Health inequalities in heart failure: A population-based study of 20-year trends in

cause-specific outcomes by gender, socioeconomic status and location of care

Lawson CA, Zaccardi F, Squire I, Ling S, Davies M, Lam C, Mamas M, Khunti K, Kadam

UT

Claire A Lawsona – PhD, Lecturer & Wellcome Trust FellowFrancesco Zaccardia –PhD, Clinical Research FellowIain Squireb – MD; Professor of Cardiovascular MedicineSuping Linga – PhD, EpidemiologistMelanie J Daviesa – MD; Professor of Diabetes MedicineCarolyn SP Lamc –  PhD, Professor of Duke-NUS Cardiovascular Academic Clinical Program Mamas Mamasd – PhD, Professor of CardiologyKamlesh Khuntia – PhD, Professor of Primary Care Diabetes & Vascular MedicineUmesh T Kadama – PhD, Professor of Primary Care & Public Health Research

a. Leicester Diabetes Centre, Leicester University, Leicestershire, UKb. NIHR Leicester BRC, Cardiovascular Research Centre, Glenfield General

Hospital, Leicester University, Leicestershire, UKc. National Heart Centre Singapore, Duke-NUS, University Medical Centre

Groningen, The George Institute for Global Healthd. Keele Cardiovascular Research Group, Centre for Prognosis Research, Keele

University, Staffordshire, UK

Word count: Abstract 299; Text 3895

3 Figures; 4 tables

Corresponding author:

Dr C. A. Lawson Diabetes Research CentreUniversity of LeicesterLeicester, LeicestershireLE5 4PW, England, UKEmail: cl417@leicester.ac.ukTelephone: +44 (0) 4782 709 590

1

2

3

4

5

6

789

101112131415161718192021222324

25

26

27

28

29

3031323334353637

Abstract:

Background: Heart failure (HF) is a global pandemic but contemporary trends in cause-

specific outcomes among different population groups are unknown.

Methods: In a retrospective cohort of incident HF patients between 1998 and 2017

(N=88,416, mean age 78-years, 49% female, median follow-up 2.4 years), cause-specific

hospitalisation and mortality were investigated.

Findings: First-year age-adjusted hospitalisation rate increases were: all-cause 28% from

97.1 (95% CI 94.3,99.9) to 124.2 (120.9,127.5); HF-specific 28% from 17·2 (16·2,18·2) to

22·1 (20·9,23·2) and non-cardiovascular 42% from 59·2 (57·2,61·2) to 83·9 (81·3,86·5) per

100 person-years. Over 70% of all admissions were for non-cardiovascular causes and annual

rate increases were faster for women (+3·9%; 2·8,4·9) than men (+1·4%; 0·6,2·1)

(P<0·0001) and for HF patients diagnosed in hospital (+2.4%; 1·4,3.3) than in the community

(+1·2%; 0·3,2·2). Annual increases for HF hospitalisations were 2·6% (1·9,3·4) for women

and 1·6% (0·6,2·6) for the most deprived, compared to stable rates in men and the most

affluent. Overall, the most deprived (adjusted Incident Rate Ratio 1·20; 1·17,1·24) and

hospital-diagnosed groups (aIRR 1·52; 1·49,1·55) had significantly higher risk of all-cause

admission, but women had significantly lower risk than men (aIRR 0·91; 0·90,0·92).

All-cause first-year age-adjusted mortality rate reduction was 6% from 24·5 (23·4,39·2) to

23·0 (22·0,24·1). Annual rate reductions were -1·4% (-2·3, -0·5) in men but stable for

women and -2·7% (-3·2, -2·2) for the community group compared to -1·1% (-1·8, -0·4) in

the hospital-diagnosed group (p<0·0001). Overall, the most deprived (adjusted Hazard Ratio

1·17;1·14,1·21) and hospital-diagnosed groups (aHR 1·46;1·43,1·48) had significantly higher

risk of all-cause mortality, but women had significantly lower risk than men (aHR

0·85;0·83,0·86).

1

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

Interpretation: Tailored management strategies and specialist HF care are needed to address

persisting and increasing inequalities for men, the most deprived and for HF patients

diagnosed in hospital and to address the worrying trends in women.

Funding: Wellcome-Trust

2

62

63

64

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

86

Research in context

Evidence before this study

Heart failure is a global pandemic affecting 23 million people worldwide and associated with

adverse clinical outcomes (high hospitalisation burden, costs and mortality). We searched

PubMed from 2000 onwards, for observational studies reporting outcome trends in HF. We

used the term ‘heart failure’ in different combinations with ‘surviv*’, ‘rate*’, ‘outcome*’,

‘mortality’, ‘death*’, ‘hospital*’, ‘prognos*’ and ‘trend*’. We reviewed the full text of

relevant articles to assess appropriateness for inclusion. Most studies were from the USA and

were focused on specific care settings (hospital or community) that are unable to ascertain HF

incidence date accurately, or on single communities, which may not reflect the general

population. No studies had reported on cause-specific outcome trends in a national sample of

incident HF patients, regardless of location of care, among different population groups.

Added value of this study

By linking national general practice, hospital and death data, we were able to report outcome

burden following newly diagnosed heart failure and make group comparisons. We found that

hospitalisation rates are increasing and, whilst mortality has improved slightly over twenty

years, risk is higher than previously reported in non-incident cohorts. Furthermore, there were

inequities in outcomes among different HF population groups, some of which persisted or

worsened over time. Risks of all outcomes were significantly higher in the older compared to

younger age groups, for men compared to women, the most deprived compared to the most

affluent and for patients who were first diagnosed with their heart failure in hospital

compared to in the community. We also observed worrying trends for women with HF, with

worsening outcomes in recent years. By investigating cause specific outcomes, we were able

to show that these disparities likely reflect the growing burden of non-cardiovascular

3

87

88

89

90

91

92

93

94

95

96

97

98

99

100

101

102

103

104

105

106

107

108

109

110

comorbidities in HF patients, requiring contemporary prevention and management

approaches.

Implications of all the available evidence

Heart failure presents a major and persistent public health challenge. Diverging outcome

trends among groups, alongside increasing non-cardiovascular outcomes, means that

disparities according to age, gender, socioeconomic status and location of care are set to

increase. Tailored management strategies that include access to specialist HF teams together

with non-cardiovascular and social interventions are urgently required to address the growing

complexity of HF patients and reduce the inequitable outcomes of care.

4

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

Introduction

Heart failure is a global pandemic affecting around twenty-three million people worldwide

including six million Americans,(1) over fifteen million Europeans and one million people in

the UK alone.(2,3) Despite improvements in HF therapies over the past two decades, mortality

risk remains high,(4) with inequalities reported among population groups according to factors

such as gender and socioeconomic status.(5-7) Hospitalisation burden is also high, accounting

for over two thirds of the estimated US$108bn spent on HF globally each year.(8)

Furthermore, the economic costs associated with HF are likely to increase exponentially over

the next decade due to a projected 46% increase in HF prevalence by 2030,(9) alongside rising

costs of healthcare in an ageing, multi-morbid population. Consequently, attempts to reduce

the significant social, health and economic burden of HF have become a top public health

priority.(10)

Prior epidemiological evidence in HF has predominantly focused on mortality in higher risk

hospitalisation cohorts,(11-13) lower risk primary care cohorts(14) or selected communities(5,6,15)

which may not represent the general HF population and limit the potential for group

comparisons. In order to develop tailored public health prevention approaches and to plan

health services effectively, contemporary population level data on emerging outcome trends

among different HF population groups are required. Information on cause-specific

hospitalisation rates in HF is particularly important to meet the projected rising demands.

Through linkage of primary care, hospital and death data we investigated differences in

cause-specific outcomes and trends among groups with incident HF according to gender,

socioeconomic status and location of care.

Methods

Study design and participants

5

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

We used Clinical Practice Research Datalink (CPRD) and Hospital Episodes Statistics (HES)

databases, to identify an incident cohort of HF patients between 1st January 1998 and 31st July

2017. The CPRD is one of the largest longitudinal linked anonymised databases in the world

and contains approximately 7% of the UK general population, representative of sex and age.

It includes patient demographics, risk factors, symptoms, clinical diagnoses, investigations,

referrals and prescriptions and has been validated for epidemiological research.(16) HES

includes all National Health Service admissions in England. The Independent Scientific

Advisory Committee provided permission for data access (protocol 18_037R).

We included all patients aged ≥ 30 years with a first code for HF in their primary care or

hospital record during the 20-years. To ascertain HF diagnostic date accurately, patients were

only included if their CPRD and HES records were eligible for linkage and they had been

registered with a CPRD primary care practice for a minimum of 1-year prior to study entry.

We used updated, clinically validated HF CPRD codes(17) and primary ICD-10 discharge

codes (S1 and S2 Table). Where patients had codes in both records, the first was assigned as

the HF diagnosis date.

Procedures

Exposure group: Patients were categorised into groups defined by age (<70 years, 70-79

years, 80-89 years and ≥ 90 years), gender, socioeconomic status and location of care.

Socioeconomic status was based on the 2010 patient level Index of Multiple Deprivation

(IMD).(18) score which was ranked into quintiles (the most affluent group; quintile 1, to the

most deprived group; quintile 5). Location of care was defined by place of HF diagnosis

(community or hospital), based on the first HF code.

6

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

Confounders: Risk factors, using the most recent value, included current smoking and alcohol

(yes or no), body mass index, systolic blood pressure, cholesterol, haemoglobin and estimated

glomerular filtration rate (eGFR) which .was calculated using the Chronic Kidney Disease

Epidemiology Collaboration formula.(19) Prescribed drugs including beta blocker,

angiotensin converting enzyme inhibitors, angiotensin II receptor blockers, aldosterone

receptor antagonists, aspirin and loop diuretics were identified by at least one prescription in

a 4-month time-window prior to HF diagnosis. Comorbidities were identified by at least one

Read or ICD-10 code in CPRD and HES respectively, at any time up to and including the HF

diagnosis date.

Outcomes: Hospitalisations included all non-elective admissions with at least one overnight

stay and occurring after but not including the HF index date. Overall hospitalisations during

follow-up were counted for each patient and further stratified by cause-specific admissions:

primary HF, other-cardiovascular and non-cardiovascular (non-CVD) admissions according

to the primary discharge code.

