Prolonged event-free survival in more Prolonged event-free survival in more complex cases of heart disease: outcome complex cases of heart disease: outcome

data from 1,226 patients from 3 randomised data from 1,226 patients from 3 randomised trials of nurse-led, multidisciplinary home-trials of nurse-led, multidisciplinary home-

based interventionbased intervention

S StewartS Stewart, JF Wiley, YK Chan, J Ball, DR Thompson & MJ , JF Wiley, YK Chan, J Ball, DR Thompson & MJ CarringtonCarrington

simon.stewart@acu.edu.au │http://mmihr.acu.edu.au

Acknowledgements

All trials were independently designed, funded (NHMRC of Australia) and conducted. A number of Investigators are also funded by the NHMRC of Australia

Trial Investigators:A/Prof. Melinda J Carrington & Dr Yih Kai Chan (ACU/Baker IDI)

Prof. John D Horowitz & Dr Gnanadevan Mahadevan (The Queen Elizabeth Hospital, South Australia)

Dr Chiew Wong (Western Hospital, Victoria)

Prof. Walter P Abhayaratna (Canberra Hospital, ACT)

Prof. Thomas H Marwick (Princess Alexander Hospital, Qld)

Prof. David R Thompson & Dr Jocasta Ball (Baker IDI/ACU)

Prof. Paul Scuffham (Griffith University, Queensland)

Prof. Garry Jennings (Baker IDI)

Prof. Peter MacDonald & Dr Phillip Newton (St Vincent’s Hospital, NSW)

Trial Statistician: Professor Adrian Esterman (Uni SA, SA)

Background & study hypotheses

Despite a wealth of evidence, the role of disease management across the full spectrum of heart disease remains unknown

Integrated program of trials of nurse-led, home-based intervention (HBI) recently completed

Prospectively tested the following hypothesis:HBI is superior to high levels of standard care in preventing recurrent hospitalisation and premature mortality overall Compared to high-level standard care, HBI is increasingly more effective as the clinical complexity (and potential to prevent poor health outcomes) increases

Composite analysis of a family of 3 trials

Stewart S & WHICH? Trial Investigators. Prolonged impact of home versus clinic-based management of chronic heart failure: Extended follow-up of a pragmatic, multicentre randomized trial cohort. International Journal of Cardiology. Jul 2014; 174(3):600-10. doi: 10.1016/j.ijcard.2014.04.164. Stewart S & SAFETY Investigators. Standard versus atrial fibrillation-specific management strategy (SAFETY) to reduce recurrent admission and prolong survival: pragmatic, multicentre, randomised controlled trial. The Lancet. 2015; 385(9970):775-784. Stewart S & NIL-CHF Study Investigators. Impact of a nurse-led home and clinic-based secondary prevention programme to prevent progressive cardiac dysfunction in high-risk individuals: the Nurse-led Intervention for Less Chronic Heart Failure (NIL-CHF) randomized controlled study. European Journal of Heart Failure. 2015 Apr 21. doi: 10.1002/ejhf.272.

Spectrum of heart disease/management

NIL-CHF Study – enrolled cardiac patients, mostly ACS & without CHF (echo confirmed)

SAFETY Trial – enrolled patients with chronic AF & without CHF (echo confirmed)

WHICH? Trial – enrolled patients with CHF (echo confirmed) with HFrEF & HFpEF

ALL cases recruited during acute hospitalisation, returning to home (metropolitan) & subject to high level standards of care

WHICH? TrialMulti-centre RCT

CONSORT compliant

1:1 blinded randomization (HrEF vs. HFpEF)

Standardized clinical managementIndependent data management/trial statistician

Blinded endpoint acquisition & adjudication

Follow-up: minimum 3 years

SAFETY TrialMulti-centre RCT

CONSORT compliant

1:1 blinded randomization (rate vs. rhythm control)Standardized clinical management

Independent data management/trial statistician

Blinded endpoint acquisition & adjudication

Follow-up: minimum 2 years

NIL-CHF StudySingle centre RCT

CONSORT compliant

1:1 blinded randomization

Standardized clinical management

Independent data management/trial statistician

Blinded endpoint acquisition & adjudication

Follow-up: minimum 3 years

CONSORT flow chart

Study Cohort   Home based Home based

interventioninterventionStandard Standard

managementmanagementMen

(n=394)

Women

(n=218)

Men

(n=415)

Women

(n=199)

Socio-demographic profileSocio-demographic profileAge (years)Age (years) 67±12 72±11 68±12 71±12

Living aloneLiving alone 152 (39%) 111 (51%) 147 (36%) 111(56%)

Risk factorsRisk factorsObese (BMI ≥ 30 kg/mObese (BMI ≥ 30 kg/m22)) 128 (35%) 80 (40%) 128 (35%) 65 (38%)

HypertensionHypertension 214 (56%) 124 (59%) 222 (58%) 105 (57%)

Current smokerCurrent smoker 87 (22%) 19 (9%) 73 (18%) 31 (16%)

Cholesterol ≥ 4 mmol/LCholesterol ≥ 4 mmol/L 125 (39%) 110 (64%) 131(42%) 87 (57%)

