an improved understanding of molecular pathogenesis.

— Richard Lilford, CLAHRC WM Director


  1. Lilford RJ. Okay Then, There is a Fourth Period of Whole-Scale Synaptic Pruning in the Grey Matter of the Brain. NIHR CLAHRC West Midlands News Blog. 13 January 2017.
  2. Lilford RJ. A Fascinating Account of the Opening up of an Area of Scientific Enquiry. NIHR CLAHRC West Midlands News Blog. 11 November 2016.
  3. Naviaux RK, Curtis B, Li K, et al. Low-dose suramin in autism spectrum disorder: a small, phase I/II, randomized clinical trial. Ann Clin Transl Neurol. 2017.

Lancet Leader on a Complex Systems Model of Evidence – an Opportunity Missed

A recent paper argued for use of models in the evaluation of complex interventions where:

  • correlations are not linear,
  • components interact,
  • feedback loops are incorporated, and
  • they adapt over time.[1]

But they leave it there – they do not say how to model the components, still less how parameters can be derived from such models for use in decision models, such as health economic models. CLAHRC WM has developed and published on the use of such models in policy and service delivery research. We show how causal chains can be mapped and how probabilities can be propagated across such causal chains.[2-4] Along with Alec Morton (University of Strathclyde) and Gavin Stewart (Newcastle University), we are leading a workshop on Bayesian causal models at the forthcoming Society of Social Medicine meeting, and will give examples of this work in forthcoming issues of the News Blog.

— Richard Lilford, CLAHRC WM Director


  1. Rutter H, Savona N, Glonti K, et al. The Need for a Complex Systems Model of Evidence for Public Health. Lancet. 2017.
  2. Watson SI & Lilford RJ. Essay 1: Integrating multiple sources of evidence: a Bayesian perspective. In: Challenges, solutions and future directions in the evaluation of service innovations in health care and public health. Southampton (UK): NIHR Journals Library, 2016.
  3. Lilford RJ, Girling AJ, Sheikh, et al. Protocol for evaluation of the cost-effectiveness of ePrescribing systems and candidate prototype for other related health information technologiesBMC Health Serv Res. 2014; 14: 314.
  4. Watson SI, Taylor CA, Chen Y-F, Lilford RJ. A Framework for the Evaluation of Service Delivery Interventions. J Health Econ. [Submitted].

A Cluster RCT of an Internet-Based Programme to Promote Activity and Reduce Postpartum Calorie Intake in Poor Hispanic Women

When I read the introduction and methods section of a research paper I often try to guess the result before I read on. In the case of the paper above [1] I guessed a null result. I guessed wrong. In this cluster RCT (12 clusters, 371 patients), a carefully designed and piloted internet-based intervention to nudge women to healthy living reduced mean mass by a statistically significant 2.3kg compared to standard care. There was no effect on exercise as assessed by a pedometer. The authors express surprise that there was ‘no’ reduction in calorie intake, but they over-interpret this finding. The variance in measured calories was very wide and the p-value was 0.06. They make the mistake of reifying the 95% limits on the confidence interval.

The 2.3kg mean intervention effect may strike some as nugatory. However, a relatively small reduction in mass can have a worthwhile metabolic and health effect, as we showed in a study of liver function tests.[2] Postpartum weight loss is important because it is associated with long-term obesity, is largely truncal, and increases risk in any subsequent pregnancy. Dr Ponnusamy Saravanan from CLAHRC WM is collaborating with Prof Kamlesh Khunti (Director of CLAHRC East Midlands) in tackling the problem.

— Richard Lilford, CLAHRC WM Director


  1. Phelan S, Hagobian T, Brannen A, et al. Effect of an Internet-Based Program on Weight Loss for Low-Income Postpartum Women: A Randomized Clinical Trial. JAMA. 2017; 312(23): 2381-91.
  2. Lilford RJ, Bentham L, Girling A, et al. Birmingham and Lambeth Liver Evaluation Testing Strategies (BALLETS): a prospective cohort study. Health Technol Assess. 2013; 17(28): 1-307.