Date and cause of death was ascertained from linked Office of National Statistics mortality

records. We coded deaths as ‘cardiovascular’ where there was an ICD-10, chapter 9 code in

the primary position and ‘non-CVD’ for all remaining deaths.

For both outcomes follow-up was until death or study end (31st July 31st).

Statistical analysis

Descriptive statistics are presented as number (%) for dichotomous data, mean (standard

deviation) for continuous data and median (25th and 75th centiles) for skewed continuous data.

7

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

Trend analyses were performed in the whole group and stratified by age groups, gender,

socioeconomic status (most affluent quintile, most deprived quintile) and location of care

(community; hospital). Given the higher risk of poor outcomes in the first year following HF

diagnosis, outcome rates were calculated separately for the first year (first-year rates) and for

the second-year onwards (subsequent-year rates). Additionally, due to the high risk of death

in the first month following HF diagnosis, first-year mortality rates were only calculated in

survivors of the first month. Rates were estimated at the mean population age using age-

adjusted negative binomial models. There were three trend measurements: (i) comparison of

outcomes rates between two 4-year time windows of HF diagnosis (at the start; 1998-2002

and the end; 2012-2015), (ii) average annual percentage change in outcome rates and (iii)

change in trends using Joinpoint Regression which is a statistical software that calculates

annual percentage rate change and tests whether any apparent change in trend is significant.

(20)

Difference in trends between population groups were investigated by fitting an interaction

term (group x HF diagnostic year) to the negative binomial models already containing age.

As interaction tests have low power, we also examined trends visually by plotting graphs of

outcome rates by year of HF diagnosis to aid interpretation.

For overall differences between groups we compared each older age group with the youngest

age group (<70 years; ref), females with males (reference), lower socioeconomic quintiles

(Q2-Q5) with the most affluent quintile (Q1; reference) and the hospital diagnosed group

with the community diagnosed group (reference). Differences in total and cause-specific

admission rates among groups were investigated using negative binomial models to estimate

incidence rate ratios (IRR) with 95% confidence intervals, adjusted for all covariates. For

time to mortality outcomes we used Royston-Parmer flexible parametric survival models to

8

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

perform age-standardised survival predictions at 1, 3 and 5-years for the whole group and

population sub-groups by time-period of diagnosis. Next, age and calendar year standardised

survival curves (all-cause mortality) and cumulative incidence curves (cause specific

mortality) were plotted.(21) Finally, adjusted Hazard ratios (HR) with 95% confidence intervals

were estimated, comparing each sub-group with their respective reference group for all-cause

and cause-specific death.

All analyses were performed using Stata-MP 14. To account for missing data in the

multivariable models (S3 Table), multiple imputations using matching variables and full-

conditional specification were performed, using MI Impute command and results were

obtained using Rubin’s rules.(22) We performed a sensitivity analysis to assess complete case

analysis, removing those with missing values imputed.

Role of the funding source

The funder of the study had no role in study design, data collection, data analysis, data

interpretation, or writing of the report. The corresponding author had full access to all of the

data and the final responsibility to submit for publication

Results

There were 88,416 incident HF patients over the 20-year study time-period, mean age 78-

years, 43,461 (49%) were female and median follow-up was 2·4 years [0·5, 5·7] (Table 1, S1

Figure). Of the study population, 36,415 (41%) patients were first diagnosed with HF in

hospital, a proportion that increased exponentially over time from 26% in 1998 to 74% in

2017 (S2 Figure).

9

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

In the first year following HF diagnosis, 40% of patients had at least one and 20% had

multiple, hospitalisations. First-year hospitalisation rates increased by 28% from 97·1 (95%

CI 94·3,99·9) per 100 person-years during the earlier time-period (1998-2001) to 124·2 (95%

CI 120·9,127·5) during the later time-period (2012-2015), but rates began to plateau after

2005 (Table 2). The steepest growth in first-year rates was for people aged over 80 years

with rates for the younger two age groups remaining stable. Subsequent-year hospitalisation

rates were lower and remained stable overall, but with reducing rates for the youngest and

increasing rates for the oldest age groups (S4 Table).

HF was listed as the primary cause in 17% of hospitalisations occurring in the first year

following HF diagnosis and 10% of hospitalisations occurring in subsequent years. First-year

HF hospitalisation rates increased from 17·2 (16·2,18·2) to 22·1 (20·9, 23·2) per 100 person-

years, at an average increase of 1·8% per annum (pa) (Table 2), whilst subsequent-year rates

remained stable (S4 Table). This steady increase was in contrast to hospitalisations for other

CVD causes which reduced by 2·4% pa after 2005 (Figure 1). Non-CVD was the primary

diagnosis in 73% of all hospitalisations·, with first-year rates increasing from 59·2 (83·9,

86·5) to 83·9 (81·3, 86·5) per 100 person-years, an average increase of 2·6% pa (Table 2).

Subsequent year rates were also high and increased by 0·8% pa, reaching 64·0 (62·3, 65·8)

per 100 person-years during 2012-2015 (S4 Table).

During follow-up, 65,063 (74%) patients with HF died with 32,623 (50·1%) cardiovascular

and 32,440 (49·9%) non-CVD deaths. Between 1998-2001 and 2012-2015, the number of

people dying within one month of their HF diagnosis reduced from 14% to 11% (Table 3). In

those surviving their first month, age-adjusted first-year mortality rates reduced only slightly

from 24·5 (23·4, 25·7) to 23·0 (22·0, 24·1) per 100 person-years, but there were differences

10

262

263

264

265

266

267

268

269

270

271

272

273

274

275

276

277

278

279

280

281

282

283

284

285

286

by age (Table 3). The youngest two age groups experienced mortality rate reductions whilst

the oldest groups had a significant increase in mortality rates. In the group aged > 90-years,

rates increased by 20%, reaching 69·6 (63·2, 76·0) per 100 person-years (Table 3, Figure 2).

Subsequent year mortality rates were lower and reduced by 16% over time, reaching 14·7

(14·3, 15·2) per 100 person-years (Table 3). Age-standardised mortality risk reduced

between the two time-periods; 1-year (32% vs 29%), 3-year (50% vs 46%) and 5-year risk

(63% vs 59%) (S5 Table).

Outcomes by gender: During 1998-2002, first-year and subsequent-year age-adjusted rates of

any and cause-specific hospitalisations were higher for men than women. However, by 2012-

2015, first-year rates for all-cause hospitalisations were significantly higher in women

(129·7; 124·7, 134·7) compared to men (119·1; 114·8, 123·5) per 100 person-years,

explained by increasing rates for women, compared to stable rates in men (Table 2, Figure

1). Women also had over twice the increase in first-year non-CVD hospitalisation rates (3·9%

pa) compared to men (1.4% pa). Similar patterns were found for subsequent-year rates. After

full co-variate adjustment, women had lower incidence rate of all cause-specific

hospitalisation types than men, with the largest reduction for CVD hospitalisations (IRR

0·85; 0·83, 0·88) (Table 4). However, when stratified by year of diagnosis, these risk

differences diminished and became statistically insignificant in later years (not shown).

In 1998-2001, first-year mortality rates per 100 person-years were higher for men (29·2;

27·4, 31·0) than women (20·7; 19·5, 22·0) but the gap diminished by 2012-2015 with rates of

23·5 (22·1, 24·9) for men and 22·0 (20·7, 23·4) for women. This was due to with a 2·6% pa

reduction in rates for men prior to 2009 compared to stable rates in women (Table 3, Figure

2). Subsequent-year rates reduced at twice the rate in men (-2·2%; -2·5, -1·9 pa) than women

11

287

288

289

290

291

292

293

294

295

296

297

298

299

300

301

302

303

304

305

306

307

308

309

310

311

(-0·8%; -1·1, -0·5 pa). Overall, age-standardised mortality risk at 1, 3 and 5 years was

approximately 3-4% higher in men than women which persisted over time (S5 Table) and

following full adjustment, women were at lower risk of all-cause deaths (HR 0·85; 95% CI

0·83,0·86), CVD deaths (HR 0·85; 95% CI 0·83, 0·88) and non-CVD deaths (HR 0·84; 95%

CI 0·82,0·87) compared to men (Table 4, S3 Figure). Again, after stratifying by year of

diagnosis (not shown), the difference between gender groups was attenuated, becoming

statistically insignificant by 2016.

Outcomes by socioeconomic status: In 1998-2002, first-year any cause hospitalisation rates

were significantly higher in the most deprived than the most affluent group. This difference

persisted over-time due to similar increasing rates in both groups (P=0·103) (Table 2, Figure

1) and subsequent-year rates showed a similar pattern. In the most recent time-period (2012-

2015), the biggest difference was for first-year non-CVD hospitalisations, with the figures for

most deprived of 101·7 (93·9, 109·5) per 100 person-years compared to 60·8 (58·1, 63·5) for

the most affluent (Table 2). In fully adjusted models, the most deprived compared to the

most affluent group had significantly higher risk of all-cause (IRR 1·20; 1·17, 1·24),

cardiovascular (IRR 1·18; 1·13, 1·19) and non-CVD hospitalisations (IRR 1·21; 1·22, 1·25).

Risk increased with quintiles of socioeconomic status from the highest (Q1) to the lowest

quintile (Q5) (Table 4).

In the most-recent time-period, first-year mortality rates per 100 person-years were higher in

the most deprived group (25·7; 22·9, 28·5) compared to the most affluent group (21·1; 19·1,

23·1) (Table 3, Figure 2). Age standardised mortality risk increased for each quintile of

socioeconomic status, from the most affluent to most deprived group. In the later time-period

risk ranged from 27% (Q1) to 32% (Q5) at 1-year, 43% to 50% at 3-years and 56 to 63% at

12

312

313

314

315

316

317

318

319

320

321

322

323

324

325

326

327

328

329

330

331

332

333

334

335

336

5-years (S5 Table). Following full adjustment, the most deprived group remained at higher

risk of all-cause death (HR 1·17; 1·14, 1·21), CVD death (HR 1·18; 1·14, 1·23) and non-

CVD death (HR 1·17; 1·13, 1·22) than the most affluent group and age-standardised survival

reduced for each quintile of socioeconomic status (Table 4, S3 Figure).

