ArticlePDF Available

Evaluation of the uptake and delivery of the NHS Health Check programme in England, using primary care data from 9.5 million people: a cross-sectional study

Authors:

Abstract and Figures

Objectives To describe the uptake and outputs of the National Health Service Health Check (NHSHC) programme in England. Design Observational study. Setting National primary care data extracted directly by NHS Digital from 90% of general practices (GP) in England. Participants Individuals aged 40–74 years, invited to or completing a NHSHC between 2012 and 2017, defined using primary care Read codes. Intervention The NHSHC, a structured assessment of non-communicable disease risk factors and 10-year cardiovascular disease (CVD) risk, with recommendations for behavioural change support and therapeutic interventions. Results During the 5-year cycle, 9 694 979 individuals were offered an NHSHC and 5 102 758 (52.6%) took up the offer. There was geographical variation in uptake between local authorities across England ranging from 25.1% to 84.7%. Invitation methods changed over time to incorporate greater digitalisation, opportunistic delivery and delivery by third-party providers. The population offered an NHSHC resembled the English population in ethnicity and deprivation characteristics. Attendees were more likely to be older and women, but were similar in terms of ethnicity and deprivation, compared with non-attendees. Among attendees, risk factor prevalence reflected population survey estimates for England. Where a CVD risk score was documented, 25.9% had a 10-year CVD risk ≥10%, of which 20.3% were prescribed a statin. Advice, information and referrals were coded as delivered to over 2.5 million individuals identified to have risk factors. Conclusion This national analysis of the NHSHC programme, using primary care data from over 9.5 million individuals offered a check, reveals an uptake rate of over 50% and no significant evidence of inequity by ethnicity or deprivation. To maximise the anticipated value of the NHSHC, we suggest continued action is needed to invite more eligible people for a check, reduce geographical variation in uptake, prioritise engagement with non-attendees and promote greater use of evidence-based interventions especially where risk is identified.
Content may be subject to copyright.
1
PatelR, etal. BMJ Open 2020;10:e042963. doi:10.1136/bmjopen-2020-042963
Open access
Evaluation of the uptake and delivery
of the NHS Health Check programme
in England, using primary care data
from 9.5 million people: a cross-
sectional study
Riyaz Patel ,1 Sharmani Barnard,2 Katherine Thompson,2 Catherine Lagord,2
Emma Clegg,2 Robert Worrall,3 Tim Evans,2 Slade Carter,2 Julian Flowers,2
Dave Roberts,3 Michaela Nuttall,2 Nilesh J Samani,4,5 John Robson,6
Matt Kearney,7 John Deaneld,1 Jamie Waterall2
To cite: PatelR, BarnardS,
ThompsonK, etal. Evaluation
of the uptake and delivery of the
NHS Health Check programme
in England, using primary care
data from 9.5 million people: a
cross- sectional study. BMJ Open
2020;10:e042963. doi:10.1136/
bmjopen-2020-042963
Prepublication history and
supplemental material for this
paper is available online. To
view these les, please visit
the journal online (http:// dx. doi.
org/ 10. 1136/ bmjopen- 2020-
042963).
RP and SB are joint rst authors.
JD and JW are joint senior
authors.
Received 21 July 2020
Revised 30 September 2020
Accepted 02 October 2020
For numbered afliations see
end of article.
Correspondence to
Dr Riyaz Patel;
riyaz. patel@ ucl. ac. uk
Original research
© Author(s) (or their
employer(s)) 2020. Re- use
permitted under CC BY.
Published by BMJ.
ABSTRACT
Objectives To describe the uptake and outputs of the
National Health Service Health Check (NHSHC) programme
in England.
Design Observational study.
Setting National primary care data extracted directly by
NHS Digital from 90% of general practices (GP) in England.
Participants Individuals aged 40–74 years, invited to or
completing a NHSHC between 2012 and 2017, dened
using primary care Read codes.
Intervention The NHSHC, a structured assessment of
non- communicable disease risk factors and 10- year
cardiovascular disease (CVD) risk, with recommendations
for behavioural change support and therapeutic
interventions.
Results During the 5- year cycle, 9 694 979 individuals
were offered an NHSHC and 5 102 758 (52.6%) took up
the offer. There was geographical variation in uptake
between local authorities across England ranging from
25.1% to 84.7%. Invitation methods changed over time
to incorporate greater digitalisation, opportunistic delivery
and delivery by third- party providers.
The population offered an NHSHC resembled the English
population in ethnicity and deprivation characteristics.
Attendees were more likely to be older and women,
but were similar in terms of ethnicity and deprivation,
compared with non- attendees. Among attendees, risk
factor prevalence reected population survey estimates for
England. Where a CVD risk score was documented, 25.9%
had a 10- year CVD risk ≥10%, of which 20.3% were
prescribed a statin. Advice, information and referrals were
coded as delivered to over 2.5 million individuals identied
to have risk factors.
Conclusion This national analysis of the NHSHC
programme, using primary care data from over 9.5 million
individuals offered a check, reveals an uptake rate of over
50% and no signicant evidence of inequity by ethnicity
or deprivation. To maximise the anticipated value of the
NHSHC, we suggest continued action is needed to invite
more eligible people for a check, reduce geographical
variation in uptake, prioritise engagement with non-
attendees and promote greater use of evidence- based
interventions especially where risk is identied.
INTRODUCTION
Cardiovascular disease (CVD) remains a
major public health priority in England.1
To address this, the government intro-
duced an ambitious programme of vascular
checks in 2009, for people aged 40–74,
delivered by England’s National Health
Service (NHS).2 NHS Health Checks
(NHSHCs) sought to address the key risk
factors driving the health and economic
burden from vascular disease,3 with early
Strengths and limitations of this study
A comprehensive national- level snapshot of
National Health Service Health Check (NHSHC) pro-
gramme, derived from primary care records, and
which underpins the recently released NHSHC data
dashboard.
Academic and public health collaboration with full
access to half a billion records for over 9.5 million
people offered an NHSHC between 2012 and 2017.
This rst data analysis reports on elements relating
to uptake, implementation, process and delivery of
NHSHCs, the sociodemographic and risk factor pro-
le of both those who did and did not attend a check
and rates of advice, referrals and statin prescriptions
delivered as part of the check.
The data were restricted to people with an NHSHC
activity code, and thus we were unable to quantify
the full eligible population to determine coverage
and the gap in programme reach.
Missing data and varying volume of completeness
of risk factor measures limit comparisons between
attendees and non- attendees.
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
2PatelR, etal. BMJ Open 2020;10:e042963. doi:10.1136/bmjopen-2020-042963
Open access
modelling suggesting that each year NHSHCs would
prevent 9500 heart attacks and strokes, 4000 new
cases of diabetes and identify at least 25 000 people
with existing undiagnosed diabetes or kidney disease
before they developed complications.2 4 Further-
more, with the same vascular risk factors increasingly
recognised as contributing to other conditions like
dementia, preventable cancers and liver disease,3 the
programme has assumed an even greater importance
in the prevention of non- communicable diseases
(NCD).5–7
Over a decade on, the NHSHC is now an embedded
systematic and nationwide detailed risk assessment,
awareness and management programme in England.
Since 2013, following legislation, local authorities
have a statutory obligation to make provision for all
eligible people to have an NHSHC every 5 years.8
However, concerns have been raised that delivery and
practical implementation of such a programme pres-
ents a paradoxical risk of increasing health inequality
if implemented in a way which does not systematically
prioritise equity of access, outputs and outcomes.
Furthermore, the absence of convincing randomised
clinical trial evidence about the effectiveness of such
programmes has further prompted ongoing scrutiny
and questions around its delivery, uptake, impact and
cost- effectiveness.9
In response, the number of studies evaluating the
delivery and impact of the NHSHC continue to grow
but have shown variable results.10 This may be a result
of heterogeneity in programme delivery, small sample
sizes, use of national data before NHSHCs were passed
into law or variation in local coding practices. In addition,
some studies have drawn conclusions from analyses of the
Clinical Practice Research Datalink or QResearch data-
bases,11 which although a representative and important
primary care research resource, are limited by being
restricted to volunteer practices using specific electronic
health record systems with some under- representation in
Northern England.11 12
To overcome some of these difficulties and provide a
contemporaneous overview of the NHSHC programme in
England, we sought to analyse the largest NHSHC national
primary care dataset to be extracted to date, drawing on
data for almost 10 million individuals and half a billion
records, specifically extracted for this purpose and one
which underpins the recently released NHSHC data dash-
board.13 A series of reports will examine the delivery of
the programme, prevention opportunities identified and
the impact of the NHSHC. The objectives of this first
paper are to describe the data extract and to provide an
overview of the programme, reporting on: (1) its uptake,
process and delivery, (2) the sociodemographic and risk
factor profiles of attendees and non- attendees and (3)
advice, referrals and statin prescriptions following the
check.
METHODS
Study setting
Public Health England (PHE) is responsible for national
oversight and implementation support of the NHSHC
programme. PHE worked with NHS Digital (NHSD) to
develop business rules for a data extract of all NHSHC
coding activity to allow England wide monitoring of the
NHSHC.14 A Data Extract Advisory Committee (DEAC)
was set up to guide use of the data extract. Full details of
the scope and composition of the committee are available
online.15
Study design
We conducted a retrospective descriptive cross- sectional
study of all individuals who were offered an NHSHC, using
individual- level participant data. We describe the data
extraction before defining the study population. The study
design and report conform to the REporting of studies
Conducted using Observational Routinely- collected Data
(RECORD) recommendations for reporting of observa-
tional studies using routinely collected data.16
Data were extracted from 6524 (90%) of the 7216
general practices (GPs) participating in the General Prac-
tice Data Extraction Service (GPES),17 after excluding
individuals who had opted out of their data being used
for purposes other than direct patient care.18
The inclusion criteria for the data extract were primary
care Read code for any one of the following NHSHC activ-
ities: invitation, completion, non- attendance, inappro-
priate, commenced or declined (prior to 1 April 2018).
Full details of the Read codes used for defining NHSHC
activity are available in online supplemental table 1.
The data extracted for each individual included sociode-
mographic characteristics, risk factors for CVD, diagnostic
tests requested following the check and interventions
including advice and referrals. CVD diagnoses and medi-
cation data were also extracted from three out of the four
GP clinical information technology system providers,
corresponding to 60% of practices. Data extraction for
all variables was restricted to time windows around the
individual’s contact with the NHSHC programme as spec-
ified in the business rules for extraction, listed in online
supplemental table 2. Data for CVD diagnoses and a
broader range of medications will be presented in subse-
quent papers.
At the time of extraction in 2018, the business rules
limited the upper age limit to 75 years for each year. Due
to the rolling nature of the programme, this resulted
in missing data for the 70–74 age group, most of whom
turned 75 during the 5- year cycle. Thus, the maximum
age of patients in the extract is 69 for the financial year
2012/2013, compared with 73 in 2016/2017. The final
extraction consisted of 12 151 896 patient records with
NHSHC activity coding recorded up until 31 March 2018.
Data management and data cleaning details are provided
in Supplementary Methods and online supplemental
table 3.
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
3
PatelR, etal. BMJ Open 2020;10:e042963. doi:10.1136/bmjopen-2020-042963
Open access
Study population
NHSHCs are offered to individuals aged 40–74 years and
without any of the following conditions: hypertension,
diabetes mellitus, familial hypercholesterolaemia, coro-
nary heart disease, heart failure, atrial fibrillation, stroke
or transient ischaemic attack, peripheral arterial disease,
chronic kidney disease and those already on statins or
known to have a 10- year CVD risk of 20%.5
The study population for this analysis was derived from
the data extract described above for any NHSHC coded
activity. From this group, individuals (1) with NHSHC
activity coded outside the study window, (2) aged <40
years at the time of activity and (3) coded by the GP as
inappropriate for an NHSHC were then additionally
excluded. The final study population thus included only
those people offered an NHSHC (invited or completed).
Figure 1 presents the study extract and population
flowchart.
Denitions and study variables
Individuals were categorised as either NHSHC attendees
if they had a Read code for a completed check within the
5- year period or a non- attendee if they did not. Uptake of
the programme was defined as the proportion of the total
study population who attended.
An index date was generated from the date of an indi-
vidual’s primary NHSHC activity to identify age and the
most relevant risk factor measurements for each patient.
Risk factor and clinical measurements were selected for
analysis if they occurred on the index date. Otherwise
Figure 1 Study extract and study population owchart. The study population inclusion dates (1 April 2012 to 31 March 2017)
reect a snapshot of the 5- year rolling programme from April 2012, when all trusts commissioning primary care in England
had implemented the programme. *NHSHC activity refers to any interaction that a patient may have had with the NHSHC
programme. This includes if a patient was invited to, commenced, completed, declined, did not attend, or was inappropriate for,
the NHSHC. More details are provided in online supplemental table 1. GP, general practices; NHSHC, National Health Service
Health Check.
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
4PatelR, etal. BMJ Open 2020;10:e042963. doi:10.1136/bmjopen-2020-042963
Open access
we took the closest recording within predefined time
windows set by the DEAC. Statin prescriptions that
occurred on or after the index date among attendees with
no data for previous statin prescription were selected. A
full list of variables, Read codes used to define variables,
time windows and coding algorithms are available in
online supplemental table 4.
Further details on study variable definitions and thresh-
olds are provided in Supplemental Methods and online
supplemental tables 4–8.
Data presentation
Statistical tests were not used for comparisons because the
amount of missing data between groups varies, preventing
meaningful comparisons and the large size of the study
population permits the identification of very small differ-
ences between groups. Instead, we highlighted the size of
differences between groups and interpreted it in relation
to the missing data. Where appropriate, we presented
data for attendees and non- attendees. Data for uptake,
invitation type and third- party provider are presented
by financial year to describe changes over time. Data on
uptake are also presented by local authority for geograph-
ical comparisons. To minimise bias, we include missing
data details in all tables and figures.
Patient and public involvement
PHE developed an information notice for patients,
including an easy read version, explaining how their
personal data would be used and the purpose of the
research project. Membership of the DEAC overseeing
the use of the NHS Health Check dataset, including
the development of this study, its design and outcomes,
includes a patient representative. Study results will not be
disseminated to individuals whose data are used but the
collective analysis presented here will be shared publicly
once published.
