Article Text

Download PDFPDF

Attention deficit hyperactivity disorder: overdiagnosed or diagnoses missed?
  1. Eric Taylor
  1. Correspondence to Professor Eric Taylor, Institute of Psychiatry, Psychology and Neuroscience, De Crespigny Park, London SE5 8AF, UK; eric.taylor{at}


The prevalence of attention deficit hyperactivity disorder is about 5% of the child population, and it carries risks for mental health, educational and personal dysfunction, antisociality and death. The diagnosis in practice is made with remarkable differences in frequency between nations. Low rates in the UK, by comparison with research estimates, suggest that the condition may often be missed in clinical practice; the reasons are considered.

  • ADHD
  • Diagnosis
  • Prevalence
  • Hyperkinetic disorder

Statistics from


The defining features of attention deficit hyperactivity disorder (ADHD) are marked and developmentally inappropriate levels of inattention and/or hyperactivity-impulsiveness that persist over time. This cluster represents a valid medical syndrome, with characteristic predictions to neurobiology, psychological function, course and treatment outcome.1 It imposes long-term risks for intellectual, emotional, social and behavioural development. Education and occupation are adversely affected; antisocial problems and substance misuse are common.

Inattention and hyperactivity-impulsiveness are distributed as dimensions in the community; in the absence of a definitely known neurobiology, there is no unchallengeable cut-off to establish what should and what should not be regarded as a disorder. The definitions of the main diagnostic schemes therefore include a requirement that function should be impaired. Many influences other than the severity of symptoms will affect the presence of educational, social and emotional dysfunction. It is therefore not surprising that estimates of prevalence vary greatly in different studies at different times and in different places. Systematic review and meta-analysis have indicated that the major influences on countries’ diagnostic rates are not the actual frequency of ADHD features, but rather the exact diagnostic criteria adopted, the measurement methods applied, the level of impairment required and the source of information used.2 ,3 It is possible that some of the differences in prevalence reflect variations in risk factors (eg, levels of exercise, screen use, obstetric practices), but if so one would have expected to see indicators in the epidemiological data, and they have not yet emerged.

This article will distinguish between the actual prevalence (taken from the application of research criteria to representative samples of the population) and the administrative prevalence (the rate with which the diagnosis is applied in practice). The former is reasonably comparable in different nations at around 5%, according to a sound meta-analysis by Polanczyk et al3 and does not seem to be rising much over time. A more recent meta-analysis has included studies from more nations (some studies being of low quality), and suggested a figure around 7% (possibly somewhat lower in Europe than America).4 The authors of the latter study argue that population rates lower than this would indicate underdiagnosis.

Administrative prevalence in the USA

Estimates in the USA vary substantially between states and at different times. In a nationally representative study in 2001–2004, using telephone interviews with parents, some 8.9% of the population aged 8–15 years met the Diagnostic and Statistical Manual of Mental Disorders (DSM)-IV criteria; less than half reported ever having received a clinical diagnosis.5 Girls were particularly likely to go undiagnosed, and poor children more likely to go untreated. The rate has been rising: the 2003 and 2007 National Surveys of Children’s Health also used telephone interviews with parents to ask if the children had ever received a diagnosis. These indicated that the percentage of children aged 4–17 years who had received a diagnosis of ADHD increased from 7.8% to 9.5% during 2003–2007.6

An extensive review by Hinshaw and Scheffler7 has examined the reasons for this ‘explosion’. They were unable to explain all the rise or the geographical differences from demographic factors and pointed to rigid school demands and the competition between schools for the results of exit examinations. Rapoport8 has brought in the notion that managed care systems may have functioned to increase medical management (and associated diagnoses) at the expense of other forms of intervention.

