Article Text


Difficulties developing evidence-based approaches in learning disabilities
  1. Patricia C Oliver, PhD DMS1,
  2. Jack Piachaud, FRCPsych MA BMBCH2,
  3. D John Done, PhD3,
  4. Adrienne Regan, FRCPsych DCH MBBS4,
  5. Sherva E Cooray, FRCPsych DPM MBBS5,
  6. Peter J Tyrer, MD FRCP FRCPsych FFPHM FMedSci6
  1. 1Faculty of Medicine, Imperial College
 London, United Kingdom
  2. 2Hertfordshire Partnership NHS Trust
 Hertfordshire, United Kingdom
  3. 3University of Hertfordshire
 Hatfield, United Kingdom
  4. 4Harrow Primary Care Trust
 Middlesex, United Kingdom
  5. 5Brent Primary Care Trust
 London, United Kingdom
  6. 6Faculty of Medicine, Imperial College
 London, United Kingdom
  1. Correspondence to: Patricia Oliver,Department of Psychological Medicine, Faculty of Medicine, Imperial College, London, UK. poliver{at}

Statistics from


Around 2% of the population have a learning disability, categorised as ‘mental retardation’ in the ICD / DSM classification systems. Mental disorders are at least two to four times more prevalent in people with learning disability compared to the general population.1 Increasing numbers of people with learning disability access generic mental health services, due to philosophies of equity of access and normalisation. Mental health practitioners may address the problem of learning disabilities directly or help people deal with indirect outcomes, such as poor self concept or even self harm.

Learning disabilities affect both adults and children and can impact on a person’s self esteem as well as their education and competency. People with learning disabilities may also have co-morbidities. Learning disability is manifested by two components: low cognitive ability and diminished social competence. Both factors are profoundly affected by social and cultural influences. Unlike other ICD 10 mental health disorders, it is difficult to specify detailed clinical diagnostic criteria for learning disability. After considering the cultural norms of a particular individual, mental health practitioners often estimate the intelligence quotient or mental age of the individual using specified levels of cognitive abilities (table).2

Levels of cognitive abilities

During the past 20 years, learning disability services have shifted from the hospital to specialist community services.3 Three key changes occurred with this shift: (1) the introduction of “care management;” (2) the creation of packages of social, educational and developmental care, and (3) allocating “physical care” to primary healthcare and “mental health care” to specialist health services. In the era of evidence-based medicine and clinical governance, each of these changes should be based on research evidence.4 Evidence-based strategies might lead to constructive change for people with learning disabilities, but evidence about effective interventions is limited and it may be difficult to extrapolate findings from research in the general population or in people with other mental difficulties. This lack of an evidence-base is of relevance to psychiatrists and psychologists practising in an era when “political correctness” and users’ views may compete with evidence gained from more rigorous sources. This editorial examines why it is difficult to obtain rigorous research evidence on learning disability interventions.

Research context

As in many other fields of mental health, evidence in the field of learning disabilities is constrained by ethical concerns over capacity and consent. Large randomised trials are not always practical or acceptable given the nature of patients and interventions, yet there is a desire to incorporate more evidence into practice.5

There are several difficulties establishing an evidence-base in learning disabilities. These include transfer of knowledge from other disciplines, mental health co-morbidities, and service delivery practicalities. Other fields of mental health may encounter similar problems. The main difference in learning disabilities is that many disciplines work side by side, some with a historical grounding in empirical research and others with a more theoretical or anecdotal evidence-base.

Almost all evidence-based practice for learning disability has been transferred directly from adult psychiatry or mental health services, but this may not always be beneficial for people with learning disabilities. This population may have special needs. For example, a comparison of intensive and standard case management in severe mental illness found no outcome differences overall,6 but outcomes in people with borderline intelligence were significantly better after intensive case management. 7,8 A similar trial compared assertive and standard community approaches in learning disability services, but the evaluators encountered several difficulties. They concluded that before learning disability services can be adequately evaluated, fundamental changes in attitudes towards research are needed in this field.9