Ethical approval
A Direction from the Secretary of State for Health and
Social Care instructed NHS Digital with the legal require-
ment to carry out the NHSHC data extract.19 This study
was subject to an internal review by the Research Support
and Governance Office in PHE to ensure that it was fully
compliant with the UK Policy Framework for Health and
Social Care Research (2017) and with all other current
regulatory requirements.
RESULTS
NHSHC uptake
Overall uptake by year
Between 1 April 2012 and 31 March 2017, 9 694 979 indi-
viduals aged 40–74 years were offered an NHSHC in
England. Of these 5 102 758 (52.6%) completed a check.
Uptake by financial year is presented in table 1. Uptake
remained >50% throughout the 5 years of programme
delivery. The number of individuals offered an NHSHC
increased from just under 1.5 million in 2012/2013 to
1.8 million the year after, plateauing at approximately 2.1
million each year after that (table 1).
Geographical variation in uptake of offers
Across England, uptake rates varied by region, as
presented in figure 2A. The highest uptake of offers over
the 5- year cycle was 84.7% and the lowest 25.1% by region.
Data for uptake by upper tier local authority are available
in online supplemental table 9. Variation in uptake in
London is shown in figure 2B. Central and north London
local authorities had higher rates of uptake, with lower
rates in the south east.
Process and delivery
Invitation frequency
Of the 9 694 979 individuals in the study population
with codes for NHSHC activity, 7 970 396 (82.2%) had
a record of at least one NHSHC invitation (see online
supplemental table 10). Table 1 presents the number
of recorded invitations for attendees and non- attendees
(recording by each financial year is available in online
supplemental table 11).
Among the 5 102 758 attendees, almost a third (32.8%)
had no invitation code recorded but still had a completed
NHSHC recorded. The remaining two- thirds (3 429 914)
had an invitation recorded, with 50.5% having one invi-
tation and 16.7% having two or more. Among these
attendees coded as invited, 590 869 (17.2%) received
an invitation on the same date as the NHSHC and
were thus assumed to be opportunistic rather than
planned. Among those with an invitation in advance of
the NHSHC (82.8%; n=2 839 045), the median number
of days between recording of their first invitation and a
completed NHSHC was 42 (IQR 21, 90) days.
Among non- attendees, 98.9% had a formal invitation
record, with a quarter (25.5%) having two or more invi-
tations. The remaining 1.1% of non- attendees had Read
codes for declining or not attending a check (see online
supplemental table 1).
Table 1 Attendance to an NHS Health Check by nancial
year among individuals aged 40–74 years in England
between April 2012 and March 2017 (N=9 694 979)
Financial
year
Individuals
offered an NHS
Health Check
Individuals
attending an
NHS Health
Check
Uptake
of offers
rate %
2012/2013 1 469 031 742 935 50.6
2013/2014 1 796 483 962 831 53.6
2014/2015 2 162 454 1 135 746 52.5
2015/2016 2 154 129 1 142 151 53.0
2016/2017 2 112 882 1 119 095 53.0
Total 9 694 979 5 102 758 52.6
NHS, National Health Service.
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
5
PatelR, etal. BMJ Open 2020;10:e042963. doi:10.1136/bmjopen-2020-042963
Open access
Invitation type
Among both attendees and non- attendees, the most
common invitation type was a letter, however, other forms
of invitations, including text messaging, increased with
each year of the programme. The online supplemental
figure 1 presents the type of invitation by financial year
among attendees and non- attendees.
Delivery
Among all attendees within the 5- year time frame, 3.0%
had a clinical code to indicate that their NHSHC was
completed by a third party. This increased gradually from
1.2% in the first year to 4.1% in the final year.
Characteristics of invitees
Sociodemographic characteristics
Table 2 presents the sociodemographic characteristics of
the study population and the characteristics of the general
population according to Office for National Statistics
modelled estimates. The population offered an NHSHC
was representative of the general population of people
aged 40–74 years in terms of sex and deprivation index
although they were younger relative to the age distribu-
tion of the general population (age <55: 62.2% vs 49.7%).
Those who were offered an NHSHC also closely resem-
bled the ethnic makeup of the general population for
most ethnicities, except for people self- reporting as white
or black Caribbean who appeared underrepresented,
although 16.7% of data for ethnicity were missing.
Attendees differed from non- attendees. More attendees
were women (54.7%) compared with non- attendees
(47.5%; general population 50.9%). There were also
notable differences by age. Most attendees were <55 years
as they constituted the largest group of eligible people,
but individuals 55 years had higher rates of attendance
after invitation. For ethnic group comparisons, a large
proportion of missing data for non- attendees (27.8%)
compared with attendees (6.8%) limits interpretation,
but where data were available and compared with the
general population, ethnic minority groups appeared
to be better represented among attendees than non-
attendees (table 2).
Deprivation indices indicate few differences between
attendees and non- attendees, except at the extreme
ends of the index of multiple deprivation spectrum,
where there were slightly more attendees from the most
affluent areas (Decile 10: 11.0% vs 10.0%) and slightly
fewer attendees from the most deprived areas (Decile 1:
8.2% vs 9.4%). Finally, although the numbers were small,
there was no evidence to indicate that people with severe
mental illness, physical or cognitive disability were under-
represented among attendees (table 2).
Risk factors
Overall, completeness of data for common risk factors
measurements including systolic blood pressure (95.8%),
smoking (95.7%), Body Mass Index (BMI) (96.3%)
and total cholesterol (93.6%) was high in attendees, in
contrast to recording of physical activity (64.5%), blood
glucose (18.2%), Haemoglobin A1C (HbA1C) (36.6%)
and alcohol (38.3%). A CVD risk score was formally docu-
mented for 79.7% of attendees (figure 3, online supple-
mental table 12). Family history data were only recorded
where a positive finding was present, making it difficult
to estimate how much data were missing or were assessed
and were negative. Completeness of most, but not all risk
factors, was lower among non- attendees, with the excep-
tion of diabetes risk measurements that were similarly low
in both groups.
Figure 4 shows the proportion of all individuals identi-
fied as having each CVD risk factor among attendees and
non- attendees and with respect to missingness of data.
Among attendees, where missingness was low, we iden-
tified 24.5% with hypertension, while 23.8% were obese
and 16% were current smokers. Where a 10- year CVD risk
score was documented in the primary care record (79.7%
of attendees), just over a quarter (25.9%) were identified
as high risk, with a score of 10%.
Interventions
Advice, information and referrals
Advice, information and referral for an intervention
following an NHSHC were recorded almost 6 million
times for all attendees and more than 2.5 million times
for individuals with elevated CVD risk factors (table 3).
Among all attendees, 16.0% were coded to have received
general lifestyle and behavioural advice, just over a fifth
were given formal advice on diet and almost a third on
physical activity. Among those whose alcohol use puts
them above low risk, more than a third were directed
to alcohol treatment services. Almost half of all current
smokers were directed to smoking cessation services and
19.6% of those who had BMI 30 were directed to weight
loss and obesity services.
Statin prescriptions
Information on a new statin prescription, occurring on or after
NHSHC completion, was available for 60.4% of all attendees
Figure 2 Variation in NHSHC uptake across (A) England and
(B) London. Uptake rates shown as % of people taking up an
offer of a check, between 2012/3 and 2016/17, by upper tier
local authority of the individuals’ usual residence. NHSHC,
National Health Service Health Check.
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
6PatelR, etal. BMJ Open 2020;10:e042963. doi:10.1136/bmjopen-2020-042963
Open access
Table 2 Sociodemographic characteristics of NHSHC invitees April 2012–March 2017 compared with ONS estimated English
population aged 40–74 at mid-2015
Sociodemographic
characteristic
ONS mid-2015
England resident
population (aged
40–74 years) n (%) NHSHC invitees n (%) Attendees n (%) Non- attendees n (%)
Sex
Male 11 200 690 (49.1) 4 724 015 (48.7) 2 311 604 (45.3) 2 412 411 (52.5)
Female 11 604 922 (50.9) 4 970 906 (51.3) 2 791 130 (54.7) 2 179 776 (47.5)
Unknown 58 (0.0) 24 (0.0) 34 (0.0)
Age group (years)
40–44 3 636 454 (15.9) 2 208 213 (22.8) 984 908 (19.3) 1 223 305 (26.6)
45–49 3 889 360 (17.1) 1 986 966 (20.5) 966 356 (18.9) 1 020 610 (22.2)
50–54 3 811 000 (16.7) 1 833 267 (18.9) 958 263 (18.8) 875 004 (19.1)
55–59 3 278 322 (14.4) 1 414 091 (14.6) 783 740 (15.4) 630 351 (13.7)
60–64 2 904 721 (12.7) 1 105 914 (11.4) 669 503 (13.1) 436 411 (9.5)
65–69 3 017 135 (13.2) 910 089 (9.4) 585 653 (11.5) 324 436 (7.1)
70–74 2 268 620 (9.9) 236 439 (2.4) 154 335 (3.0) 82 104 (1.8)
Ethnic group
White 20 383 677 (89.4) 6 946 824 (71.7) 4 067 864 (79.7) 2 878 960 (62.7)
Indian 524 313 (2.3) 202 004 (2.1) 136 598 (2.7) 65 406 (1.4)
Pakistani 291 546 (1.3) 137 222 (1.4) 89 970 (1.8) 47 252 (1)
Bangladeshi 101 926 (0.4) 46 802 (0.5) 34 863 (0.7) 11 939 (0.3)
Black African 314 107 (1.4) 147 462 (1.5) 94 539 (1.9) 52 923 (1.2)
Black Caribbean 271 649 (1.2) 79 987 (0.8) 53 621 (1.1) 26 366 (0.6)
Chinese 121 129 (0.5) 44 730 (0.5) 27 360 (0.5) 17 370 (0.4)
Other Asian 302 667 (1.3) 125 853 (1.3) 79 354 (1.6) 46 499 (1)
Other group 494 599 (2.2) 239 024 (2.5) 142 621 (2.8) 96 403 (2.1)
Not stated 104 136 (1.1) 31 319 (0.6) 72 817 (1.6)
Missing 1 620 935 (16.7) 344 649 (6.8) 1 276 286 (27.8)
Deprivation index (IMD decile)
Most deprived 1 914 356 (8.4) 853 547 (8.8) 420 547 (8.2) 433 000 (9.4)
2 1 999 183 (8.8) 896 809 (9.3) 472 647 (9.3) 424 162 (9.2)
3 2 083 743 (9.1) 904 131 (9.3) 477 140 (9.4) 426 991 (9.3)
4 2 202 902 (9.7) 921 244 (9.5) 477 516 (9.4) 443 728 (9.7)
5 2 304 663 (10.1) 974 023 (10) 509 715 (10.0) 464 308 (10.1)
6 2 402 719 (10.5) 991 135 (10.2) 517 381 (10.1) 473 754 (10.3)
7 2 443 073 (10.7) 1 044 505 (10.8) 547 909 (10.7) 496 596 (10.8)
8 2 458 761 (10.8) 1 034 751 (10.7) 547 016 (10.7) 487 735 (10.6)
9 2 491 679 (10.9) 1 045 098 (10.8) 565 872 (11.1) 479 226 (10.4)
Least deprived 2 504 533 (11.0) 1 022 539 (10.5) 563 798 (11.0) 458 741 (10.0)
Missing 7197 (0.1) 3217 (0.1) 3980 (0.1)
Patient characteristics
Deaf n/a 321 (0.0) 171 (0.0) 150 (0.0)
Blind n/a 13 405 (0.1) 7224 (0.1) 6181 (0.1)
Severe mental illness n/a 111 878 (1.2) 59 351 (1.2) 52 527 (1.1)
Learning disability n/a 39 612 (0.4) 21 535 (0.4) 18 077 (0.4)
Dementia n/a 7521 (0.1) 3060 (0.1) 4461 (0.1)
Continued
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
7
PatelR, etal. BMJ Open 2020;10:e042963. doi:10.1136/bmjopen-2020-042963
Open access
(n=3 079 705, see the Methods section). Overall, a statin was
prescribed for 8.2% of these attendees. Stratifying this group
by CVD risk revealed that a statin was prescribed in 20.3% of
those with a 10- year CVD risk score 10% and in 39.1% of
those with a CVD risk score of 20%. Among the 1 910 919
individuals with a CVD risk score <10%, 3.3% received a new
statin prescription, while in the remaining 504 374 with no
CVD risk score recorded, 11.0% were prescribed a statin (see
online supplemental table 13).
Assuming similar rates of statin prescription nationally,
we estimate that of the 5 102 758 attendees in this study,
up to 418 000 may have received a new statin prescription,
with over half of these (n~2 13 000) prescribed to those
identified at the NHSHC visit as being at >10% risk of
CVD events.
DISCUSSION
In the largest nationwide study of the NHS Health Check
programme, using primary care data, we find that the
checks have been offered to over 9.5 million people
during a 5- year cycle up to 2017, with 52% of people
taking up the offer. While we noted geographical varia-
tion in uptake rates and an age and sex bias for atten-
dance, we found little evidence of inequality in who
was offered or who received an NHSHC by ethnicity or
deprivation indices. Where an NHSHC was delivered, risk
factors were identified at a similar rate to population esti-
mates, with advice and referrals offered over 2.5 million
times to those with risk factors, along with 20% of those
at highest risk receiving a new statin prescription as per
guidelines. These insights into the evolving process and
delivery of the NHSHC programme will support efforts to
further enhance the value of the programme, especially
for improving uptake rates, targeting those at greatest risk
Sociodemographic
characteristic
ONS mid-2015
England resident
population (aged
40–74 years) n (%) NHSHC invitees n (%) Attendees n (%) Non- attendees n (%)
Rheumatoid arthritis n/a 74 281 (0.8) 38 104 (0.7) 36 177 (0.8)
Total 22 805 612 9 694 979 5 102 758 4 592 221
IMD, index of multiple deprivation; NHSHC, National Health Service Health Check; ONS, Ofce for National Statistics.
Table 2 Continued
Figure 3 Completion of risk factor measurements for
attendees and non- attendees (2012/13–2016/17). Proportion
of available and missing data for each risk factor related
measurements are shown here. Note these are available
measurements within the time frame of the data extract
(see Supplemental Methods). Family history not shown as
coded only as yes with unknown negative/missing data.
See also online supplemental table 12 for the completeness
values. AUDIT- C, Alcohol Use Disorders Identication Test-
Consumption; BP, blood pressure; CVD, cardiovascular
disease; HbA1C, haemoglobin A1c; HDL, high- density
lipoproteins; GPPAQ, General Practice Physical Activity
Questionnaire.
