This website is intended for healthcare professionals

NICE dementia guideline/guidance updated

Posted in Clinical Review Article on 12th Aug 2018


AJ Larner, Cognitive Function Clinic,Walton Centre for Neurology and Neurosurgery, Liverpool, L9 7LJ, UK.
Correspondence to:
Conflict of interest statement: None declared
Provenance and peer review: Submitted and internally reviewed
Date first submitted: 12/9/18
Acceptance date: 12/9/18
Published online first: 16/8/18

NICE issued a guideline on dementia (henceforward NG97) in June 2018,1 updating the original guidance (sic) on this subject issued jointly with the Social Care Institute for Excellence in 2006.2  NG97 is a substantial document, more than 400 pages in length, with 136 recommendations and 16 appendices.  This reviewer admits to not having read every single page of this material (Appendix P alone runs to more than 600 pages, summarising an immense amount of data which might be profitably mined by others) – fortunately briefer summaries, of guidance, are available– but has chosen to focus on those areas likely to be most pertinent to neurologists with an interest in cognitive disorders, namely investigation (cognitive testing, imaging, biomarkers; Chapter 5) and treatment (dementia drugs, other medications; Chapters 11 and 14).  (In the interests of full disclosure, it should be noted that a number of publications emanating from this reviewer’s clinic are both included in and excluded from the guideline.) Specialists in other areas may wish to focus on other chapters, such as care planning, inpatient care, supporting informal carers, and staff training.

For cognitive testing in primary care, specific recommendations on suitable screening instruments are given (no similar recommendations are given for secondary care settings).  The guideline advocates use of brief validated tools, the specified instruments being the 10-point cognitive screener (10-CS) and the six-item cognitive impairment test (6CIT).  The former was unknown to me, and as far as I can ascertain there is only a single publication,4 hence no validation in independent patient cohorts.  10-CS may indeed be a very good test (although NICE judge the index study to have an overall serious risk of bias) but pending further data it might be difficult to understand how it can be recommended, other than on the basis of opinion of what constitutes good practice.5  The explanation relates to the committee’s decision to base recommendations on the use of likelihood ratios (LRs), on which metric 10-CS scores highly (in the “very large increase in probability” range).  Pragmatically, however, 10-CS has never been mentioned in referrals from primary care directed to this author’s dedicated secondary care cognitive disorders clinic, unlike 6CIT which appears to be the most frequently used cognitive screening instrument in primary care in this catchment area, based on information contained in referral letters.6  However, the negative scoring of 6CIT (higher scores worse) is associated, in our experience, with errors in scoring and reporting in about a quarter of referrals from primary care.6  The guideline finds no place for the General Practitioner Assessment of Cognition (GPCOG), apparently due to lack of data (p103), other than to acknowledge it as the most cost effective when compared to MMSE and 6CIT (p95).  Other systematic reviews have preferred GPCOG because it assesses recall and visuospatial skills, and incorporates an informant interview;7 it is frequently mentioned in referrals from primary care to our clinic, as is the MMSE.6  MoCA is reported to be “not well tolerated by people with suspected dementia” (p100), but TYM is recommended (p110) despite both cited studies emanating from secondary care settings.  Longer (i.e. more time consuming) tests “did not appear to be more effective at detecting dementia than shorter and simpler tests” (p102); although this generalisation may be true in primary care, there is some evidence to the contrary in secondary care.8

For imaging, functional studies with FDG-PET or SPECT are advocated in suspected Alzheimer’s disease and frontotemporal dementia undiagnosed by other methods, although neither achieved very large LRs, and FP-CIT-SPECT or MIBG cardiac scintigraphy for suspected dementia with Lewy bodies.  Specialist input to interpret imaging data is recommended.  Concerning biomarkers, amyloid PET is not discussed (but is one of the research recommendations, p113) but CSF biomarker studies are recommended in suspected Alzheimer’s disease although again the LRs were not spectacular.

Guidelines for treatment with cholinesterase inhibitors and memantine have been updated, and generally speaking these are more liberal/less restrictive than previous documents from NICE, but whether this is a consequence of evidence or cost (a previous concern of one senior committee member9) is less apparent to this reviewer: “if cost containment had been a motivating factor in restricting prescribing to people with specialist experience of Alzheimer’s disease, this was no longer such a substantial concern” (p197) because all the drugs have switched from proprietary to generic status.  Hence, with a recommendation from a specialist, memantine may now be started in primary care.  Furthermore, slavish adherence to MMSE scores to determine prescription decisions is now eschewed: “health professionals should not rely solely on cognition scores” (p198), and the importance of considering the “overall benefit of treatment” (p211) is emphasised.  A corollary is that disease severity should not be used as a reason for drug discontinuation (p212).

