making the case for cardiac rehabilitation: modelling potential impact on readmissions

NHSNHS Improvement

Heart

Making the case for cardiac rehabilitation: modelling potential impact on readmissions

HEART

LUNG

CANCER

DIAGNOSTICS

STROKE March 2013

Acknowledgements

Authors:

Michael Kaiser, Managing Director, MCK Healthcare Consultancy

Mel Varvel, National Improvement Lead, NHS Improvement - Heart

Patrick Doherty, Associate Clinical Lead for Cardiac Rehabilitation, NHS Improvement - Heart

Advice and Support:

Julie Harries, Director, NHS Improvement - Heart

Professor Sally Singh, Head of Cardiac and Pulmonary Rehabilitation, University Hospitals of Leicester

Tracey Ellison, Senior Data Analyst, National Cancer Services Analysis Team

Gary Shield, Senior Costing Analyst, National Institute for Health and Clinical Excellence

Michael Carpenter, Independent Consultant

National Audit of Cardiac Rehabilitation Team, University of York

Payments by Results Team, Department of Health

South East Public Health Observatory

Executive summary

Introduction

Methodology

Results

Conclusions and recommendations

Appendices

Appendix 1 – List of ‘in scope’ (eligible) patients

Appendix 2 – Assumptions

Appendix 3 – Frequently Asked Questions

References

Contents

4

5

6

8

11

12

12

18

19

21


3

Achieving an uptake rate for cardiacrehabilitation of 65% in England amongall eligible patients could release over£30 million per year in savings whichcould be reinvested in rehabilitation and re-ablement.

‘‘’’

For more than a decade NHS Improvement and its predecessors (the Coronary HeartDisease (CHD) Collaborative and the Heart Improvement Programme) have beenundertaking focused work with a wide range of partners aimed at increasing the accessto cardiac rehabilitation (CR) and improving the equity of provision and uptake of CRservices on the basis that it saves lives and improves quality of life.

This report summarises the findings of a short study, commissioned by NHS Improvement, whichmodels the relationship between uptake of CR and unplanned cardiac readmission rates bothnationally and at commissioner level.

The primary purpose of the study was to examine the Quality, Innovation, Productivity andPrevention (QIPP) potential of CR and to establish whether the benefits of CR outweigh the costs interms of the potential impact on readmissions alone.

Over and above the well-documented, positive effects of rehabilitation on mortality, morbidity andquality of life, the results suggest that increasing the uptake of ’gold standard’ CR has the potentialto reduce cardiac-related readmissions and deliver significant financial savings.

The outputs of this analysis and the approach adopted by it should be used by commissioners andproviders to make a cogent business case for CR in the new commissioning landscape and to assistwith the achievement of the outcomes specified in the NHS Outcomes Framework.

Executive summary


4

Coronary heart disease (CHD) is the most commoncause of premature death in the UK. It is also a leadingcause of disability. CHD accounted for 80,000 deathsin the UK in 2010 which is the equivalent to one in fivemen and one in eight women. Approximately 2.5million Britons are living with heart disease.

Cardiac rehabilitation (CR) is a professionallysupervised, menu-based programme incorporating anumber of components which, taken together, ensurethat people who have a cardiac condition, event ortreatment are able to recover and get back to‘everyday life’ as quickly and fully as possible.

The overall aim of CR is to provide all eligible patientswith a person-centred service, which optimises theirhealth and well-being, enhances their quality of lifeand minimises the risk of recurrent cardiac events.

Current national clinical guidelines and qualitystandards (NICE CG48, NICE CG94, NICE CG108 andNICE QS9) which recommend CR for specific cardiacconditions and treatments are based on a wealth ofresearch evidence demonstrating the positive outcomesof CR, including, but not limited to:

• A 26% relative reduction in cardiac mortality over five years according to an analysis of more than 48 randomised trials (Taylor et al, 2004);

• A reduction in cardiac-related morbidity; and• An improvement in functional capacity and quality of life.

Participation in CR has been shown to result in areduction in anxiety and depression and an increase inphysical activity and involvement in smoking cessationprogrammes. In addition, evidence suggests that CRprovides support for patients to return to work andfosters the development of self-management skills aswell as having a positive impact on overall health andwell-being.