Figure 4 Proportion of attendees and non- attendees
with common CVD risk factors. Denitions as per online
supplemental table 6) and include: high cholesterol=total
cholesterol >5 mmol/L or cholesterol ratio >4; high blood
pressure=systolic ≥140 or diastolic pressure ≥90 mm Hg;
obesity=body mass index≥30 kg/m2; alcohol>low risk=Alcohol
Use Disorders Identication Test- Consumption (AUDIT C)
score ≥8; low physical activity=General Practice Physical
Activity Questionnaire (GPPAQ) moderate inactive or inactive;
possible diabetes= haemoglobin A1C (HbA1C) ≥48 mmol/
mol or Fasting Blood Glucose (FBG) >7 mmol/L; current
smoker=current smoking; high CVD risk score=10- year
CVD risk score ≥10%. *Family history is predominantly only
recorded if present so accurate information on its absence
is unavailable. See also online supplemental table 6 for more
detailed information. CVD, cardiovascular disease.
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
8PatelR, etal. BMJ Open 2020;10:e042963. doi:10.1136/bmjopen-2020-042963
Open access
and maximising the use of available NCD and CVD risk
reduction interventions.
Our key finding of a 52% uptake rate is slightly higher
than previous studies, reporting around 48%.10 This may
be due to the larger, more nationally representative and
contemporary data to which we had access, supported by
the finding that uptake rates have steadily increased since
2012. Furthermore, we also found wide geographical vari-
ation, across the country and in London, possibly due to
differing coding practices or invitation methods, which
could skew findings from smaller studies or explain discor-
dance with other reports of NHSHC activity.20 However,
an important difference that precludes direct comparison
with other studies reporting on NHSHC reach is that our
study was restricted to people who had an NHSHC code
in their GP records, indicating either an invitation or
completion of a check. As such we were unable to quan-
tify coverage of the programme, that is, how many eligible
people were offered a check. Estimates from PHE, based
on Office for National Statistics data minus the estimated
number of people on existing disease registers suggests
an eligible population of ~15.5 million.20 Using this
number and based on 5.1 million having had a check, we
estimate that a further 6.5 million in the same 5 year cycle
would need to complete an NHSHC to achieve the orig-
inal programme aspiration of 75% coverage.4 8
Some NHSHC providers have raised concerns that the
programme may paradoxically increase health inequality
by only attracting the worried well with more affluent and
white people.21 Reassuringly the data do not show gross
differences in the offering or uptake of the programme.
First, those who were offered an NHSHC closely resemble
the population of England, as measured through census
data, with no differences by sex, ethnicity or depriva-
tion indices. They were slightly younger overall, but
this is likely because eligibility for an NHSHC falls with
comorbidities which are frequently age related.5 Second,
although missing data on ethnicity limit definitive conclu-
sions, ethnic minorities such as those from South Asian
backgrounds were equally if not more represented as
reported by others.22 23 Furthermore, although there
were small differences at the extremes of deprivation
deciles, overall, there was no gross bias towards greater
attendance by increasing affluence and previous mixed
findings are likely due to regional variation,22–24 while the
similar uptake rates in those with physical disability or
serious mental illness also indicate that the programme
is equitably delivered. There was however a notable bias
towards more women and older people attending for an
NHSHC compared with non- attendees, a finding also
observed by others.10 11 22 23
Of note, despite older people being more likely to
attend than not attend after having an offer of an NHSHC,
proportionally 57% of all attendees were <55 years, which
is higher than reports from other national evaluations of
the programme.11 This could be because our data were
limited for the age 70–74 group or that more older people
are excluded having been identified with comorbidities
earlier in the programme cycle when these other studies
reported. However, it may also indicate that younger
people are motivated to understand their CVD risk and
engage with care providers to address their longer term
and lifetime risk, a finding we previously observed with
the use of digital risk assessment tool.25 The potential
benefits of this earlier engagement with CVD risk will
need to be evaluated over the longer term.
An important benefit of the NHSHC programme has
been improvements in risk factor and behaviour data
recording, which can guide patient interventions and
inform regional resource priorities. For core data items
such as smoking status, data completeness was as high as
96%, while for alcohol and physical activity (measures
that are legally required as part of the NHSHC but not
needed to calculate a person’s 10- year CVD risk) was close
to 65%. This contrasts with the high degree of missing
data among non- attendees for most risk factors. The
exception being blood glucose and HbA1C measure-
ments which were similarly complete at low levels for both
non- attendees and attendees. This may be because these
tests are only performed in attendees at high diabetes
risk, combined with parallel current or historical efforts
to establish and maintain a diabetes disease register
outside of the NHSHC. Where risk factors were recorded,
they reveal that prevalence in attendees is close to those
in the wider UK population.3 26 A 10- year risk score was
documented in 79.7% of all attendees. We anticipate that
in the remaining ~20%, practitioners may have estimated
Table 3 Number and proportion of attendees that were
coded as received advice, information or a referral following
their NHSHC among all attendees and attendees with CVD
risk factors
Intervention type All attendees n (%)
Attendees with
the CVD risk
factor above
threshold for
intervention n (%)
Alcohol
consumption
792 761 (15.5) 46 611 (38.4)
Diet 1 189 986 (23.3) 766 521 (25.1)
Physical activity 1 501 103 (29.4) 434 326 (39.3)
General lifestyle/
behaviours
814 611 (16.0) 211 571 (20.1)
Smoking
cessation
865 913 (17) 467 119 (57.3)
Weight loss and
obesity
821 414 (16.1) 599 380 (19.6)
Diabetes
prevention
programme
4551 (0.1) 3348 (0.9)
Total 2 501 565 (49.0) 565 047 (53.7)
Thresholds dened in online supplemental table 8.
CVD, cardiovascular disease; NHSHC, National Health Service
Health Check.
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
9
PatelR, etal. BMJ Open 2020;10:e042963. doi:10.1136/bmjopen-2020-042963
Open access
the score using an online or other tool not integrated into
the clinical system, which may have meant that the score
was discussed but not recorded, although it is possible
some may not have calculated it at all. Overall, where a
score was recorded over a quarter of all attendees were
calculated to have a 10 year CVD risk score of 10%, the
current threshold set by the National Institute of Clinical
Excellence (NICE) to consider preventative interventions
such as statin prescription.27 Indeed, we found that 20%
of this population was newly prescribed a statin following
the NHSHC. This figure was even higher at nearly 40%
for those with a 10- year CVD risk score of 20%, an older
NICE threshold for statin prescription. This is an encour-
aging finding, being higher than in earlier studies and
approaching the national ambition of 45% for statin use
in this very high risk group.11 28 Our data also suggest
that the NHSHC encounter prompted relevant non-
statin interventions with over 2.5 million people with
risk factors being coded as having received advice, infor-
mation or referrals. We note however that these figures
may be an underestimate being entirely dependent on
coding practices and availability of services by region. For
example, the low referral rates for the diabetes preven-
tion programme) are partly explained by the programme
launching relatively recently in 2016 and also due to
variation in its availability across England and the poor
recording of referrals to the programme in the primary
care record as reported by others.29
Limitations
Despite being the largest national evaluation of the
NHSHC programme, our study has some important
limitations. First, our data were restricted to people
with an NHSHC activity code, and thus we were unable
to quantify the full eligible population to determine
coverage and the gap in programme reach. Although this
is an aspiration for future analyses, it will require access to
GP records for much of the population, raising important
data governance and handling challenges. Second,
we had substantial missing data, especially for the non-
attendees, limiting our ability to make robust conclusions
about differences in characteristics and risk between
these groups. Also, our data extract did not include infor-
mation on 10% of practices in GPES, which could have
introduced a degree of bias in our estimates if the reasons
for missing data were not random and related to partic-
ipation in the NHSHC programme. Third, important
information on those >70 years was limited due to a
business rule that led to loss of older people once they
turned 75 for each year of the data extract. However, the
proportionally smaller number of older people eligible
for an NHSHC means our results are unlikely to have
been impacted significantly. Fourth, prescription data
were only available from 60% of practices. The estimate
for statin prescriptions derived from the available data
however is likely to be representative. Finally, we used a
Read code to identify whether an NHSHC took place.
This, of course, does not provide any indication as to the
extent or quality of the conversations around risk or the
suitability of information given, on which the full impact
and value of an NHSHC are likely to depend.
Clinical implications
This analysis provides a national- level overview of the
NHSHC programme, against which local authorities and
healthcare providers can benchmark local achievements.
Used with the NHSD dashboard, this will enable local
CVD risk strategies to be developed, to increase the invita-
tion of eligible individuals not yet invited for an NHSHC
as well as targeting those who still do not attend even after
invitation.13 Importantly, we show that a national preven-
tion programme to tackle NCDs is possible and popula-
tion health can be targeted through routine healthcare. It
represents a systematic approach to switching the conver-
sation from illness to preventing disease and appears to
have good engagement from the public so far. From the
data, we observe that in England, there remains a major
challenge for reducing risk factors that impact multiple
long- term chronic conditions. The programme appears to
have been successful at promoting advice and guideline-
based interventions. Although assessing the efficacy of
these interventions on individual- level behaviour change
is challenging, further analysis of this large dataset will
explore the impact on available metrics such as diagnosis
rates and clinical outcomes.
CONCLUSION
In this large- scale analysis of the NHSHC programme
using national primary care data, we found that in
recent years, over half of all people offered a check have
completed one. Although there was substantial varia-
tion between local authorities in uptake rates, we found
little or no evidence of inequity in invitation processes
or uptake. Furthermore, the programme has identified
a high burden of risk among attendees, with correspond-
ingly encouraging levels of guideline- driven advice, refer-
rals and statin prescriptions for the primary prevention
of CVD. However, to achieve fully the anticipated bene-
fits of the NHSHC programme, we highlight a need for
continued efforts to invite more of the eligible popula-
tion for an NHSHC, reduce geographical variation in
uptake of offers, prioritise those who are not attending
and to maximise the use of evidence- based interventions
to support risk reduction. Subsequent research should
provide more insight into how different delivery models
influence outcomes.
Author afliations
1Institute of Cardiovascular Science, University College London, London, UK
2Public Health England, London, UK
3NHS Digital, Leeds, UK
4Department of Cardiovascular Sciences, University of Leicester, Leicester, UK
5NIHR Leicester Biomedical Research Centre, Gleneld Hospital, Leicester, UK
6Centre for Primary Care and Public Health, Queen Mary University of London,
London, UK
7UCL Partners, London, UK
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
10 PatelR, etal. BMJ Open 2020;10:e042963. doi:10.1136/bmjopen-2020-042963
Open access
Acknowledgements We would like to thank colleagues from PHE and NHS
Digital who supported this work. We would also like to thank the patient and public
representatives involved with this work, for their input.
Contributors All authors contributed to conception of the study, study design,
overall analysis plan and critically reviewed the nal manuscript. Specically in
addition, RP, SB and KT contributed to the statistical analysis plan, review of results
and drafted and revised the nal paper; SB, CL, EC, TE and RW obtained and
analysed all data and contributed to drafting of the nal manuscript; SC, JF and DR
supported data extraction for the analysis and review of the nal manuscript; MN,
NJS, JR critically reviewed and edited the paper; MK, JD, JW conceived the study;
contributed to the analysis plan and critically reviewed the nal manuscript.
Funding RP (FS/14/76/30933) and JD (BHF chair) were funded by the BHF. Data
extraction and analysis were funded by PHE.
Map disclaimer The depiction of boundaries on the map(s) in this article does not
imply the expression of any opinion whatsoever on the part of BMJ (or any member
of its group) concerning the legal status of any country, territory, jurisdiction or
area or of its authorities. The map(s) are provided without any warranty of any kind,
either express or implied.
Competing interests None declared.
Patient consent for publication Not required.
Ethics approval The review also covered all ethical considerations. No ethical
issues were identied and thus review by an ethics committee was not required
(Personal communication between Katherine Thomson & PHE Research Support
Governance Ofce, 2019).
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement Data are available upon reasonable request. All
data relevant to the study are included in the article or uploaded as supplemental
information. The legal basis for the data extract was a Secretary of State for Health
and Social Care Direction. With DEAC approval PHE and NHS Digital have set up
a process for dealing with information requests relating to the pseudonymised
primary care data used in this paper. The purpose for using this data must be for
the scope of work relating to the evaluation of the NHS Health Check in line with the
requirements of the Direction.
This content has been supplied by the author(s). It has not been vetted by BMJ
Publishing Group Limited (BMJ) and may not have been peer- reviewed. Any
opinions or recommendations discussed are solely those of the author(s) and are
not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any
reliance placed on the content. Where the content includes any translated material,
BMJ does not warrant the accuracy and reliability of the translations (including but
not limited to local regulations, clinical guidelines, terminology, drug names and
drug dosages), and is not responsible for any error and/or omissions arising from
translation and adaptation or otherwise.
Open access This is an open access article distributed in accordance with the
Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits
others to copy, redistribute, remix, transform and build upon this work for any
purpose, provided the original work is properly cited, a link to the licence is given,
and indication of whether changes were made. See:https:// creativecommons. org/
licenses/ by/ 4. 0/.
ORCID iD
RiyazPatel http:// orcid. org/ 0000- 0003- 4603- 2393
REFERENCES
1 GBD 2017 Causes of Death Collaborators. Global, regional, and
national age- sex- specic mortality for 282 causes of death in 195
countries and territories, 1980-2017: a systematic analysis for the
global burden of disease study 2017. Lancet 2018;392:1736–88.
2 Department of Health. Putting prevention rst: vascular checks,
risk assessment and management, 2008. Available: https://www.
healthcheck. nhs. uk/ seecmsle/? id= 1302 [Accessed Dec 2019].
3 GBD 2017 Risk Factor Collaborators. Global, regional, and national
comparative risk assessment of 84 behavioural, environmental
and occupational, and metabolic risks or clusters of risks for 195
countries and territories, 1990-2017: a systematic analysis for the
global burden of disease study 2017. Lancet 2018;392:1923–94.
4 Department of Health. Economic modelling for vascular vhecks,
2009. Available: www. healthcheck. nhs. uk/ document. php? o= 225
[Accessed Feb 2020].
5 Public Health England. NHS health check best practice guidance
2019. Available: https://www. healthcheck. nhs. uk/ seecmsle/? id=
1474 [Accessed Feb 2020].
