- Email: [email protected]
Improving identification and management of familial hypercholesterolaemia in primary care: Pre- and post-intervention study
Accepted Manuscript Improving identification and management of familial hypercholesterolaemia in primary care: Pre- and post-intervention study Stephen Weng, Joe Kai, Jennifer Tranter, Jo Leonardi-Bee, Nadeem Qureshi PII:
S0021-9150(18)30222-3
DOI:
10.1016/j.atherosclerosis.2018.04.037
Reference:
ATH 15490
To appear in:
Atherosclerosis
Received Date: 16 March 2018 Revised Date:
18 April 2018
Accepted Date: 27 April 2018
Please cite this article as: Weng S, Kai J, Tranter J, Leonardi-Bee J, Qureshi N, Improving identification and management of familial hypercholesterolaemia in primary care: Pre- and post-intervention study, Atherosclerosis (2018), doi: 10.1016/j.atherosclerosis.2018.04.037. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT 1
Improving identification and management of familial
2
hypercholesterolaemia in primary care: Pre- and post-intervention study
3
5 6 7
1
8
School of Medicine, University of Nottingham, UK;
NIHR Research Fellow, Division of Primary Care, NIHR School for Primary Care Research,
10
2
11
University of Nottingham, UK;
M AN U
9
Division of Primary Care, NIHR School for Primary Care Research, School of Medicine,
12 13
3
14
University of Nottingham, UK.
TE D
Division of Primary Care, NIHR School for Primary Care Research, School of Medicine,
15 16
4
17
Nottingham, UK.
AC C
EP
Division of Epidemiology and Public Health, School of Medicine, University of
18 19
5
20
University of Nottingham, UK.
21
RI PT
Stephen Weng1, Joe Kai2, Jennifer Tranter3, Jo Leonardi-Bee4, Nadeem Qureshi5
SC
4
Division of Primary Care, NIHR School for Primary Care Research, School of Medicine,
22 23
Correspondence:
24
Stephen Weng
25
Division of Primary Care 1
26
1403 The Tower
27
University Park
28
University of Nottingham
29
NG7 2RD, UK
30
[email protected]
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
31
2
ACCEPTED MANUSCRIPT 32
ABSTRACT
33 Background and aims: Familial hypercholesterolaemia (FH) is a major cause of premature
35
heart disease but remains unrecognised in most patients. This study investigated if a
36
systematic primary care-based approach to identify and manage possible FH improves
37
recommended best clinical practice.
RI PT
34
38
Methods: Pre- and post-intervention study in six UK general practices (population 45,033)
40
which invited patients with total cholesterol >7.5 mmol/L to be assessed for possible FH.
41
Compliance with national guideline recommendations to identify and manage possible FH
42
(repeat cholesterol; assess family history of heart disease; identify secondary causes and
43
clinical features; reduce total & LDL-cholesterol; statin prescribing; lifestyle advice) was
44
assessed by calculating the absolute difference in measures of care pre- and six months post-
45
intervention.
TE D
M AN U
SC
39
46
Results: The intervention improved best clinical practice in 118 patients consenting to
48
assessment (of 831 eligible patients): repeat cholesterol test (+75.4%, 95% CI 66.9-82.3);
49
family history of heart disease assessed (+35.6%, 95% CI 27.0-44.2); diagnosis of secondary
50
causes (+7.7%, 95% CI 4.1-13.9), examining clinical features (+6.0%, 95% CI 2.9-11.7). For
51
32 patients diagnosed with possible FH using Simon-Broome criteria, statin prescribing
52
significantly improved (18.8%, 95% CI 8.9-35.3) with non-significant mean reductions in
53
cholesterol post-intervention (total: -0.16 mmol/L, 95% CI -0.78-0.46; LDL: -0.12 mmol/L,
54
95% CI -0.81-0.57).
AC C
EP
47
55
3
ACCEPTED MANUSCRIPT Conclusions: Within six months, this simple primary care intervention improved both
57
identification and management of patients with possible FH, in line with national evidence-
58
based guidelines. Replicating and sustaining this approach across the country could lead to
59
substantial improvement in health outcomes for these individuals with very high
60
cardiovascular risk.
RI PT
56
61 Keywords
63
primary health care; familial hypercholesterolaemia; family health; medical genetics
SC
62
64
AC C
EP
TE D
M AN U
65
4
ACCEPTED MANUSCRIPT 66
INTRODUCTION
67 Familial hypercholesterolaemia (FH) is the most common autosomal dominant disorder with
69
at least 1 in 500 individuals affected in the general population.1 Over 80% of an estimated
70
120,000 individuals in the United Kingdom remain undiagnosed, with similar rates of under-
71
diagnosis across Europe, which results in major lost opportunities to prevent premature heart
72
disease and death.2 Left untreated, coronary heart disease (CHD) will develop in
73
approximately 50% of men with FH by the age of 50 and 30% of women with FH by the age
74
of 60.3 This results in a 100-fold increase in mortality risk compared to the general
75
population.4,5 This can be very effectively reduced by high intensity lipid lowering treatment,
76
resulting in a 48% reduction in CHD mortality.6
M AN U
SC
RI PT
68
77
Despite established national clinical guidelines in several countries 7-9, people with raised
79
cholesterols are not recognised as having FH. In the UK, cholesterol levels are well captured
80
in primary care electronic health records, offering the opportunity to improve recognition and
81
quality of care of this condition. The English National Institute for Health and Care
82
Excellence (NICE) guidelines introduced in 2008, 8 recommend general practitioners (GPs)
83
use Simon-Broome diagnostic criteria to identify FH.10 These state individuals with
84
cholesterol levels over 7.5 mmol/L and a relevant family history of premature CHD be
85
diagnosed as “possible FH” and be referred for specialist assessment and confirmation of
86
diagnosis.
AC C
EP
TE D
78
87 88
Combining a search of general practice electronic medical records for patients with elevated
89
cholesterol and systematic collection of family history may improve identification of possible
90
FH. 11-13 This should be accompanied by better recording of clinical features of FH and
5
ACCEPTED MANUSCRIPT excluding secondary causes of raised cholesterol.8 Further, proactive case ascertainment
92
should also improve lipid management. This study aimed to assess whether such a proactive
93
approach to identify possible familial hypercholesterolaemia (FH) in primary care improves
94
best practice in accord with national guideline recommendations on identifying and managing
95
FH. 8
RI PT
91
96 97
MATERIALS AND METHODS
99
SC
98 Study design
M AN U
100 101
This was a pre- and post-intervention study over 17 months, which ended in Aug 2015.
102
Ethical approval granted by the NRES Committee West Midlands – Solihull (Reference
103
12/WM/0322).
TE D
104
The baseline (pre-intervention period) was defined as the 6-month period from when a
106
participants’ total cholesterol was found to be ≥ 7.5 mmol/L after the release of the NICE FH
107
guidelines in November 2008. The intervention was first introduced at various time points
108
from March 2013 (first practice recruited) to January 2014 (last practice recruited). During
109
this period, participants with total cholesterol levels ≥ 7.5 mmol/L were invited to complete a
110
family history questionnaire. If the family history was positive, the patients was then offered
111
further clinical assessment over a six month period (post-intervention period). This process is
112
shown in Fig. 1.
AC C
EP
105
113 114 115 6
ACCEPTED MANUSCRIPT 116
Participants
117 All general practices in Nottingham (East Midlands, UK) were invited to participate in this
119
study. We received initial expressions of interest from 18 general practices. Six of these
120
practices responded (33.3%) following provision of further study information, and all these
121
were recruited (four inner city, one suburban, one rural).
122
RI PT
118
Patients from each of the six practices were eligible if they were aged 18 years or over, with a
124
total cholesterol level greater than 7.5 mmol/L documented in their primary care electronic
125
health records (EHRs). Participants with a confirmed diagnosis of familial
126
hypercholesterolaemia were excluded. Full details of participant recruitment are reported
127
elsewhere.14 Written informed consent was obtained from each patient included in the study
128
as outlined in the study protocol which conforms to the ethical guidelines of the 1975
129
Declaration of Helsinki and has been prior approved by the Institution's ethics committee on
130
research on humans.
133
M AN U
TE D
Intervention and procedure
EP
132
AC C
131
SC
123
134
The intervention consisted of an educational session with general practitioners and practice
135
nurses which lasted an hour on case-identification and assessment of FH using Simon-
136
Broome diagnostic criteria recommended in the current NICE guidelines.8 The computer-
137
based reminder message that appears when they accessed eligible participants’ records in
138
patient consultations was also demonstrated.
139
7
ACCEPTED MANUSCRIPT A baseline data extraction from electronic health records (EHR) was then completed to
141
identify eligible participants and their electronic health records were tagged with a reminder
142
message whenever the record was accessed in a consultation. In line with NICE FH
143
guidelines 8 , this prompted their GP to opportunistically assess for secondary causes of
144
raised cholesterol, ask participants to complete Family History and Symptom Questionnaires
145
(FHSQ) and repeat lipid profile, Four months into the intervention, eligible participants not
146
yet invited opportunistically by practitioners were contacted through postal invitation by their
147
practice.
SC
RI PT
140
148
Postal invitations included requests to conduct a follow-up fasting lipid test and complete the
150
FHSQ. After return of the FHSQ, a research assistant collated fasting lipid results, current
151
family history information, statin therapy, and notified GPs and participants if there was any
152
possibility of a FH diagnosis based on NICE Simon-Broome criteria.8 Post-intervention, we
153
then assessed GP compliance to national (NICE) guideline recommendations on best clinical
154
practice for FH identification and management.
TE D
M AN U
149
157 158
Outcome measures
AC C
156
EP
155
159
Outcome measures for compliance to NICE guideline recommendations were assessed for all
160
recruited participants at the end of the pre-intervention period (Fig.1, Time 2) and at the end
161
of the intervention period (post-intervention) (Fig.1, Time 4), including: (i) Assessment for
162
FH in patients with elevated cholesterol (ii) Management of patients who have been
163
diagnosed with possible FH. (see Table 1; further detail is available in Supplementary
164
Data). These outcome measures were also extracted from general practice EHRs at the end of
8
ACCEPTED MANUSCRIPT 165
each six month period. Extraction from EHRs was performed using automated data extraction
166
via medical codes and manual clinical record review for free-text entries and hospital
167
correspondence.
168 For the assessment of FH in patients with elevated cholesterols, the primary outcome measure
170
was defined as the difference in the proportion of patients with any family history of coronary
171
heart disease (CHD) documented in their EHRs between pre- and post-intervention periods.