Outcomes according to location of care: First-year hospitalisation rates for any cause (per

100 person-years) were significantly higher in the hospital-diagnosed than the community-

diagnosed group, with similar growth over time reaching 163·2 (157·7, 169·0) and 88·3

(84·9, 91·6) respectively by 2012-2015. Annual rates between the two groups began to

diverge after 2004 reducing at the same rate for the community group as they increased for

the hospital group (1% pa) (Table 2, Figure 1). The hospital group experienced a consistent

2·4 % pa increase in first-year non-CVD hospitalisations, whilst rates did not increase for the

community group after 2004.· When adjusted for all covariates the hospital group had

significantly higher risk of cardiovascular (incidence rate ratio [IRR] 1·77; 1·72, 1·82) and

non-CVD (IRR 1·40; 1·37, 1·43) hospitalisations than the community group (Table 4).

First-year mortality rates in the hospital group (34·3; 32·3, 36·2) were more than double the

rates in the community group (14·2; 13·2, 15·1) per 100 person-years and this gap widened

over time (p<0·0001). Age-adjusted mortality rates reduced in the community group (-2·7%;

-3·2, -2·2 pa) at twice the rate of the hospital group (-1·1%; -1·8,-0·4 pa) (Table 3, Figure

2). A similar trend was identified for subsequent-year rates (S4 Table).

During 2012-2015, age-standardised mortality risks remained significantly higher in the

hospital-diagnosed compared to community-diagnosed group: 1-year risk (34 vs 24%), 3-year

risk (53 vs 40%) and 5-year (66 vs 52%) (S5 Table). Overall, median survival time in the

13

337

338

339

340

341

342

343

344

345

346

347

348

349

350

351

352

353

354

355

356

357

358

359

360

361

community group (4·2; 1·0, 9·5 years) was twice that of the hospital group (1·9; 0·2, 5·6)

(Table 4, S3 Figure). In the fully adjusted models, the hospital-diagnosed group had

significantly higher risk of CVD deaths (HR 1·55; 95% CI 1·51, 1·59) and non-CVD deaths

(HR 1·37; 95% CI 1·34, 1·40) than the community group (Table 4).

For all groups, risk of cardiovascular death was higher than non-CVD deaths for the first few

years following HF diagnosis but the two causes converged in later years with non-CVD

causes taking over at around 3-5 years (S4 Figure). Effects remained relatively stable

following full case analysis (S6 Table).

Discussion

Our study provides a comprehensive and contemporary assessment of outcomes in patients

with incident HF in community and specialist hospital settings. In a large, nationally

representative sample over 20 years, there were specific diverging trends and inequalities in

hospitalisations and mortality. Non-elective hospitalisation burden has increased, particularly

for non-CVD causes, which account for over two-thirds of all admissions. HF remains

associated with a high risk of death, with only minimal improvement in risk over the last 20

years. Inequalities and poor outcomes have persisted, particularly for the oldest age groups,

men, the most deprived and for patients diagnosed in hospital and new worrying trends for

women have emerged.

Mortality risk following HF diagnosis has reduced only minimally over the past two decades.

Whilst one and five-year mortality risk reduced by 3-4%, this small reduction is likely in part

to reflect advances in diagnostic technology to identify milder HF over this time. Despite the

risk reduction, our 1 and 5-year figures are similar or worse than those at the turn of the

14

362

363

364

365

366

367

368

369

370

371

372

373

374

375

376

377

378

379

380

381

382

383

384

385

386

century in the Framingham (29% at one year) and Rochester (48% at five years) cohorts.(23, 24)

Hospital cohorts in the USA,(25) Scotland,(26) and recently in Demark(27) reported slightly

lower 1-year risk of 29%, 33% and 30% compared to the 34% risk in our hospital diagnosed

group. Similarly, community cohorts from UK(14) have recently reported 1-year risk of 24%

compared to the 27% overall 1-year risk in our community diagnosed group. Like most prior

HF epidemiological evidence from ‘hospital only’ or ‘primary care only’ cohorts, these

cohorts were not able to ascertain HF diagnosis date. By linking community and hospital

records we were able to identify incident HF, which likely accounts for the higher mortality

risk in our study and indicates that the true risk burden for HF patients may be higher than

previously reported. The persisting high risk in HF possibly reflects the ageing multi-morbid

population, alongside a survival plateau in the era of effective, evidence-based treatment for

HF (selective beta-blockers and angiotensin converting enzyme inhibitors). The rise in the

number of non-CVD hospitalisations as well as the 50% deaths for non-CVD causes would

support this hypothesis.

We observed a new and important trend of increasing HF hospitalisations. Prior evidence on

cause-specific trends grouped HF admissions with cardiovascular admissions, reporting

stable or reducing rates over time.(5) By partitioning out HF from other-cardiovascular

admissions we identified a new trend, not previously reported. This is particularly important

given the growth of day case diuretic units that should reduce the number of acute HF

patients needing hospitalisation and meaning that our findings may underestimate the

worsening acute HF problem. Increased HF hospitalisations were particularly pronounced in

women and possibly indicates later presentation of more severe HF in older multi-morbid

patients. Other potential triggers may be the lack of evidenced based therapies for HF with

15

387

388

389

390

391

392

393

394

395

396

397

398

399

400

401

402

403

404

405

406

407

408

409

410

preserved ejection fraction which is increasing in prevalence(28) or improved survival of more

severe HF patients with contemporary HF therapies.

Hospital-diagnosed HF patients had significantly worse outcome rates over time compared to

the community group, a finding shared by previous reports.(29,30) Our findings reveal a

worrying exponential increase in these patients reaching 79% of the HF population by 2017.

Furthermore, our study showed that the steeply rising burden of non-CVD hospitalisations

were predominantly focused to the hospital-diagnosed group. Death rates were also

significantly higher for the hospital-diagnosed group with slower rates of decline than in the

community group. Delay in diagnosis, higher comorbidity burden and lower prescribing of

first-line preventive medications in community care, leading to more severe HF at the point

of diagnosis, may part explain the worse prognosis in this group. Recent developments

suggests that growth of dedicated HF specialists has better outcomes for hospitalised patients

(UK National Heart Failure Audit)(31) and recent policy recommends that all patients with HF

have access to a multidisciplinary team.(32) It is also important to remember that “evidence-

based” treatments in HF are relevant only to those patients with reduced ejection fraction, and

some of the apparent under-use of these treatments may be explained by the lack of evidence

for the use of these treatments in HFpEF. A recent report shows that around half of these

hospital diagnosed patients present earlier to their primary care with HF symptoms,(33)

demonstrating an urgent need for specialist support for earlier diagnosis.

Prior HF studies have shown worse outcomes for men than women.(6,34) Our findings show

that steeper increases in HF and non-CVD hospitalisation rates for women as well as slower

reduction in mortality rates in women compared to men has resulted in .similar outcomes

between the genders in recent years, a finding also reported recently in Denmark and

16

411

412

413

414

415

416

417

418

419

420

421

422

423

424

425

426

427

428

429

430

431

432

433

434

435

Ontario(27,29) These trends likely reflect the increasing prevalence of non-CVD comorbidities

in women. However, the attenuation of gender difference over time, even after taking account

of comorbidities, as well as greater increasing rate of HF hospitalisations over time in

women, suggest worsening HF severity or lack of effective HF therapies in women compared

to men. These unfavourable trends in women may also relate to the increasing prevalence of

HF with preserved ejection fraction (HFpEF) in women (for which there is still, to date, no

proven therapy to improve outcomes).(12,35)

The most deprived HF group compared to the most affluent group had consistently poor

outcomes, with 20% higher risk even after adjustment, an inequality that has persisted over

the last 20 years. Previously, deprivation status has not been associated with cause-specific

hospitalisations,(7) but our study highlights a new finding of increased risk across all cause-

specific outcomes in the most deprived HF patients. Whilst higher burden of comorbidities

and worse lifestyle risk factors may partly explain this increased risk, the persisting

difference following full adjustment for these factors suggests other social and healthcare

factors may be at play.(36) Potential delays in health seeking behaviour,(37) with higher severity

of comorbidities and HF at the point of diagnosis, may be one explanation and points to the

preventable nature of HF. Lower prevalence of cancer in the most deprived group opposes

the trends found for other comorbidities, but is perhaps unsurprising considering the

persistent socioeconomic inequalities in cancer survival.(38)

To our knowledge, this is the largest incident HF study to report trends in cause-specific

hospitalisations and mortality in general population groups. .By linking large, nationally

representative databases, we were able to ascertain the incident date of HF and follow

patients for up to 20 years for cause-specific hospitalisations and death. However, the use of

17

436

437

438

439

440

441

442

443

444

445

446

447

448

449

450

451

452

453

454

455

456

457

458

459

460

routinely collected data means HF diagnosis can be subject to misclassification and

measurement error, which may change over time. Diagnosis of HF in the community is

clinically defined and we cannot rule out that the lower risk in the community group may in

part be due to misclassification. However, in the UK, HF diagnosis has been enhanced by the

introduction of echocardiography as part of performance incentives(39) and national HF audits

and HF diagnoses have been shown to have high precision in CPRD.(17) The outcomes

measured in the study are also dependent on HF severity or type but routine data did not

provide echocardiographic data or specific HF phenotypes. To account for severity, a range

of comorbidity and medications were included in our analyses, but further investigation on

specific HF phenotypes with outcomes is required. Accuracy of clinical recording and

diagnoses within the CPRD have been found to be valid for a range of morbidities(16) and we

also used primary ICD codes for cause-specific outcomes and clinically validated code sets

which have high precision.(17) Changes in diagnostic procedures over time, such as more

advanced echocardiography to detect milder forms of HF, may have influenced our findings

and be partly responsible for the small reduction in mortality risk over time. This bias would

only serve to temper the overall mortality risk in the later time-period as well as suppress

increasing hospitalisation rates, meaning that the overall burden of HF could be even higher

than reported.

Conclusions

Population inequalities in outcomes of HF present a major and persistent public health

challenge. Diverging trends among groups, alongside increasing non-CVD outcomes, means

that disparities according to age, gender, socioeconomic status and location of care are set to

increase. Tailored management strategies that include access to specialist HF teams together

18

461

462

463

464

465

466

467

468

469

470

471

472

473

474

475

476

477

478

479

480

481

482

483

484

with non-CVD and social interventions are urgently required to address the growing

complexity of HF patients and reduce the inequitable outcomes of care.