6 NHS. NHS long term plan 2018. Available: https://www. england. nhs.
uk/ long- term- plan/ [Accessed Feb 2020].
7 Department for Health and Social Care. Advancing our health:
prevention in the 2020’s online, 2019. Available: https://www. gov. uk/
government/ consultations/ advancing- our- health- prevention- in- the-
2020s/ advancing- our- health- prevention- in- the- 2020s- consultation-
document [Accessed Mar 2020].
8 Department of Health and Social Care. The local authorities (public
health functions and entry to premises by local Healthwatch
representatives) regulations London, 2013. Available: http://www.
legislation. gov. uk/ uksi/ 2013/ 351/ regulation/ 4/ made [Accessed Dec
2019].
9 Capewell S, McCartney M, Holland W. NHS Health Checks- a naked
emperor? J Public Health 2015;37:187–92.
10 Martin A, Saunders CL, Harte E, etal. Delivery and impact of the
NHS health check in the rst 8 years: a systematic review. Br J Gen
Pract 2018;68:e449–59.
11 Robson J, Dostal I, Sheikh A, etal. The NHS health check
in England: an evaluation of the rst 4 years. BMJ Open
2016;6:e008840.
12 Herrett E, Gallagher AM, Bhaskaran K, etal. Data resource
prole: clinical practice research Datalink (CPRD). Int J Epidemiol
2015;44:827–36.
13 NHS Digital. Nhs health check programme: interactive dashboard
2019. Available: https:// digital. nhs. uk/ data- and- information/ data-
tools- and- services/ data- services/ general- practice- data- hub/ nhs-
health- check- programme [Accessed Feb 2020].
14 NHS Digital. Nhs health checks business rules NHS digital, 2018.
Available: https:// digital. nhs. uk/ services/ general- practice- gp-
collections/ service- information/ nhs- health- checks- business- rules
[Accessed Feb 2020].
15 NHS Digital. Data extract advisory committee to the NHS health
check data extract, 2018. Available: https://www. healthcheck. nhs.
uk/ commissioners- and- providers/ governance/ data- extract- advisory-
committe- deac/ [Accessed Feb 2020].
16 Benchimol EI, Smeeth L, Guttmann A, etal. The reporting of studies
conducted using observational Routinely- collected health data
(record) statement. PLoS Med 2015;12:e1001885.
17 NHS Digital. Privacy notice: NHS health check for adults aged 40-74
years: NHS digital, 2019. Available: https:// digital. nhs. uk/ services/
general- practice- gp- collections/ service- information/ nhs- health-
checks
18 NHS Digital. General practice extraction service (GPES), 2019.
Available: https:// digital. nhs. uk/ services/ general- practice- extraction-
service [Accessed Feb 2020].
19 NHS Digital. Direction for the NHS health check for adults aged
40-74 years data extraction 2018. Available: https:// digital. nhs. uk/
about- nhs- digital/ corporate- information- and- documents/ directions-
and- data- provision- notices/ public- health- england- directions/
direction- for- the- nhs- health- check- for- adults- aged- 40- 74- years-
data- extraction [Accessed Feb 2020].
20 Public Health England. Public health outcome framework: NHS
health check indicators: public health England; 2019. Available:
https:// ngertips. phe. org. uk/ prole/ nhs- health- check- detailed
[Accessed Feb 2020].
21 Usher- Smith JMA, Harte E, MacLure C, etal. NHS health check
programme rapid evidence synthesis, 2017. Available: www.
healthcheck. nhs. uk/ commissioners_ and_ providers/ evidence/
[Accessed Feb 2020].
22 Attwood S, Morton K, Sutton S. Exploring equity in uptake of the
NHS health check and a nested physical activity intervention trial. J
Public Health 2016;38:560–8.
23 Dalton ARH, Bottle A, Okoro C, etal. Uptake of the NHS health
checks programme in a deprived, culturally diverse setting: cross-
sectional study. J Public Health 2011;33:422–9.
24 Cochrane T, Gidlow CJ, Kumar J, etal. Cross- sectional review of
the response and treatment uptake from the NHS health checks
programme in Stoke on Trent. J Public Health
2013;35:92–8.
25 Patel RS, Lagord C, Waterall J, etal. Online self- assessment of
cardiovascular risk using the Joint British Societies (JBS3)- derived
heart age tool: a descriptive study. BMJ Open
2016;6:e011511.
26 NHS Digital. Health survey for England 2018. Available: https://
digital. nhs. uk/ data- and- information/ publications/ statistical/ health-
survey- for- england/ 2018 [Accessed Feb 2020].
27 National Institute for Health and Care Excellence. Cardiovascular
disease: risk assessment and reduction, including lipid modication
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
11
PatelR, etal. BMJ Open 2020;10:e042963. doi:10.1136/bmjopen-2020-042963
Open access
2014. Available: https://www. nice. org. uk/ guidance/ cg181 [Accessed
Feb 2020].
28 Public Health England. Health matters: preventing cardiovascular
disease, 2019. Available: https://www. gov. uk/ government/
publications/ health- matters- preventing- cardiovascular- disease/
health- matters- preventing- cardiovascular- disease# cvd- ambitions-
and- secondary- prevention [Accessed May 2020].
29 Barron E, Clark R, Hewings R, etal. Progress of the healthier
you: NHS diabetes prevention programme: referrals, uptake and
participant characteristics. Diabet Med 2018;35:513–8.
copyright. on November 18, 2020 at UCL Library Services. Protected byhttp://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2020-042963 on 5 November 2020. Downloaded from
1
Supplementary Materials
An evaluation of the uptake and delivery of the NHS Health Check Programme in England,
using primary care data from 9.5 million people: A cross-sectional study
Contents
Supplementary Methods ........................................................................................................................ 2
Supplementary Figures ........................................................................................................................... 4
Supplementary Figure 1 - Invitation type for first invitation record by year of invitation among
attendees and non-attendees ............................................................................................................ 4
Supplementary Tables ............................................................................................................................ 5
Supplementary Table 1: Read codes for NHS Health Check activity codes and prioritisation rules
for definition of primary contact with programme ............................................................................ 5
Supplementary Table 2: Data extraction rules ................................................................................... 6
Supplementary Table 3: Plausible ranges for risk factor measurements ........................................... 7
Supplementary Table 4: Order of priority for selecting metrics in time window around patient’s
index date ........................................................................................................................................... 8
Supplementary Table 5: Derived Ethnic Group Categories .............................................................. 16
Supplementary Table 6: Categories for risk factors - Risk factors by binary cut points ................... 17
Supplementary Table 7: Rules for conflicting risk factors measurements ....................................... 18
Supplementary Table 8: Intervention risk thresholds for action ...................................................... 18
Supplementary Table 9: Data for attendance by UTLA .................................................................... 19
Supplementary Table 10: Number of invitations recorded for attendees and non-attendees ........ 22
Supplementary Table 11: Invitations by financial year ..................................................................... 23
Supplementary Table 12: Completeness of risk factor measurement ............................................. 23
Supplementary Table 13: Statin prescription rates .......................................................................... 24
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance
Supplemental material placed on this supplemental material which has been supplied by the author(s) BMJ Open
doi: 10.1136/bmjopen-2020-042963:e042963. 10 2020;BMJ Open, et al. Patel R
2
Supplementary Methods
Data Management and Cleaning
The data extract was stored within a Structured Query Language (SQL) database and processed using
queries within SQL Server Management Studio. Duplicate patient records were removed. Implausible
values were re-coded as missing values. Plausible ranges for risk factors, Supplementary Table 3, were
defined by DEAC.
Definitions and Study Variables
Individuals were categorised as either NHSHC attendees if they had a Read code for a completed check
within the 5-year period, or a non-attendee if they did not. Further details are provided in
Supplementary Table 1. Uptake of the programme was defined as the proportion of the total study
population who attended.
An index date was generated from the date of an individual’s primary NHSHC activity to identify age
and the most relevant risk factor measurements for each patient. Risk factor and clinical
measurements were selected for analysis if they occurred on the index date, otherwise we took the
closest recording within pre-defined time windows set by the DEAC. A full list of variables, Read codes
used to define variables, time windows and coding algorithms is available in Supplementary Table 4.
An individual’s age in years was estimated based on year of birth and index date and presented in five-
year intervals. We derived an ethnic group variable with the aim of generating fewer categories while
still representing important ethnic groups for CVD (Supplementary Table 5). We also included Index
of Multiple Deprivation (IMD) (2015) national deciles matched at Lower Super Output Area (LSOA)
level based on the patient’s postcode of residence at the time of data extraction.1 ONS April 2019
upper tier local authority (UTLA) boundaries were used.2 Gender was reported as coded in the
extract (Male; Female). Learning difficulty, serious mental illness (SMI), blindness, deafness,
rheumatoid arthritis and dementia (present/absent) are reported as binary variables.
We present the following risk factors as binary variables, using cut-points defined in consultation with
DEAC, Supplementary Table 6; obesity (BMI>30kg/m2), blood pressure (derived from systolic
(>=140mmHg) or diastolic blood pressure (>=90mmHg), cholesterol (total cholesterol >5mmol/L or
cholesterol ratio >4), blood glucose (fasting plasma glucose >=7mmol/L or HbA1C>=48mmol/mol),
smoking (current), physical activity (general practice physical activity questionnaire = moderately
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance
Supplemental material placed on this supplemental material which has been supplied by the author(s) BMJ Open
doi: 10.1136/bmjopen-2020-042963:e042963. 10 2020;BMJ Open, et al. Patel R
3
inactive or inactive), alcohol intake and behaviour (Audit C score >=8), CVD risk score (10 year risk
>=10%) and family history of CVD before 60 years. Rules for conflicting measures for the same patient
on the same day are available in Supplementary Table 7.
Among attendees, we considered invitations in the 365 days prior to the index date. Time to
attendance was derived from the number of days between first recorded invitation and the index
date. Invitation type for attendees was grouped into three categories: advanced invitation (invitation
recorded prior to date of NHSHC), opportunistic invitation (invitation recorded same date as NHSHC)
and missing invitation (invitation not recorded but NHSHC completed). Among non-attendees for
whom the primary contact was an invitation, we considered invitations in the 365 days after the index
date. The provider delivering the NHSHC (GP staff; third party) was reported as a binary variable.
Among attendees, we present data for delivery of advice, information or referral for diet, alcohol,
physical activity, smoking, weight loss and general lifestyle, referrals for diabetes prevention and
prescriptions for statins (present/absent) as binary variables. Statin prescribing data was made
available by three out of four GP clinical IT system providers, and subsequently a Read code was
attached to 60.4% of attendees in the dataset. We present data for any statin prescription on or after
the date of NHSHC activity, as individuals with current statin prescriptions would not be eligible for an
invitation to the NHSHC. We also present these data among attendees with a risk profile indicating
that intervention was appropriate. We defined appropriate thresholds for action of intervention
through consultation with the DEAC advisory board. These are available in Supplementary Table 8.
REFERENCES
1. Office for National Statistics. English indices of deprivation 2015 2015 [Available from:
https://www.gov.uk/government/statistics/english-indices-of-deprivation-2015.
2. Office for National Statistics. Counties and Unitary Authorities (April 2019) Boundaries EW BFC
2019 [updated November 2019. Available from:
https://geoportal.statistics.gov.uk/datasets/counties-and-unitary-authorities-april-2019-
boundaries-ew-bfc accessed December 2019.
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance
Supplemental material placed on this supplemental material which has been supplied by the author(s) BMJ Open
doi: 10.1136/bmjopen-2020-042963:e042963. 10 2020;BMJ Open, et al. Patel R
4
Supplementary Figures
Supplementary Figure 1 - Invitation type for first invitation record by year of invitation among attendees and non-attendees
5
Supplementary Tables
Supplementary Table 1: Read codes for NHS Health Check activity codes and
prioritisation rules for definition of primary contact with programme
Orde
r
Clinical
NHSHC
activity code
Read V2 clinical
codes
(date
introduced)
CTV3 clinical
codes (date
introduced)
Reported
grouping
Criteria
1
Inappropriate
9NSH.
(01/10/2013)
Xaaac
(01/10/2013)
Excluded from
study
Patient has a code recorded as
being inappropriate for an NHS
Health Check in the data extract
2
Completed
8BAg.
(01/04/2010)
8BAg0
(01/10/2012)
XaRBQ
(01/04/2010)
XaZPq
(01/10/2012)
Attendee
Patient has a completed NHS
Health Check code recorded in
the 5-year period
Index date: date of patient’s first
completed check code
3
Declined
8IAx.
(01/04/2011)
XaX8h
(01/04/2011)
Non-attendee
Patient has a declined NHS
Health Check code recorded in
the 5-year period
Index date: date of patient’s first
declined code
4
Did not attend
9NiS.
(01/04/2010)
XaRAA
(01/04/2010)
Non-attendee
Patient has an NHS Health Check
not attended code recorded in the
5-year period
Index date: date of patient’s first
non-attendance code
5
Commenced
8CV9.
(01/04/2016)
Xaeab
(01/04/2016)
Non-attendee
Patient has a commenced NHS
Health Check code recorded in
the 5-year period (and no
completed/did not attend/declined
code recorded in the following 8
weeks)
Index date: date of patient’s first
commenced code
6
Invitation
9mC.., 9mC0.,
9mC1., 9mC2.,
9mC3., 9mC4.,
(01/04/2010)
9mC5., 9mC6.