The recommendations on antipsychotics for dementia-associated agitation, aggression, distress and psychosis are familiar i.e. avoid if at all possible.  For depression psychological treatments are to be considered, and antidepressants should not be routinely offered.  For sleep problems, a significant issue, especially for many carers, melatonin is firmly vetoed (p323), despite a recent more positive meta-analysis which is not cited.10

Overall, these guidelines are to be welcomed.  The committee is to be congratulated on the immense amount of work and analysis which has evidently gone in to their deliberations.  Elements in the original guidance which were viewed as tendentious and provoked objection (e.g. single point of referral omitting any substantive role for neurologists, a “one size fits all” approach)11 not least in the pages of this journal,12 are no longer in evidence – whether this reflects the inclusion of a (co-opted) neurologist on the committee is uncertain, but if so, take a bow Dr Jeremy Isaacs!

What will be the effect of these guidelines? Will they have any impact on practice?  Experience with the previous guidance suggests that they may attract plenty of commentary, but little by way of analysis of actual effects.13,14  If NICE guidelines, adoption of which is often de facto mandatory rather than optional (hence a possible Orwellian use of language), are considered as interventions, then their effects should surely merit some kind of evaluation as for any other clinical intervention.  For example, it would have been interesting to research how many people were harmed, or what additional costs were incurred (e.g. for nursing home placement), by previous NICE guidance restricting use of cholinesterase inhibitors.


  1. National Institute for Health and Care Excellence. Assessment, management and support for people living with dementia and their carers. NICE Guideline 97. Methods, evidence and recommendations. London: NICE, 2018(
  2. National Institute for Health and Clinical Excellence/Social Care Institute for Excellence. Dementia: supporting people with dementia and their carers in health and social care. NICE Clinical Guidance 42. London: NICE, 2006.
  3. Pink J, O’Brien J, Robinson L, Longson D, on behalf of the Guideline Committee. Dementia: assessment, management, and support: summary of updated NICE guidance. BMJ 2018;361:k2438.
  4. Apolinario D, Lichtenthaler DG, Magaldi RM et al. Using temporal orientation, category fluency, and word recall for detecting cognitive impairment: the 10-point cognitive screener (10-CS). Int J Geriatr Psychiatry 2016;31:4-12.
  5. Larner AJ. Re: Dementia: assessment, management, and support: summary of updated NICE guidance. (19 July 2018).
  6. Cannon P, Larner AJ. Errors in the scoring and reporting of cognitive screening instruments administered in primary care. Neurodegener Dis Manag 2016;6:271-276.
  7. Creavin S, Wisniewski S, Noel-Storr A et al. Cognitive tests to help diagnose dementia in symptomatic people in primary care and the community. Br J Gen Pract 2018;68:149-150.
  8. Larner AJ. Performance-based cognitive screening instruments: an extended analysis of the time versus accuracy trade-off. Diagnostics (Basel) 2015;5:504-512.
  9. O’Brien JT. NICE and anti-dementia drugs: a triumph of health economics over clinical wisdom? Lancet Neurol 2006;5:994-996.
  10. Wang YY, Zheng W, Ng CH, Ungyari GS, Wei W, Xiang YT. Meta-analysis of randomized, double-blind, placebo-controlled trials of melatonin in Alzheimer’s disease. Int J Geriatr Psychiatry 2017;32:50-57.
  11. Larner AJ. Integrated care pathways in dementia: a challenge to National Institute for Health and Clinical Excellence/Social Care Institute for Excellence guidance. J Integrated Care Pathways 2007;11:95-99.
  12. Doran M, Larner AJ. NICE/SCIE dementia guidance: time to reconsider. Adv Clin Neurosci Rehabil 2008;8(1):34-35.
  13. Larner AJ. Impact of the National Institute for Health and Clinical Excellence and Social Care Institute for Excellence’s dementia guidelines in a neurology-led memory clinic. Clin Med 2009;9:197-198.
  14. Menon R, Larner AJ. Use of cognitive screening instruments in primary care: the impact of national dementia directives (NICE/SCIE, National Dementia Strategy). Fam Pract 2011;28:272-276.