Acknowledgements

The research was supported by Leicester-Wellcome Trust ISSF Fellowship [Reference

204801/Z/16/Z]. FZ is funded with an unrestricted educational grant from the NIHR

CLAHRC East Midlands to the University of Leicester.

The study sponsors had no role in the design and conduct of the study; collection,

management, analysis, and interpretation of the data; preparation, review, or approval of the

manuscript; and decision to submit the manuscript for publication. The views and opinions

expressed therein are those of the authors and do not necessarily reflect those of the NIHR

(UK) or Wellcome Trust.

Data

CL had full access to all the data in the study and takes responsibility for the integrity of the

data and the accuracy of the data analysis. This study is based in part on data from the

Clinical Practice Research Datalink obtained under licence from the UK Medicines and

Healthcare products Regulatory Agency. However, the interpretation and conclusions

contained in this report are those of the author/s alone. Data access is through permissions

from CPRD only.

Author contributions

CAL: conception, design, acquisition, analysis, interpretation of data, writing; FZ: design,

analysis, interpretation, critical revision; IS: design, interpretation, critical revision; SL:

19

485

486

487

488

489

490

491

492

493

494

495

496

497

498

499

500

501

502

503

504

505

506

507

508

509

acquisition of data, interpretation, critical revision; MD: interpretation, critical revision; CL:

figures, interpretation, critical revision; MM: design, interpretation, critical revision; KK:

conception, design, acquisition, interpretation, critical revision; UK: conception, design,

analysis, interpretation of data, writing, critical revision

Declarations of Interest

Dr. Khunti reports personal fees from Amgen , personal fees from Astrazeneca, personal fees

from Bayer, personal fees from NAPP, personal fees from Lilly , personal fees from Merck

Sharp & Dohme , personal fees from Novartis, personal fees from Novo Nordisk , personal

fees from Roche, personal fees from Berlin-Chemie AG / Menarini Group, personal fees

from Sanofi-Aventis , personal fees from Servier, personal fees from Boehringer Ingelheim,

grants from Pfizer , grants from Boehringer Ingelheim, grants from AstraZeneca, grants from

Novartis, grants from Novo Nordisk, grants from Sanofi-Aventis, grants from Lilly, grants

from Merck Sharp & Dohme, grants from Servier, outside the submitted work; .

MJD has acted as consultant, advisory board member and speaker for Novo Nordisk, Sanofi–

Aventis, Lilly, Merck Sharp & Dohme, Boehringer Ingelheim, AstraZeneca and Janssen, an

advisory board member for Servier and as a speaker for Mitsubishi Tanabe Pharma

Corporation and Takeda Pharmaceuticals International Inc. She has received grants in support

of investigator and investigator initiated trials from Novo Nordisk, Sanofi–Aventis, Lilly,

Boehringer Ingelheim and Janssen.

All other authors have reported that they have no relationships relevant to the contents of this

paper to disclose.

References

20

510

511

512

513

514

515

516

517

518

519

520

521

522

523

524

525

526

527

528

529

530

531

532

533

1. Bui AL, Horwich TB, Fonarow GC. Epidemiology and risk profile of heart failure. Nature Reviews Cardiology. 2011;8(1):30-41.

2. Cardiovascular disease statistics [Internet]. London: British Heart Foundation; 2015. Available from: https://www.bhf.org.uk/publications/statistics/cvd-stats-2015 [Accessed 10th September 2018).

3. Ponikowski P, Voors AA, Anker SD, Bueno H, Cleland JGF, Coats AJS, et al. 2016 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure. European Heart Journal. 2016;37(27):2129-200.

4. Mamas MA, Sperrin M, Watson MC, Coutts A, Wilde K, Burton C, et al. Do patients have worse outcomes in heart failure than in cancer? A primary care-based cohort study with 10-year follow-up in Scotland. European Journal Heart Failure. 2017;19(9):1095-104.

5. Gerber Y, Weston SA, Redfield MM, Chamberlain AM, Manemann SM, Jiang R, et al. A contemporary appraisal of the heart failure epidemic in Olmsted County, Minnesota, 2000 to 2010. JAMA Internal Medicine. 2015 Jun;175(6):996-1004.

6. Akwo EA, Kabagambe EK, Wang TJ, Harrell FE,Jr, Blot WJ, Mumma M, et al. Heart Failure Incidence and Mortality in the Southern Community Cohort Study. Circulation Heart Failure. 2017 Mar;10(3):10.1161/CIRCHEARTFAILURE.116.003553.

7. Witte KK, Patel PA, Walker AMN, Schechter CB, Drozd M, Sengupta A, et al. Socioeconomic deprivation and mode-specific outcomes in patients with chronic heart failure. Heart 2018;104:993–998. doi:10.1136/heartjnl-2017-312539

8. Cook C, Cole G, Asaria P, Jabbour R, Francis DP. The annual global economic burden of heart failure. International Journal of Cardiology. 2014;171(3):368-76.

9. Benjamin EJ, Blaha MJ, Chiuve SE, Cushman M, Das SR, Deo R, et al. Heart Disease and Stroke Statistics-2017 Update: A Report From the American Heart Association. Circulation. 2017 Mar 7;135(10):e603.

10. Schocken DD, Benjamin EJ, Fonarow GC, Krumholz HM, Levy D, Mensah GA, et al. Prevention of Heart Failure. Circulation. 2008;117(19):2544-65.

11. Chen J, Normand ST, Wang Y,Krumholz HM. NAtional and regional trends in heart failure hospitalization and mortality rates for medicare beneficiaries, 1998-2008. Journal Of The American Medical Association. 2011 October 19;306(15):1669-78.

12. Chang PP, Wruck LM, Shahar E, Rossi JS, Loehr LR, Russell SD, et al. Trends in Hospitalizations and Survival of Acute Decompensated Heart Failure in Four US Communities (2005-2014): The Atherosclerosis Risk in Communities (ARIC) Study Community Surveillance. Circulation. 2018;138:12–24

13. Christiansen MN, Kober L, Weeke P, Vasan RS, Jeppesen JL, Smith JG, et al. Age-Specific Trends in Incidence, Mortality, and Comorbidities of Heart Failure in Denmark, 1995 to 2012. Circulation. 2017;135(13):1214-23.

21

534535

536537538

539540541

542543544

545546547

548549550

551552553

554555

556557558

559560

561562563

564565566567

568569570

14. Taylor CJ, Ordóñez-Mena JM, Roalfe AK, Lay-Flurrie S, Jones NR, Marshall T, et al. Trends in survival after a diagnosis of heart failure in the United Kingdom 2000-2017: population based cohort study. British Journal of Medicine. 2019;364:l223.

15. Roger VL, Weston SA, Redfield MM, Hellermann-Homan JP, Killian J, Yawn BP, et al. Trends in heart failure incidence and survival in a community-based population. Journal of The American Medical Association. 2004;292(3):344-50.

16. Herrett E, Gallagher A,M., Bhaskaran K, Forbes H, Mathur R, van Staa T, et al. Data Resource Profile: Clinical Practice Research Datalink (CPRD). International Journal of Epidemiology. 2015;44(3):827-36.

17. Maru S, Koch GG, Stender M, Clark D, Gibowski L, Petri H, et al. Antidiabetic drugs and heart failure risk in patients with type 2 diabetes in the U.K. primary care setting. Diabetes Care. 2005;28(1):20-6.

18. English Indices of Deprivation 2010: Guidance Doument [Online] [Internet]. 2011; Available from: https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/6871/1871208.pdf [Accessed June 8th 2015].

19. Valente MA, Hillege HL, Navis G, Voors AA, Dunselman PH, van Veldhuisen DJ, et al. The Chronic Kidney Disease Epidemiology Collaboration equation outperforms the Modification of Diet in Renal Disease equation for estimating glomerular filtration rate in chronic systolic heart failure. European Journal Heart Failure. 2014;16(1):86-94.

20. Rea F, Pagan E, Compagnoni M, Cantarutti A, Pugni P, Bagnardi V, et al. Joinpoint regression analysis with time-on-study as time-scale. Application to three Italian population-based cohort studies. Epidemiology Biostatistics and Public Health. 2017;14(3).

21. Putter H, Fiocco M, Geskus RB. Tutorial in biostatistics: competing risks and multi-state models. Statistics in Medicine. 2007;26(11):2389-430.

22. Rubin DB. Multiple Imputation for Nonresponse in Surveys. New York: Wiley. New York: Wiley; 1987.

23. Levy D, Kenchaiah S, Larson MG, Benjamin EJ, Kupka MJ, Ho K, et al. Long-term trends in the incidence of and survival with heart failure. New England Journal of Medicine. 2002;347(18):1397.

24. Roger VL, Weston SA, Redfield MM, Hellermann-Homan JP, Killian J, Yawn BP, et al. Trends in heart failure incidence and survival in a community-based population. Journal of The American Medical Association. 2004 Jul 21;292(3):344-50.

25. Chen J, Normand SL, Wang Y, Krumholz HM. National and regional trends in heart failure hospitalization and mortality rates for Medicare beneficiaries, 1998-2008. Journal of The American Medical Association. 2011;306(15):1669-78.

26. Jhund PS, Macintyre K, Simpson CR, Lewsey JD, Stewart S, Redpath A, et al. Long-term trends in first hospitalization for heart failure and subsequent survival between 1986 and 2003: a population study of 5.1 million people. Circulation. 2009;119(4):515-23.

22

571572573

574575576

577578579

580581582

583584585

586587588589

590591592

593594

595596

597598599

600601602

603604605

606607608

27. Schmidt M, Ulrichsen SP, Pedersen L, Botker HE, Sorensen HT. Thirty-year trends in heart failure hospitalization and mortality rates and the prognostic impact of co-morbidity: a Danish nationwide cohort study. European Journal Heart Failure. 2016;18(5):490-9.

28. Shah A, Udeoji D, Baraghoush A, Bharadwaj P, Yennurajalingam S, Schwarz E.      An Evaluation of the Prevalence and Severity of Pain and Other Symptoms in Acute Decompensated Heart Failure. Journal of Palliative Medicine. 2013;16(1):87-90.