172
Other outcome measures related to diagnostic assessment include the difference; between
173
pre- and post-intervention periods in proportions of: patients with repeated cholesterol tests,
174
diagnoses of secondary causes, biomarkers for secondary causes measured (TSH, HbA1c,
175
serum creatinine, liver function tests), diagnoses of arcus senilis or xanthalasma. GPs in the
176
study were advised to specifically assess for arcus cornealis and tendon xanthoma but code in
177
the electronic health records as arcus senilis and generic xanthalasma due to limited
178
availability of relevant NHS read codes.
TE D
M AN U
SC
RI PT
169
179
For the management of those diagnosed with possible FH, the outcomes were change in total
181
cholesterol, LDL cholesterol, as well as, difference in proportions of patients taking statin
182
treatment (any and high potency), given advice for managing diet and weight, and given
183
advice to stop smoking.
AC C
184
EP
180
185
Sample size calculation
186
For the primary outcome measure, a sample size of 118 patients from the six practices would
187
allow an absolute increase of 10% in the proportion of patients with family history of CHD
188
recorded in EHRs to be determined with 95% confidence intervals to within ±7% (i.e. 95%
189
CI 3% to 17%), assuming a baseline of 3.7% from a previous study. 15 9
ACCEPTED MANUSCRIPT 190 Statistical analysis
192
A descriptive analysis was conducted presenting numbers and percentages for all categorical
193
outcome variables, with continuous variables of total cholesterol and LDL-cholesterol
194
presented as mean and standard deviation, stratified by gender. To evaluate effects of the
195
intervention on compliance to NICE guideline recommendations, we calculated the change in
196
absolute difference for proportions with 95% confidence intervals (Wald’s method based on
197
normal approximation) 16 for categorical outcomes and mean difference with 95% confidence
198
intervals 17 for continuous cholesterol outcomes, between pre- and post-intervention periods.
199
Continuous cholesterol variables were investigated for normality and if needed, were log-
200
transformed for better fit. All analyses were conducted in STATA 14 MP4.
M AN U
SC
RI PT
191
201 Ethical approval
203
Approved by NRES Committee West Midlands – Solihull (Reference 12/WM/0322).
TE D
202
204 RESULTS
206
Study participants
AC C
207
EP
205
208
In the six participating practices, there were 45,033 total patients registered on the baseline
209
date. Baseline searches identified a total of 831 eligible patients who were at least 18 years of
210
age who had total cholesterol greater than 7.5 mmol/L, and without a previous diagnosis of
211
FH (1.84% of practice population). Six patient were excluded at baseline as they had a
212
previous diagnosis of FH. During the 6 month intervention period, 127 of the 831 eligible
213
patients (15.3%) consented to participate, two were excluded (one declined lipid test, one had
214
a previous diagnosis of FH). Of the 125 eligible patients, 118 patients’ medical records could 10
ACCEPTED MANUSCRIPT 215
be accessed (7 left the practice during study period). Thirty-two (27.1%) of the remaining 118
216
eligible patients met NICE Simon-Broome diagnostic criteria for possible FH, with two
217
patients confirmed with definite FH. Further study details on participants with possible FH
218
are provided elsewhere.14
220
RI PT
219
Characteristics of study participants in the pre-intervention period
221
Comparing males and females (Table 2), in the 118 patients with baseline data, there were
223
more females than males (61% vs. 39%), and the males were slightly older. The mean total
224
cholesterol was also slightly higher in females, as was mean LDL-cholesterol and proportion
225
with elevated cholesterol (total or LDL).
M AN U
SC
222
226
Considering potential opportunities for diagnosis of FH, after the release of NICE FH
228
guidelines (CG71) in November 2008, median date when a total cholesterol ≥ 7.5 mmol/L
229
was documented was 10 May 2009 for men and 8 February 2010 for women. During the pre-
230
intervention period, cholesterol testing was repeated in only 1 man (2%) and 11 women
231
(15%), whilst the proportion with family history documented (both complete and any family
232
history of CHD was small in both genders, with similar levels of documentation in men and
233
women (any family history assessed: 6.5% of men compared to 5.6% of women; complete
234
family history assessed: 2.2% of men compared to 2.7% of women). Considering workup of
235
patients, during six months pre-intervention period, clinical features of FH (arcus or
236
xanthelasma) and diagnosis of secondary cause were uncommon, whilst investigation of
237
secondary causes was better recorded. Further, during this pre-intervention period, statins
238
were only prescribed for 8 men (17%) and 10 women (14%); of which 2 men (4%) and 4
239
women (6%) were prescribed high potency statins. Considering lifestyle advice, dietary or
AC C
EP
TE D
227
11
ACCEPTED MANUSCRIPT 240
weight management advice was given to around 47% of both genders, with smoking
241
cessation advice offered to around 10% of patients.
242 243 Guideline compliance in assessing individuals for familial hypercholesterolaemia
245
Compliance to guideline recommendations for assessing FH at the end of the pre-intervention
246
and intervention periods (post-intervention) are shown in Table 3. First, considering
247
identification of FH, over a third of the 118 patients (35.6%, 95% CI 27.0 to 44.2) had a
248
family history of coronary heart disease documented. There was also a smaller but significant
249
increase (6.8%, 95% CI 3.5 to 12.8) in documenting a complete family history (age, coronary
250
heart disease, and degree of relation). There was a much greater increase in the proportion of
251
patients receiving repeat cholesterol tests (75.4%, 95% CI 66.9 to 82.3), whilst there was a
252
small but significant increase in examination for clinical features: with 7 more patients (6.0%,
253
95% CI 2.9 to 11.7) having arcus senilus or xanthalasma documented in records. There were
254
similar levels of increasing diagnosis of secondary causes (7.7%, 95% CI 4.1 to 13.9) and
255
investigating causes of raised cholesterol (from 6.8% for liver function tests to 12.7% for
256
serum creatinine).
EP
AC C
257
TE D
M AN U
SC
RI PT
244
258
Guideline compliance in managing possible familial hypercholesterolaemia
259
Compliance to guideline recommendations for managing patients diagnosed with possible FH
260
at the end of the pre-intervention and intervention periods (post-intervention) are shown in
261
Table 4.
262 263
Considering the possible impact of intervention on cholesterol management, compared to pre-
264
intervention period, total cholesterol and LDL cholesterol at the end of the intervention
12
ACCEPTED MANUSCRIPT period was not significantly reduced in these 32 patients with possible FH (total cholesterol:
266
mean difference: -0.16 mmol/L, 95% CI -0.78 to 0.46; LDL cholesterol: mean difference: -
267
0.12 mmol/L, 95% CI -0.81 to 0.57). However, there was a large and significant increase in
268
statin prescribing for these 32 patients (18.8%, 95% CI 8.9 to 35.3). This difference was
269
smaller but remained significant for prescribing a high potency statin (9.4%, 95% CI 3.2 to
270
24.2). Lifestyle risk reducing advice for coronary heart disease, as indicated by diet or weight
271
management and smoking cessation, showed little recorded improvement pre- and post-
272
intervention.
SC
RI PT
265
273 DISCUSSION
M AN U
274 275
This study suggests that a simple intervention in primary care can improve assessment and
277
identification of possible familial hypercholesterolaemia among patients with high
278
cholesterol. These patients accounted for almost 2% of the whole general practice population,
279
highlighting the scale and need for assessment. Despite short follow-up period, the study
280
demonstrated some improvement in management of patients with possible FH. Sustaining
281
this effort over a longer period of time could potentially yield even greater gains.
EP
AC C
282
TE D
276
283
Considering compliance with national recommendations on assessing patients for FH, repeat
284
cholesterol testing and family history recording improved considerably, with more modest
285
improvements in recommended completeness of family history recording and assessment of
286
clinical signs. The intervention had more limited impact on investigation and diagnosis of
287
secondary causes of FH.
288
13
ACCEPTED MANUSCRIPT Considering management of possible FH, statin prescribing improved greatly, with more
290
modest improvement in high potency statin prescribing. Together with the short (six month)
291
follow-up period, this may explain the smaller than expected improvements in cholesterol
292
levels. There was also room for improvement in lifestyle risk-reducing advice. However, the
293
baseline lifestyle advice given was quite high.
294
RI PT
289
Whilst there are no randomised controlled trials for improving identification of possible FH
296
using systematic approaches in primary care, there have been several observational studies.
297
18-22
298
interpretative comments by specialists 23 , and searches of family practice EHRs. 20-22 The
299
latter similar to the approach used in this study. Although previous approaches have shown
300
such strategies are feasible and promising, none has assessed intervention performance
301
against national evidence-based guideline recommendations on the identification and initial
302
management of FH.
SC
295
TE D
303
M AN U
These have used nurse review of patient records19, telephone feedback from specialists18,
Strengths
305
This study was successful in measuring outcomes for patients on all specified FH guideline
306
recommendations, and is one of the first experimental studies to demonstrate changes in best
307
practice within six months of intervention in primary care. This included significant increases
308
in repeating cholesterol testing, assessing for family history of heart disease, diagnosing
309
secondary causes, and examining for clinical features of FH, while also increasing statin
310
prescribing for those diagnosed with possible FH. Although there was no significant decrease
311
in mean total cholesterol or LDL-cholesterol within the brief follow up period, this simple
312
intervention nevertheless identified individuals with possible FH who were previously
313
unrecognised and who are benefiting from lipid lowering therapies.
AC C
EP
304
14
ACCEPTED MANUSCRIPT 314 315
The intervention was successfully implemented in general practices from populations with
316
higher levels of socio-economic deprivation, with variation of uptake from 8-15% from the
317
four inner city practices, 14% from the suburban practice, and 27% from the urban practice.
318
14
319
and poorer management of patients with increased cardiovascular risk25 are more common in
320
this context, improvements in compliance to guideline recommendations were still
321
demonstrated.
SC
RI PT
While lower rates of FH diagnosis, higher levels of premature coronary heart disease 24,
322
Finally, we were able to combine outcome data from both coded EHR searches and free text
324
in manual review of the clinical records to enhance the completeness of the outcome
325
measures. This process is consistent with previous studies on FH identification using both
326
EHRs and manual review. 20,22
TE D
327
M AN U
323
Limitations
329
Within the constraints of the short study period, a relatively modest number of eligible
330
patients were engaged and agreed to assessment, but this was sufficient for the specified
331
study sample size. The lower uptake of eligible patients with elevated cholesterol was also
332
due to purposeful recruitment in areas of low deprivation (4/6 practices). It is recognised that
333
patients from these deprived UK areas may face greater challenges in engaging with
334
preventive and primary health care in the context of multiple morbidities and competing
335
health and other life priorities. 26
AC C
EP
328
336
15
ACCEPTED MANUSCRIPT 337
The six-month follow-up period did not allow for full review and access to secondary care
338
outcomes for the 32 patients who were diagnosed with possible FH. This includes confirming
339
the diagnosis by genetic testing for the monogenic mutation.