Good evidence is definitely needed. Although most mental health workers agree that people with learning disabilities require specialist services, the nature and delivery of these services is far from clear. There have been many changes in recent years, including the expansion of community teams, reduction of hospital beds and development of specialist in-patient assessment services with controlled access to beds. These changes are by no means uniform – and we have little idea of their impact.10 We need common goals and outcomes to evaluate smaller locally accessible community services, but this is difficult in the complex and multidisciplinary field of learning disabilities.11 It is also difficult to determine the prevalence of mental disorders in people with learning disability.12 There is continuing debate as to whether behaviour disorder constitutes an environmental problem or a developmental disorder, for instance.13

All of these contextual factors make it difficult to find good evidence in the field of learning disabilities. There are also some practical problems with the quality of evidence in this field. The main practical difficulties involve ethical considerations, methodological problems, and service capacity.9

Ethical dilemmas

Randomisation is now better accepted in many mental health disciplines, but it is a major ethical issue in the field of learning disability. Many purchasers and service providers feel that it would be wrong to deprive a person of a “necessary” service. There may also be concerns that well established relationships between service users and carers would be wrongly compromised by randomising patients.

Campbell et al suggest that complex interventions can be evaluated using a systematic controlled trial methodology.14,15 But some learning disabilities providers believe that services do not need to be evaluated or the most effective interventions trialled, because their value is self-evident. In our experience with the TACTILD trial, providers who did not take part “on ethical grounds” often said that denial of treatment was unethical, as ipso facto, it must be effective.

Informed consent may also be a problem, with participants being less competent to provide consent than the general population. This problem permeates many other mental health fields, such as work with youths and older people. However, among older people and youth, problems with informed consent may be limited to particular life stages. A person with learning disability will have difficulty providing informed consent throughout their lifetime. Similarly, relatives are often involved in learning disabilities services, either as the primary carer or as a significant contact. This means that, as in studies of young people, a second layer of informed consent is often needed.


A wide range of disciplines are involved in the field of learning disability, some of which do not have a strong research tradition. There is a greater reliance on “expert opinion” than in some other mental health fields. There is also significant variation in resources and practice between teams. Even the definitions of different treatments may differ between practice teams. Services rarely define their practices operationally.

Another practical difficulty is the relatively long time period needed to detect changes in learning disabilities. Due to the low prevalence of certain conditions, sample sizes may also be considered small by traditional standards. The variety of disciplines involved in learning disability services and the variation and complexity of the services offered comprises a unique difficulty when implementing evidence-based medicine.

Service capacity

Currently, new learning disability services emerge only where there are strong supportive voices. This “expert opinion” is not necessarily based on high quality evidence.16 This is of increasing importance in the clinical governance era, where services with no scientific basis for their continuation may be withdrawn. Learning disabilities workers often acknowledge the need for high quality research, but resources are limited. It may not be justified to simply transfer the results of evidence-based practice from standard psychiatric services to learning disability services. It is likely to be more appropriate to develop evidence-based practices specifically for the learning disabled population, especially for those with a moderate or severe level of learning disability.

Evidence needs

Learning disability spans the boundaries between health and social care; specialist health and more general health services, and between child and adult services. As in other fields of mental health, an evidence-based philosophy is needed to develop and maintain high quality services. Questions of cost-effectiveness and cost-benefit also need to be addressed. Adequate research funding is fundamental to this process, necessitating a change in service culture. The need and intention to evaluate must become the norm, not the exception. Without open questioning of current practice, learning disability services will be governed by opinion and product champions instead of a formal evidence-base.

Many mental health professionals have embraced evidence-based philosophies, but the field of learning disabilities has a little “catching up” to do. Promoting and developing rigorous academic research takes time. Despite the problems outlined here, the task is by no means impossible, provided there is ongoing multidisciplinary professional enthusiasm, co-operation and collaboration between users, carers, academic institutions, and health agencies.


Some of the ideas in this piece are based on the article: Oliver PC, Piachaud J, Done J et al. Difficulties in conducting a randomized controlled trial of health service interventions in intellectual disability: implications for evidence-based practice. Journal of Intellectual Disability Research 2002; 46: 340–5. We thank the North Thames Region R & D for original funding and Helen Bond, Senior Library Assistant, Hertfordshire Partnership NHS Trust, for her assistance throughout the TACTILD project.


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.