Administrative prevalence in Europe

In the UK, analysis of a primary care database indicated that administrative prevalence in children aged 6–17 years had risen from 0.19% in 1998 to 0.51% in 2009.9 When families were surveyed to ask if they had ever received a diagnosis from any source, the rate was about 1.2%10 This was in contrast to a research prevalence in children aged 8–12 years, ascertained from a representative population survey in 1999, of about 2.5%.11

The rise could well have reflected the impacts of National Institute for Health and Care Excellence (NICE) guidance, public awareness and continuing professional education. Since 2007, however, prevalence and incidence appear not to have increased and may even have fallen a little.9 The conservative implication is that more than half of the affected children at any one time have never been identified as such—or at least that, if they were, the information does not reach the family. In Germany, lifetime administrative prevalence has been found in several community surveys to be rather higher. The rate was about 6.5% in a study by Döpfner et al,12 in which research prevalence at ages 7–17 years was 5% for symptom presence, 2.2% for full diagnosis and 1% for hyperkinetic disorder Other surveys in Germany have given administrative rates of 4.8% at ages 3–17 years13 and 5% for ages 7–12 years and 1.3% for ages 13–19 years.14 In France, a representative population survey found that 3.5% of children had been treated for inattention and/or hyperactivity.15 The research prevalence was also 3.5%, although the two groups were not identical.

Consequences of underdiagnosis

One may ask whether missing a few cases matters much.16 There are after all some effective interventions (such as parent training for oppositional behaviour and good classroom management) that are not contingent upon a diagnosis; but their value for the core problems of inattention and impulsiveness is decidedly limited.17

Specific interventions can also help. Psychoeducation for families about the nature of the disorder is effective.18 Information for teachers has a measurable short-term benefit for the children's progress.19 Both are dependent on recognition that ADHD is present. Medication is valuable over periods of a year or more, especially for the worst affected, and can create an opportunity for educational and remedial help to become feasible.1 Some young people may obtain long-term benefit from drug therapy.

Furthermore, as with other long-term psychological disabilities such as autism spectrum disorders, a diagnosis can be useful in liberating the problem-solving abilities in the family and enhancing the children's understanding of their predicament, and that of their families. Without a diagnosis, less helpful labels may be applied such as ‘idle’, ‘difficult’ and ‘naughty’. Most parents in a study of those seeking individual help with their children were chiefly interested in interventions that would make them feel more informed about their child’s problems and understanding them.20

Reasons for missing cases in the UK

The unusually low figure for identification in the UK is in keeping with the low rate of prescribing for ADHD which, although rising, is still much less than implied by NICE recommendations and only about one-tenth of the rate in the USA.21

The apparent underdiagnosis in the UK is paralleled by underprovision of specialist service. Of those with a research diagnosis of ‘hyperkinetic disorder’, less than half had received a formal assessment.22 To set this in context, the same can be said of most mental health disorders in childhood and adolescence; and indeed has been said by an authoritative report to government from the Children's Society.23 Of the approximately 10% of the child population screening positive for a mental health disorder, about half had not been in contact with any service, while only one-fifth had seen specialised mental health or paediatric services.20 When telephoned, parents gave several reasons for lack of contact, of which the most common, in two-thirds of respondents, was a lack of information about what services were available.

Public attitudes

Some underdiagnosis may come from public misunderstandings:

  • Stigma can make parents and teachers both to use the labels.

  • Teacher training seldom includes material about the recognition of inattention or poor impulse control.

  • A notion that medication will always follow diagnosis will, although false, deter many from seeking a diagnosis.

  • An excessive use of both diagnosis and medication in the USA has been widely described in the media, and has created a misunderstanding that the same has been happening in the UK.

Lack of recognition by health services

Professional under-recognition may come about for many reasons.

  • Most children with ADHD have other problems too24 and many parents come to their family doctor because of concern about those associated problems. A survey in the UK general practices screened for ADHD features and found that referral to a secondary service was conditioned by parental beliefs.25 If they had presented hyperactivity as the problem, referral to a secondary service (paediatric or mental health) was generally prompt and accurate. If they had not described the problem, but attended because of their child's bad behaviour, uncontrollable tempers or unhappiness, then referral was usually to community services of parent training or counselling. Those services would be appropriate to the problem presented, but not necessarily to an underlying problem of ADHD.

  • Even at specialist level, recognition could be affected by diagnostic overshadowing. A US-UK study of diagnosis developed panels of expert diagnosticians from both countries, and presented them with case histories and vignettes of children attending clinics from both countries.26 All clinicians applied both the DSM-III and International Statistical Classification of Diseases (ICD)-9 criteria, the standard versions at that time. When DSM-III was used, multiple diagnoses were applied and the diagnosis of ADHD was applied at a similar rate by both sets of clinicians. When ICD-9 was used, with its expectation of a single diagnosis in most cases, the UK clinicians were much less likely to diagnose ADHD and much more likely to diagnose conduct disorder. The likely explanation was that the features of both conditions were recognised, but priority was given to one condition or the other according to the training and background of diagnosticians.