(01/10/2015)
XaRBR, XaR9z,
XaRBS, XaRBT,
XaRBU, XaRBV
(01/04/2010)
Xad0C, Xad0D,
(01/10/2015)
Non-attendee
Patient has an invitation to attend
an NHS Health Check code
recorded in the 5-year period
(and no follow up (non-invitation)
code recorded within the following
6 months)
Index date: date of patient’s first
invitation code
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance
Supplemental material placed on this supplemental material which has been supplied by the author(s) BMJ Open
doi: 10.1136/bmjopen-2020-042963:e042963. 10 2020;BMJ Open, et al. Patel R
6
Supplementary Table 2: Data extraction rules
7
Supplementary Table 3: Plausible ranges for risk factor measurements
Risk factor
Plausible measurement range
(inclusive unless stated)
Alcohol risk score
(AUDIT; AUDITC; FAST)
0 40
Blood pressure - systolic
70 300 mmHg
Blood pressure - diastolic
20 150 mmHg
BMI
12 90 kg/m^2
Cholesterol total
1 40 (exclusive)
Cholesterol HDL
0.5 5
Cholesterol ratio
0.2 80
Fasting Plasma Glucose (FPG)
0 (exclusive) 100
HbA1c
20 195 mmol/mol
Height
100 230 cm
CVD risk score
0 100
Weight
20 250 kg
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance
Supplemental material placed on this supplemental material which has been supplied by the author(s) BMJ Open
doi: 10.1136/bmjopen-2020-042963:e042963. 10 2020;BMJ Open, et al. Patel R
8
Supplementary Table 4: Order of priority for selecting metrics in time window around patient’s index date
Metric
First priority
Second priority
Third priority
Derivation / other
prioritisation rules
Clinical codes (Read
V2)
Clinical codes (CTV3)
Patient characteristics
Ethnic
group
Ethnic group
recorded in
patient’s GPES
profile at time of
data extraction
(31/3/2018)
Most recent ethnic
group recorded via a
clinical code (looking
over whole data
extract)
n/a
n/a
9S...% , 9T...% , 9t...% ,
9i...%
XaBEN%
Blindness
On index date
Anytime before index
date (most proximal to
index date used)
n/a
n/a
6689. , 6688. , 668D. ,
668C.
6689.% , XaW0l ,
XaCGX% , XaLMz
Deafness
On index date
Anytime before index
date (most proximal to
index date used)
n/a
n/a
F599. , F591B , F591E ,
F59A. , F5919
XaRE4 , XaZuB , XaZuE ,
XaaLf , XaRE5 , Xa0PN
Dementia
On index date
Anytime before index
date (most proximal to
index date used)
n/a
n/a
Eu02.% , E00..% , Eu01.%
, E02y1 , E012.% ,
Eu00.% , E041. , Eu041 ,
F110.- F112. , F116. ,
F118. , F21y2 , A410. ,
A411.%
X002w% (excluding
X003E , X003F , X001T) ,
Eu02.% , XE1Xt , E00z. ,
E02y1
Learning
Disability
On index date
Anytime before index
date (most proximal to
index date used)
n/a
n/a
E3...% , Eu7..% , Eu814 ,
Eu815 , Eu816 , Eu817 ,
Eu81z , 918e. , Eu818
E3...% , XaQZ4 , XaQZ3 ,
XaKYb , XaREt , XaREu ,
Eu81z , XaaiS , Xabk1
Severe
Mental
Illness
On index date
Anytime before index
date (most proximal to
index date used)
n/a
n/a
E10..% , E110.% , E111.%
, E1124 , E1134 , E114.-
E117z , E11y.%
(excluding E11y2) , E11z.
, E11z0 , E11zz , E12..% ,
E13..% (excluding E135.)
, E2122 , Eu2..% , Eu30.%
X00S6% (excluding
Xa9B0% , E14..%) ,
X00SL , X00SM% ,
X00SJ% , XSGon , E11z. ,
E11z0 , E11zz , XE1ZZ ,
XE1Ze , XaX54 , XaX53 ,
E130. , E1124 , E1134
9
, Eu31.% , Eu323 , Eu328
, Eu333 , Eu32A , Eu329
CVD risk factors
Family
history of
CVD
On index date
Anytime before index
date (most proximal to
index date used)
Anytime after index date
(most proximal to index
date used)
n/a
12C.. , 12C2. , 12C3. ,
12C4. , 12C5. , 12CA. ,
12CB. , 12CC. , 12CD. ,
12CE. , 12CF. , 12CG. ,
12CH. , 12CI. , 12CL. ,
12CM. , 12CN. , 12CP. ,
12CV. , 12CW. , 12CZ.
XaP9K , XaP9M , ZV174
, XE24Z , XaLQq ,
Xa6aj% , XM1Jg ,
XM1Jw% , XaP9K ,
XaP9M
Rheumatoi
d arthritis
On index date
Anytime before index
date (most proximal to
index date used)
Attendees: n/a
Non-
attendees: Anytime afte
r index date (most
proximal to index date
used)
n/a
N040.% , N041. , N042.%
(excluding N0420) ,
N047. , N04X. , N04y0 ,
N04y2 , Nyu11 , Nyu12 ,
Nyu1G , Nyu10 , G5yA. ,
G5y8.
N040.% , XE1DU , X705I
, G5y8.
Alcohol
AUDIT/AU
DIT-
C/FAST
On index date
Most proximal score to
index date for each of
AUDIT, AUDIT-C and
FAST used.
Attendees: Up to 365
days before index date
Non-attendees:
Anytime before index
date
Most proximal score to
index date for each of
AUDIT, AUDIT-C and
FAST used.
Attendees: Up to 90
days after index date
Non-attendees: Anytime
after index date
No AUDIT-C/FAST/AUDIT
score available: risk
factor is missing
AUDIT-C or FAST
assessment is positive,
but no AUDIT score
available: risk factor is
missing
AUDIT-C (and/or) FAST
assessment is negative:
risk factor is low risk
AUDIT score available
and greater than or
equal to 8: risk factor is
high risk
38D4. (AUDIT-C),
388u. (FAST),
38D3. (AUDIT)
XaORP (AUDIT-C),
XaNO9 (FAST),
XM0aD (AUDIT)
10
Blood
pressure
On index date
Systolic and diastolic
BP recordings recorded
most proximal to index
date used.
Attendees: Up to 365
days before index date
Non-attendees:
Anytime before index
date
Systolic and diastolic BP
recordings recorded
most proximal to index
date used.
Attendees: Up to 90
days after index date
Non-attendees: Anytime
after index date
On examination (O/E)
readings considered
only.
Systolic BP or Diastolic
BP is unavailable: risk
factor is missing
246..% (excluding 2460. ,
2468. , 246H. , 246I. ,
246K. , 246L. , 246M. ,
246h. , 246i. , 246j. ,
246k. , 246n.% , 246o.%)
X773t% (excluding XaI9f
, XaI9g , XaZvo , XaZxj ,
X779b , X779R , X779T ,
X779W , XaYai , XaYg8 ,
XaYg9 , Xabhx , Xac5K ,
Xac5L , Xaedn%) ,
246..% (excluding 2460.
, 2468. , XaCFN , XaCFO)
Blood
glucose
On index date
HbA1c and Fasting
Plasma Glucose
recorded most
proximal to index date
considered.
Attendees: Up to 365
days before index date
Non-attendees:
Anytime before index
date
HbA1c and Fasting
Plasma Glucose
recorded most proximal
to index date
considered.
Attendees: Up to 90
days after index date
Non-attendees: Anytime
after index date
HbA1c:
42W5. , 42W50 , 42W51
Fasting Plasma Glucose:
44g1.
HbA1c:
XaPbt , Xaezd , Xaeze
Fasting Plasma Glucose:
44g1.
Body mass
index
On index date
Most proximal to index
date used.
Attendees: Up to 365
days before index date
Non-attendees:
Anytime before index
date
Most proximal to index
date used.
Attendees: Up to 90
days after index date
Non-attendees: Anytime
after index date
If BMI is unavailable but
height and weight are,
BMI is calculated (BMI =
kg/m^2)
Height and weight are
not used if BMI is
available
BMI:
22K..% (excluding
22K9.% , 22KA.)
Weight:
22A..% (excluding 22A7.-
22A9.) , 9NSa. , 8IAH.
Height:
229..% (excluding 2296.)
, 9NSZ. , 8IHM.
BMI:
22K..% (excluding
XaVwA% , X76CN ,
XaZMj) , Xa7wG%
Weight:
22A..% , 22AA. , X76C3 ,
XaesG , XaQ7T
Height:
11
229..% (excluding
2296.) , XaesF , Xaef4
Cholestero
l (ratio)
On index date
Most proximal to index
date used.
Attendees: Up to 365
days before index date
Non-attendees:
Anytime before index
date
Most proximal to index
date used.
Attendees: Up to 90
days after index date
Non-attendees: Anytime
after index date
If cholesterol ratio is
unavailable but total and
HDL cholesterol are, the
cholesterol ratio is
calculated (ratio =
total/HDL)
Total and HDL
cholesterol are not used
if cholesterol ratio is
available
Cholesterol:
44O5. , 44PH. , 44P5. ,
44PF. , 44PJ. , 44P.. ,
44OE. , 44P1. , 44P2. ,
44P3. , 44P4. , 44PK. ,
44PZ. , 44l2. , 44lF. ,
44lG. , 662a.
HDL cholesterol:
44P5. , 44PB. , 44PC. ,
44d3. , 44d2.
Cholesterol:
XaFs9 , XSK14 , 44P5. ,
44PF , 44PJ. , XalRd ,
XE2eD% , 44P1. , 44P2. ,
44P3. , 44P4. , 44PH. ,
XaERR , XaEUq , XaEUr ,
X772L
HDL cholesterol:
X772M , 44P5. , 44PB. ,
44PC. , XaEVr , 44d3. ,
44d2.
Physical
activity
(GPPAQ)
On index date
Most proximal to index
date used.
Attendees: Up to 365
days before index date
Non-attendees:
Anytime before index
date
Most proximal to index
date used.
Attendees: Up to 90
days after index date
Non-attendees: Anytime
after index date
n/a
138b. , 138a. , 138Y. ,
138X. , 38Dh.
XaPPE , XaPPD , XaPPB ,
XaPP8 , XaXX5
CVD
risk score
On index date
QRISK/QRISK2 and
Framingham risk score
recorded most
proximal to index date
used.
Attendees: Up to 365
days before index date
QRISK/QRISK2 and
Framingham risk score
recorded most proximal
to index date used.
Attendees: Up to 90
days after index date
QRISK or QRISK2 score
recorded most proximal
to index date is used if
available.
If QRISK and QRISK2
unavailable, Framingham
score is used.
QRISK/QRISK2:
8IEL., 8IEV., 38DF., 38DP.
Framingham:
38DR.
QRISK/QRISK2:
XaYzy, XaZdA, XaPBq,
XaQVY
Framingham:
XaQaG
12
Non-attendees:
Anytime before index
date
Non-attendees: Anytime
after index date
Smoking
status
On index date
Most proximal to index
date used.
Attendees: Up to 365
days before index date
Non-attendees:
Anytime before index
date
Most proximal to index
date used.
Attendees: Up to 90
days after index date
Non-attendees: Anytime
after index date
Lookup used to map
smoking status to binary
categories: Non-smoker;
Current smoker
Non-smoker:
1371, 137A., 137l.,
137N., 137O., 137S.,
Current smoker:
137.., 137C., 137e.,
137h., 137m., 137P.,
137Q., 137R., 137V.,
137X., 137Y.,
Non-smoker:
1371, 1377, 1378, 1379,
137B., 137F., 137K.,
137T., Ub0p1, Ub1na,
Xa1bv, XaQ8V, XE0oj,
XE0ok, XE0ol, XE0om,
XE0on, XE0op, XE0oh
Current smoker:
1372, 1373, 1374, 1375,
1376, 137D., 137G.,
137J., 137Z., Ub1tI,
Ub1tJ, Ub1tK, Ub1tR,
Ub1tS, Ub1tU, Ub1tW,
XaIIu, XaIkW, XaIkX,
XaIkY, XaItg, XaJX2,
XaLQh, XaWNE, XaZIE,
XE0oq, XE0or
Interventions attendees only
Advice,
informatio
n, referral
ALCOHOL
On index date
Up to 365
days after index date
n/a
n/a
Advice, information and
any brief intervention
given on alcohol usage:
67H0. , 67A5. , 8CAM. ,
8CAM0 , 8CAv. , 8CE1. ,
9k1A. , 8IAF. , 8IAt. ,
9k11. , 9k14. , ZV6D6 ,
6792. , 8CdK.
Referral regarding
alcohol usage:
Advice, information and
any brief intervention
given on alcohol usage:
XaJIr , Xa1dA , 67A5. ,
XaFvp , XaXan , XaPmB ,
8CE1. , XaPPv , XaPty ,
XaX4S , XaKAC , XaKAo ,
ZV6D6 , 6792. , Xac6H
Referral regarding
alcohol usage:
13
8HkG. , 8H7p. , 8HHe.
XaYWV , XaIPn , XaKUg ,
XaPna , XaORR
Advice,
informatio
n, referral
DIET
On index date
Up to 365
days after index date
n/a
n/a
Advice, signposting or
information on diet:
67H7. , 8CA4. , 8CA40 ,
6799.
Referral regarding diet:
8H76. , 8H760 , 8HHE.
Advice, signposting or
information on diet:
XaQaU , 8CA4. , XaXTD ,
Xa2jQ , XE0i1 , Xa2hD ,
6799.
Referral regarding diet:
XaBSz , XaAhZ , XaAha ,
XaJSp , XaAdX , XaAdY ,
XaAdZ
Advice,
informatio
n, referral
LIFESTYLE
On index date
Up to 365
days after index date
n/a
n/a
67H..% , 8Hlu.
XaEFY% , Xaam2
Advice,
informatio
n, referral
PHYSICAL
ACTIVITY
On index date
Up to 365
days after index date
n/a
n/a
Advice, signposting or
information on physical
activity:
67H2. , 8CA5. , 9Oq3. ,
6798. , 8CA52 , 8Cd4. ,
8IAv. , 8HBN.
Referral regarding
physical activity:
8H7q. , 8H7q0 , 8HHc. ,
8HkX. , 8BAH.
Advice, signposting or
information on physical
activity:
XaJIt , Xa1dN , 8CA5. ,
XM18T , XaPjx , 6798. ,
XabFV , XaREx , XaX5H ,
XaREy
Referral regarding
physical activity:
XaIPu , XaR5C , XaKRq ,
XaREh , XaCmH
Advice,
informatio
n, referral
On index date
Up to 365
days after index date
n/a
n/a
Support and refer Stop
Smoking
Service/Advisor:
Support and refer Stop
Smoking
Service/Advisor:
14
SMOKING
8CAL. , 8HTK. , 8HkQ. ,
8H7i. , 8IAj. , 8IEK. ,
9N2k. , 13p50 , 9Ndf. ,
9Ndg. , 8T08. , 8IEo.
Advice, signposting or
information on smoking:
67H1. , 8CAL. , 67A3. ,
8CAg. , 6791. , 8IAj. ,
8CdB.