29. Sun LY, Tu JV, Sherrard H, Rodger N, Coutinho T, Turek M, et al. Sex-Specific Trends in Incidence and Mortality for Urban and Rural Ambulatory Heart Failure Patients in Eastern Ontario from 1994 to 2013. Journal of Cardiac Failure. 2018;24(9):568-574

30. Koudstaal S, Pujades-Rodriguez M, Denaxas S, Gho, J. M. I. H., Shah AD, Yu N, et al. Prognostic burden of heart failure recorded in primary care, acute hospital admissions, or both: a population-based linked electronic health record cohort study in 2.1 million people. European Journal Heart Failure. 2017;19(9):1119-27.

31. University College London / British Society for Heart Failure National Institute for Cardiovascular Outcomes Research (NICOR). National Heart Failure Audit 2016/17 Summary Report. University College London: National Institute for Cardiovascular Outcomes Research; 2018.

32. National Confidential Enquiry into Patient Outcome and Death, (NCEPOD). Acute Heart Failure: Failure to Function. London: NCEPOD; 2018.

33. Bottle A, Kim D, Aylin P, Cowie MR, Majeed A, Hayhoe B. Routes to diagnosis of heart failure: observational study using linked data in England. Heart. 2018;104(7):600-5.

34. Gerber Y, Weston SA, Enriquez-Sarano M, Berardi C, Chamberlain AM, Manemann SM, et al. Mortality Associated With Heart Failure After Myocardial Infarction: A Contemporary Community Perspective. Circulation Heart Failure. 2016;9(1):e002460.

35. Bhatia RS, Tu JV, Lee DS, Austin PC, Fang J, Haouzi A, et al. Outcome of heart failure with preserved ejection fraction in a population-based study. New England Journal of Medicine. 2006;355(3):260-9.

36. Hawkins NM, Jhund PS, McMurray JJ, Capewell S. Heart failure and socioeconomic status: accumulating evidence of inequality. European Journal Heart Failure. 2012;14(2):138-46.

37. McAlister FA, Murphy NF, Simpson CR, Stewart S, MacIntyre K, Kirkpatrick M, et al. Influence of socioeconomic deprivation on the primary care burden and treatment of patients with a diagnosis of heart failure in general practice in Scotland: population based study. British Journal of Medicine. 2004;328(7448):1110.

38. Exarchakou A, Rachet B, Belot A, Maringe C, Coleman MP. Impact of national cancer policies on cancer survival trends and socioeconomic inequalities in England, 1996-2013: population based study. British Journal of Medicine. 2018;360.

23

609610611

612613614

615616617

618619620621

622623624625

626627

628629

630631632

633634635

636637638

639640641642

643644645

39. Stephen Gillam, Nicholas Steel. The Quality and Outcomes Framework—where next? British Journal of Medicine. 2013;346.

24

646647

 

25

648

649

Table 1 Patient characteristics by gender, socio-economic status and location of care

Gender Socioeconomic status Location of care

CharacteristicsAll Men Women Q1 (most affluent) Q2 Q3 Q4 Q5 (most deprived) Community Hospital

(N=88,416) (N=44,955) (N=43,461) (N=16,481) (N=20,468) (N=18,910) (N=17,542) (N=14,861) (N=52,001) (N=36,415)

Age in years 77.8 (11.3) 75.3 (11.5) 80.3 (10.5) 79.0 (10.8) 78.7 (10.8) 78.2 (11.2) 76.9 (11.5) 75.6 (12.0) 76.9 (11.2) 79.0 (11.4)

<70 years 18168 (21) 12,178 (27) 5,990 (14) 2,884 (17) 3,659 (18) 3,664 (19) 3984 (23) 3953 (27) 11,627 (22) 6,541 (18)

70 to 79 years 25878 (29) 14,379 (32) 11,499 (26) 4,638 (28) 5,804 (28) 5,416 (29) 5336 (30) 4653 (31) 16,357 (31) 9,521 (26)

80 to 89 years 33152 (37) 14,956 (33) 18,196 (42) 6,542 (40) 8,104 (40) 7,310 (39) 6287 (36) 4837 (33) 18,697 (36) 14,455 (40)

90 years and over 1 1218 (13) 3,442 (8) 7,776 (18) 2,417 (15) 2,901 (14) 2,520 (13) 1935 (11) 1418 (10) 5,320 (10) 5,898 (16)

Female 43,461 (49) - - 7,804 (47) 9,889 (48) 9,258 (49) 8879 (51) 7552 (51) 24,700 (47) 18,761 (52)

Most affluent group 16,481 (19) 8,677 (19) 7,804 (18) - - - - - 10,011 (19) 6,470 (18)

Most deprived group 14,861 (17) 7,309 (16) 7,552 (17) - - - - - 8,463 (16) 6,398 (18)

Hospital diagnosis 36,415 (41) 17,654 (39) 18,761 (43) 6,470 (39) 8,384 (41) 7,662 (41) 7441 (42) 6398 (43) - -

Beta blocker 26,627 (30) 14,222 (32) 12,405 (29) 5,322 (32) 6,225 (30) 5,619 (30) 5224 (30) 4209 (28) 17,134 (33) 9,493 (26)

ACE inhibitor 33,083 (37) 18,350 (41) 14,733 (34) 6,040 (37) 7,717 (38) 7,054 (37) 6633 (38) 5596 (38) 21,978 (42) 11,105 (31)

ARB 9,166 (10) 4,354 (10) 4,812 (11) 2,009 (12) 2,171 (11) 1,905 (10) 1736 (10) 1332 (9) 5,541 (11) 3,625 (10)

ACE or ARB 40,983 (46) 22,031 (49) 18,952 (44) 7,795 (47) 9,567 (47) 8,715 (46) 8129 (46) 6723 (45) 26,689 (51) 14,294 (39)

AA 5,767 (7) 3,108 (7) 2,659 (6) 1,154 (7) 1,348 (7) 1,200 (6) 1148 (7) 908 (6) 3,432 (7) 2,335 (6)

Aspirin 33,908 (38) 18,697 (42) 15,211 (35) 6,068 (37) 7,719 (38) 7,303 (39) 6855 (39) 5911 (40) 21,404 (41) 12,504 (34)

Diuretic (loop) 43,332 (49) 21,412 (48) 21,920 (50) 7,987 (48) 10,113 (49) 9,274 (49) 8568 (49) 7307 (49) 28,139 (54) 15,193 (42)

Comorbidities

Number; mean (SD) 4.1 (2.0) 4.0 (2.0) 4.2 (2.1) 3.9 (2.0) 4.0 (2.0) 4.1 (2.0) 4.2 (2.1) 4.2 (2.1) 3.7 (2.0) 4.6 (2.1)

<2 8,634 (10) 4,760 (11) 3,874 (9) 1,730 (10) 2,068 (10) 1860 (10) 1601 (9) 1,343 (9) 6,386 (12) 2,248 (6)

2 to 3 27,676 (31) 14,798 (33) 12,878 (30) 5,416 (33) 6,476 (32) 6041 (32) 5307 (30) 4,387 (30) 18,495 (36) 9,181 (25)

4 to 5 31,205 (35) 15,785 (35) 15,420 (35) 5,834 (35) 7,396 (36) 6566 (35) 6193 (35) 5,171 (35) 17,699 (34) 13,506 (37)

>5 20,901 (24) 9,612 (21) 11,289 (26) 3,501 (21) 4,528 (22) 4443 (23) 4441 (25) 3,960 (27) 9,421 (18) 11,480 (32)

26

IHD 43,983 (50) 25,111 (56) 18,872 (43) 7,979 (48) 9,891 (48) 9356 (49) 8919 (51) 7,779 (52) 24,165 (46) 19,818 (54)

MI 23,481 (27) 14,795 (33) 8,686 (20) 4,311 (26) 5,266 (26) 4984 (26) 4734 (27) 4,162 (28) 13,197 (25) 10,284 (28)

AF 34,884 (39) 17,909 (40) 16,975 (39) 6,795 (41) 8,361 (41) 7481 (40) 6715 (38) 5,483 (37) 17,332 (33) 17,552 (48)

Hypertension 56,908 (64) 27,768 (62) 29,140 (67) 10,603 (64) 13,135 (64) 12068 (64) 11416 (65) 9,614 (65) 30,915 (59) 25,993 (71)

Diabetes 21,565 (24) 11,872 (26) 9,693 (22) 3,439 (21) 4,745 (23) 4462 (24) 4688 (27) 4,202 (28) 10,675 (21) 10,890 (30)

Stroke 10,686 (12) 5,556 (12) 5,130 (12) 1,911 (12) 2,386 (12) 2316 (12) 2178 (12) 1,878 (13) 5,668 (11) 5,018 (14)

Anaemia 11,167 (13) 4,419 (10) 6,748 (16) 19,42 (12) 2,,488 (12) 2415 (13) 2272 (13) 2,027 (14) 5,318 (10) 5,849 (16)

Obesity 22,207 (25) 11,319 (25) 10,888 (25) 3,457 (21) 4,725 (23) 4706 (25) 4910 (28) 4,380 (29) 12,475 (24) 9,732 (27)

CKD 37,850 (56) 17,244 (49) 20,606 (63) 7,361 (56) 9,062 (57) 8198 (56) 7349 (56) 5,821 (54) 20,021 (53) 17,829 (60)

COPD 16,438 (19) 9,229 (21) 7,209 (17) 2,282 (14) 3,299 (16) 3276 (17) 3688 (21) 3,859 (26) 8,900 (17) 7,538 (21)

Asthma 16,512 (19) 7,911 (18) 8,601 (20) 2,726 (17) 3,631 (18) 3371 (18) 3450 (20) 3,313 (22) 9,488 (18) 7,024 (19)

Depression 20,060 (23) 7,997 (18) 12,063 (28) 3,392 (21) 4,310 (21) 4301 (23) 4273 (24) 3,754 (25) 11,461 (22) 8,599 (24)

Osteoarthritis 32,580 (37) 13,753 (31) 18,827 (43) 6,097 (37) 7,612 (37) 6985 (37) 6333 (36) 5,490 (37) 18,462 (36) 14,118 (39)