340 The primary limitation of this exploratory study is its non-randomised and uncontrolled
342
design; therefore the observed changes in compliance to NICE guideline recommendations
343
must be interpreted with caution, as they may not be directly attributable to the intervention.
344
Other potential external influences could include the recently introduced UK Department of
345
Health vascular check (CVD risk assessment programmes) 27 or release of revised NICE lipid
346
modification guidelines that briefly highlight the importance of identifying FH 28 29, though
347
there is no evidence of their major impact on routine primary care practice given continuous
348
under-diagnosis of FH in primary care. 2
M AN U
SC
RI PT
341
349
Finally, in line with NICE guideline recommendations, only those patients who have ever had
351
a cholesterol above the threshold (> 7.5mmol) were assessed. However, from our search of
352
EHRs, we observed some patients below this threshold were prescribed statins prior to their
353
cholesterol result entry. These individuals could have had possible FH. Systematic
354
approaches to account for treatment effects as well as other relevant clinical criteria are
355
currently being developed. 22,30-33
357 358
EP
AC C
356
TE D
350
Conclusion
359 360
This study shows the promise of a simple intervention to improve identification and
361
management of possible familial hypercholesterolaemia in line with clinical guideline
16
ACCEPTED MANUSCRIPT recommendations. If this simple intervention was rolled out across primary care, this could
363
lead to a substantial improvement in recognition of possible FH leading to improved health
364
outcomes for these individuals through targeted specialist follow-up and management.
365
Further testing of this intervention is merited, using a cluster-randomised controlled design,
366
with a longer study period to improve patient engagement and assessment of longer-term
367
outcomes.
368
CONFLICT OF INTEREST
369
Nadeem Qureshi is a member of the National Institute for Health and Care Excellence
370
(NICE) familial hypercholesterolaemia guideline development group (CG71). Stephen Weng
371
is a member of the Clinical Practice Research Datalink Independent Scientific Advisory
372
Committee under the UK Medicines and Health Regulatory Agency. The remaining authors
373
have no conflicts of interest to disclose.
374
M AN U
SC
RI PT
362
FINANCIAL SUPPORT
376
This work was supported by the Division of Primary, University of Nottingham and funded
377
by National Institute for Health Research School of Primary Care Research (NIHR-SPCR
378
FR-2)
380
EP AC C
379
TE D
375
381
AUTHOR CONTRIBUTIONS
382
SW wrote the first draft of the manuscript, conducted the primary analysis. NQ and JK
383
conceived the study, developed the study design, and contributed to the analysis. JT recruited
384
patients, extracted the data, and conducted the manual review of patient records. JLB
385
contributed to the primary analysis and verified the statistical methods. All study authors
386
have contributed to writing and finalising the final submission version of the manuscript.
17
ACCEPTED MANUSCRIPT 387 ACKNOWLEDGEMENTS
389
This paper presents independent research funded by the National Institute for Health
390
Research School for Primary Care Research (NIHR-SPCR) [FR2-95]. The views expressed
391
are those of the author(s) and not necessarily those of the NIHR, the NHS or the Department
392
of Health. The authors would like to thank other members of the FAMCHOL study steering
393
group (Professor Carol Coupland, Professor Steve Humphries, Professor Denise Kendrick, Dr
394
Daniel Lasserson, Professor Elizabeth Murray, Professor Andrew Neil, Mr Phil Rowlands),
395
and Mr Tim Morell and Dr John Robinson (PRIMIS) for their technical support in extracting
396
electronic health data from general practices.
M AN U
SC
RI PT
388
AC C
EP
TE D
397
18
ACCEPTED MANUSCRIPT 398
REFERENCES
399 1.
Nordestgaard BG, Chapman MJ, Humphries SE, et al. Familial hypercholesterolaemia
401
is underdiagnosed and undertreated in the general population: guidance for clinicians to
402
prevent coronary heart disease: consensus statement of the European Atherosclerosis Society.
403
Eur Heart J 2013; 34(45): 3478-90a.
404
2.
405
management of familial hypercholesterolaemia: what does it mean to primary care? Br J Gen
406
Pract 2009; 59(567): 773-6.
407
3.
408
Lancet 1969; 294(7635): 1380-2.
409
4.
410
hypercholesterolaemia. British Medical Journal 1991; 303(6807): 893-6.
411
5.
412
hypercholesterolaemia: implications for clinical management. Atherosclerosis 1999; 142(1):
413
105-12.
414
6.
415
mortality in statin-treated patients with heterozygous familial hypercholesterolaemia: a
416
prospective registry study. European Heart Journal 2008; 29(21): 2625-33.
417
7.
418
diagnosis and management of pediatric and adult patients: clinical guidance from the
419
National Lipid Association Expert Panel on Familial Hypercholesterolemia. Journal of
420
clinical lipidology 2011; 5(3 Suppl): S1-8.
421
8.
422
familial hypercholesterolaemia (CG71). 2008. https://www.nice.org.uk/guidance/cg71.
RI PT
400
SC
Qureshi N, Humphries SE, Seed M, Rowlands P, Minhas R. Identification and
M AN U
Slack J. Risks of ischaemic heart-disease in familial hyperlipoproteinaemic states. The
Simon-Broome Register Group. Risk of fatal coronary heart disease in familial
TE D
Simon-Broome Register Group. Mortality in treated heterozygous familial
AC C
EP
Neil A, Cooper J, Betteridge J, et al. Reductions in all-cause, cancer, and coronary
Goldberg AC, Hopkins PN, Toth PP, et al. Familial hypercholesterolemia: screening,
National Institute for Health and Care Excellence. Identification and management of
19
ACCEPTED MANUSCRIPT 423
9.
Reiner Z, Catapano AL, De Backer G, et al. ESC/EAS Guidelines for the
424
management of dyslipidaemias: the Task Force for the management of dyslipidaemias of the
425
European Society of Cardiology (ESC) and the European Atherosclerosis Society (EAS).
426
European Heart Journal 2011; 32(14): 1769-818.
427
10.
428
Heterozygous Familial Hypercholesterolemia: A HuGE Prevalence Review. American
429
Journal of Epidemiology 2004; 160(5): 407-20.
430
11.
431
enquiry to cardiovascular disease risk assessment in primary care: a matched-pair, cluster
432
randomized trial. Ann Intern Med 2012; 156(4): 253-62.
433
12.
434
in risk assessment and primary prevention of coronary heart disease in primary care:
435
ADDFAM study protocol. BMC Health Services Research 2009; 9(1): 184.
436
13.
437
evaluating a new family history tool. Family Practice 2005; 22(6): 663-9.
438
14.
439
identification of familial hypercholesterolaemia in general practice: intervention development
440
study. BMJ Open 2016; 6(5).
441
15.
442
ASSIGN cardiovascular disease risk score on a UK cohort of patients from general practice.
443
Heart (British Cardiac Society) 2010.
444
16.
445
Statistical Science 2001; 16(2): 101-17.
446
17.
447
Matched-Pairs t-Test. Optometry & Vision Science 1993; 70(1): 79-83.
RI PT
Austin MA, Hutter CM, Zimmern RL, Humphries SE. Genetic Causes of Monogenic
M AN U
SC
Qureshi N, Armstrong S, Dhiman P, et al. Effect of adding systematic family history
Qureshi N, Armstrong S, Saukko P, et al. Realising the potential of the family history
TE D
Qureshi N, Bethea J, Modell B, et al. Collecting genetic information in primary care:
AC C
EP
Qureshi N, Weng S, Tranter J, El-Kadiki A, Kai J. Feasibility of improving
de la Iglesia B, Potter JF, Poulter NR, Robins MM, Skinner J. Performance of the
Brown LD, Cai TT, DasGupta A. Interval Estimation for a Binomial Proportion.
De Land P, Chase W. Statistics Notebook: Entry IV.B: (1) One-Sample t-Test and (2)
20
ACCEPTED MANUSCRIPT 448
18.
Bell DA, Hooper AJ, Edwards G, et al. Detecting familial hypercholesterolaemia in
449
the community: impact of a telephone call from a chemical pathologist to the requesting
450
general practitioner. Atherosclerosis 2014; 234(2): 469-72.
451
19.
452
hypercholesterolaemia in primary care? Heart, lung & circulation 2014; 23(12): 1153-7.
453
20.
454
familial hypercholesterolaemia in primary care: an informatics-based approach in one
455
primary care centre. Heart (British Cardiac Society) 2008; 94(6): 754-8.
456
21.
457
hypercholesterolaemia in primary health care: a community based prospective study of three
458
methods. Heart, lung & circulation 2015; 24(3): 250-6.
459
22.
460
screening tool for identifying risk of familial hypercholesterolaemia in general practice.
461
Heart (British Cardiac Society) 2016.
462
23.
463
suggesting specialist referral increase the detection of familial hypercholesterolaemia.
464
Pathology; 48(5): 463-6.
465
24.
466
premature coronary heart disease mortality in Great Britain, 1994–2008: a time trend
467
ecological study. BMJ Open 2012; 2(3).
468
25.
469
in blood pressure, cholesterol, body mass index and smoking among participants of a UK
470
primary care-based cardiovascular risk factor screening programme: both narrowing and
471
widening in cardiovascular risk factor inequalities. Heart (British Cardiac Society) 2006;
472
92(9): 1198-206.
RI PT
Bell DA, Kirke AB, Barbour R, et al. Can patients be accurately assessed for familial
SC
Gray J, Jaiyeola A, Whiting M, Modell M, Wierzbicki AS. Identifying patients with
M AN U
Kirke AB, Barbour RA, Burrows S, et al. Systematic detection of familial
TE D
Troeung L, Arnold-Reed D, Chan She Ping-Delfos W, et al. A new electronic
EP
Bender R, Edwards G, McMahon J, et al. Interpretative comments specifically
AC C
McCartney D, Scarborough P, Webster P, Rayner M. Trends in social inequalities for
Lyratzopoulos G, Heller RF, McElduff P, Hanily M, Lewis P. Deprivation and trends
21
ACCEPTED MANUSCRIPT 473
26.
Mercer SW, Zhou Y, Humphris GM, et al. Multimorbidity and Socioeconomic
474
Deprivation in Primary Care Consultations. The Annals of Family Medicine 2018; 16(2): 127-
475
31.
476
27.
477
management. London, UK. 2008.
478
http://www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAndGuida
479
nce/DH_090277.
480
28.
481
assessment and reduction, including lipid modification (CG181). 2014.
482
https://www.nice.org.uk/guidance/cg181.
483
29.