  • Diagnostic overshadowing can take another form in children with mixed neurodevelopmental problems. The component of autism spectrum (ASD) or that of intellectual disability (ID) in such cases is often evident before ADHD features can be seen as age-inappropriate, and ASD or ID may therefore dominate subsequent management. Mixed neurodevelopmental cases are often referred to paediatric services rather than Child and Adolescent Mental Health (CAMHS); some CAMHS professionals appear to lack experience and confidence in assessment.

    Controversies about the diagnostic criteria may also play a part. Limitations of the main diagnostic schemes (DSM-5 and ICD-10) have been widely canvassed. ICD-10 refers not to ‘ADHD’ but to a subtype—‘hyperkinetic disorder’ (HKD),27 which is less prevalent. Clinicians may therefore be applying the more restrictive concept of HKD, yet using the words of ADHD.

  • ADHD is not in fact an aetiological diagnosis, but some suppose it to be so. Such a belief can polarise professional attitudes. Those who see it as a neurological disease may be reluctant to recognise it in people without other neurological problems, or to suppose that another brain disorder rules out ADHD as a useful diagnosis. Conversely, those whose experience is with children showing the scars of neglect and abuse may suppose that a psychosocial cause of problems rules out the possibility that the form of the problems is ADHD.

Clinical implications

The very variable rates of recognition of ADHD imply an unmet need for reliability in clinical diagnosis. The possible reasons given above for low rates of diagnosis in the UK argue for several service improvements besides the correction of underprovision of services, whether in CAMHS or specialist paediatrics. Better information for parents and teachers about ADHD could be expected to overcome stigma and create better informed referrals. Screening in the population as a whole is not yet an evaluated way to improve outcomes: it does not necessarily make for more or more appropriate referrals,19 and if it did it would risk overloading already stretched services. Nevertheless, screening within existing services of children presented for problems of behaviour or learning could well overcome the existing filter at that level. Foreman et al28 found some years ago that a questionnaire screen in a child mental health service enhanced the recognition of ADHD features, and subsequent diagnostic rates rose from 2% to 25% of the clinic cases.

Several screening questionnaires are available, and many are widely used (eg, Conners, Swanson, Nolan and Pelham Teacher and Parent Rating Scale (SNAP)-IV and Strengths and Difficulties Questionnaire, the latter two being available free online).29 They offer advantages of saving professional time and of systematic enquiry about a range of common problems. By comparison with clinical assessment, however, they tend to overidentify cases—especially if they focus solely on symptoms and not on impairment. They were not developed as means of making a diagnosis and are not suitable for that purpose.

Nevertheless, cost-effective methods for recognition of ADHD features in paediatric practice can well include rating scales from parents and schools, gathered at an early stage after referral. Guides to a subsequent full assessment are published,30 and include detailed interview with parents (focussing on the behaviours they have seen rather than their overall ratings), and with the child. They assume the presence of clinicians with specialist training. Attempts have therefore been made to develop measures of gathering necessary information that reduce the time required by specialist professionals. The Development and Well-Being Assessment is a package of questionnaires, online interview and expert clinician judgement made on the basis of the information obtained.31 Formal evaluation has suggested that it does not add much to the efficiency of a CAMHS clinical assessment32 Nevertheless, it has reasonable predictive power33 and deserves more evaluation in settings where time or personnel constraints forbid detailed assessment, and as an economical method of gathering extensive information after referral and before definitive assessment.

Advances in diagnostic methods might perhaps in the future make it possible for specialist services to screen more children with automated neurocognitive testing, and thus economise on scarce specialist time. Different children with ADHD, however, have very different patterns of neurocognitive alterations.34 At present, no test is either necessary or sufficient for the diagnosis. For the moment, clinical utility maintains observable behaviour as the basis of diagnosis. Clinical specialists are still required, and their training should improve to fit them better for the task.


The author is grateful for discussions with colleagues of the European ADHD Guidelines Group.



  • Competing interests The author has chaired a guideline development group for the National Institute for Health and Care Excellence and written chapters for textbooks.

  • Provenance and peer review Commissioned; externally peer reviewed.

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.

Linked Articles

  • Atoms
    Martin Ward Platt