Cancer 20,500 (23) 10,733 (24) 9,767 (22) 4,227 (26) 5,121 (25) 4444 (24) 3773 (22) 2,900 (20) 11,360 (22) 9,140 (25)

Dementia 3,993 (5) 1,502 (3) 2,491 (6) 7,28 (4) 9,35 (5) 896 (5) 790 (5) 633 (4) 1,927 (4) 2,066 (6)

Smoking 17,408 (21) 9,967 (23) 7,441 (18) 2,553 (16) 3,619 (19) 3554 (20) 3838 (23) 3,811 (27) 11,101 (23) 6,307 (18)

Alcohol 53,789 (71) 31,526 (79) 22,263 (61) 10,890 (77) 12,894 (74) 11503 (71) 10175 (67) 8,245 (64) 32,266 (73) 21,523 (68)

BMI (kg/m2) 26.8 (23.6, 30.8) 27.0 (24.1, 30.6) 26.6 (22.9, 31.2) 26.3 (23.4, 29.8) 26.6 (23.5, 30.4) 26.8 (23.6, 30.8) 27.2 (23.8, 31.4) 27.4 (23.8, 31.8) 26.8 (23.6, 30.6) 26.9 (23.6, 31.2)

Systolic BP (mm/Hg) 137.9 (21.7) 135.7 (20.9) 140.2 (22.2) 137.3 (21.2) 138.2 (21.7) 138.0 (22.0) 137.8 (21.8) 138.2 (21.6) 138.8 (21.8) 136.7 (21.5)

Cholesterol (mmol/L) 4.5 (3.8, 5.4) 4.3 (3.6, 5.1) 4.8 (4.1, 5.7) 4.5 (3.8, 5.4) 4.5 (3.8, 5.4) 4.6 (3.8, 5.4) 4.5 (3.8, 5.4) 4.5 (3.7, 5.3) 4.6 (3.9, 5.5) 4.4 (3.7, 5.3)

Haemoglobin (g/dL) 13.0 (1.9) 13.4 (2.0) 12.5 (1.7) 13.0 (1.9) 13.0 (1.9) 12.9 (1.9) 12.9 (1.9) 13.0 (1.9) 13.1 (1.8) 12.7 (1.9)

eGFR (ml/min/m2) 57.5 (20.9) 60.3 (21.4) 54.5 (19.9) 57.2 (20.1) 56.9 (20.4) 57.2 (20.6) 57.8 (21.2) 59.0 (22.4) 59.2 (20.3) 55.4 (21.4)

Data are reported as number of patients (%) or mean with standard deviation or median with 25th and 75th centiles. Q, quintile; ACE, angiotensin-converting enzyme; ARB, angiotensin receptor blocker; AA, aldosterone antagonist (spironolactone or eplerenone)· IHD, ischaemic heart disease; MI, myocardial infarction; AF, atrial fibrillation; CKD, chronic kidney disease; COPD, chronic obstructive pulmonary disease; BMI, body mass index; BP, blood pressure; eGFR, estimated glomerular filtration rate

27

650

Table 2 Predicted rates of admissions during the first year following HF diagnosis by population groups and calendar year

Predicted rateper 100 person-years (95% CI)

Relative diff. (%)a

PInteractionb

Average annual percent change per year (95% CI)c

Slope change: annual percent change before and after trend change (95% CI)d

1998-2001 2012-2015 Before Break After

All admissions

All 97·1 (94·3, 99·9) 124·2 (120·9, 127·5) 28 1·8 (1·2, 2·4) 3·7 (2·4, 5·2) 2005 0·4 (-0·2, 1·1)

<70 years 94·1 (88·2, 100·0) 89·0 (83·9, 94·0) -5 Ref -0·2 (-1·2, 0·8) -0·2 (-1·2, 0·8) N/A -0·2 (-1·2, 0·8)

70 to 79 years 91·9 (87·4, 96·3) 107·8 (102·3, 113·3) 17 <0·0001 0·9 (-0·1, 2·0) 3·1 (0·8, 5·4) 2005 -0·6 (-1·7, 0·6)

80 to 89 years 98·4 (93·6, 103·2) 144·6 (138·3, 150·8) 47 <0·0001 3·0 (2·4, 3·7) 6·3 (4·5, 8·1) 2004 1·3 (0·8, 1·8)

90 years and over 102·0 (92·1, 111·9) 174·1 (161·6, 186·7) 71 <0·0001 4·4 (2·7, 6·2) 10·9 (5·8, 16·3) 2004 1·1 (-0·2, 2·3)

Men 109·9 (105·4, 114·4) 119·1 (114·8, 123·5) 8 Ref 0·5 (-0·1, 1·0) 0·5 (-0·1, 1·0) N/A 0·5 (-0·1, 1·0)

Women 85·7 (82·3, 89·2) 129·7 (124·7, 134·7) 51 <0·0001 3·0 (2·1, 3·9) 6·0 (4·3, 7·7) 2006 0·4 (-0·7, 1·5)

Most affluent (Q1) 85·9 (79·8, 92·1) 109·7 (103·1, 116·3) 28 Ref 1·5 (0·0, 3·1) 4·0 (0·0, 8·2) 2004 0·2 (-1·1, 1·5)

Most deprived (Q5) 113·2 (105·8, 120·6) 145·3 (135·6, 155·1) 28 0·103 1·9 (0·9, 2·9) 1·9 (0·9, 2·9) N/A 1·9 (0·9, 2·9)

Community diagnosis 80·0 (77·3, 82·7) 88·3 (84·9, 91·6) 10 Ref 0·7 (-0·2, 1·6) 4·6 (2·2, 6·9) 2004 -1·3 (-2·1, -0·5)

Hospital diagnosis 145·0 (137·6, 152·5) 163·2 (157·3, 169·0) 13 0·058 0·9 (0·3, 1·5) 0·9 (0·3, 1·5) N/A 0·9 (0·3, 1·5)

HF admissions

All 17·2 (16·2, 18·2) 22·1 (20·9, 23·2) 28 1·8 (0·7, 2·9) 0·8 (-0·1, 1·7) 2011 5·2 (0·7, 9·9)

<70 14·5 (12·6, 16·3) 14·1 (12·5, 15·7) -3 Ref -0·5 (-2·7, 1·9) -6·5 (-10·1, -2·8) 2015 5·3 (1·7, 8·9)

70 to 79 16·4 (14·9, 17·9) 18·1 (16·3, 20·0) 10 0·761 0·1 (-0·9, 1·0) 0·1 (-0·9, 1·0) N/A 0·1 (-0·9, 1·0)

80 to 89 18·8 (17·1, 20·6) 26·9 (24·7, 29·1) 43 <0·0001 2·4 (1·7, 3·2) 2·4 (1·7, 3·2) N/A 2·4 (1·7, 3·2)

>90 17·8 (14·4, 21·2) 35·2 (30·4, 39·9) 98 <0·0001 3·7 (1·4, 6·0) 3·7 (1·4, 6·0) N/A 3·7 (1·4, 6·0)

Men 20·3 (18·7, 21·9) 22·4 (20·8, 24·1) 10 Ref 0·6 (-0·9, 2·1) -1·9 (-3·4, -0·4) 2009 5·3 (1·5, 9·3)

Women 14·6 (13·4, 15·8) 21·6 (20·0, 23·3) 48 <0·0001 2·6 (1·9, 3·4) 2·6 (1·9, 3·4) N/A 2·6 (1·9, 3·4)

Most affluent 15·1 (13·0, 17·3) 19·1 (16·9, 21·3) 26 Ref 1·2 (-0·3, 2·8) 1·2 (-0·3, 2·8) N/A 1·2 (-0·3, 2·8)

28

Most deprived 20·5 (18·0, 23·1) 26·2 (22·9, 29·5) 28 0·597 1·6 (0·6, 2·6) 1·6 (0·6, 2·6) N/A 1·6 (0·6, 2·6)

Community diagnosis 11·3 (10·5, 12·1) 12·3 (11·3, 13·2) 9 Ref 0·2 9-0·9, 1·3) 0·2 9-0·9, 1·3) N/A 0·2 9-0·9, 1·3)

Hospital diagnosis 35·5 (32·2, 38·7) 32·9 (30·8, 35·1) -7 0·403 -1·6 (-3·7, 0·6) -8·5 (-14·4, -2·1) 2003 1·4 (-0·5, 3·4)

Other CVD admissions

All 20·3 (19·3, 21·3) 22·5 (21·4, 23·5) 11 0·7 (-0·4, 1·8) 5·3 (2·9, 7·9) 2005 -2·4 (-3·7, -1·1)

<70 23·0 (20·2, 25·9) 17·8 (15·6, 20·1) -23 Ref -1·7 (-2·6, -0·7) -1·7 (-2·6, -0·7) N/A -1·7 (-2·6, -0·7)

70 to 79 19·9 (18·3, 21·5) 21·7 (19·8, 23·6) 9 0·003 0·0 (-2·1, 2·2) 4·5 (-0·2, 9·3) 2005 -3·0 (-5·5, -0·4)

80 to 89 19·0 (17·3, 20·8) 24·7 (22·6, 26·8) 30 <0·0001 2·4 (1·2, 3·7) 10·3 (6·7, 14·1) 2004 -1·6 (-2·6, 0·6)

>90 17·1 (13·6, 20·6) 27·9 (23·7, 32·2) 63 <0·0001 3·3 (-0·1, 6·8) 9·5 (3·7, 15·7) 2007 -3·4 (-7·9, 1·4)

Men 22·6 (21·1, 24·2) 21·9 (20·5, 23·3) -3 Ref -0·1 (-1·8, 1·6) 5·5 (1·1, 10·1) 2004 -3·0 (-4·6, -1·4)

Women 18·2 (16·9, 19·4) 23·1 (21·6, 24·7) 27 <0·0001 1·8 (0·4, 3·2) 6·7 (3·6, 9·9) 2005 -1·5 (-3·0, -0·0)

ost affluent 18·6 (16·4, 20·9) 21·2 (19·0, 23·4) 14 Ref 0·8 (-0·8, 2·5) 7·2 (2·6, 12·0) 2004 -2·5 (-4·0, -0·9)