Ua1Nz , XaFw9 , XaQT5
, XaItC , XaIye , XaW0h ,
XaX5W , XaX5X , XaRFh
, XaREz , XaaDy , XaaDx
Advice, signposting or
information on
smoking:
XaJIs , Ua1Nz , 67A3. ,
Ua1O0 , XaLD4 , 6791. ,
XaRFh , XaXnG
Advice,
informatio
n, referral
WEIGHT
On index date
Up to 365
days after index date
n/a
n/a
Advice, signposting or
information on weight
management:
67I9. , 8CA40 , 8Cd7. ,
66CQ. , 679P. , 8CdC. ,
8IAu.
Referral regarding
weight management:
8HHH. , 8HHH1 , 8HHH0
, 8H4n.
Advice, signposting or
information on weight
management:
XaADJ , Xa1dF , XaX5F ,
XaX5k , XaKHd , XaXnI ,
XaX5G
Referral regarding
weight management:
XaJSu , XaZKe , XaXZ9 ,
XaZKi
Diabetes
Prevention
Programm
e referral
On index date
Up to 365
days after index date
n/a
n/a
679m4,
679m0, 679m1, 679m2
XaeDH,
XaeCw, XaeCz, XaeD0
Statin
prescriptio
ns
On index date
Up to 365
days after index date
n/a
n/a
bxi..% , bxg..% , bxe..% ,
bxk..% , bxd..%
DM+D codes (EMIS):
134489001,
319996000,
319997009,
320000009,
bxi..% , x01R2% ,
x01R3% , bxk..% ,
bxd..%
15
320006003,
320012008,
320013003,
320014009,
320029006,
320030001,
320031002,
408036003,
408037007,
409108001,
4896711000001108
16
Supplementary Table 5: Derived Ethnic Group Categories
Ethnic group
Subgroups (with ONS codes)
White
A = White British
B = Irish
C = Any other White background
T = White: Gypsy or Irish Traveller
Indian
H = Indian
Pakistani
J = Pakistani
Bangladeshi
K = Bangladeshi
Black African
N = African
Black Caribbean
M = Caribbean
Chinese
R = Chinese
Other Asian
L = Any other Asian background
Other Ethnic Group
D = White and Black Caribbean
E = White and Black African
F = White and Asian
G = Any other mixed background
P = Any other Black background
S = Any other ethnic group
W = Other ethnic group: Arab
Unknown
X = Unknown/No information
Z = Not stated
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance
Supplemental material placed on this supplemental material which has been supplied by the author(s) BMJ Open
doi: 10.1136/bmjopen-2020-042963:e042963. 10 2020;BMJ Open, et al. Patel R
17
Supplementary Table 6: Categories for risk factors - Risk factors by binary cut points
Risk factors by binary risk cut-offs
Risk factor
High risk
threshold/
cutpoint
Risk
category
Attendees n (%)
Non-attendees
n(%)
Total
Alcohol >
Low Risk
Full AUDIT score
8 or more
Missing
3,150,667 (61.7)
3,823,634 (83.3)
6,974,301
Low risk
1,830,799 (35.9)
714,947 (15.6)
2,545,746
High risk
121,292 (2.4)
53,640 (1.2)
174,932
Possible
Diabetes
HbA1C ≥ 48 or
FPG ≥ 7
Missing
2,558,719 (50.1)
2,590,405 (56.4)
5,149,124
Low risk
2,460,489 (48.2)
1,885,332 (41.1)
4,345,821
High risk
83,550 (1.6)
116,484 (2.5)
200,034
High Blood
Pressure
Systolic BP ≥ 140
or Diastolic BP ≥
90
Missing
217,714 (4.3)
1,086,797 (23.7)
1,304,511
Low risk
3,636,511 (71.3)
2,404,097 (52.4)
6,040,608
High risk
1,248,533 (24.5)
1,101,327 (24)
2,349,860
Obesity
BMI ≥ 30
Missing
187,402 (3.7)
2,064,936 (45)
2,252,338
Low risk
3,700,522 (72.5)
1,755,019 (38.2)
5,455,541
High risk
1,214,834 (23.8)
772,266 (16.8)
1,987,100
High
Cholesterol
Total cholesterol
>5mmol/L or
Ratio > 4
Missing
282,100 (5.5)
2,286,595 (49.8)
2,568,695
Low risk
1,519,485 (29.8)
696,458 (15.2)
2,215,943
High risk
3,301,173 (64.7)
1,609,168 (35.0)
4,910,341
CVD risk
score
10 or more
Missing
1,036,820 (20.3)
3,197,683 (69.6)
4,234,503
Low risk
3,014,556 (59.1)
979,685 (21.3)
3,994,241
High risk
1,051,382 (20.6)
414,853 (9)
1,466,235
Family
history of
CVD
Clinical code
present for a CVD
event before 60
years old in a first
degree relative
No
4,910,543 (96.2)
4,561,766 (99.3)
9,472,309
Yes
192,215 (3.8)
30,455 (0.7)
222,670
Physical
Activity
GPPAQ
“moderately
inactive” or
“inactive”
Missing
1,812,161 (35.5)
3,952,015 (86.1)
5,764,176
Low risk
2,184,515 (42.8)
392,263 (8.5)
2,576,778
High risk
1,106,082 (21.7)
247,943 (5.4)
1,354,025
Smoking
Current smoker
Missing
221,351 (4.3)
1,296,474 (28.2)
1,517,825
Low risk
4,066,412 (79.7)
2,325,196 (50.6)
6,391,608
High risk
814,995 (16)
970,551 (21.1)
1,785,546
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance
Supplemental material placed on this supplemental material which has been supplied by the author(s) BMJ Open
doi: 10.1136/bmjopen-2020-042963:e042963. 10 2020;BMJ Open, et al. Patel R
18
Supplementary Table 7: Rules for conflicting risk factors measurements
Rules for processing conflicting risk factor measurements for the same patient on the same day
Risk factor
Rule applied
Smoking status;
Physical activity status
(from GPPAQ)
Records deleted if descriptive statuses are
conflicting (e.g. “smoker” and “non-
smoker” recorded on the same day)
Blood pressure
Record with lowest systolic measurement
taken
BMI; height; weight;
QRISK/QRISK2 score;
Framingham score; total
cholesterol; HDL
cholesterol; Cholesterol
ratio; HbA1c; FPG
Measurements recoded as missing
(unclear which is correct)
Supplementary Table 8: Intervention risk thresholds for action
Intervention
type
Advice or Information given
High risk threshold for action
Advice,
information
or referral
Alcohol usage
Alcohol: FULL AUDIT 8 or more
Diet
Overweight (BMI ≥ 25)
Physical activity
GPPAQ “moderately inactive” or
“inactive”
Lifestyle/Counselling
CVD risk score 10 or more
Smoking cessation
Current smoker
Weight management
Overweight (BMI ≥ 25)
Diabetes
referral
Diabetes Prevention Programme (DPP)
referral
Blood glucose: RAISED risk
HbA1C ≥ 42 and < 48 or FPG ≥ 5.5 and <
7
Statin
prescription
Statins prescribed
CVD risk score 10 or more
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance
Supplemental material placed on this supplemental material which has been supplied by the author(s) BMJ Open
doi: 10.1136/bmjopen-2020-042963:e042963. 10 2020;BMJ Open, et al. Patel R
19
Supplementary Table 9: Data for attendance by UTLA
Number of NHS Health Check invitees and attendees with attendance rate by Upper Tier Local
Authority of patient’s residence
UTLA Code
UTLA
Invitees
Attendees
Attendance
rate
Lower
95% CI
Upper
95% CI
E10000014
Hampshire
179,937
152,318
84.7
84.5
84.8
E09000030
Tower Hamlets
42,098
34,660
82.3
82.0
82.7
E09000028
Southwark
41,938
33,536
80.0
79.6
80.3
E09000025
Newham
51,556
40,706
79.0
78.6
79.3
E09000012
Hackney
37,636
29,713
78.9
78.5
79.4
E08000001
Bolton
64,013
49,792
77.8
77.5
78.1
E09000001
City of London
1,176
910
77.4
74.9
79.7
E08000017
Doncaster
19,869
14,736
74.2
73.6
74.8
E06000053
Isles of Scilly
482
353
73.2
69.1
77.0
E09000022
Lambeth
35,757
26,172
73.2
72.7
73.7
E09000010
Enfield
38,337
27,370
71.4
70.9
71.8
E09000005
Brent
68,977
48,573
70.4
70.1
70.8
E08000002
Bury
31,309
21,979
70.2
69.7
70.7
E09000002
Barking and
Dagenham
36,578
25,402
69.4
69.0
69.9
E09000026
Redbridge
51,865
35,942
69.3
68.9
69.7
E06000021
Stoke-on-Trent
55,178
37,866
68.6
68.2
69.0
E06000008
Blackburn with
Darwen
17,852
12,192
68.3
67.6
69.0
E08000030
Walsall
49,943
33,947
68.0
67.6
68.4
E09000023
Lewisham
26,396
17,838
67.6
67.0
68.1
E08000016
Barnsley
51,420
34,550
67.2
66.8
67.6
E09000009
Ealing
61,109
40,012
65.5
65.1
65.9
E06000039
Slough
16,191
10,600
65.5
64.7
66.2
E09000017
Hillingdon
45,539
29,447
64.7
64.2
65.1
E08000007
Stockport
44,540
28,763
64.6
64.1
65.0
E08000005
Rochdale
36,853
22,967
62.3
61.8
62.8
E09000015
Harrow
29,691
18,476
62.2
61.7
62.8
E06000047
County Durham
120,544
73,877
61.3
61.0
61.6
E09000019
Islington
38,209
23,415
61.3
60.8
61.8
E08000033
Calderdale
41,631
25,247
60.6
60.2
61.1
E09000031
Waltham Forest
50,680
30,720
60.6
60.2
61.0
E08000034
Kirklees
97,779
59,189
60.5
60.2
60.8
E10000029
Suffolk
147,142
89,051
60.5
60.3
60.8
E09000032
Wandsworth
57,469
34,442
59.9
59.5
60.3
E08000025
Birmingham
178,771
106,909
59.8
59.6
60.0
E06000036
Bracknell Forest
19,697
11,778
59.8
59.1
60.5
E10000019
Lincolnshire
200,192
119,037
59.5
59.2
59.7
E06000046
Isle of Wight
24,068
14,251
59.2
58.6
59.8
E08000004
Oldham
34,227
20,184
59.0
58.4
59.5
E06000031
Peterborough
44,281
26,027
58.8
58.3
59.2
E06000025
South
Gloucestershire
59,350
34,683
58.4
58.0
58.8
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance
Supplemental material placed on this supplemental material which has been supplied by the author(s) BMJ Open
doi: 10.1136/bmjopen-2020-042963:e042963. 10 2020;BMJ Open, et al. Patel R
20
E09000014
Haringey
29,867
17,448
58.4
57.9
59.0
E08000022
North Tyneside
40,154
23,434
58.4
57.9
58.8
E06000013
North Lincolnshire
24,121
13,870
57.5
56.9
58.1
E10000017
Lancashire
218,451
125,262
57.3
57.1
57.5
E06000005
Darlington
27,163
15,546
57.2
56.6
57.8
E06000011
East Riding of
Yorkshire
12,161
6,894
56.7
55.8
57.6
E10000003
Cambridgeshire
116,035
65,679
56.6
56.3
56.9
E08000018
Rotherham
7,953
4,476
56.3
55.2
57.4
E06000016
Leicester
40,169
22,547
56.1
55.6
56.6
E06000034
Thurrock
32,083
17,982
56.0
55.5
56.6
E09000018
Hounslow
44,165
24,579
55.7
55.2
56.1
E10000006
Cumbria
120,237
65,183
54.2
53.9
54.5
E06000040
Windsor and
Maidenhead
21,114
11,418
54.1
53.4
54.7
E06000057
Northumberland
75,940
40,859
53.8
53.4
54.2
E10000034
Worcestershire
141,667
76,000
53.6
53.4
53.9
E10000012
Essex
331,942
178,015
53.6
53.5
53.8
E10000024
Nottinghamshire
198,187
106,221
53.6
53.4
53.8
E09000024
Merton
43,144
23,114
53.6
53.1
54.0
E06000022
Bath and North
East Somerset
44,466
23,810
53.5
53.1
54.0
E06000004
Stockton-on-Tees
35,341
18,857
53.4
52.8
53.9
E08000014
Sefton
48,044
25,630
53.3
52.9
53.8
E08000026
Coventry
64,356
34,306
53.3
52.9
53.7
E06000002
Middlesbrough
23,037
12,243
53.1
52.5
53.8
E08000019
Sheffield
80,302
42,628
53.1
52.7
53.4
E10000007
Derbyshire
197,165
104,520
53.0
52.8
53.2
E08000035
Leeds
174,645
92,288
52.8
52.6
53.1
E06000003
Redcar and
Cleveland
25,185
13,304
52.8
52.2
53.4
E08000015
Wirral
80,558
42,456
52.7
52.4
53.0
E10000027
Somerset
75,851
39,814
52.5
52.1
52.8
E10000015
Hertfordshire
200,153
104,948
52.4
52.2
52.7
E09000016
Havering
42,627
22,305
52.3
51.9
52.8
E06000012
North East
Lincolnshire
38,004
19,816
52.1
51.6
52.6
E08000029
Solihull
32,476
16,930
52.1
51.6
52.7
E10000013
Gloucestershire
137,245
71,077
51.8
51.5
52.1
E06000045
Southampton
33,058
17,102
51.7
51.2
52.3
E06000038
Reading
8,400
4,338
51.6
50.6
52.7
E06000027
Torbay
31,524
16,268
51.6
51.1
52.2
E06000024
North Somerset
40,162
20,498
51.0
50.5
51.5
E06000001
Hartlepool
12,989
6,616
50.9
50.1
51.8
E09000027
Richmond upon
Thames
33,597
17,021
50.7
50.1
51.2
E06000033
Southend-on-Sea
48,006
24,182
50.4
49.9
50.8
E06000054
Wiltshire
114,656
57,526
50.2
49.9
50.5
E10000031
Warwickshire
102,623
51,428
50.1
49.8
50.4
E09000029
Sutton
24,049
11,959
49.7
49.1
50.4
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance
Supplemental material placed on this supplemental material which has been supplied by the author(s) BMJ Open
doi: 10.1136/bmjopen-2020-042963:e042963. 10 2020;BMJ Open, et al. Patel R
21
E10000025
Oxfordshire
175,246
87,139
49.7
49.5
50.0
E06000056
Central
Bedfordshire
73,732
36,607
49.6
49.3
50.0
E08000021
Newcastle upon
Tyne
32,888
16,287
49.5
49.0
50.1
E10000021
Northamptonshire
155,686
76,979
49.4
49.2
49.7
E09000003
Barnet
52,312
25,849
49.4
49.0
49.8
E08000006
Salford
34,274
16,934
49.4
48.9
49.9
E06000019
Herefordshire,
County of
37,499
18,421
49.1
48.6
49.6
E06000018
Nottingham
52,693
25,880
49.1
48.7
49.5
E06000043
Brighton and Hove
33,275
16,336
49.1
48.6
49.6
E06000030
Swindon
18,496
9,078
49.1
48.4
49.8
E06000023
Bristol, City of
58,017
28,467
49.1
48.7
49.5
E09000033
Westminster
48,724
23,723
48.7
48.2
49.1
E06000051
Shropshire
67,337
32,700
48.6
48.2
48.9
E08000028
Sandwell
39,552
19,164
48.5
48.0
48.9
E06000042
Milton Keynes
63,247
30,510
48.2
47.9
48.6
E08000036
Wakefield
61,543
29,680
48.2
47.8
48.6
E06000010
Kingston upon
Hull, City of
17,074
8,219
48.1
47.4
48.9
E06000055
Bedford
31,728
15,205
47.9
47.4
48.5
E06000049
Cheshire East
52,794
25,264
47.9
47.4
48.3
E10000011
East Sussex
118,596
56,747
47.8
47.6
48.1
E08000009
Trafford
38,971
18,629
47.8
47.3
48.3
E06000044
Portsmouth
25,966
12,359
47.6
47.0
48.2
E06000059
Dorset
51,066
24,250
47.5
47.1
47.9
E08000023
South Tyneside
33,636
15,962
47.5
46.9
48.0
E10000030
Surrey
74,960
35,532
47.4
47.0
47.8
E06000015
Derby
62,407
29,315
47.0
46.6
47.4
E06000032
Luton
48,454
22,742
46.9
46.5
47.4
E08000008
Tameside
42,845
20,077
46.9
46.4
47.3
E10000008
Devon
105,836
49,495
46.8
46.5
47.1
E09000013
Hammersmith and
Fulham
43,237
20,205
46.7
46.3
47.2
E09000007
Camden
44,662
20,798
46.6
46.1
47.0
E10000023
North Yorkshire
160,704
74,128
46.1