484
lipoprotein cholesterol, ischaemic heart disease, and stroke: systematic review and meta-
485
analysis. BMJ 2003; 326(7404): 1423.
486
30.
487
system to augment the opportunistic detection of individuals with familial
488
hypercholesterolaemia from a community laboratory. Clinical Chimica Acta 2015; 448: 18-
489
21.
490
31.
491
applications. Clinical Lipidology 2015; 10(4): 295-8.
492
32.
493
hypercholesterolaemia in primary care using electronic audit and nurse-led clinics. Journal of
494
evaluation in clinical practice 2015; 22(3): 341-8.
495
33.
496
of familial hypercholesterolaemia in primary care: Derivation and validation of the familial
RI PT
Department of Health. Putting prevention first—vascular checks: risk assessment and
M AN U
SC
National Institute for Health and Care Excellence. Cardiovascular disease: risk
Law MR, Wald NJ, Rudnicka AR. Quantifying effect of statins on low density
EP
TE D
Bell DA, Edwards G, Hooper AJ, et al. The potential role of an expert computer
AC C
Bell DA, Watts GF. Screening for familial hypercholesterolemia: primary care
Green P, Neely D, Humphries SE, et al. Improving detection of familial
Weng SF, Kai J, Andrew Neil H, Humphries SE, Qureshi N. Improving identification
22
ACCEPTED MANUSCRIPT 497
hypercholesterolaemia case ascertainment tool (FAMCAT). Atherosclerosis 2015; 238(2):
498
336-43.
AC C
EP
TE D
M AN U
SC
RI PT
499
23
ACCEPTED MANUSCRIPT 500
Figure 1. Study process: pre- and during intervention periods.
501 502 Table 1. Outcome measures for compliance to NICE Familial Hypercholesterolaemia
504
Guideline (CG71) recommendations
Assessment for diagnosing familial hypercholesterolaemia (NICE CG71 Section 1.1)
SC
Proportion of patients with any family history of coronary heart disease assessed
RI PT
503
Proportion of patients with complete family history (age, condition, degree of relation) of coronary heart disease assessed Proportion of patients with repeat cholesterol test
M AN U
Proportion of patients diagnosed with secondary cause (diabetes, hypothyroidism, chronic kidney disease, liver disease) Proportion of patients with TSH assessment (proxy for hypothyroidism investigations) Proportion of patients with HbA1c assessment (proxy for diabetes investigation)
Proportion of patients with serum creatinine assessment (proxy for chronic kidney disease investigations) Proportion of patients with bilirubin, ALP, or gamma GT assessment (proxy for liver disease)
TE D
Proportion of patients with arcus senilis or xanthalasma diagnosed
Management of diagnosed possible familial hypercholesterolaemia (NICE (CG71 Section 1.3) Total cholesterol levels (mmol/L)
EP
LDL cholesterol levels (mmol/L)
Proportion of patients prescribed any statin treatment
AC C
Proportion of patients prescribed high potency statin treatment (Simvastatin ≥ 80 mg/day; Atorvastatin ≥ 20 mg/day; Rosuvastatin ≥ 10 mg/day)
Proportion of patients given dietary or weight management advice Proportion of patients given smoking cessation advice
505 506
24
ACCEPTED MANUSCRIPT 507
Table 2. Descriptive characteristics for recruited patients in pre-intervention period, extracted
508
from electronic health records (EHRs) and manual review. Variables are means and standard
509
deviations unless otherwise specified.
510 Female
(n = 46)
(n = 72)
58 (9.0)
56 (7.5)
3 (6.5)
4 (5.6)
1 (2.2)
2 (2.7)
1 (2.2)
11 (15)
3 (6.5)
2 (2.7)
28 (61)
37 (51)
16 (35)
23 (32)
38 (83)
65 (90)
37 (80)
62 (86)
0 (0.0)
1 (2.2)
Mean total cholesterol (mmol/L)
6.1 (1.6)
6.6 (1.5)
Mean LDL cholesterol (mmol/L)
3.8 (1.7)
4.0 (1.5)
8 (17)
10 (14)
Prescribed high potency statins g (%)
2 (4.3)
4 (5.6)
Given dietary or weight management advice (%)
22 (48)
34 (47)
4 (8.7)
8 (11)
Family history of coronary heart disease assessed (%) Complete family history a of coronary heart disease assessed (%)
Diagnosed with secondary cause b (%) TSH assessed c (%) HbA1c assessed d (%) Serum creatinine assessed e (%) Bilirubin, ALP, or gamma GT assessed f (%)
M AN U
Repeat cholesterol test (%)
SC
Age (years)
RI PT
Male
Patient characteristics
AC C
EP
Prescribed statins (%)
TE D
Arcus senilis or xanthalasma diagnosed (%)
Given smoking cessation advice (%) a
Complete family history is defined when age, condition, and degree of relation to patient is documented.
b
Diabetes, hypothyroidism, chronic kidney disease, liver disease.
c
Proxy for secondary investigations for hypothyroidism.
d
Proxy for secondary investigations for diabetes.
e
Proxy for secondary investigations for chronic kidney disease.
f
Proxy for secondary investigations for liver disease.
g
Simvastatin ≥ 80 mg/day, atorvastatin ≥ 20 mg /day, rosuvastatin ≥ 10 mg/day.
25
ACCEPTED MANUSCRIPT 511
Table 3. Change in compliance (absolute difference) to outcomes for assessing familial
512
hypercholesterolaemia in 118 patients with elevated cholesterol Pre-
Post-
intervention
intervention
(Time 2)
(Time 4)
7 (5.9%)
49 (41.5%)
35.6%*
27.0 , 44.2
3 (2.5%)
11 (9.3%)
6.8%*
3.5 , 12.8
12 (10.2%)
101 (85.6%)
75.4%*
66.9 , 82.3
5 (4.2%)
14 (11.9%)
7.7%*
4.1 , 13.9
54 (45.8%)
69 (58.5%)
12.7%*
6.7 , 18.7
39 (33.1%)
51 (43.2%)
10.1%*
5.9 , 16.9
103 (87.3%)
113 (95.8%)
8.5%*
4.7 , 14.9
99 (83.9%)
107 (90.7%)
6.8%*
3.5 , 12.8
1 (0.8%)
8 (6.8%)
6.0%*
2.9 , 11.7
NICE compliance outcomes for diagnosis of possible
No. with any family history of coronary heart disease (%) No. with any complete family history a of coronary heart
No. with repeat cholesterol test (%)
No. with TSH assessed c (%) No. with HbA1c assessed d (%) No. with serum creatinine e (%) No. with bilirubin, ALP, or gamma GT f (%) No. with arcus senilis or xanthalasma diagnosed (%)
M AN U
No. diagnosed with secondary cause b (%)
difference
interval (%)
SC
disease assessed (%)
95% confidence
RI PT
familial hypercholesterolaemia (n = 118)
Absolute
Significant p < 0.05
a
Complete family history is defined when age, condition, and degree of relation to patient is documented
b
Diabetes, hypothyroidism, chronic kidney disease, liver disease
c
Proxy for secondary investigations for hypothyroidism
d
Proxy for secondary investigations for diabetes
e
Proxy for secondary investigations for chronic kidney disease
f
Proxy for secondary investigations for liver disease
513
AC C
EP
TE D
*
26
ACCEPTED MANUSCRIPT 514
Table 4. Change in compliance to outcomes (absolute difference) for management of 32
515
patients diagnosed with possible familial hypercholesterolaemia Pre-
Post-
intervention
intervention
(Time 2)
(Time 4)
17 (53.1%)
23 (71.9%)
3 (9.4%)
6 (18.8%)
10 (31.3%)
11 (34.4%)
11 (34.4%)
11 (34.4%)
Pre-
Post-
intervention
intervention
(Time 2)
(Time 4)
Mean Total cholesterol in mmol/L (SD)
7.23 (1.19)
Mean LDL cholesterol in mmol/L (SD)
4.58 (1.32)
familial hypercholesterolaemia (n = 32)
No. prescribed any statins (%) No. prescribed high potency statins a (%)
(%)
EP
516
interval (%)
27
18.8%*
8.9 , 35.3
9.4%*
3.2 , 24.2
3.1%
0.01 , 15.7
0%
0 , 10.7
Mean
95% confidence
difference
interval
7.07 (1.22)
-0.16
-0.78 , 0.46
4.46 (1.36)
-0.12
-0.81 , 0.57
Simvastatin ≥ 80 mg/day, atorvastatin ≥ 20 mg /day, rosuvastatin ≥ 10 mg/day
AC C
a
TE D
* Significant p < 0.05
difference
M AN U
No. given smoking cessation advice (%)
95% confidence
SC
No. given dietary or weight management advice
Absolute
RI PT
NICE compliance outcomes for management of
ACCEPTED MANUSCRIPT 517
Supplementary
518
Hypercholesterolaemia Guideline (CG71) recommendations
NICE Guideline
Data.
Outcome
measures
for
compliance
to
NICE
Familial
Measure
RI PT
NICE (CG71) Section 1.1 Assessment for Diagnosing Familial Hypercholesterolaemia Section 1.1.8 Proportion of patients with any family history of coronary heart disease assessed Section 1.1.8 Proportion of patients with complete family history (age, condition, degree of relation) Proportion of patients with repeat cholesterol test Proportion of patients diagnosed with secondary cause (diabetes, hypothyroidism,
Section 1.1.2 Section 1.1.2 Section 1.1.2
Proportion of patients with TSH assessment (proxy for hypothyroidism investigations) Proportion of patients with HbA1c assessment (proxy for diabetes investigation) Proportion of patients with serum creatinine assessment (proxy for chronic kidney
Section 1.1.2
Proportion of patients with bilirubin, ALP, or gamma GT assessment (proxy for liver
SC
Section 1.1.6 Section 1.1.2
Section 1.3.2 Section 1.3.2
M AN U
Section 1.1.3 Proportion of patients with arcus senilis or xanthalasma diagnosed NICE (CG71) Section 1.3 Management of Diagnosed Possible Familial Hypercholesterolaemia Section 1.3.1 Total cholesterol levels (mmol/L) Section 1.3.1 LDL cholesterol levels (mmol/L) Section 1.3.1.2 Proportion of patients prescribed any statin treatment Section 1.3.1.3 Proportion of patients prescribed high potency statin treatment (Simvastatin = 80 Proportion of patients given dietary or weight management advice Proportion of patients given smoking cessation advice
AC C
EP
TE D
519
28
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT Highlights Heterozygote familial hypercholesterolaemia (FH) is a common genetic disorder that leads to premature coronary heart disease.
•
We have evaluated a simple pro-active primary care-based intervention inviting patients with raised cholesterols to be assessed for possible FH.