Most deprived 23·0 (20·5, 25·5) 23·6 (20·8, 26·3) 3 0·942 -0·4 (-2·8, 2·1) 2·9 (0·2, 5·7) 2008 -4·9 (-9·9, 0·3)

Community diagnosis 17·2 (16·2, 18·3) 18·4 (17·2, 19·7) 7 0·5 (-0·9, 1·9) 5·9 (2·9, 9·0) 2005 -3·2( -4·7, -1·5)

Hospital diagnosis 29·3 (26·7, 31·9) 27·1 (25·4, 28·8) -8 0·003 -0·5 (-1·7, 0·8) 3·2 (-0·0, 6·6) 2004 -2·5 (-3·7, -1·2)

Non CVD admissions

All 59·2 (57·2, 61·2) 83·9 (81·3, 86·5) 42 2·6 (1·9, 3·3) 4·2 (2·8, 5·6) 2006 1·2 (0·3, 2·1)

<70 57·1 (51·8, 62·5) 60·5 (55·0, 65·9) 6 Ref 0·6 (-0·9, 2·2) 0·6 (-0·9, 2·2) N/A 0·6 (-0·9, 2·2)

70 to 79 55·7 (52·5, 58·9) 72·1 (67·8, 76·5) 29 0·006 1·9 (0·9, 2·8) 3·7 (1·6, 5·9) 2005 0·6 (-0·4, 1·6)

80 to 89 60·0 (56·3, 63·6) 98·7 (93·4, 104·0) 65 <0·0001 3·4 (2·8, 4·0) 3·4 (2·8, 4·0) N/A 3·4 (2·8, 4·0)

>90 66·7 (58·5, 74·8) 114·7 (103·6, 125·7) 72 <0·0001 4·2 (2·8, 5·6) 8·3 (4·9, 11·8) 2005 1·4 (0·1, 2·8)

Men 66·3 (63·1, 69·6) 79·4 (76·1, 82·8) 20 Ref 1·4 (0·6, 2·1) 1·4 (0·6, 2·1) N/A 1·4 (0·6, 2·1)

Women 52·9 (50·4, 55·5) 88·9 (84·9, 92·8) 68 <0·0001 3·9 (2·8, 4·9) 7·1 (5·1, 9·3) 2006 1·0 (-0·2, 2·3)

Most affluent (Q1) 50·9 (46·6, 55·2) 73·4 (68·3, 78·6) 44 Ref 2·2 (1·3, 3·1) 2·2 (1·3, 3·1) N/A 2·2 (1·3, 3·1)

29

Most deprived (Q5) 70·0 (64·7, 75·4) 101·7 (93·9, 109·5) 45 0·167 2·8 (1·4, 4·3) 2·8 (1·4, 4·3) N/A 2·8 (1·4, 4·3)

Community diagnosis 51·6 (49·5, 53·6) 60·8 (58·1, 63·5) 18 Ref 1·2 (0·3, 2·2) 4·2 (1·6, 6·8) 2004 -0·4 (-1·2, 0·5)

Hospital diagnosis 80·4 (75·4, 85·3) 109·1 (104·5, 113·6) 36 <0·0001 2·4 (1·4, 3·30) 2·4 (1·4, 3·30) N/A 2·4 (1·4, 3·30)

With the exception of age groups, all predictions are at the mean population age (78 years). Follow up was until death or study end. Relative diff., relative difference; py, person-years; CI, confidence interval. a relative percentage difference in admission rates (per 100 person-years) between the first and second diagnosis time periods, calculated by 100*([time-period 2 – time period 1] / time-period 1].b P value for the difference in trend lines between groups. Estimated by fitting an interaction term between calendar year and group in the negative binomial models also containing age. Estimated by fitting an interaction term between calendar year and group in the Poisson models also containing age. As interaction tests have low power, p-values should be interpreted along with the graphical trends (Figure 1)c Average annual percentage change in rates (per 100 person-years) for each increasing year of diagnosisd Change in slope estimated using Joinpoint regression.

30

651

Table 3 Predicted rates of death following HF diagnosis by population groups and calendar year

Deaths in first month N (%) Predicted rate per 100 py (95% CI)Relative diff. (%)a

P interactionb

Average annual percent change per year (95% CI)c

Slope changeAnnual percent change before and after break (95% CI)d

1998-2001 2012-2015 1998-2001 2012-2015 Before Break After

Rates during the first year following heart failure diagnosis in those who survived the first month

All 2,397 (14) 1,967 (11) 24·5 (23·4, 25·7) 23·0 (22·0, 24·1) -6 -0·5 (-0·9, -0·2) -0·5 (-0·9, -0·2) N/A -0·5 (-0·9, -0·2)

<70 years 328 (10) 186 (5) 13·6 (12·1, 15·1) 8·9 (7·9, 10·0) -35 -3·6 (-4·9, -2·3) -3·6 (-4·9, -2·3) N/A -3·6 (-4·9, -2·3)

70 to 79 years 684 (11) 301 (7) 21·8 (20·2, 23·4) 15·8 (14·4, 17·3) -28 0·113 -2·3 (-3·0, -1·6) -2·3 (-3·0, -1·6) N/A -2·3 (-3·0, -1·6)

80 to 89 years 1,004 (16) 901 (13) 32·6 (30·4, 34·8) 34·1 (31·9, 36·3) 5 <0·0001 0·3 (-0·1, 0·8) 0·3 (-0·1, 0·8) N/A 0·3 (-0·1, 0·8)

90 years and over 381 (21) 579 (22) 58·1 (51·5, 64·7) 69·6 (63·2, 76·0) 20 <0·0001 1·2 (0·0, 2·4) 1·2 (0·0, 2·4) N/A 1·2 (0·0, 2·4)

Men 1,103 (13) 926 (10) 29·2 (27·4, 31·0) 23·5 (22·1, 24·9) -20 -1·4 (-2·3, -0·5) -2·6 (-3·4, -1·7) 2009 0·8 (-1·7, 3·3)

Women 1,294 (14) 1,041 (12) 20·7 (19·5, 22·0) 22·0 (20·7, 23·4) 6 <0·0001 0·3 (-0·5, 1·1) 0·3 (-0·5, 1·1) N/A 0·3 (-0·5, 1·1)

Most affluent (Q1) 375 (13) 367 (10) 23·8 (21·3, 26·2) 21·1 (19·1, 23·1) -11 -0·8 (-1·2, -0·4) -0·8 (-1·2, -0·4) N/A -0·8 (-1·2, -0·4)

Most deprived (Q5) 476 (14) 310 (11) 27·6 (24·8, 30·4) 25·7 (22·9, 28·5) -7 0·46 -0·2 (-1·2, 0·8) -0·2 (-1·2, 0·8) N/A -0·2 (-1·2, 0·8)

Community diagnosis 1,449 (11) 606 (7) 20·8 (19·7, 21·8) 14·2 (13·2, 15·1) -32 -2·7 (-3·2, -2·2) -2·7 (-3·2, -2·2) N/A -2·7 (-3·2, -2·2)

Hospital diagnosis 948 (19) 1,361 (15) 38·8 (35·8, 41·8) 34·3 (32·3, 36·2) -12 <0·0001 -1·1 (-1·8, -0·4) -1·1 (-1·8, -0·4) N/A -1·1 (-1·8, -0·4)

Rates after the first year following heart failure diagnosis in survivors of the first year

Deaths in first year N (%) Predicted rate per 100 py (95% CI)Relative diff· (%)a

P interactionb

Average annual percent change per year (95% CI)c

Slope changeAnnual percent change before and after break (95% CI)d

1998-2001 2012-2015 1998-2001 2012-2015 1998-2001 2012-2015 1998-2001

All 5,475 (31) 5,143 (29) 17·6 (17·3, 18·0) 14·7 (14·3, 15·2) -16 -1·3 (-1·6, -1·1) -1·3 (-1·6, -1·1) N/A -1·3 (-1·6, -1·1)

<70 years 682 (20) 479 (13) 8·0 (7·7, 8·4) 5·4 (4·9, 5·9) -33 -2·9 (-3·6, -2·1) -2·9 (-3·6, -2·1) N/A -2·9 (-3·6, -2·1)

70 to 79 years 1,638 (27) 872 (19) 14·6 (14·1, 15·0) 11·3 (10·5, 12·0) -23 0·001 -1·9 (-2·2, -1·6) -1·9 (-2·2, -1·6) N/A -1·9 (-2·2, -1·6)

80 to 89 years 2,274 (36) 2,363 (35) 23·7 (23·0, 24·5) 20·8 (19·8, 21·7) -12 <0·0001 -0·9 (-1·3, -0·6) -0·9 (-1·3, -0·6) N/A -0·9 (-1·3, -0·6)

90 years and over 881 (49) 1,429 (54) 38·3 (35·8, 40·8) 41·7 (38·8, 44·7) 9 <0·0001 0·4 (-0·1, 1·0) 0·4 (-0·1, 1·0 N/A 0·4 (-0·1, 1·0

31

Men 2,630 (31) 2,501 (27) 20·2 (19·6, 20·7) 15·1 (14·4, 15·8) -25 -2·2 (-2·5, -1·9) -2·2 (-2·5, -1·9) N/A -2·2 (-2·5, -1·9)

Women 2,845 (31) 2,642 (31) 15·8 (15·4, 16·2) 14·2 (13·5, 14·8) -10 <0·0001 -0·8 (-1·1, -0·5) -0·8 (-1·1, -0·5) N/A -0·8 (-1·1, -0·5)

Most affluent 908 (31) 1,000 (28) 16·2 (15·5, 16·9) 13·1 (12·2, 14·0) -19 -1·4 (-2·1, -0·7) -1·4 (-2·1, -0·7) N/A -1·4 (-2·1, -0·7)

Most deprived 1,056 (32) 782 (29) 20·2 (19·4, 21·1) 16·9 (15·5, 18·2) -16 0·600 -1·3 (-1·8, -0·9) -1·3 (-1·8, -0·9) N/A -1·3 (-1·8, -0·9)

Community diagnosis 3,452 (27) 1,656 (19) 16·7 (16·3, 17·1) 12·0 (11·5, 12·5) -28 -2·4 (-2·7, -2·1) -2·4 (-2·7, -2·1) N/A -2·4 (-2·7, -2·1)

Hospital diagnosis 2,023 (31) 3,487 (38) 21·4 (20·6, 22·2) 18·7 (17·9, 19·5) -13 <0·0001 -1·0 (-1·3, -0·7) -1·0 (-1·3, -0·7) N/A -1·0 (-1·3, -0·7)

Crude death rates are reported for the first month and first year following HF diagnosis as N (%); number and percentage. Death rates are reported for (i) the first year in people who survived the first month following diagnosisand (ii) for subsequent years in people who survived the first year following HF diagnosis. With the exception of age groups, all predictions are at the mean population age (78 years). Relative diff., relative difference; py, person-years; CI, confidence interval, Q; quintile.a relative percentage difference in death rates (per 100 person-years) between the first and second diagnosis calendar time periods, calculated by 100*([time-period 2 – time period 1] / time-period 1].b P value for the difference in trend lines between groups. Estimated by fitting an interaction term between calendar year and group in the Poisson models also containing age. As interaction tests have low power, p-values should be interpreted along with the graphical trends (Figure 2)c Average annual percentage change in rates (per 100 person-years) for each increasing year of diagnosisd Change in slope estimated using Joinpoint regression.