Recent research studies by Davies et al (2010), Heran et al (2011) and Lam et al (2011) also suggest that thedelivery of a comprehensive CR service has thepotential to reduce unplanned cardiac readmissions by28-56%. This is supported by a significant and robustbut unpublished trial undertaken by a research groupat the University Hospitals of Leicester NHS Trust whichevidences a 30% reduction in readmissions with anuptake rate of 65%.

Introduction

Despite the significant benefits associated with CR,the National Audit of Cardiac Rehabilitation (NACR)2012 continues to highlight that uptake of CR isinequitable and poor – as low as 14% of eligiblepatients for some conditions in some regions. Theaverage uptake among the main treatment groups inEngland remains below 50% (NACR, 2012) and fallsfar short of the National Service Framework forCoronary Heart Disease (2000) target that more than85% of people discharged from hospital after acutemyocardial infarction or after coronaryrevascularisation should be offered CR.

In addition, variation also exists in:

• Readmission rates;• Costs of service;• Quality of services provided;• Quantity of services delivered;• Outcomes; and• Payment mechanisms.

There is no doubt that high quality CR saves lives andimproves the quality of life. Evidence also suggeststhat CR is cost-effective. However, making theeconomic case for such a complex service ischallenging.

There is increasing pressure on the NHS to continue todeliver quality whilst improving productivity andeliminating waste. In addition there is a newcommissioning landscape for the NHS. Therefore, it ismore important than ever to ensure that each poundspent on CR offers value for money and that the useof NHS resources is optimised.

With this in mind, NHS Improvement set out toestablish the costs and benefits of CR by modellingthe potential impact of increasing uptake onunplanned cardiac readmissions. The primary aim wasto enable commissioners to demonstrate the Quality,Innovation, Productivity and Prevention (QIPP)potential of good quality CR and to help make acompelling business case for local services.


5

The modelling was based closely on key findings fromthe aforementioned research studies and indicatesthat achieving a 65% uptake of ‘gold standard’ CR –which meets British Association for CardiovascularPrevention and Rehabilitation (BACPR) standards andincludes BACPR core components – would result in a30% reduction in unplanned cardiac readmissions.

The starting point of the analysis was therefore toestablish the number and cost of cardiac admissionsand emergency cardiac readmissions in a baselineyear. A list of the specific ‘in scope’ cardiac diagnosisand treatment codes included in the analysis,denoting eligibility for CR, was derived from theDepartment of Health (DH) Commissioning Pack forCardiac Rehabilitation and the National Institute forHealth and Clinical Excellence (NICE) CardiacRehabilitation Commissioning Guide (CMG40). Thislist can be found in Appendix 1.

Cardiac activity at commissioner1 level was obtainedvia Hospital Episode Statistics (HES) data and used tocalculate regional and national figures.

The HES data that formed the basis of the analysisincluded cardiac spells from 2009/10 and any cardiacspells for the same patients in the previous 12 monthsand the following 12 months.

Any spells with a cardiac admission in the 12 monthsprior to the 2009/10 admission were not regarded asa new spell and therefore not counted. This enabledcalculation of the readmission rates within a 12 monthperiod. Patients who died prior to discharge were alsosubtracted from the remaining new spells to result inthe cohort of patients eligible for CR.

According to the NICE Cardiac RehabilitationCommissioning Guide (CMG40), the indicative cost ofdelivering a good quality CR service is £477 perpatient2. Multiplying this figure by 65% of the eligiblecohort provided an understanding of the cost todeliver CR to 65% of eligible patients. The figure of65% was chosen on the basis of uptake achieved in

Methodology

one of the key trials underpinning the modelling.

Effectively representing the midway point betweenthe current national average and the target set out inthe National Service Framework, this was also deemedto be a reasonable, realistic and achievable uptakerate given the current levels.

For the purpose of this analysis, a ‘readmission’ wasdefined as any emergency cardiac admission in the 12 months following the 2009/10 ’new’ spell.Readmissions were divided into those which occurredwithin 30 days of discharge, and readmissions whichoccurred outside 30 days but within 12 months ofdischarge. This was in view of the current interest inpreventing readmissions within 30 days and thefinancial penalties for providers under Payment byResults. Readmission rates were then calculated atPrimary Care Trust (PCT), regional and national levels.

Readmissions were counted as a total and at a patientlevel in order to understand the number of patientsthat had readmitted in addition to the number ofreadmissions that had occurred.

The Department of Health’s Cardiac RehabilitationCommissioning Pack gives the average weighted costof a cardiac readmission as £3,637. Multiplying thiscost by the number of readmissions gave an indicationof the total cost of cardiac readmissions at PCT,regional and national levels.