Clinical profileClinical profileLVEF (%)LVEF (%) 50±17 56±14 53±17 60±15

Type 2 diabetesType 2 diabetes 123 (31%) 57 (26%) 133 (32%) 53 (27%)

Renal failureRenal failure 100 (26%) 78 (37%) 122 (31%) 69 (37%)

Charlson IndexCharlson Index 5·0±2·9 5·2±2·5 5·1±2·7 5·0±2·5

Depressive symptomsDepressive symptoms 155 (40%) 89 (41%) 142 (35%) 77 (39%)

Clinical complexity scoreClinical complexity score 3·8±1·7 5·2±1·6 3·8±1·6 5·2±1·5

In-hospital managementIn-hospital managementMedian length of stayMedian length of stay 4·5 (2·0-8·0) 4·0 (2·0-8) 4·0 (2·0-8) 4·0 (2·0-8)

Coronary revascularisationCoronary revascularisation 96 (24%) 25 (11%) 110 (27%) 23 (12%)

Primary Discharge DiagnosisPrimary Discharge DiagnosisAcute coronary syndromeAcute coronary syndrome 81 (21%) 32 (15%) 84 (20%) 31 (16%)

Acute heart failure Acute heart failure 72 (18%) 37 (17%) 65 (16%) 30 (15%)

Stable CADStable CAD 63 (16%) 19 (9%) 57 (14%) 20 (10%)

Atrial FibrillationAtrial Fibrillation 59 (15%) 56 (26%) 66 (16%) 60 (30%)

Typically older patient cohort

~1/3 women (older)

Full-spectrum of heart disease

Multimorbidity & high clinical complexity

Appropriate levels of treatment

Well-matched for all baseline profiling

Clinical Complexity Score

Comprising a combination of clinical, functional and socio-demographic variables

Generalized linear model with multiple imputations & boot-strapping of baseline profiling10-15 key variables important in explaining days alive and out-of-hospital

Sensitivity analyses - clinical complexity score versus other tools (e.g. Charlson & MAGGIC) & mortality versus hospital stay

Recurrent hospital stayUnplanned Hospital Stay (All-Cause)

HBI accumulated 7469 days of hospital stay from 1336 unplanned hospitalisations Standard management group accumulated 10448 days from 1412 hospitalisations

  HBI(n=612)

Standard (n=614)

Median (IQR) Rate per 100 days follow-up

P

Total follow-up (days) 747,827 737,090    Median follow-up (days) 1321 (922 – 1571) 1295 (868 – 1587)   0·568

Unplanned admission days Median admissions/patient Median length of stay/patient

1336 (7649)1 (0 – 3)

3 (0 – 14)

1412 (10448)1 (0 – 3)

4 (0 – 18)

 0·08 (0 – 0·32) vs. 0·12 (0 – 0·35)0·22 (0 – 1·33) vs. 0·36 (0 – 2·10)

 0·0830·011

Cardiovascular admission days Median admissions/patient

Median length of stay/patient

735 (3334)0 (0 – 2)0 (0 – 6)

842 (4488)1 (0 – 3)1 (0 – 7)

 0 (0 – 0·16) vs. 0·07 (0 = 0·19)0 (0 – 0·55) vs. 0·11 (0 – 0·76)

 0·0470·039

All admissions (days in hospital) Median admissions/patient Median length of stay/patient

1868 (9975)2 (1 – 4)

6 (1 – 20)

2017 (13563)2 (1 – 5)7 (1 - 25)

 0·15 (0·05 – 0·42) vs. 0·18 (0·06 – 0·52)0·49 (0·07 – 2·10) vs. 0·67 (0·11 – 3·10)

 0·1170·017

All-cause mortalityAdjusted HR 0·56, 95% CI 0·41-0·78; p=0·001 for HBI versus standard management

Days alive & out-of-hospitalHBI achieved mean of 1210±463 days event-free (90·1%, 95% CI 88·2-92·0). Standard management achieved 1184±494 days event-free (87·2%,95% CI 85·1-89·3) - p=0·02.However……

At low clinical complexity HBI conferred worse event-free survivalAt high clinical complexity HBI conferred better event-free survivalSimilar pattern noted in relation to all-cause mortality/survival

Limitations

Pragmatic trials (non-blinding of participants)

No formal study power calculations

More detailed justification of clinical complexity score requiredMechanism(s) of effect need to be explored:

Why would HBI increase events/mortality at low complexity (clinical cascade)?

What are the benefits of HBI at increased complexity?

Post-hoc/composite analysis of individual trials

Historical context of clinical management

SummaryFirst reported individual trial analysis to examine benefits of HBI across full spectrum of heart diseaseOver long-term follow-up, HBI was associated with:

Significantly less hospital stay (~200 days/100 patients)

Significantly better survival (~5 deaths/100 patients)

Significantly prolonged days alive & out-of-hospital (~1100 days/100 patients)

Consistent across all 3 trials - HBI should be preserved for clinically complex cases to avoid harm

Pending confirmation, these data support careful application of HBI beyond a single cardiac diagnosis