Skip to main content

Dementia: 65+ Estimated Diagnosis Rate

Independently assured by the Information Governance Board (IGB)

Published by
NHS Digital
Assurance Date
14 Apr 2017
Reporting level
Reporting period
Review date
14 Apr 2020
Based on data from
GP Extraction Service (GPES), National Health Applications and Infrastructure Services (NHAIS), Cognitive Function and Ageing Studies II (CFAS II)
Geographic Coverage


This indicator has several purposes:

  • to provide insight into the number of dementia diagnoses for those aged 65 and over, when compared to the estimated number of dementia sufferers based on the age and sex characteristics of the population under consideration
  • NHS England use it to monitor the performance of individual organisations when compared to the national target of 66.7% of estimated dementia cases being diagnosed
  • to provide Public Health England, Local Authorities, other organisations and the public with information about the estimated rate of diagnosis of dementia in general practice
  • to allow Clinical Commissioning Groups (CCGs) to compare their indicator value to the national target, and, where population characteristics are similar, to other CCGs


The indicator value is the estimated diagnosis rate for dementia in persons aged 65 and over. This is defined as the number diagnosed with dementia divided by the number estimated to have dementia given the characteristics of the population. It is expressed per 100 persons estimated to have dementia by the second Cognitive Function and Ageing Study (CFAS II).

For this indicator, the reference population is the CFAS II sample. The subject population is all 65+ patients under the care of a given organisation extracted from the General Practice Extraction Service (GPES).

Applying the estimated age and sex-specific 65+ prevalence rates of the reference population to the age and sex structure of the subject population, provides the number of people 65+ estimated to have dementia within the subject population regardless of diagnosis. Dividing the actual number of cases recorded for said organisation (again extracted from GPES) by the estimated number provides the estimated diagnosis rate.

How this indicator is calculated


GP Extraction Servce (GPES)

National Health Applications and Infrastructure Services (NHAIS), commonly known as the Exeter system

Cognitive Function and Ageing Studies II (CFAS II)


Monthly count of the number of registered patients aged 65+ with a recorded diagnosis of dementia at any point in their clinical history, as of the last day of the month.


Number of people aged 65+ estimated to have dementia, given the age and sex distribution of the registered patients and the Cognitive Function and Ageing Studies II (CFAS II) reference rates.

These values are calculated using indirect standardisation. The bandings used are:

65 to 69 years 1.2% (95% CI: 0.6–2.3)
70 to 74 years 3.0% (2.0–4.4)
75 to 79 years 5.2% (3.8–7.0)
80 to 84 years 10.6% (8.2–13.7)
85 to 89 years 12.8% (9.0–18.0)
≥90 years 17.1% (10.6–26.4)

65 to 69 years 1.8% (0.9–3.6)
70 to 74 years 2.5% (1.6–3.9)
75 to 79 years 6.2% (4.5–8.4)
80 to 84 years 9.5% (7.3–12.3)
85 to 89 years 18.1% (14.5–22.2)
≥90 years 35.0% (28.4–42.3)

Further details on how this is calculated can be found in the indicator specification document.


The numerator is divided by the denominator and multiplied by 100 to provide a percentage indicator value.

Confidence intervals are estimated by simulation. The observed counts are subject to the usual random variation assumptions and are assumed to be observations from an underlying Poisson distribution.

The expected number of cases is calculated as a sum of age-specific expected numbers, each of which is calculated by applying Cognitive Function and Ageing Studies II (CFAS II) published prevalence proportions to the local age-sex specific population. The published proportions have confidence intervals, and these need to be reflected in the confidence intervals for the overall proportion.


There are many factors that affect the true prevalence and the recorded prevalence of dementia which cannot be factored into the calculation., This means areas fall into a spectrum of greater or lesser compatibility with the Cognitive Function and Ageing Studies II (CFAS II) findings. The findings themselves are based on a limited sample from three areas scaled up to England level.

The indicator can only report on people aged 65 and over because reference rates are only available for these ages in the CFAS II reference population. CFAS II reference rates are given with confidence intervals expressing the levels of uncertainty involved.

The GP Extract Service (GPES) extract is not instantaneous; it runs over a number of days known as the 'extract window'. Depending on the length of this window, GPES may not manage to collect data for all potential practices. There is, therefore, limited potential for the misreporting of dementia diagnoses. However, the extent of any such problem will be indirectly quantified by the inclusion of the coverage figures.

The extract only includes practices that are defined as a ‘GP Practice’ in the organisational reference data (TRUD). Practices defined as walk-in centres, out of hours clinics, or prison prescribing cost centres are excluded, as are ‘shared’ and ‘dormant’ GP practices.

Data are only extracted where the diagnosis of dementia is ‘coded’, i.e. a Read/CVT3 code is on the patient record.

There is no implication that the results can be scaled down to small areas. The original cohort in the CFAS II research consisted of 7,796 people. The smallest organisation in the indicator is larger than this.

Interpretation Guidelines

The indicator gives the ratio of the actual to estimated number of dementia diagnoses, expressed as a percentage.

Values close to 100 per cent are desirable as they indicate that the observed value is close to the estimated value, indicating by proxy that most people who have dementia have also received a diagnosis.

Indicator values significantly lower than 100 per cent indicate the possible under-diagnosis of dementia whilst values over 100 per cent indicate possible over-diagnosis.