Based on the evidence showing a potential reductionin the cardiac readmission rate of 30% as a result of arobust CR service being completed by 65% of eligiblepatients, a new readmission rate was calculated. Thismodelled readmission rate was then multiplied by thenumber of new spells to give a potential new numberof readmissions. This allowed the number of cardiacreadmissions saved and the associated financialsavings at PCT, regional and national levels to bemodelled.


1The modelling is based on analysis of 2009/10 Hospital Episode Statistics (HES) data is therefore split by Primary Care Trust (PCT) rather than Clinical Commissioning Group (CCG).

2Please note that the tariff is based on pay costs (including staff on costs) only. Full details are given in Appendix 2.

6


The model assumed the same ratio of <30 day: 30-365 day cardiac readmissions as the HES baselinefigures to highlight how the potential financial savingscould be split between organisations within a healthcommunity. It was understood that this would notgive a definitive result, but an indication of potentialsavings.

Due to the complexity of current commissioningarrangements, the difficulties associated withobtaining the true costs of CR at PCT level for theentire country and robust figures on current uptake,the model assumes a 0% uptake. However, it doesnot take into account the current cost or outcomes ofany existing CR service delivery.

A full list of assumptions relating to this methodologyis included in Appendix 2.

Further information on the methodology is included inthe form of some Frequently Asked Questions and canbe found in Appendix 3.

7

Based on a 0% uptake increasing to a 65% uptakerate for CR, the cost-benefit analysis shows a potentialreduction in emergency cardiac readmissions inEngland of 28,782 in the 12 months following theinitial cardiac admission.

In terms of financial impact, this equates to apotential saving of £30,646,085 nationally, taking intoconsideration the costs to deliver a complete CRservice to 65% of patients – a rate significantly higherthan the current figure.

Results

The tables below provide the key results of theanalysis undertaken.

Table 1 highlights the number of cardiac readmissionsoccurring within 12 months of the 2009/10admissions, the readmission rate that this equates toand the financial cost of these cardiac readmissions.


Table 1: Cardiac readmissions occurring within 12 months of the 2009/10 admissions, readmissionrate and financial cost of these cardiac readmissions by Strategic Health Authority region

Region

East Midlands

East of England

London

North East

North West

South Central

South East

South West

West Midlands

Yorkshire and the Humber

England

Cohort of patients eligiblefor CR

21,710

26,604

28,412

14,304

35,546

16,256

19,455

26,451

25,324

24,719

238,781

No. of cardiacreadmissions from2009/10 admissions

8,752

10,309

12,562

6,211

15,316

5,297

7,008

9,790

10,847

9,848

95,940

Readmissionrate in 2009/10

35%

34%

39%

38%

37%

28%

31%

32%

37%

35%

35%

Cost ofreadmissions2009/10

£31,831,024

£37,493,833

£45,687,994

£22,589,407

£55,704,292

£19,265,189

£25,488,096

£35,606,230

£39,450,539

£35,817,176

£348,933,780

Source: Hospital Episode Statistics, the NHS Information Centre for Health and Social Care. Analysis provided by the National Cancer Services Analysis Team (NatCanSAT) www.natcansat.nhs.uk

8


Table 2 shows the first results of the analysis. Thisincludes the number of cardiac readmissions thatpotentially would have occurred within 12 months ofthe 2009/10 admissions had a ‘gold standard’ CRprogramme been in place.

The table also shows the resulting, reducedreadmission rate together with the cost associatedwith the reduced number of readmissions, the cost ofdelivering the CR service and the total cost of theservice and the readmissions.

Table 2: Modelled cardiac readmissions within 12 months of the 2009/10 admissions, ‘new’readmission rate and associated costs assuming a ‘gold standard’ CR programme with 65% uptake had been in place

Region

East Midlands

East of England

London

North East

North West

South Central

South East

South West

West Midlands

Yorkshire andthe Humber

England

Cohort of patientseligiblefor CR

21,710

26,604

28,412

14,304

35,546

16,256

19,455

26,451

25,324

24,719

238,781

No. of cardiacreadmissions using newmodel

6,126

7,216

8,793

4,348

10,721

3,708

4,906

6,853

7,593

6,894

67,158

Readmissionrate usingnew model

24%

24%

27%

27%

26%

20%

22%

23%

26%

24%

24%

Total cost ofnew service(readmissions+ service cost)