•
We found large improvements in best practice for identification and management of familial hypercholesterolaemia indicating the potential of this approach.
AC C
EP
TE D
M AN U
SC
RI PT
•
S0021-9150(18)30222-3
DOI:
10.1016/j.atherosclerosis.2018.04.037
Reference:
ATH 15490
To appear in:
Atherosclerosis
Received Date: 16 March 2018 Revised Date:
18 April 2018
Accepted Date: 27 April 2018
Please cite this article as: Weng S, Kai J, Tranter J, Leonardi-Bee J, Qureshi N, Improving identification and management of familial hypercholesterolaemia in primary care: Pre- and post-intervention study, Atherosclerosis (2018), doi: 10.1016/j.atherosclerosis.2018.04.037. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT 1
Improving identification and management of familial
2
hypercholesterolaemia in primary care: Pre- and post-intervention study
3
5 6 7
1
8
School of Medicine, University of Nottingham, UK;
NIHR Research Fellow, Division of Primary Care, NIHR School for Primary Care Research,
10
2
11
University of Nottingham, UK;
M AN U
9
Division of Primary Care, NIHR School for Primary Care Research, School of Medicine,
12 13
3
14
University of Nottingham, UK.
TE D
Division of Primary Care, NIHR School for Primary Care Research, School of Medicine,
15 16
4
17
Nottingham, UK.
AC C
EP
Division of Epidemiology and Public Health, School of Medicine, University of
18 19
5
20
University of Nottingham, UK.
21
RI PT
Stephen Weng1, Joe Kai2, Jennifer Tranter3, Jo Leonardi-Bee4, Nadeem Qureshi5
SC
4
Division of Primary Care, NIHR School for Primary Care Research, School of Medicine,
22 23
Correspondence:
24
Stephen Weng
25
Division of Primary Care 1
26
1403 The Tower
27
University Park
28
University of Nottingham
29
NG7 2RD, UK
30
[email protected]
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
31
2
ACCEPTED MANUSCRIPT 32
ABSTRACT
33 Background and aims: Familial hypercholesterolaemia (FH) is a major cause of premature
35
heart disease but remains unrecognised in most patients. This study investigated if a
36
systematic primary care-based approach to identify and manage possible FH improves
37
recommended best clinical practice.
RI PT
34
38
Methods: Pre- and post-intervention study in six UK general practices (population 45,033)
40
which invited patients with total cholesterol >7.5 mmol/L to be assessed for possible FH.
41
Compliance with national guideline recommendations to identify and manage possible FH
42
(repeat cholesterol; assess family history of heart disease; identify secondary causes and
43
clinical features; reduce total & LDL-cholesterol; statin prescribing; lifestyle advice) was
44
assessed by calculating the absolute difference in measures of care pre- and six months post-
45
intervention.
TE D
M AN U
SC
39
46
Results: The intervention improved best clinical practice in 118 patients consenting to
48
assessment (of 831 eligible patients): repeat cholesterol test (+75.4%, 95% CI 66.9-82.3);
49
family history of heart disease assessed (+35.6%, 95% CI 27.0-44.2); diagnosis of secondary
50
causes (+7.7%, 95% CI 4.1-13.9), examining clinical features (+6.0%, 95% CI 2.9-11.7). For
51
32 patients diagnosed with possible FH using Simon-Broome criteria, statin prescribing
52
significantly improved (18.8%, 95% CI 8.9-35.3) with non-significant mean reductions in
53
cholesterol post-intervention (total: -0.16 mmol/L, 95% CI -0.78-0.46; LDL: -0.12 mmol/L,
54
95% CI -0.81-0.57).
AC C
EP
47
55
3
ACCEPTED MANUSCRIPT Conclusions: Within six months, this simple primary care intervention improved both
57
identification and management of patients with possible FH, in line with national evidence-
58
based guidelines. Replicating and sustaining this approach across the country could lead to
59
substantial improvement in health outcomes for these individuals with very high
60
cardiovascular risk.
RI PT
56
61 Keywords
63
primary health care; familial hypercholesterolaemia; family health; medical genetics
SC
62
64
AC C
EP
TE D
M AN U
65
4
ACCEPTED MANUSCRIPT 66
INTRODUCTION
67 Familial hypercholesterolaemia (FH) is the most common autosomal dominant disorder with
69
at least 1 in 500 individuals affected in the general population.1 Over 80% of an estimated
70
120,000 individuals in the United Kingdom remain undiagnosed, with similar rates of under-
71
diagnosis across Europe, which results in major lost opportunities to prevent premature heart
72
disease and death.2 Left untreated, coronary heart disease (CHD) will develop in
73
approximately 50% of men with FH by the age of 50 and 30% of women with FH by the age
74
of 60.3 This results in a 100-fold increase in mortality risk compared to the general
75
population.4,5 This can be very effectively reduced by high intensity lipid lowering treatment,
76
resulting in a 48% reduction in CHD mortality.6
M AN U
SC
RI PT
68
77
Despite established national clinical guidelines in several countries 7-9, people with raised
79
cholesterols are not recognised as having FH. In the UK, cholesterol levels are well captured
80
in primary care electronic health records, offering the opportunity to improve recognition and
81
quality of care of this condition. The English National Institute for Health and Care
82
Excellence (NICE) guidelines introduced in 2008, 8 recommend general practitioners (GPs)
83
use Simon-Broome diagnostic criteria to identify FH.10 These state individuals with
84
cholesterol levels over 7.5 mmol/L and a relevant family history of premature CHD be
85
diagnosed as “possible FH” and be referred for specialist assessment and confirmation of
86
diagnosis.
AC C
EP
TE D
78
87 88
Combining a search of general practice electronic medical records for patients with elevated
89
cholesterol and systematic collection of family history may improve identification of possible
90
FH. 11-13 This should be accompanied by better recording of clinical features of FH and
5
ACCEPTED MANUSCRIPT excluding secondary causes of raised cholesterol.8 Further, proactive case ascertainment
92
should also improve lipid management. This study aimed to assess whether such a proactive
93
approach to identify possible familial hypercholesterolaemia (FH) in primary care improves
94
best practice in accord with national guideline recommendations on identifying and managing
95
FH. 8
RI PT
91
96 97
MATERIALS AND METHODS
99
SC
98 Study design
M AN U
100 101
This was a pre- and post-intervention study over 17 months, which ended in Aug 2015.
102
Ethical approval granted by the NRES Committee West Midlands – Solihull (Reference
103
12/WM/0322).
TE D
104
The baseline (pre-intervention period) was defined as the 6-month period from when a
106
participants’ total cholesterol was found to be ≥ 7.5 mmol/L after the release of the NICE FH
107
guidelines in November 2008. The intervention was first introduced at various time points
108
from March 2013 (first practice recruited) to January 2014 (last practice recruited). During
109
this period, participants with total cholesterol levels ≥ 7.5 mmol/L were invited to complete a
110
family history questionnaire. If the family history was positive, the patients was then offered
111
further clinical assessment over a six month period (post-intervention period). This process is
112
shown in Fig. 1.
AC C
EP
105
113 114 115 6
ACCEPTED MANUSCRIPT 116
Participants
117 All general practices in Nottingham (East Midlands, UK) were invited to participate in this
119
study. We received initial expressions of interest from 18 general practices. Six of these
120
practices responded (33.3%) following provision of further study information, and all these
121
were recruited (four inner city, one suburban, one rural).
122
RI PT
118
Patients from each of the six practices were eligible if they were aged 18 years or over, with a
124
total cholesterol level greater than 7.5 mmol/L documented in their primary care electronic
125
health records (EHRs). Participants with a confirmed diagnosis of familial
126
hypercholesterolaemia were excluded. Full details of participant recruitment are reported
127
elsewhere.14 Written informed consent was obtained from each patient included in the study
128
as outlined in the study protocol which conforms to the ethical guidelines of the 1975
129
Declaration of Helsinki and has been prior approved by the Institution's ethics committee on
130
research on humans.
133
M AN U
TE D
Intervention and procedure
EP
132
AC C
131
SC
123
134
The intervention consisted of an educational session with general practitioners and practice
135
nurses which lasted an hour on case-identification and assessment of FH using Simon-
136
Broome diagnostic criteria recommended in the current NICE guidelines.8 The computer-
137
based reminder message that appears when they accessed eligible participants’ records in
138
patient consultations was also demonstrated.
139
7
ACCEPTED MANUSCRIPT A baseline data extraction from electronic health records (EHR) was then completed to
141
identify eligible participants and their electronic health records were tagged with a reminder
142
message whenever the record was accessed in a consultation. In line with NICE FH
143
guidelines 8 , this prompted their GP to opportunistically assess for secondary causes of
144
raised cholesterol, ask participants to complete Family History and Symptom Questionnaires
145
(FHSQ) and repeat lipid profile, Four months into the intervention, eligible participants not
146
yet invited opportunistically by practitioners were contacted through postal invitation by their
147
practice.
SC
RI PT
140
148
Postal invitations included requests to conduct a follow-up fasting lipid test and complete the
150
FHSQ. After return of the FHSQ, a research assistant collated fasting lipid results, current
151
family history information, statin therapy, and notified GPs and participants if there was any
152
possibility of a FH diagnosis based on NICE Simon-Broome criteria.8 Post-intervention, we
153
then assessed GP compliance to national (NICE) guideline recommendations on best clinical
154
practice for FH identification and management.
TE D
M AN U
149
157 158
Outcome measures
AC C
156
EP
155
159
Outcome measures for compliance to NICE guideline recommendations were assessed for all
160
recruited participants at the end of the pre-intervention period (Fig.1, Time 2) and at the end
161
of the intervention period (post-intervention) (Fig.1, Time 4), including: (i) Assessment for
162
FH in patients with elevated cholesterol (ii) Management of patients who have been
163
diagnosed with possible FH. (see Table 1; further detail is available in Supplementary
164
Data). These outcome measures were also extracted from general practice EHRs at the end of
8
ACCEPTED MANUSCRIPT 165
each six month period. Extraction from EHRs was performed using automated data extraction
166
via medical codes and manual clinical record review for free-text entries and hospital
167
correspondence.
168 For the assessment of FH in patients with elevated cholesterols, the primary outcome measure
170
was defined as the difference in the proportion of patients with any family history of coronary
171
heart disease (CHD) documented in their EHRs between pre- and post-intervention periods.
172
Other outcome measures related to diagnostic assessment include the difference; between
173
pre- and post-intervention periods in proportions of: patients with repeated cholesterol tests,
174
diagnoses of secondary causes, biomarkers for secondary causes measured (TSH, HbA1c,
175
serum creatinine, liver function tests), diagnoses of arcus senilis or xanthalasma. GPs in the
176
study were advised to specifically assess for arcus cornealis and tendon xanthoma but code in
177
the electronic health records as arcus senilis and generic xanthalasma due to limited
178
availability of relevant NHS read codes.