45.9
46.4
E09000004
Bexley
41,045
18,789
45.8
45.3
46.3
E08000003
Manchester
36,987
16,930
45.8
45.3
46.3
E10000028
Staffordshire
99,238
45,042
45.4
45.1
45.7
E08000013
St. Helens
35,045
15,868
45.3
44.8
45.8
E08000011
Knowsley
31,100
14,066
45.2
44.7
45.8
E06000058
Bournemouth,
Christchurch and
Poole
43,888
19,839
45.2
44.7
45.7
E06000020
Telford and
Wrekin
34,384
15,444
44.9
44.4
45.4
E06000009
Blackpool
28,193
12,621
44.8
44.2
45.3
Unknown
Unknown
7,197
3,217
44.7
43.6
45.9
E10000002
Buckinghamshire
136,674
61,016
44.6
44.4
44.9
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance
Supplemental material placed on this supplemental material which has been supplied by the author(s) BMJ Open
doi: 10.1136/bmjopen-2020-042963:e042963. 10 2020;BMJ Open, et al. Patel R
22
E10000032
West Sussex
90,033
40,022
44.5
44.1
44.8
E06000006
Halton
26,863
11,753
43.8
43.2
44.3
E06000052
Cornwall
48,099
20,877
43.4
43.0
43.8
E06000050
Cheshire West
and Chester
40,408
17,537
43.4
42.9
43.9
E06000035
Medway
60,300
26,064
43.2
42.8
43.6
E10000020
Norfolk
161,582
69,173
42.8
42.6
43.1
E06000017
Rutland
6,741
2,862
42.5
41.3
43.6
E09000006
Bromley
75,672
31,841
42.1
41.7
42.4
E10000016
Kent
347,229
145,984
42.0
41.9
42.2
E09000008
Croydon
29,612
12,399
41.9
41.3
42.4
E09000011
Greenwich
32,488
13,547
41.7
41.2
42.2
E06000014
York
20,330
8,385
41.2
40.6
41.9
E08000027
Dudley
78,489
32,316
41.2
40.8
41.5
E06000026
Plymouth
28,855
11,707
40.6
40.0
41.1
E08000012
Liverpool
99,029
40,074
40.5
40.2
40.8
E10000018
Leicestershire
172,437
69,666
40.4
40.2
40.6
E08000024
Sunderland
47,131
18,370
39.0
38.5
39.4
E09000020
Kensington and
Chelsea
35,607
13,811
38.8
38.3
39.3
E06000007
Warrington
48,004
18,287
38.1
37.7
38.5
E08000031
Wolverhampton
32,226
12,091
37.5
37.0
38.0
E08000010
Wigan
53,620
19,638
36.6
36.2
37.0
E09000021
Kingston upon
Thames
32,087
11,529
35.9
35.4
36.5
E06000041
Wokingham
5,010
1,621
32.4
31.1
33.7
E08000037
Gateshead
49,663
14,497
29.2
28.8
29.6
E06000037
West Berkshire
16,235
4,376
27.0
26.3
27.6
E08000032
Bradford
82,669
20,791
25.1
24.9
25.4
Supplementary Table 10: Number of invitations recorded for attendees and non-
attendees
Number of invitations
Attendees n(%)
Non-attendees n(%)
0
1,672,844 (32.8)
51,739 (1.1)
1
2,577,581 (50.5)
3,369,517 (73.4)
2
677,783 (13.3)
783,472 (17.1)
> 2
174,550 (3.4)
387,493 (8.4)
TOTAL
5,102,758 (100.0)
4,592,221 (100.0)
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance
Supplemental material placed on this supplemental material which has been supplied by the author(s) BMJ Open
doi: 10.1136/bmjopen-2020-042963:e042963. 10 2020;BMJ Open, et al. Patel R
23
Supplementary Table 11: Invitations by financial year
Proportion of attendees and non-attendees with an invitation recorded
Year
Attendees with
invitation
% attendees
Non-attendees
with invitation
% non-
attendees
2012/13
468,766
63.1
718,527
99.0
2013/14
619,559
64.3
824,429
98.9
2014/15
763,444
67.2
1,016,155
99.0
2015/16
790,731
69.2
999,178
98.7
2016/17
787,414
70.4
982,193
98.8
TOTAL
3,429,914
67.2
4,540,482
98.9
Supplementary Table 12: Completeness of risk factor measurement
Percentage of NHSHC attendees and non-attendees with recorded risk factor measurements
(restricted to 15-month window around index date for attendees and unrestricted for non-
attendees)
Group
CVD risk score
Body Mass Index
Physical
Activity (GPPAQ)
Alcohol (Audit C)
Fasting glucose
HbA1C
Smoking Status
Cholesterol (HDL)
Cholesterol
(total)
Diastolic BP
Systolic BP
Atten
dees
79.7%
96.3%
64.5%
38.3%
18.2%
36.6%
95.7%
87.2%
93.6%
95.7%
95.8%
Non-
atten
dees
30.4%
55.0%
13.9%
16.7%
15.1%
37.5%
71.8%
47.3%
50.0%
76.3%
76.3%
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance
Supplemental material placed on this supplemental material which has been supplied by the author(s) BMJ Open
doi: 10.1136/bmjopen-2020-042963:e042963. 10 2020;BMJ Open, et al. Patel R
24
Supplementary Table 13: Statin prescription rates
New statin (any dose) prescriptions among the subset (60.4%) of NHSHC attendees in whom
medication data was available
Group
Attendees (n)
Prescribed a statin (n)
Proportion (%)
CVD score <10%
1,910,919
63,227
3.3
10-19.9%
532,046
83,279
15.7
20%
132,366
51,691
39.1
No CVD score
504,374
55,630
11.0
Overall total
3,079,705
253,827
8.2
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance
Supplemental material placed on this supplemental material which has been supplied by the author(s) BMJ Open
doi: 10.1136/bmjopen-2020-042963:e042963. 10 2020;BMJ Open, et al. Patel R
... Considering the trends in the NHSHC uptake [15][16][17][18], there has been a steady increase in the number of patients attending appointments. ...
... Ethnicity was cumulatively compared with the UK 2011 census, as this is the closest to the eligible studies data collection period [19]. The ratio of each ethnic group within our review was compared with the ratio of the same ethnicity residing in the UK population (2011 census) and presented as a percentage (called degree of representation- Table 2, Ref. [15][16][17][18][19][20][21][22][23][24]). ...
... The characteristics of the seven papers eligible for the review are presented in Table 3 (Ref. [15][16][17][18][20][21][22]). The seven studies comprised 6,622,374 patients, 80.2% of whom were white, with 3.1% being Black, 5.1% South Asian, 0.5% being Chinese, 4.2% categorised as other, and 6.9% with missing ethnicity data. ...
Article
Full-text available
Background Cardiovascular diseases (CVD) affect around 7.6 million people in the UK, disproportionately affecting the minority ethnic community. In 2009, the UK's National Health Service (NHS) launched a Health Check (NHSHC) scheme to improve early diagnosis of various clinical conditions, including CVD, by screening patients for associated risk factors. This systematic review investigated the engagement of minority ethnic groups with these services. Methods Seven studies identified patient demographics of NHSHC attendees using the Preferred Reporting Items for Systematic And Meta Analysis-Diagnostic Test Accuracy (PRISMA-DTA) guidelines and accessing Ovid (MEDLINE), PubMed and Web of Science databases. Results The screening was either by invitation or opportunistic at other appointments with their doctor. Engagement with the service was highest among the South Asian patients (21%–68%), but lowest amongst Chinese patients (12%–61%). Further, engagement was lower among those screened following a formal invitation than those seen opportunistically. However, a greater proportion of patients were screened opportunistically than by invitation. Conclusions Overall, we found that the NHSHC is not being utilised adequately for all patients at high risk of CVD, particularly White and Chinese patients. It highlights the critical role of primary care could play to improve patient engagement with the service.
... 14 Those attending Health Checks include a higher proportion of women, those who are younger and those from the most affluent areas, although ethnic minority groups are also well represented. 15 They are also more likely than non-attendees to be diagnosed with type 2 diabetes, hypertension and chronic kidney disease and to receive treatment such as statins and antihypertensives. 16 The NHS is committed to digital transformation of health and social care, including specific aims for digital health checks and risk-based screening in England by 2028. ...
... One of the primary aims of this study was to assess the extent to which the DHC is effective at engaging those groups not been reached by the F2F Health Check. Previous research has found that those less likely to attend health checks are male, older and from more deprived areas, 15 and we saw similar demographics in this study among both F2F and DHC non-completers. Of those invited to DHC, those choosing DHC rather than F2F were disproportionately male, white and from more affluent areas. ...
... This phase of the online tool is designed to mirror and extend the advice and signposting to further information part of the F2F health check, which is recommended by the NHS Health Check Best Practice Guidance. 13 However, it is not clear that this always takes place, with a pre-COVID-19 national study reporting that only 16% of health check attendees were coded as having received general lifestyle and behavioural advice, 15 although they report high levels of missing data and this will likely vary depending on how the F2F programme is delivered within a local authority; note that our survey of local users reported much higher levels of advice received for F2F participants (around 80%). However, even when advice is offered, current guidance recommends the delivery of only 'very brief advice' to 'extended brief intervention', depending on an individual need. ...
Article
Full-text available
Objectives To compare the uptake, effectiveness and costs of a digital version of the National Health Service (NHS) Health Check (DHC) to the standard face-to-face NHS Health Check (F2F). Participants and setting A random sample of 9000 patients aged 40–74 eligible for an NHS Health Check in Southwark, England, between January and April 2023. Intervention and design The DHC was an online tool with a health assessment section, an advice and support section, and a section on how to obtain and update follow-up physical measures (blood pressure, cholesterol, glycated haemoglobin (HbA1c)). 6000 patients from GP records were randomly allocated to receive a DHC invitation and 3000 to receive an F2F invitation. Those invited to DHC were able to choose F2F if they preferred. Outcomes The primary outcome was the uptake of any type of health check, either a completed F2F appointment or completion of the DHC health assessment section, along with demographics and data on appointments, medications and referrals within the study period. QRISK3 and QDiabetes risk scores were calculated. Management and operation costs were estimated for F2F and DHC pathways. Results Excluding participants who moved away or died, the DHC uptake to the health assessment section was 21% (1189/5705), with a further 3% (198/5705) choosing F2F, compared with 11% (305/2900) for F2F completion (p<0.001). The DHC uptake was lower among those from Black (14%) and Mixed (13%) compared with White (29%) ethnicities (p<0.001), and there was no evidence of higher DHC uptake among groups less likely to engage in NHS Health Checks. Of those who completed the health assessment, 60% (714) completed the support section, and 7% (84) completed the provision and updating of physical measures. Appointments, medications and referrals were lower among DHC service users than among F2F users (p<0.001). The estimated total management and operation costs for F2F were £154.80 per user, compared with total management and operation costs for DHC of £68.48 per user for health assessment only, £134.46 including the support section and £1479.01 per user with completed physical measures. Conclusions The study suggests that a choice of Health Check pathways may potentially reduce pressures on the NHS. Cholesterol and HbA1c were not generally known, and the options to obtain and update these measures require further development for the DHC to be considered a viable comparable alternative to the F2F service for estimating cardiovascular disease and diabetes risk. Strategies are still needed to reach those groups not currently engaging with NHS Health Checks. Registration This study was registered on the Open Science Framework: https://osf.io/y87zt .
... This presents an opportunity to access a broader patient population who may not have signs and symptoms of chronic disease, are at an earlier stage in the disease course and may not have recently attended their general medical practitioner (GMP) or accessed health screening elsewhere. Additionally, success of the NHS Health Check programme has been modest; with only about half (52.6%) of those invited attending (Patel et al., 2020). Therefore, the remainder of those invited may benefit from receiving health screening in alternative settings, such as in dental settings. ...
... By comparison, the first year of the NHS Healthcheck, a highly promoted national screening service for adult cardiovascular risk, achieved a 33.9% uptake. 29 Furthermore, COVID-19 interrupting the service is likely to have impeded engagement. ...