£29,012,903

£34,494,253

£40,790,737

£20,247,540

£50,014,041

£18,525,805

£23,873,690

£33,125,494

£35,467,083

£32,736,149

£318,287,694


Cost ofreadmissionsusing newmodel

£22,281,717

£26,245,683

£31,981,596

£15,812,585

£38,993,004

£13,485,632

£17,841,667

£24,924,361

£27,615,377

£25,072,023

£244,253,645

Cost ofservice based onnew model(for 65%uptake)

£6,731,186

£8,248,570

£8,809,141

£4,434,955

£11,021,037

£5,040,173

£6,032,023

£8,201,133

£7,851,706

£7,664,126

£74,034,049

9

Finally, Table 3 calculates the difference betweentables 1 and 2 – effectively representing the resultingsavings. It highlights a potential reduction of 28,782readmissions and despite a cost of £74,034,049 todeliver a robust CR service in England; a saving of£30,646,086 could be achieved nationally.


Table 3: Modelled reduction in cardiac readmissions and associated financial savings as a result ofdelivering a ‘gold standard’ CR service to 65% of eligible patients

Region

East Midlands

East of England

London

North East

North West

South Central

South East

South West

West Midlands

Yorkshire and the Humber

England

Cohort of patientseligible forCR

21,710

26,604

28,412

14,304

35,546

16,256

19,455

26,451

25,324

24,719

238,781

Total numericreduction inreadmissions

2,626

3,093

3,769

1,863

4,595

1,589

2,102

2,937

3,254

2,954

28,782

Readmissionrate withnew model

24%

24%

27%

27%

26%

20%

22%

23%

26%

24%

24%

Financialsavings fromreadmissions

£9,549,307

£11,248,150

£13,706,398

£6,776,822

£16,711,288

£5,779,557

£7,646,429

£10,681,869

£11,835,162

£10,745,153

£104,680,135


Total financialsavings fromnew serviceincludingreadmissionsavings

£2,818,121

£2,999,580

£4,897,257

£2,341,867

£5,690,251

£739,384

£1,614,406

£2,480,736

£3,983,456

£3,081,027

£30,646,086

10


Overall the results of the modelling demonstrate thatthe benefits of providing good quality CR outweighthe costs as stated in the NICE Commissioning Guidefor CR (CMG40). The analysis adds to existingevidence that CR saves lives and reduces readmissionsby indicating that it also has the potential to savemoney. This should help to ensure that CR is a priorityin the new commissioning landscape.

Aside from the results, this analysis sets a precedentfor establishing the costs and benefits of CR bymodelling the potential impact of improving qualityand increasing uptake on unplanned cardiacreadmissions. In doing so it supports the increasingfocus on outcomes-based commissioning.

Taking into account the various caveats andassumptions, it should assist commissioners todemonstrate the Quality, Innovation, Productivity andPrevention (QIPP) potential of good quality CR andmake a compelling business case for local services. Itwill also help to demonstrate the contribution thatgood quality CR has to the achievement of theoutcomes specified in the NHS Outcomes Framework.

The results of the analysis effectively represent asignificant new evidence base and provide greaterimpetus to the need to increase the quality of CRservices and to further improve uptake across alleligible cardiac patients.

In addition, the analysis points to further savings thatcould be made by considering the current costs todeliver a CR service at a local level. Further qualityimprovements and financial gains might achieved byredesigning existing services and/or exploringinnovative ways of delivering rehabilitation, forexample web-based CR, or ‘generic’ rehabilitation inwhich service provision is targeted at commondisability rather than the primary organ disease andcombines both generalist and specialist input asappropriate. CR could also be used as a primaryprevention intervention for those at high risk ofcardiovascular disease (CVD).

Conclusions and recommendations

Providers in particular may also be interested inexploring the potential impact of early rehabilitationon readmissions within 30 days of discharge in orderto avoid financial penalties under Payment by Results(PbR). The advent of a post-discharge tariff for CR inPbR in 2012-13 reflects emerging evidence whichpoints to the role that CR may have in avoidingreadmissions in the first few weeks followingdischarge. The results of this analysis suggest that25% of cardiac readmissions currently occur withinthe first 30 days following discharge nationally,although this varies considerably across PCTs. It maytherefore pay dividends for providers andcommissioners to work together to explore earlyrehabilitation interventions which may impactfavourably on patient outcomes and use of resourcesin the first 30 days after discharge from hospital.