TE D
M AN U
SC
RI PT
169
179
For the management of those diagnosed with possible FH, the outcomes were change in total
181
cholesterol, LDL cholesterol, as well as, difference in proportions of patients taking statin
182
treatment (any and high potency), given advice for managing diet and weight, and given
183
advice to stop smoking.
AC C
184
EP
180
185
Sample size calculation
186
For the primary outcome measure, a sample size of 118 patients from the six practices would
187
allow an absolute increase of 10% in the proportion of patients with family history of CHD
188
recorded in EHRs to be determined with 95% confidence intervals to within ±7% (i.e. 95%
189
CI 3% to 17%), assuming a baseline of 3.7% from a previous study. 15 9
ACCEPTED MANUSCRIPT 190 Statistical analysis
192
A descriptive analysis was conducted presenting numbers and percentages for all categorical
193
outcome variables, with continuous variables of total cholesterol and LDL-cholesterol
194
presented as mean and standard deviation, stratified by gender. To evaluate effects of the
195
intervention on compliance to NICE guideline recommendations, we calculated the change in
196
absolute difference for proportions with 95% confidence intervals (Wald’s method based on
197
normal approximation) 16 for categorical outcomes and mean difference with 95% confidence
198
intervals 17 for continuous cholesterol outcomes, between pre- and post-intervention periods.
199
Continuous cholesterol variables were investigated for normality and if needed, were log-
200
transformed for better fit. All analyses were conducted in STATA 14 MP4.
M AN U
SC
RI PT
191
201 Ethical approval
203
Approved by NRES Committee West Midlands – Solihull (Reference 12/WM/0322).
TE D
202
204 RESULTS
206
Study participants
AC C
207
EP
205
208
In the six participating practices, there were 45,033 total patients registered on the baseline
209
date. Baseline searches identified a total of 831 eligible patients who were at least 18 years of
210
age who had total cholesterol greater than 7.5 mmol/L, and without a previous diagnosis of
211
FH (1.84% of practice population). Six patient were excluded at baseline as they had a
212
previous diagnosis of FH. During the 6 month intervention period, 127 of the 831 eligible
213
patients (15.3%) consented to participate, two were excluded (one declined lipid test, one had
214
a previous diagnosis of FH). Of the 125 eligible patients, 118 patients’ medical records could 10
ACCEPTED MANUSCRIPT 215
be accessed (7 left the practice during study period). Thirty-two (27.1%) of the remaining 118
216
eligible patients met NICE Simon-Broome diagnostic criteria for possible FH, with two
217
patients confirmed with definite FH. Further study details on participants with possible FH
218
are provided elsewhere.14
220
RI PT
219
Characteristics of study participants in the pre-intervention period
221
Comparing males and females (Table 2), in the 118 patients with baseline data, there were
223
more females than males (61% vs. 39%), and the males were slightly older. The mean total
224
cholesterol was also slightly higher in females, as was mean LDL-cholesterol and proportion
225
with elevated cholesterol (total or LDL).
M AN U
SC
222
226
Considering potential opportunities for diagnosis of FH, after the release of NICE FH
228
guidelines (CG71) in November 2008, median date when a total cholesterol ≥ 7.5 mmol/L
229
was documented was 10 May 2009 for men and 8 February 2010 for women. During the pre-
230
intervention period, cholesterol testing was repeated in only 1 man (2%) and 11 women
231
(15%), whilst the proportion with family history documented (both complete and any family
232
history of CHD was small in both genders, with similar levels of documentation in men and
233
women (any family history assessed: 6.5% of men compared to 5.6% of women; complete
234
family history assessed: 2.2% of men compared to 2.7% of women). Considering workup of
235
patients, during six months pre-intervention period, clinical features of FH (arcus or
236
xanthelasma) and diagnosis of secondary cause were uncommon, whilst investigation of
237
secondary causes was better recorded. Further, during this pre-intervention period, statins
238
were only prescribed for 8 men (17%) and 10 women (14%); of which 2 men (4%) and 4
239
women (6%) were prescribed high potency statins. Considering lifestyle advice, dietary or
AC C
EP
TE D
227
11
ACCEPTED MANUSCRIPT 240
weight management advice was given to around 47% of both genders, with smoking
241
cessation advice offered to around 10% of patients.
242 243 Guideline compliance in assessing individuals for familial hypercholesterolaemia
245
Compliance to guideline recommendations for assessing FH at the end of the pre-intervention
246
and intervention periods (post-intervention) are shown in Table 3. First, considering
247
identification of FH, over a third of the 118 patients (35.6%, 95% CI 27.0 to 44.2) had a
248
family history of coronary heart disease documented. There was also a smaller but significant
249
increase (6.8%, 95% CI 3.5 to 12.8) in documenting a complete family history (age, coronary
250
heart disease, and degree of relation). There was a much greater increase in the proportion of
251
patients receiving repeat cholesterol tests (75.4%, 95% CI 66.9 to 82.3), whilst there was a
252
small but significant increase in examination for clinical features: with 7 more patients (6.0%,
253
95% CI 2.9 to 11.7) having arcus senilus or xanthalasma documented in records. There were
254
similar levels of increasing diagnosis of secondary causes (7.7%, 95% CI 4.1 to 13.9) and
255
investigating causes of raised cholesterol (from 6.8% for liver function tests to 12.7% for
256
serum creatinine).
EP
AC C
257
TE D
M AN U
SC
RI PT
244
258
Guideline compliance in managing possible familial hypercholesterolaemia
259
Compliance to guideline recommendations for managing patients diagnosed with possible FH
260
at the end of the pre-intervention and intervention periods (post-intervention) are shown in
261
Table 4.
262 263
Considering the possible impact of intervention on cholesterol management, compared to pre-
264
intervention period, total cholesterol and LDL cholesterol at the end of the intervention
12
ACCEPTED MANUSCRIPT period was not significantly reduced in these 32 patients with possible FH (total cholesterol:
266
mean difference: -0.16 mmol/L, 95% CI -0.78 to 0.46; LDL cholesterol: mean difference: -
267
0.12 mmol/L, 95% CI -0.81 to 0.57). However, there was a large and significant increase in
268
statin prescribing for these 32 patients (18.8%, 95% CI 8.9 to 35.3). This difference was
269
smaller but remained significant for prescribing a high potency statin (9.4%, 95% CI 3.2 to
270
24.2). Lifestyle risk reducing advice for coronary heart disease, as indicated by diet or weight
271
management and smoking cessation, showed little recorded improvement pre- and post-
272
intervention.
SC
RI PT
265
273 DISCUSSION
M AN U
274 275
This study suggests that a simple intervention in primary care can improve assessment and
277
identification of possible familial hypercholesterolaemia among patients with high
278
cholesterol. These patients accounted for almost 2% of the whole general practice population,
279
highlighting the scale and need for assessment. Despite short follow-up period, the study
280
demonstrated some improvement in management of patients with possible FH. Sustaining
281
this effort over a longer period of time could potentially yield even greater gains.
EP
AC C
282
TE D
276
283
Considering compliance with national recommendations on assessing patients for FH, repeat
284
cholesterol testing and family history recording improved considerably, with more modest
285
improvements in recommended completeness of family history recording and assessment of
286
clinical signs. The intervention had more limited impact on investigation and diagnosis of
287
secondary causes of FH.
288
13
ACCEPTED MANUSCRIPT Considering management of possible FH, statin prescribing improved greatly, with more
290
modest improvement in high potency statin prescribing. Together with the short (six month)
291
follow-up period, this may explain the smaller than expected improvements in cholesterol
292
levels. There was also room for improvement in lifestyle risk-reducing advice. However, the
293
baseline lifestyle advice given was quite high.
294
RI PT
289
Whilst there are no randomised controlled trials for improving identification of possible FH
296
using systematic approaches in primary care, there have been several observational studies.
297
18-22
298
interpretative comments by specialists 23 , and searches of family practice EHRs. 20-22 The
299
latter similar to the approach used in this study. Although previous approaches have shown
300
such strategies are feasible and promising, none has assessed intervention performance
301
against national evidence-based guideline recommendations on the identification and initial
302
management of FH.
SC
295
TE D
303
M AN U
These have used nurse review of patient records19, telephone feedback from specialists18,
Strengths
305
This study was successful in measuring outcomes for patients on all specified FH guideline
306
recommendations, and is one of the first experimental studies to demonstrate changes in best
307
practice within six months of intervention in primary care. This included significant increases
308
in repeating cholesterol testing, assessing for family history of heart disease, diagnosing
309
secondary causes, and examining for clinical features of FH, while also increasing statin
310
prescribing for those diagnosed with possible FH. Although there was no significant decrease
311
in mean total cholesterol or LDL-cholesterol within the brief follow up period, this simple
312
intervention nevertheless identified individuals with possible FH who were previously
313
unrecognised and who are benefiting from lipid lowering therapies.
AC C
EP
304
14
ACCEPTED MANUSCRIPT 314 315
The intervention was successfully implemented in general practices from populations with
316
higher levels of socio-economic deprivation, with variation of uptake from 8-15% from the
317
four inner city practices, 14% from the suburban practice, and 27% from the urban practice.
318
14
319
and poorer management of patients with increased cardiovascular risk25 are more common in
320
this context, improvements in compliance to guideline recommendations were still
321
demonstrated.
SC
RI PT
While lower rates of FH diagnosis, higher levels of premature coronary heart disease 24,
322
Finally, we were able to combine outcome data from both coded EHR searches and free text
324
in manual review of the clinical records to enhance the completeness of the outcome
325
measures. This process is consistent with previous studies on FH identification using both
326
EHRs and manual review. 20,22
TE D
327
M AN U
323
Limitations
329
Within the constraints of the short study period, a relatively modest number of eligible
330
patients were engaged and agreed to assessment, but this was sufficient for the specified
331
study sample size. The lower uptake of eligible patients with elevated cholesterol was also
332
due to purposeful recruitment in areas of low deprivation (4/6 practices). It is recognised that
333
patients from these deprived UK areas may face greater challenges in engaging with
334
preventive and primary health care in the context of multiple morbidities and competing
335
health and other life priorities. 26
AC C
EP
328
336
15
ACCEPTED MANUSCRIPT 337
The six-month follow-up period did not allow for full review and access to secondary care
338
outcomes for the 32 patients who were diagnosed with possible FH. This includes confirming
339
the diagnosis by genetic testing for the monogenic mutation.
