Intended for healthcare professionals

Rapid response to:

Observations Open letter to the prime minister and chief medical officer for England

There is no evidence base for proposed dementia screening

BMJ 2012; 345 doi: (Published 27 December 2012) Cite this as: BMJ 2012;345:e8588

Rapid Response:

Re: There is no evidence base for proposed dementia screening

We are writing in response to the Department of Health’s consultation on ‘early’ or ‘timely’ diagnosis of dementia in primary care and the debate this and other policy initiatives have exposed in England. Similar policy change has been promoted in the United States (US) with the introduction of a Medicare Annual Wellness Visit (2011) which includes an annual cognitive screen.

Our first point is around the use of the term ‘early’ diagnosis, which appears to have entered public discourse without discussion of what it actually means. The prevailing view from many patients, their representatives and professionals is that a diagnosis should be made as soon as possible in every individual case. This is understandably driven by personal and professional experiences of delays in access to diagnosis and appropriate support. But timely diagnosis for that person in their particular setting is undeniably current good clinical practice, if not universally available. This ‘timely’ diagnosis is not necessarily the same as ‘early’. There is currently no high quality evidence that diagnosis before the usual point of clinical presentation leads to long term improvements for people with dementia and their families. Early diagnosis is driven by the ambition to ‘close’ the gap between those people estimated to have dementia without a diagnosis and those in contact with services who have a diagnosis. This has implicit acceptance that all these people would benefit from earliest possible diagnosis. The Nuffield Bioethics Report from 2009 provides a useful moral argument in this area; the report stated that there is a ‘distinction between early and timely diagnosis’ and that if a person’s wellbeing is not enhanced by receiving a diagnosis, it should not be forced upon them.

Whether the policies being consulted on and currently implemented constitute an introduction of screening or aggressive opportunistic case finding in order to meet targets is important. The loose use of terms such as ‘early’ and ‘screening’, alongside policies that advocate widespread use of dementia orientated tests in people presenting with other conditions, seems perilously close to slipping into population screening. Hitherto, the UK has prided itself internationally on its rational screening policy, rigorously applied to other major chronic diseases. The UK approach has been that screening is not introduced until the weight of evidence on effectiveness, cost benefit, social and health service training implications and system and societal readiness is fully worked through, including screening and diagnostic methods and therapeutic evidence. Why therefore have we not done this with dementia?

Of particular concern is the lack of acknowledgement in current policy development and implementation of potential harms of premature diagnosis. The usual argument in favour of very early diagnosis is the apparently self-evident question ‘Why would you not do this?’ However, there are potential adverse effects, many of which have been articulated by others (and are familiar to those who work in screening). They include: diversion of resources from activities of proven value; misclassification of substantial numbers of people through application of screening tests which cannot have 100% sensitivity and 100% specificity; introduction of testing in unprepared settings to people - who do not fully understand what screening means, have had no chance to consider the benefits or harms of being tested, which must include possibility of overdiagnosis and overtreatment (viz. the breast cancer screening discussions), and have not therefore had the opportunity to give informed consent; raising expectations of support and services that cannot be provided and may be of limited benefit; and raising further the levels of anxiety in the population, particularly among older people.

This not to say that screening for dementia in one or more guises should not be introduced at some point in the future. Our concern is that we do not know and will not know what is appropriate and better than current practice unless these changes are made within an evaluative and research based approach. To address the current gaps in knowledge requires rigorous research of different types: formal trials, experimental roll out with evaluation, implementation research and qualitative research, all of which play unique and complementary roles and which must be integrated to create a sound evidence base.

The current piecemeal implementation of dementia policies will not provide this foundation. The concerns we have expressed here include clinicians’ reports from their experiences. We support entirely improved support for GPs to recognise dementia and to be aware of those groups at higher risk (including learning disability, stroke and Parkinson’s disease), with appropriate support for referral. However any move towards systematic screening of these groups should also be underpinned by evidence that it helps more than harms. Existing evidence and ongoing research should be fully considered in policy development. The Cochrane Collaboration has a review group focused on dementia. The Diagnostic Test Accuracy and ALOIS programmes of the Cochrane Dementia and Cognitive Improvement Group (Oxford) are systematically exploring the existing research base for early tests and how robust they are in clinical and population settings. The Alzheimer’s Society and BUPA are supporting the existing evidence base for population screening for dementia as well as the creation of a dementia research evidence map. This is complemented by the current work of the James Lind Alliance with the Alzheimer’s Society seeking what the public and professionals think are the priorities for research into dementia and the gaps. We in the UK are not alone over debating these issues: the National Institute on Aging in the US has recently funded a new randomised controlled dementia screening study in primary care (due to report 2016) and the US Preventive Services Task Force is updating its 2003 recommendation regarding dementia screening.

As a group of clinical and applied researchers we urge governments, charities, the academic community and others to be more coordinated in order to put the policy cart after the research horse. Dementia screening should neither be recommended nor routinely implemented unless and until there is robust evidence to support it. The UK can play a unique role in providing the evidence base to inform the ageing world in this area, whilst making a positive difference to the lives of individuals and their families in the future.


Carol Brayne, Professor Public Health Medicine, University of Cambridge

Anthony Bayer, Professor Geriatric Medicine, University of Cardiff

Malaz Boustani, Associate Professor of Medicine, Indiana University

Linda Clare, Professor of Clinical Psychology and Neuropsychology, Bangor University

Sarah Cullum, Consultant Old Age Psychiatrist & Honorary Senior Lecturer, University of Bristol

Tom Dening, Professor Dementia Research, University of Nottingham

Leon Flicker, Professor Geriatric Medicine, University of Western Australia

Chris Fox, Senior Clinical Lecturer Old Age Psychiatry, University of East Anglia

Claire Goodman, Professor Health Care Research, University of Hertfordshire

Cornelius Katona, Honorary Professor Mental Health Sciences, University College London

Louise Lafortune, Senior Research Associate, University of Cambridge

Jill Manthorpe, Professor Social Work, Kings College, University of London

Angelique Mavrodavis, Academic Clinical Fellow in Public Health Medicine, University of Warwick

Rupert McShane, Consultant Old Age Psychiatry, Cochrane Dementia and Cognitive Improvement Group, University of Oxford

Greta Rait, Clinical Senior Lecturer Primary Care and Population Health, University College London

Marcus Richards, Professor Psychology in Epidemiology, University College London

Louise Robinson, Professor Primary Care and Ageing, Newcastle University

Dementia: ethical issues. Nuffield Council on Bioethics (2009).

Screening for dementia. U.S. Preventive Services task Force (2003).

Competing interests: No competing interests

27 February 2013
Carol Brayne
Professor of Public Health Medicine
Anthony Bayer, Malaz Boustani, Linda Clare, Sarah Cullum, Tom Dening, Leon Flicker, Chris Fox, Claire Goodman, Cornelius Katona, Louise Lafortune, Jill Manthorpe, Angelique, Mavrodavis, Rupert McShane, Greta Rait, Marcus Richards, Louise Robinson
University of Cambridge
Department of Public Health and Primary Care, University Forvie Site, Robinson Way, Cambridge, CB20SR