As evidence suggests that CR may impact on all causemortality and morbidity, work could also beundertaken to understand the impact that theimproved delivery of CR has on non-cardiac pathwayssuch-as stroke and chronic obstructive pulmonarydisorder (COPD). In addition, further analysis could becarried out to understand the potential effect ofimproved CR on demand in Accident and Emergencydepartments across the country given the reducednumber of emergency readmissions implied by themodelling.

11

In order to determine what comprised a ‘cardiac admission’ or ‘cardiac readmission’, the following codes wereused. These are codes referred to in the Department of Health’s Cardiac Rehabilitation Commissioning Pack.

Appendix 1 - List of ‘in scope’ (eligible) patients


Diagnosis Codes:

ICD10 Code

I 110

I 130

I 132

I 255

I 420

I 429

I 500

I 501

I 509

I 515

I 21

I 210

I 211

I 212

I 213

I 214

I 219

I 22

I 220

I 221

I 228

I 229

Group

Chronic HF

Chronic HF

Chronic HF

Chronic HF

Chronic HF

Chronic HF

Chronic HF

Chronic HF

Chronic HF

Chronic HF

MI

MI

MI

MI

MI

MI

MI

MI

MI

MI

MI

MI

Description

Hypertensive heart disease with (congestive) heart failure

Hypertens heart and renal dis with (conges) heart failure

Hyper heart and renal dis both (cong) heart and renal failure

Ischaemic cardiomyopathy

Dilated cardiomyopathy

Cardiomyopathy, unspecified

Congestive heart failure

Left ventricular failure

Heart failure, unspecified

Myocardial degeneration

Acute myocardial infarction, unspecified

Acute transmural myocardial infarction of anterior wall

Acute transmural myocardial infarction of inferior wall

Acute transmural myocardial infarction of other sites

Acute transmural myocardial infarction of unspecified site

Acute subendocardial myocardial infarction

Acute myocardial infarction, unspecified

Subsequent myocardial infarction of unspecified site

Subsequent myocardial infarction of anterior wall

Subsequent myocardial infarction of inferior wall

Subsequent myocardial infarction of other sites

Subsequent myocardial infarction of unspecified site

12


OPCS Code

K401

K402

K403

K404

K408

K409

K411

K412

K413

K414

K418

K419

K421

K422

K423

Group

CABG

CABG

CABG

CABG

CABG

CABG

CABG

CABG

CABG

CABG

CABG

CABG

CABG

CABG

CABG

Description - Level 1

Saphenous Vein Graft Replacement of Coronary Artery






Other Autograft Replacement ofCoronary Artery






Allograft Replacement of CoronaryArtery



Procedure Codes:


Saphenous Vein Graft Replacementof One Coronary Artery

Saphenous Vein Graft Replacementof Two Coronary Arteries

Saphenous Vein Graft Replacementof Three Coronary Arteries

Saphenous Vein Graft Replacementof Four or More Coronary Arteries

Other Specified

Unspecified

Autograft Replacement of OneCoronary Artery NEC

Autograft Replacement of TwoCoronary Arteries NEC

Autograft Replacement of ThreeCoronary Arteries NEC

Autograft Replacement of Four orMore Coronary Arteries NEC

Other Specified

Unspecified

Allograft Replacement of OneCoronary Artery

Allograft Replacement of TwoCoronary Arteries

Allograft Replacement of ThreeCoronary Arteries

13


OPCS Code

K424

K428

K429

K431

K432

K433

K434

K438

K439

K441

K442

K448

K449

K451

K452

Group

CABG

CABG

CABG

CABG

CABG

CABG

CABG

CABG

CABG

CABG

CABG

CABG

CABG

CABG

CABG





Prosthetic Replacement of CoronaryArtery






Other Replacement of CoronaryArtery




Connection of Thoracic Artery ToCoronary Artery

Connection of Thoracic Artery ToCoronary Artery

Procedure Codes:


Allograft Replacement of Four orMore Coronary Arteries

Other Specified

Unspecified

Prosthetic Replacement of OneCoronary Artery

Prosthetic Replacement of TwoCoronary Arteries

Prosthetic Replacement of ThreeCoronary Arteries

Prosthetic Replacement of Four orMore Coronary Arteries

Other Specified

Unspecified

Replacement of Coronary ArteriesUsing Multiple Methods

Revision of Replacement ofCoronary Artery

Other Specified

Unspecified

Double Anastomosis of MammaryArteries To Coronary Arteries

Double Anastomosis of ThoracicArteries To Coronary Arteries

14


OPCS Code

K453

K454

K455

K456

K458

K459

K461

K462

K463

K464

K465

K468

K469

K491

K492

Group

CABG

CABG

CABG

CABG

CABG

CABG

CABG

CABG

CABG

CABG

CABG

CABG

CABG

PCI

PCI


Connection of Thoracic ArteryTo Coronary Artery






Other Bypass of CoronaryArtery







Transluminal BalloonAngioplasty of Coronary Artery

Transluminal BalloonAngioplasty of Coronary Artery

Procedure Codes:


Anastomosis of Mammary Artery To LeftAnterior Descending Coronary Artery

Anastomosis of Mammary Artery ToCoronary Artery NEC

Anastomosis of Thoracic Artery ToCoronary Artery NEC

Revision of Connection of Thoracic ArteryTo Coronary Artery

Other Specified

Unspecified

Double Implantation of MammaryArteries Into Heart

Double Implantation of Thoracic ArteriesInto Heart NEC

Implantation of Mammary Artery IntoHeart NEC

Implantation of Thoracic Artery Into HeartNEC

Revision of Implantation of ThoracicArtery Into Heart

Other Specified

Unspecified

Percutaneous Transluminal BalloonAngioplasty of One Coronary

Percutaneous Transluminal BalloonAngioplasty of Multiple Co

15


OPCS Code

K493

K494

K498

K499

K501

K502

K503

K504

K508

K509

K751

K752

K753

K754

K758

Group

PCI

PCI

PCI

PCI

PCI

PCI

PCI

PCI

PCI

PCI

PCI

PCI

PCI

PCI

PCI


Transluminal Balloon Angioplasty ofCoronary Artery




Other Therapeutic TransluminalOperations/Coronary Arte



Other Therapeutic TransluminalOperations on Coronary



Percutaneous Translum Balloon Angioplasty And Stenting

Percutaneous Translum BalloonAngioplasty And Stenting



Stenting

Procedure Codes:


Percutaneous Transluminal BalloonAngioplasty of Bypass Graf

Percutaneous Translum CuttingBalloon Angioplasty of Coronar

Other Specified

Unspecified

Percutaneous Transluminal LaserCoronary Angioplasty

Percutaneous Transluminal CoronaryThrombolysis Using Strept

Percutaneous Transluminal Injectionof Therapeutic Substance

Percutaneous TransluminalAtherectomy of Coronary Artery

Other Specified

Unspecified

Percutaneous Translum BalloonAngioplasty Ins 1-2 Drug

Percutaneous Translum BalloonAngioplasty Ins 3 Or More Drug

Percutaneous Translum BalloonAngioplasty Ins 1-2 Stents

Percutaneous Translum BalloonAngioplasty Ins 3 Or More Stents

Other Specified

16

Cardiac rehabilitation: a QIPP analysis

OPCS Code

K759

K591

K592

K596

K598

K599

Group

PCI

ICD

ICD

ICD

ICD

ICD



Cardioverter DefibrillatorIntroduced Through the Vein





Procedure Codes:


Unspecified

Implantation of Cardioverter DefibrillatorUsing One Electro

Implantation of Cardioverter DefibrillatorUsing Two Electro

Implantation of cardioverter defibrillatorusing three electrode leads

Other Specified

Unspecified

17

Data Sources:• All HES data has been provided by the National Cancer Services Analysis Team (NatCanSAT) courtesyof the NHS Information Centre for Health and Social Care.

• The model uses the CR cost from the NICE Cardiac Rehabilitation Commissioning Guide (CMG 40) at £477 per patient. This cost includes pay and staff- related on-costs. However, it does not include any capital-related costs. Please refer to the NICE Commissioning Guide for specific assumptions associated with this cost calculation.

• The cost of the CR service (as per the NICE Commissioning Guide) does not take into account London weighting.

• The model uses the average cardiac readmission cost from the Department of Health’s Commissioning Pack for Cardiac Rehabilitation (£3,637 per patient). Please refer to the DH Commissioning Pack for specific assumptions associated with this cost calculation.