340 The primary limitation of this exploratory study is its non-randomised and uncontrolled
342
design; therefore the observed changes in compliance to NICE guideline recommendations
343
must be interpreted with caution, as they may not be directly attributable to the intervention.
344
Other potential external influences could include the recently introduced UK Department of
345
Health vascular check (CVD risk assessment programmes) 27 or release of revised NICE lipid
346
modification guidelines that briefly highlight the importance of identifying FH 28 29, though
347
there is no evidence of their major impact on routine primary care practice given continuous
348
under-diagnosis of FH in primary care. 2
M AN U
SC
RI PT
341
349
Finally, in line with NICE guideline recommendations, only those patients who have ever had
351
a cholesterol above the threshold (> 7.5mmol) were assessed. However, from our search of
352
EHRs, we observed some patients below this threshold were prescribed statins prior to their
353
cholesterol result entry. These individuals could have had possible FH. Systematic
354
approaches to account for treatment effects as well as other relevant clinical criteria are
355
currently being developed. 22,30-33
357 358
EP
AC C
356
TE D
350
Conclusion
359 360
This study shows the promise of a simple intervention to improve identification and
361
management of possible familial hypercholesterolaemia in line with clinical guideline
16
ACCEPTED MANUSCRIPT recommendations. If this simple intervention was rolled out across primary care, this could
363
lead to a substantial improvement in recognition of possible FH leading to improved health
364
outcomes for these individuals through targeted specialist follow-up and management.
365
Further testing of this intervention is merited, using a cluster-randomised controlled design,
366
with a longer study period to improve patient engagement and assessment of longer-term
367
outcomes.
368
CONFLICT OF INTEREST
369
Nadeem Qureshi is a member of the National Institute for Health and Care Excellence
370
(NICE) familial hypercholesterolaemia guideline development group (CG71). Stephen Weng
371
is a member of the Clinical Practice Research Datalink Independent Scientific Advisory
372
Committee under the UK Medicines and Health Regulatory Agency. The remaining authors
373
have no conflicts of interest to disclose.
374
M AN U
SC
RI PT
362
FINANCIAL SUPPORT
376
This work was supported by the Division of Primary, University of Nottingham and funded
377
by National Institute for Health Research School of Primary Care Research (NIHR-SPCR
378
FR-2)
380
EP AC C
379
TE D
375
381
AUTHOR CONTRIBUTIONS
382
SW wrote the first draft of the manuscript, conducted the primary analysis. NQ and JK
383
conceived the study, developed the study design, and contributed to the analysis. JT recruited
384
patients, extracted the data, and conducted the manual review of patient records. JLB
385
contributed to the primary analysis and verified the statistical methods. All study authors
386
have contributed to writing and finalising the final submission version of the manuscript.
17
ACCEPTED MANUSCRIPT 387 ACKNOWLEDGEMENTS
389
This paper presents independent research funded by the National Institute for Health
390
Research School for Primary Care Research (NIHR-SPCR) [FR2-95]. The views expressed
391
are those of the author(s) and not necessarily those of the NIHR, the NHS or the Department
392
of Health. The authors would like to thank other members of the FAMCHOL study steering
393
group (Professor Carol Coupland, Professor Steve Humphries, Professor Denise Kendrick, Dr
394
Daniel Lasserson, Professor Elizabeth Murray, Professor Andrew Neil, Mr Phil Rowlands),
395
and Mr Tim Morell and Dr John Robinson (PRIMIS) for their technical support in extracting
396
electronic health data from general practices.
M AN U
SC
RI PT
388
AC C
EP
TE D
397
18
ACCEPTED MANUSCRIPT 398
REFERENCES
399 1.
Nordestgaard BG, Chapman MJ, Humphries SE, et al. Familial hypercholesterolaemia
401
is underdiagnosed and undertreated in the general population: guidance for clinicians to
402
prevent coronary heart disease: consensus statement of the European Atherosclerosis Society.
403
Eur Heart J 2013; 34(45): 3478-90a.
404
2.
405
management of familial hypercholesterolaemia: what does it mean to primary care? Br J Gen
406
Pract 2009; 59(567): 773-6.
407
3.
408
Lancet 1969; 294(7635): 1380-2.
409
4.
410
hypercholesterolaemia. British Medical Journal 1991; 303(6807): 893-6.
411
5.
412
hypercholesterolaemia: implications for clinical management. Atherosclerosis 1999; 142(1):
413
105-12.
414
6.
415
mortality in statin-treated patients with heterozygous familial hypercholesterolaemia: a
416
prospective registry study. European Heart Journal 2008; 29(21): 2625-33.
417
7.
418
diagnosis and management of pediatric and adult patients: clinical guidance from the
419
National Lipid Association Expert Panel on Familial Hypercholesterolemia. Journal of
420
clinical lipidology 2011; 5(3 Suppl): S1-8.
421
8.
422
familial hypercholesterolaemia (CG71). 2008. https://www.nice.org.uk/guidance/cg71.
RI PT
400
SC
Qureshi N, Humphries SE, Seed M, Rowlands P, Minhas R. Identification and
M AN U
Slack J. Risks of ischaemic heart-disease in familial hyperlipoproteinaemic states. The
Simon-Broome Register Group. Risk of fatal coronary heart disease in familial
TE D
Simon-Broome Register Group. Mortality in treated heterozygous familial
AC C
EP
Neil A, Cooper J, Betteridge J, et al. Reductions in all-cause, cancer, and coronary
Goldberg AC, Hopkins PN, Toth PP, et al. Familial hypercholesterolemia: screening,
National Institute for Health and Care Excellence. Identification and management of
19
ACCEPTED MANUSCRIPT 423
9.
Reiner Z, Catapano AL, De Backer G, et al. ESC/EAS Guidelines for the
424
management of dyslipidaemias: the Task Force for the management of dyslipidaemias of the
425
European Society of Cardiology (ESC) and the European Atherosclerosis Society (EAS).
426
European Heart Journal 2011; 32(14): 1769-818.
427
10.
428
Heterozygous Familial Hypercholesterolemia: A HuGE Prevalence Review. American
429
Journal of Epidemiology 2004; 160(5): 407-20.
430
11.
431
enquiry to cardiovascular disease risk assessment in primary care: a matched-pair, cluster
432
randomized trial. Ann Intern Med 2012; 156(4): 253-62.
433
12.
434
in risk assessment and primary prevention of coronary heart disease in primary care:
435
ADDFAM study protocol. BMC Health Services Research 2009; 9(1): 184.
436
13.
437
evaluating a new family history tool. Family Practice 2005; 22(6): 663-9.
438
14.
439
identification of familial hypercholesterolaemia in general practice: intervention development
440
study. BMJ Open 2016; 6(5).
441
15.
442
ASSIGN cardiovascular disease risk score on a UK cohort of patients from general practice.
443
Heart (British Cardiac Society) 2010.
444
16.
445
Statistical Science 2001; 16(2): 101-17.
446
17.
447
Matched-Pairs t-Test. Optometry & Vision Science 1993; 70(1): 79-83.
RI PT
Austin MA, Hutter CM, Zimmern RL, Humphries SE. Genetic Causes of Monogenic
M AN U
SC
Qureshi N, Armstrong S, Dhiman P, et al. Effect of adding systematic family history
Qureshi N, Armstrong S, Saukko P, et al. Realising the potential of the family history
TE D
Qureshi N, Bethea J, Modell B, et al. Collecting genetic information in primary care:
AC C
EP
Qureshi N, Weng S, Tranter J, El-Kadiki A, Kai J. Feasibility of improving
de la Iglesia B, Potter JF, Poulter NR, Robins MM, Skinner J. Performance of the
Brown LD, Cai TT, DasGupta A. Interval Estimation for a Binomial Proportion.
De Land P, Chase W. Statistics Notebook: Entry IV.B: (1) One-Sample t-Test and (2)
20
ACCEPTED MANUSCRIPT 448
18.
Bell DA, Hooper AJ, Edwards G, et al. Detecting familial hypercholesterolaemia in
449
the community: impact of a telephone call from a chemical pathologist to the requesting
450
general practitioner. Atherosclerosis 2014; 234(2): 469-72.
451
19.
452
hypercholesterolaemia in primary care? Heart, lung & circulation 2014; 23(12): 1153-7.
453
20.
454
familial hypercholesterolaemia in primary care: an informatics-based approach in one
455
primary care centre. Heart (British Cardiac Society) 2008; 94(6): 754-8.
456
21.
457
hypercholesterolaemia in primary health care: a community based prospective study of three
458
methods. Heart, lung & circulation 2015; 24(3): 250-6.
459
22.
460
screening tool for identifying risk of familial hypercholesterolaemia in general practice.
461
Heart (British Cardiac Society) 2016.
462
23.
463
suggesting specialist referral increase the detection of familial hypercholesterolaemia.
464
Pathology; 48(5): 463-6.
465
24.
466
premature coronary heart disease mortality in Great Britain, 1994–2008: a time trend
467
ecological study. BMJ Open 2012; 2(3).
468
25.
469
in blood pressure, cholesterol, body mass index and smoking among participants of a UK
470
primary care-based cardiovascular risk factor screening programme: both narrowing and
471
widening in cardiovascular risk factor inequalities. Heart (British Cardiac Society) 2006;
472
92(9): 1198-206.
RI PT
Bell DA, Kirke AB, Barbour R, et al. Can patients be accurately assessed for familial
SC
Gray J, Jaiyeola A, Whiting M, Modell M, Wierzbicki AS. Identifying patients with
M AN U
Kirke AB, Barbour RA, Burrows S, et al. Systematic detection of familial
TE D
Troeung L, Arnold-Reed D, Chan She Ping-Delfos W, et al. A new electronic
EP
Bender R, Edwards G, McMahon J, et al. Interpretative comments specifically
AC C
McCartney D, Scarborough P, Webster P, Rayner M. Trends in social inequalities for
Lyratzopoulos G, Heller RF, McElduff P, Hanily M, Lewis P. Deprivation and trends
21
ACCEPTED MANUSCRIPT 473
26.
Mercer SW, Zhou Y, Humphris GM, et al. Multimorbidity and Socioeconomic
474
Deprivation in Primary Care Consultations. The Annals of Family Medicine 2018; 16(2): 127-
475
31.
476
27.
477
management. London, UK. 2008.
478
http://www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAndGuida
479
nce/DH_090277.
480
28.
481
assessment and reduction, including lipid modification (CG181). 2014.
482
https://www.nice.org.uk/guidance/cg181.
483
29.
484
lipoprotein cholesterol, ischaemic heart disease, and stroke: systematic review and meta-
485
analysis. BMJ 2003; 326(7404): 1423.
486
30.