32

652

Table 4: Associations between population groups and outcomes

Hospital admissions Incidence rate ratio with 95% confidence intervals

Follow-up All CVD Non-CVDGroup Total

admissions Total py Median[IQR] Unadjusted Adjusteda Unadjusted Adjusteda Unadjusted Adjusteda

All 228,113 330,137 2·4 [0·5, 5·7]

<70 years 57,164 106,267 4·8 [ 1·6·, 9·2] 1·0 1·0 1·0 1·0 1·0 1·0

70 to 79 years 77,526 115,053 3·3 [0·9, 6·9] 1·31 (1·27, 1·34) 1·21 (1·18, 1·24) 1·29 (1·24, 1·34) 1·18 (1·13, 1·23) 1·31 (1·27, 1·35) 1·23 (1·20, 1·27)

80 to 89 years 76,565 91,903 1·8 [0·3, 4·2] 1·64 (1·60, 1·68) 1·47 (1·43, 1·51) 1·65 (1·59, 1·71) 1·45 (1·39, 1·52) 1·64 (1·59, 1·68) 1·49 (1·45, 1·54)

90 years and above 16,858 16,914 0·7 [0·1, 2·3] 1·94 (1·87, 2·01) 1·72 (1·65, 1·79) 1·89 (1·79, 1·99) 1·67 (1·57, 1·77) 1·95 (1·88, 2·03) 1·74 (1·67, 1·82)

Male (ref) 116,829 174,160 2·5 [0·5, 5·9] 1·0 1·0 1·0 1·0 1·0 1·0

Female 111,284 155,977 2·2 [0·4, 5·4] 1·05 (1·03,1·07) 0·91 (0·90,0·92) 0·97 (0·94, 1·00) 0·85 (0·83, 0·88) 1·09 (1·07,1·11) 0·90 (0·88,0·92)

Socioeconomic statusQ1 Most affluent (ref) 37,567 62,395 2·4 [0·5, 5·8] 1·0 1·0 1·0 1·0 1·0 1·0

Q2 50,412 76,813 2·4 [0·5, 5·7] 1·08 (1·05, 1·11) 1·05 (1·02, 1·08) 1·10 (1·05, 1·15) 1·06 (1·02, 1·11) 1·08 (1·04, 1·11) 1·05 (1·02, 1·08)

Q3 47,541 70,313 2·3 [0·4, 5·7] 1·13 (1·10, 1·17) 1·10 (1·07, 1·13) 1·14 (1·09, 1·19) 1·10 (1·05, 1·15) 1·13 (1·09, 1·17) 1·09 (1·06, 1·12)

Q4 48,043 65,182 2·4 [0·4, 5·6] 1·20 (1·16, 1·23) 1·14 (1·11, 1·18) 1·16 (1·11, 1·22) 1·12 (1·08, 1·17) 1·22 (1·18, 1·26) 1·16 (1·12, 1·19)

Q5 Most deprived 44,228 54,986 2·3 [0·4, 5·6] 1·34 (1·32,1·35) 1·20 (1·17,1·24) 1·23 (1·17, 1·29) 1·18 (1·13,1·19) 1·33 (1·28,1·37) 1·22 (1·21,1·25)

Community (ref) 135,050 234,304 3·3 [0·9, 6·9] 1·0 1·0 1·0 1·0 1·0 1·0

Hospital 93,063 95,833 1·3 [0·2, 3·8] 1·76 (1·73, 1·80) 1·52 (1·49, 1·55) 2·04 (1·98, 2·10) 1·77 (1·72, 1·82) 1·64 (1·63, 1·66) 1·40 (1·37, 1·43)

Deaths Hazard ratios with 95% confidence intervals

All CVD Non-CVDTotal deaths

N (%)Survival time

Mean (95% CI)Age at death

Median [IQR] Unadjusted Adjustedb Unadjusted Adjustedb Unadjusted Adjustedb

All 65,063 (74) 3·2 [3·1, 3·2] 84 [78, 90]

<70 years 8,859 (49) 9·0 [8·8, 9·3] 67 [62, 71] 1·0 1·0 1·0 1·0 1·0 1·0

70 to 79 years 18,517 (72) 4·5 [4·5, 4·6] 79 [76, 82] 1·88 (1·84, 1·93) 1·68 (1·64, 1·73) 1·77 (1·70, 1·83) 1·60 (1·54, 1·66) 2·00 (1·93, 2·08) 1·73 (1·66, 1·81)

80 to 89 years 27,568 (83) 2·2 [2·1, 2·2] 87 [84, 89] 3·27 (3·19, 3·36) 2·66 (2·58, 2·74) 3·15 (3·04, 3·26) 2·59 (2·49, 2·70) 3·33 (3·20, 3·46) 2·61 (2·49, 2·73)

90 years and above 10,119 (90·2) 0·8 [0·8, 0·9] 94 [ 92, 96] 5·59 (5·42, 5·76) 4·08 (3·93, 4·22) 5·30 (5·08, 5·52) 3·95 (3·76, 4·15) 5·84 (5·57, 6·12) 4·05 (3·82, 4·29)

Male (ref) 31,819 (71) 3·5 [3·4, 3·5] 82 [75, 88] 1·0 1·0 1·0 1·0 1·0 1·0

Female 33,244 (76) 2·9 [2·8, 2·9] 86 [80, 91] 1·15 (1·13, 1·17) 0·85 (0·83, 0·86) 1·11 (1·09, 1·13) 0·85 (0·83, 0·88) 1·21 (1·18, 1·24) 0·84 (0·82,0·87)

Socioeconomic statusQ1 Most affluent (ref) 11,645 (71) 3·4 [3·3, 3·5] 86 [80, 90] 1·0 1·0 1·0 1·0 1·0 1·0

33

Q2 15,093 (74) 3·2 [3·3, 3·5] 85 [79, 90] 1·05 (1·03, 1·08) 1·05 (1·03, 1·08) 1·07 (1·04, 1·11) 1·09 (1·05, 1·12) 1·02 (0·98, 1·06) 1·01 (0·97, 1·05)

Q3 14,065 (74) 3·1 [3·0, 3·2] 85 [89, 90] 1·07 (1·04, 1·10) 1·07 (1·05, 1·10) 1·11 (1·07, 1·15) 1·12 (1·09, 1·16) 1·03 (1·00, 1·08) 1·02 (0·98, 1·06)

Q4 12,971 (74) 3·1 [3·0, 3·2] 84 [77, 89] 1·06 (1·04, 1·09) 1·11 (1·09, 1·14) 1·06 (1·02, 1·10) 1·12 (1·09, 1·17) 1·06 () 1·02, 1·11) 1·10 (1·06, 1·14)

Q5 Most deprived 11,149 (75) 3·1 [2·9, 3·2] 82 [75, 88] 1·08 (1·05, 1·11) 1·17 (1·14, 1·21) 1·06 (1·02, 1·10) 1·18 (1·14, 1·23) 1·12 (1·07, 1·16) 1·17 (1·13, 1·22)

Community (ref) 37,734 (73) 4·2 [4·2, 4·3] 84 [78, 90] 1·0 1·0 1·0 1·0 1·0 1·0

Hospital 27,329 (75) 1·9 [1·8, 1·9] 85 [78, 90] 1·55 (1·53, 1·58) 1·46 (1·43, 1·48) 1·72 (1·69, 1·76) 1·55 (1·51, 1·59) 1·28 (1·25, 1·31) 1·37 (1·34, 1·40)

IRR; Incidence rate ratio, CI; Confidence interval, py; person-years, IQR; interquartile range, Q; quintilea adjusted for age gender socioeconomic status, ethnicity, place of diagnosis, calendar year, beta-blocker, angiotensin converting enzyme inhibitor, angiotensin receptor blocker; aldosterone antagonist, aspirin, loop diuretic, number of comorbidities, ischaemic heart disease, myocardial infarction, atrial fibrillation, hypertension, diabetes, stroke, anaemia, obesity, chronic kidney disease, chronic obstructive pulmonary disease, asthma, depression, osteoarthritis, cancer and dementia, smoking, alcohol, body mass index, systolic blood pressure, cholesterol, haemoglobin and estimated glomerular filtration rate.

34

653

Figure 1

35

654

655

Figure 2

36

656

657

Figure legends

Figure 1 Trends in predicted 1-year rates of cause-specific hospitalisations at mean age

of 78 years

Predicted admission rates at mean population age (78 years) per 100 person-years by cause

between 1998 and 2015. Follow up was until death or study end. Spikes indicate 95% CI. P-

values are for the test of difference in trends. As interaction tests have low power, p-values

should be interpreted along with the graphical trends.

Figure 2. First year mortality rates by population group and calendar year of diagnosis

Predicted mortality rates at mean population age (78 years) per 100 person-years between

1998 and 2015. Spikes indicate 95% CI. Rates were calculated in survivors of the first month

following HF diagnosis. P-values are for the test of difference in trends. As interaction tests

have low power, p-values should be interpreted along with the graphical trends.

37

658

659

660

661

662

663

664

665

666

667

668

669

670

671

672