Article
Full-text available
Background Early intervention for unmet needs is essential to improve health. Clear inequalities in healthcare use and outcomes exist. The Children and Young People’s Health Partnership (CYPHP) model of care uses population health management methods to (1) identify and proactively reach children with asthma, eczema and constipation (tracer conditions); (2) engage these families, with CYPHP, by sending invitations to complete an online biopsychosocial Healthcheck Questionnaire; and (3) offer early intervention care to those children found to have unmet health needs. We aimed to understand this model’s effectiveness to improve equitable access to care. Methods We used primary care and CYPHP service-linked records and applied the same methods as the CYPHP’s population health management process to identify children aged <16 years with a tracer condition between 1 April 2018 and 30 August 2020, those who engaged by completing a Healthcheck and those who received early intervention care. We applied multiple imputation with multilevel logistic regression, clustered by general practitioner (GP) practice, to investigate the association of deprivation and ethnicity, with children’s engagement and receiving care. Results Among 129 412 children, registered with 70 GP practices, 15% (19 773) had a tracer condition and 24% (4719) engaged with CYPHP’s population health management system. Children in the most deprived, compared with least deprived communities, had 26% lower odds of engagement (OR 0.74; 95% CI 0.62 to 0.87). Children of Asian or black ethnicity had 31% lower odds of engaging, compared with white children (0.69 (0.59 to 0.81) and 0.69 (0.62 to 0.76), respectively). However, once engaged with the population health management system, black children had 43% higher odds of receiving care, compared with white children (1.43 (1.15 to 1.78)), and children from the most compared with least deprived communities had 50% higher odds of receiving care (1.50 (1.01 to 2.22)). Conclusion Detection of unmet needs is possible using population health management methods and increases access to care for children from priority populations with the highest needs. Further health system strengthening is needed to improve engagement and enhance proportionate universalist access to healthcare. Trial registration number ClinicalTrials.gov Registry ( NCT03461848 ).
... 11 By contrast, the incidence of myocardial infarction has decreased over recent decades, 12 in association with national programmes of vascular checks to address key risk factors for ischaemic heart disease. 13 This new study reinforces the principle that analogous primary prevention programmes for atrial fibrillation are required to stem the apparent rise in incidence, associated disease burden, and cost. 2 14 Unfortunately, the evidence base for primary prevention of atrial fibrillation predominantly relies on observational data and post-hoc analyses of data from randomised clinical trials where atrial fibrillation was not prespecified as a primary or secondary endpoint, and occurrence was not systematically collected. 15 As a consequence, international guidelines do not provide specific recommendations for interventions to reduce the risk of newly onset atrial fibrillation. ...
Article
This critical review paper examines the health inequalities faced by individuals with Severe Mental Illness (SMI) in the United Kingdom; highlighting the disproportionate burden of physical health conditions such as respiratory disorders, cardiac illnesses, diabetes and stroke amongst this population. These conditions contribute to a significantly higher rate of premature mortality in individuals with SMI, with two‐thirds of these deaths deemed preventable. Despite the National Health Service (NHS) acknowledging the need to address these health inequalities, the mortality gap between those with and without SMI continues to widen. Additionally, there is limited engagement from service users in annual physical health checks, a concern that this paper addresses by identifying several barriers and providing recommendations to improve access and engagement in physical health checks. This review emphasises the focus on primary care systems as a critical point for addressing health disparities in individuals with SMI. Also, it highlights the need for primary care services to be more adaptive and integrated, playing a key role in managing the physical health of patients with SMI through regular health checks, flexible service delivery, and enhanced coordination with secondary care. Effectively supporting individuals with SMI requires tailored, integrated primary care interventions that address both psychological and physical health challenges, considering diverse demographic needs across the UK.
Article
Purpose Preventative health services are keen to identify how to engage men and increase their participation, thus improving health, well-being and life expectancy over time. Prior research has shown general gender norms are a key reason for men’s avoidance of these services, yet there is little investigation of specific gender norms. Furthermore, masculinity has not been examined as a factor associated with customer vulnerability. This paper aims to identify the relationship between gender norm segments for men, likely customer vulnerability over time and subjective health and well-being. Design/methodology/approach Adult males ( n = 13,891) from an Australian longitudinal men’s health study were classified using latent class analysis. Conditional growth mixture modelling was conducted at three timepoints. Findings Three masculinity segments were identified based on masculine norm conformity: traditional self-reliant , traditional bravado and modern status . All segments had likely customer experience of vulnerability. Over time, the likely experience was temporary for the modern status segment but prolonged for the traditional self-reliant and traditional bravado segments. The traditional self-reliant segment had low subjective health and low overall well-being over time. Practical implications Practitioners can tailor services to gender norm segments, enabling self-reliant men to provide expertise and use the “Status” norm to reach all masculinity segments. Originality/value The study of customer vulnerability in a group usually considered privileged identifies differential temporal experiences based on gender norms. The study confirms customer vulnerability is temporal in nature; customer vulnerability changes over time from likely to actual for self-reliant men.
Article
Background and Aims A cardiovascular disease polygenic risk score (CVD-PRS) can stratify individuals into different categories of cardiovascular risk, but whether the addition of a CVD-PRS to clinical risk scores improves the identification of individuals at increased risk in a real-world clinical setting is unknown. Methods The Genetics and the Vascular Health Check Study (GENVASC) was embedded within the UK National Health Service Health Check (NHSHC) programme which invites individuals between 40–74 years of age without known CVD to attend an assessment in a UK general practice where CVD risk factors are measured and a CVD risk score (QRISK2) is calculated. Between 2012–2020, 44,141 individuals (55.7% females, 15.8% non-white) who attended an NHSHC in 147 participating practices across two counties in England were recruited and followed. When 195 individuals (cases) had suffered a major CVD event (CVD death, myocardial infarction or acute coronary syndrome, coronary revascularisation, stroke), 396 propensity-matched controls with a similar risk profile were identified, and a nested case-control genetic study undertaken to see if the addition of a CVD-PRS to QRISK2 in the form of an integrated risk tool (IRT) combined with QRISK2 would have identified more individuals at the time of their NHSHC as at high risk (QRISK2 10-year CVD risk of ≥10%), compared with QRISK2 alone. Results The distribution of the standardised CVD-PRS was significantly different in cases compared with controls (cases mean score .32; controls, −.18, P = 8.28×10−9). QRISK2 identified 61.5% (95% confidence interval [CI]: 54.3%–68.4%) of individuals who subsequently developed a major CVD event as being at high risk at their NHSHC, while the combination of QRISK2 and IRT identified 68.7% (95% CI: 61.7%–75.2%), a relative increase of 11.7% (P = 1×10−4). The odds ratio (OR) of being up-classified was 2.41 (95% CI: 1.03–5.64, P = .031) for cases compared with controls. In individuals aged 40–54 years, QRISK2 identified 26.0% (95% CI: 16.5%–37.6%) of those who developed a major CVD event, while the combination of QRISK2 and IRT identified 38.4% (95% CI: 27.2%–50.5%), indicating a stronger relative increase of 47.7% in the younger age group (P = .001). The combination of QRISK2 and IRT increased the proportion of additional cases identified similarly in women as in men, and in non-white ethnicities compared with white ethnicity. The findings were similar when the CVD-PRS was added to the atherosclerotic cardiovascular disease pooled cohort equations (ASCVD-PCE) or SCORE2 clinical scores. Conclusions In a clinical setting, the addition of genetic information to clinical risk assessment significantly improved the identification of individuals who went on to have a major CVD event as being at high risk, especially among younger individuals. The findings provide important real-world evidence of the potential value of implementing a CVD-PRS into health systems.
Article
Atherosclerotic cardiovascular disease (ASCVD) remains a leading cause of morbidity and mortality worldwide, highlighting the urgent need for advancements in risk assessment and management strategies. Although significant progress has been made recently, identifying and managing apparently healthy individuals at a higher risk of developing atherosclerosis and those with subclinical atherosclerosis still poses significant challenges. Traditional risk assessment tools have limitations in accurately predicting future events and fail to encompass the complexity of the atherosclerosis trajectory. In this review, we describe novel approaches in biomarkers, genetics, advanced imaging techniques, and artificial intelligence that have emerged to address this gap. Moreover, polygenic risk scores and imaging modalities such as coronary artery calcium scoring, and coronary computed tomography angiography offer promising avenues for enhancing primary cardiovascular risk stratification and personalised intervention strategies. On the other hand, interventions aiming against atherosclerosis development or promoting plaque regression have gained attention in primary ASCVD prevention. Therefore, the potential role of drugs like statins, ezetimibe, proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitors, omega-3 fatty acids, antihypertensive agents, as well as glucose-lowering and anti-inflammatory drugs are also discussed. Since findings regarding the efficacy of these interventions vary, further research is still required to elucidate their mechanisms of action, optimize treatment regimens, and determine their long-term effects on ASCVD outcomes. In conclusion, advancements in strategies addressing atherosclerosis prevention and plaque regression present promising avenues for enhancing primary ASCVD prevention through personalised approaches tailored to individual risk profiles. Nevertheless, ongoing research efforts are imperative to refine these strategies further and maximise their effectiveness in safeguarding cardiovascular health.
Article
Full-text available
Background The Global Burden of Diseases, Injuries, and Risk Factors Study (GBD) 2017 comparative risk assessment (CRA) is a comprehensive approach to risk factor quantification that offers a useful tool for synthesising evidence on risks and risk–outcome associations. With each annual GBD study, we update the GBD CRA to incorporate improved methods, new risks and risk–outcome pairs, and new data on risk exposure levels and risk–outcome associations. Methods We used the CRA framework developed for previous iterations of GBD to estimate levels and trends in exposure, attributable deaths, and attributable disability-adjusted life-years (DALYs), by age group, sex, year, and location for 84 behavioural, environmental and occupational, and metabolic risks or groups of risks from 1990 to 2017. This study included 476 risk–outcome pairs that met the GBD study criteria for convincing or probable evidence of causation. We extracted relative risk and exposure estimates from 46 749 randomised controlled trials, cohort studies, household surveys, census data, satellite data, and other sources. We used statistical models to pool data, adjust for bias, and incorporate covariates. Using the counterfactual scenario of theoretical minimum risk exposure level (TMREL), we estimated the portion of deaths and DALYs that could be attributed to a given risk. We explored the relationship between development and risk exposure by modelling the relationship between the Socio-demographic Index (SDI) and risk-weighted exposure prevalence and estimated expected levels of exposure and risk-attributable burden by SDI. Finally, we explored temporal changes in risk-attributable DALYs by decomposing those changes into six main component drivers of change as follows: (1) population growth; (2) changes in population age structures; (3) changes in exposure to environmental and occupational risks; (4) changes in exposure to behavioural risks; (5) changes in exposure to metabolic risks; and (6) changes due to all other factors, approximated as the risk-deleted death and DALY rates, where the risk-deleted rate is the rate that would be observed had we reduced the exposure levels to the TMREL for all risk factors included in GBD 2017. Findings In 2017, 34·1 million (95% uncertainty interval [UI] 33·3–35·0) deaths and 1·21 billion (1·14–1·28) DALYs were attributable to GBD risk factors. Globally, 61·0% (59·6–62·4) of deaths and 48·3% (46·3–50·2) of DALYs were attributed to the GBD 2017 risk factors. When ranked by risk-attributable DALYs, high systolic blood pressure (SBP) was the leading risk factor, accounting for 10·4 million (9·39–11·5) deaths and 218 million (198–237) DALYs, followed by smoking (7·10 million [6·83–7·37] deaths and 182 million [173–193] DALYs), high fasting plasma glucose (6·53 million [5·23–8·23] deaths and 171 million [144–201] DALYs), high body-mass index (BMI; 4·72 million [2·99–6·70] deaths and 148 million [98·6–202] DALYs), and short gestation for birthweight (1·43 million [1·36–1·51] deaths and 139 million [131–147] DALYs). In total, risk-attributable DALYs declined by 4·9% (3·3–6·5) between 2007 and 2017. In the absence of demographic changes (ie, population growth and ageing), changes in risk exposure and risk-deleted DALYs would have led to a 23·5% decline in DALYs during that period. Conversely, in the absence of changes in risk exposure and risk-deleted DALYs, demographic changes would have led to an 18·6% increase in DALYs during that period. The ratios of observed risk exposure levels to exposure levels expected based on SDI (O/E ratios) increased globally for unsafe drinking water and household air pollution between 1990 and 2017. This result suggests that development is occurring more rapidly than are changes in the underlying risk structure in a population. Conversely, nearly universal declines in O/E ratios for smoking and alcohol use indicate that, for a given SDI, exposure to these risks is declining. In 2017, the leading Level 4 risk factor for age-standardised DALY rates was high SBP in four super-regions: central Europe, eastern Europe, and central Asia; north Africa and Middle East; south Asia; and southeast Asia, east Asia, and Oceania. The leading risk factor in the high-income super-region was smoking, in Latin America and Caribbean was high BMI, and in sub-Saharan Africa was unsafe sex. O/E ratios for unsafe sex in sub-Saharan Africa were notably high, and those for alcohol use in north Africa and the Middle East were notably low. Interpretation By quantifying levels and trends in exposures to risk factors and the resulting disease burden, this assessment offers insight into where past policy and programme efforts might have been successful and highlights current priorities for public health action. Decreases in behavioural, environmental, and occupational risks have largely offset the effects of population growth and ageing, in relation to trends in absolute burden. Conversely, the combination of increasing metabolic risks and population ageing will probably continue to drive the increasing trends in non-communicable diseases at the global level, which presents both a public health challenge and opportunity. We see considerable spatiotemporal heterogeneity in levels of risk exposure and risk-attributable burden. Although levels of development underlie some of this heterogeneity, O/E ratios show risks for which countries are overperforming or underperforming relative to their level of development. As such, these ratios provide a benchmarking tool to help to focus local decision making. Our findings reinforce the importance of both risk exposure monitoring and epidemiological research to assess causal connections between risks and health outcomes, and they highlight the usefulness of the GBD study in synthesising data to draw comprehensive and robust conclusions that help to inform good policy and strategic health planning.
Article
Full-text available
This paper estimated mortality for 282 causesof death in 195 countries from 1980 to 2017, adding 18 causes to its estimates compared to GBD 2016. In 2017, the GBD study added numerous data sources, including 127 country-years of vital registration data and 502 country-years of cancer registry data.