487
system to augment the opportunistic detection of individuals with familial
488
hypercholesterolaemia from a community laboratory. Clinical Chimica Acta 2015; 448: 18-
489
21.
490
31.
491
applications. Clinical Lipidology 2015; 10(4): 295-8.
492
32.
493
hypercholesterolaemia in primary care using electronic audit and nurse-led clinics. Journal of
494
evaluation in clinical practice 2015; 22(3): 341-8.
495
33.
496
of familial hypercholesterolaemia in primary care: Derivation and validation of the familial
RI PT
Department of Health. Putting prevention first—vascular checks: risk assessment and
M AN U
SC
National Institute for Health and Care Excellence. Cardiovascular disease: risk
Law MR, Wald NJ, Rudnicka AR. Quantifying effect of statins on low density
EP
TE D
Bell DA, Edwards G, Hooper AJ, et al. The potential role of an expert computer
AC C
Bell DA, Watts GF. Screening for familial hypercholesterolemia: primary care
Green P, Neely D, Humphries SE, et al. Improving detection of familial
Weng SF, Kai J, Andrew Neil H, Humphries SE, Qureshi N. Improving identification
22
ACCEPTED MANUSCRIPT 497
hypercholesterolaemia case ascertainment tool (FAMCAT). Atherosclerosis 2015; 238(2):
498
336-43.
AC C
EP
TE D
M AN U
SC
RI PT
499
23
ACCEPTED MANUSCRIPT 500
Figure 1. Study process: pre- and during intervention periods.
501 502 Table 1. Outcome measures for compliance to NICE Familial Hypercholesterolaemia
504
Guideline (CG71) recommendations
Assessment for diagnosing familial hypercholesterolaemia (NICE CG71 Section 1.1)
SC
Proportion of patients with any family history of coronary heart disease assessed
RI PT
503
Proportion of patients with complete family history (age, condition, degree of relation) of coronary heart disease assessed Proportion of patients with repeat cholesterol test
M AN U
Proportion of patients diagnosed with secondary cause (diabetes, hypothyroidism, chronic kidney disease, liver disease) Proportion of patients with TSH assessment (proxy for hypothyroidism investigations) Proportion of patients with HbA1c assessment (proxy for diabetes investigation)
Proportion of patients with serum creatinine assessment (proxy for chronic kidney disease investigations) Proportion of patients with bilirubin, ALP, or gamma GT assessment (proxy for liver disease)
TE D
Proportion of patients with arcus senilis or xanthalasma diagnosed
Management of diagnosed possible familial hypercholesterolaemia (NICE (CG71 Section 1.3) Total cholesterol levels (mmol/L)
EP
LDL cholesterol levels (mmol/L)
Proportion of patients prescribed any statin treatment
AC C
Proportion of patients prescribed high potency statin treatment (Simvastatin ≥ 80 mg/day; Atorvastatin ≥ 20 mg/day; Rosuvastatin ≥ 10 mg/day)
Proportion of patients given dietary or weight management advice Proportion of patients given smoking cessation advice
505 506
24
ACCEPTED MANUSCRIPT 507
Table 2. Descriptive characteristics for recruited patients in pre-intervention period, extracted
508
from electronic health records (EHRs) and manual review. Variables are means and standard
509
deviations unless otherwise specified.
510 Female
(n = 46)
(n = 72)
58 (9.0)
56 (7.5)
3 (6.5)
4 (5.6)
1 (2.2)
2 (2.7)
1 (2.2)
11 (15)
3 (6.5)
2 (2.7)
28 (61)
37 (51)
16 (35)
23 (32)
38 (83)
65 (90)
37 (80)
62 (86)
0 (0.0)
1 (2.2)
Mean total cholesterol (mmol/L)
6.1 (1.6)
6.6 (1.5)
Mean LDL cholesterol (mmol/L)
3.8 (1.7)
4.0 (1.5)
8 (17)
10 (14)
Prescribed high potency statins g (%)
2 (4.3)
4 (5.6)
Given dietary or weight management advice (%)
22 (48)
34 (47)
4 (8.7)
8 (11)
Family history of coronary heart disease assessed (%) Complete family history a of coronary heart disease assessed (%)
Diagnosed with secondary cause b (%) TSH assessed c (%) HbA1c assessed d (%) Serum creatinine assessed e (%) Bilirubin, ALP, or gamma GT assessed f (%)
M AN U
Repeat cholesterol test (%)
SC
Age (years)
RI PT
Male
Patient characteristics
AC C
EP
Prescribed statins (%)
TE D
Arcus senilis or xanthalasma diagnosed (%)
Given smoking cessation advice (%) a
Complete family history is defined when age, condition, and degree of relation to patient is documented.
b
Diabetes, hypothyroidism, chronic kidney disease, liver disease.
c
Proxy for secondary investigations for hypothyroidism.
d
Proxy for secondary investigations for diabetes.
e
Proxy for secondary investigations for chronic kidney disease.
f
Proxy for secondary investigations for liver disease.
g
Simvastatin ≥ 80 mg/day, atorvastatin ≥ 20 mg /day, rosuvastatin ≥ 10 mg/day.
25
ACCEPTED MANUSCRIPT 511
Table 3. Change in compliance (absolute difference) to outcomes for assessing familial
512
hypercholesterolaemia in 118 patients with elevated cholesterol Pre-
Post-
intervention
intervention
(Time 2)
(Time 4)
7 (5.9%)
49 (41.5%)
35.6%*
27.0 , 44.2
3 (2.5%)
11 (9.3%)
6.8%*
3.5 , 12.8
12 (10.2%)
101 (85.6%)
75.4%*
66.9 , 82.3
5 (4.2%)
14 (11.9%)
7.7%*
4.1 , 13.9
54 (45.8%)
69 (58.5%)
12.7%*
6.7 , 18.7
39 (33.1%)
51 (43.2%)
10.1%*
5.9 , 16.9
103 (87.3%)
113 (95.8%)
8.5%*
4.7 , 14.9
99 (83.9%)
107 (90.7%)
6.8%*
3.5 , 12.8
1 (0.8%)
8 (6.8%)
6.0%*
2.9 , 11.7
NICE compliance outcomes for diagnosis of possible
No. with any family history of coronary heart disease (%) No. with any complete family history a of coronary heart
No. with repeat cholesterol test (%)
No. with TSH assessed c (%) No. with HbA1c assessed d (%) No. with serum creatinine e (%) No. with bilirubin, ALP, or gamma GT f (%) No. with arcus senilis or xanthalasma diagnosed (%)
M AN U
No. diagnosed with secondary cause b (%)
difference
interval (%)
SC
disease assessed (%)
95% confidence
RI PT
familial hypercholesterolaemia (n = 118)
Absolute
Significant p < 0.05
a
Complete family history is defined when age, condition, and degree of relation to patient is documented
b
Diabetes, hypothyroidism, chronic kidney disease, liver disease
c
Proxy for secondary investigations for hypothyroidism
d
Proxy for secondary investigations for diabetes
e
Proxy for secondary investigations for chronic kidney disease
f
Proxy for secondary investigations for liver disease
513
AC C
EP
TE D
*
26
ACCEPTED MANUSCRIPT 514
Table 4. Change in compliance to outcomes (absolute difference) for management of 32
515
patients diagnosed with possible familial hypercholesterolaemia Pre-
Post-
intervention
intervention
(Time 2)
(Time 4)
17 (53.1%)
23 (71.9%)
3 (9.4%)
6 (18.8%)
10 (31.3%)
11 (34.4%)
11 (34.4%)
11 (34.4%)
Pre-
Post-
intervention
intervention
(Time 2)
(Time 4)
Mean Total cholesterol in mmol/L (SD)
7.23 (1.19)
Mean LDL cholesterol in mmol/L (SD)
4.58 (1.32)
familial hypercholesterolaemia (n = 32)
No. prescribed any statins (%) No. prescribed high potency statins a (%)
(%)
EP
516
interval (%)
27
18.8%*
8.9 , 35.3
9.4%*
3.2 , 24.2
3.1%
0.01 , 15.7
0%
0 , 10.7
Mean
95% confidence
difference
interval
7.07 (1.22)
-0.16
-0.78 , 0.46
4.46 (1.36)
-0.12
-0.81 , 0.57
Simvastatin ≥ 80 mg/day, atorvastatin ≥ 20 mg /day, rosuvastatin ≥ 10 mg/day
AC C
a
TE D
* Significant p < 0.05
difference
M AN U
No. given smoking cessation advice (%)
95% confidence
SC
No. given dietary or weight management advice
Absolute
RI PT
NICE compliance outcomes for management of
ACCEPTED MANUSCRIPT 517
Supplementary
518
Hypercholesterolaemia Guideline (CG71) recommendations
NICE Guideline
Data.
Outcome
measures
for
compliance
to
NICE
Familial
Measure
RI PT
NICE (CG71) Section 1.1 Assessment for Diagnosing Familial Hypercholesterolaemia Section 1.1.8 Proportion of patients with any family history of coronary heart disease assessed Section 1.1.8 Proportion of patients with complete family history (age, condition, degree of relation) Proportion of patients with repeat cholesterol test Proportion of patients diagnosed with secondary cause (diabetes, hypothyroidism,
Section 1.1.2 Section 1.1.2 Section 1.1.2
Proportion of patients with TSH assessment (proxy for hypothyroidism investigations) Proportion of patients with HbA1c assessment (proxy for diabetes investigation) Proportion of patients with serum creatinine assessment (proxy for chronic kidney
Section 1.1.2
Proportion of patients with bilirubin, ALP, or gamma GT assessment (proxy for liver
SC
Section 1.1.6 Section 1.1.2
Section 1.3.2 Section 1.3.2
M AN U
Section 1.1.3 Proportion of patients with arcus senilis or xanthalasma diagnosed NICE (CG71) Section 1.3 Management of Diagnosed Possible Familial Hypercholesterolaemia Section 1.3.1 Total cholesterol levels (mmol/L) Section 1.3.1 LDL cholesterol levels (mmol/L) Section 1.3.1.2 Proportion of patients prescribed any statin treatment Section 1.3.1.3 Proportion of patients prescribed high potency statin treatment (Simvastatin = 80 Proportion of patients given dietary or weight management advice Proportion of patients given smoking cessation advice
AC C
EP
TE D
519
28
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT Highlights Heterozygote familial hypercholesterolaemia (FH) is a common genetic disorder that leads to premature coronary heart disease.
•
We have evaluated a simple pro-active primary care-based intervention inviting patients with raised cholesterols to be assessed for possible FH.
•
We found large improvements in best practice for identification and management of familial hypercholesterolaemia indicating the potential of this approach.
AC C
EP
TE D
M AN U
SC
RI PT
•