• All reference data was extracted from Hospital Episode Statistics admitted patient care records - inpatient or day case episodes between 1st April 2008 and 31st December 2011 (inclusive).

Additional Savings:• The current service cost for CR has not been calculated for each PCT or region. Instead, the model refers to a 0% uptake rate and is modelled on a CR service based on the NICE CMG40 cost per patient. The current cost of CR represents an additional saving.

• The model looks at the impact of commissioning a new CR service for 65% of the eligible patients in place of any existing service.

• Non-cardiac readmissions as a result of cardiac primary admission have not been included in any of the modelling and could be regarded as potential additional savings.

Data Information:• Spells have been classified by year and by the year of the last episode in the spell. Only 2009/2010 spells have been analysed.

• Only patients resident in England are included in theanalysis. Patients living outside England, attending an English trust were therefore excluded.

• Spells were included if any episode in that spell contained a relevant cardiac diagnosis/procedure in any diagnosis/procedure position; spells were counted once only.

Appendix 2 - Assumptions

• All admission methods are included in the original cardiac admission data (i.e. elective, emergency and transfers).

• The Elective Admission Method Group for the original cardiac admissions includes both elective admissions and transfers.

• Spells include all patient classifications. Examples include, but are not limited to day case episodes, ordinary admission, inpatients and regular day attendees.

• Patients who had a previous admission (for a relevant procedure/diagnosis) within the previous year were identified as subsequent spells.

• 'New Spells' were those spells who did not have a previous admission within the last year (of any admission method).

• Spells were identified as being "Discharged Alive" ifthe discharge method for that spell was not 'Died' or 'Baby Still Born'.

• The 'Cohort' patients were those 'new spells' (i.e. that did not have a previous relevant admission within one year) and where the spell was dischargedalive.

• The 'Cohort' patients were checked to see if they had a subsequent emergency readmission (for a relevant procedure/diagnosis) within one year following discharge

• The type of spell - i.e. 'New', 'Subsequent', and ‘Cohort’ was determined by analysing all relevant codes together (not by individual groupings).

Additional Information:• CR was not coded as a separate service in 2009/10, nor did it have its own tariff.

• A split in health outcomes and associated finances between provider and commissioner has been madeassuming even impacts on readmission rates. This does not affect the total figures stated for health outcomes or associated finances.

A second model was undertaken looking at theimpact of commissioning a new CR service foradditional eligible patients to result in a 65% uptakerate.

Further information can be obtained by contacting NHS Improvement - email:[email protected]


18


1. What is the definition of ‘uptake’ used in themodelling?‘Uptake’ in the modelling is synonymous with‘completion’ of cardiac rehabilitation (CR) as this is thedefinition used by existing national guidance and thetrigger for payment under the Payment by Results(PbR) post-discharge tariff for CR. The readmissionsstudy conducted at the University Hospitals Leicester(unpublished) compared eligible patients who did nottake up the offer of CR, those who had an initialassessment but did not attend CR, and those whoaccepted the offer and completed the course. Thestudy concluded that if 65% of eligible patientscompleted a CR programme (i.e. 65% uptake), areduction of 30% in cardiac emergency re-admissionscould be achieved. NB. The National Audit of CardiacRehabilitation (NACR) defines uptake by the presenceof a Phase 3 start date on the audit database.

2. Why have you chosen to model 65% uptake?The figure of 65% was chosen as a conservative butrealistic uptake ‘target’. It is above the currentnational average (from NACR) but below theestimates on the take-up and referral of CR providedby the National Institute for Health and ClinicalExcellence (NICE) commissioning guide for CR(CMG40) topic advisory group (which were based onbest practice and assumed to be the proportions thatcould be achieved given optimal service design). Theunpublished readmissions study from UniversityHospitals Leicester on which the modelling is (in part)based demonstrated that a 65% uptake amongeligible patients (c. 3000) is achievable.

3. What evidence is there to suggest that CR canlead to a 30% reduction in cardiac readmissions?The figure of 30% is the estimated reduction in therisk of being admitted to hospital per heart failurepatient as a result of rehabilitation used in the NICEcommissioning and benchmarking tool for CR services(CMG40, 2011). This reduction is supported by recentMeta-analyses- see Heran et al (2011) in TheCochrane Database of Systematic Reviews 2011, Issue7 and Davies et al (2010) in The Cochrane Databaseof Systematic Reviews 2010, Issue 4. Lam et al (2011)also found a significant reduction in 60-day hospitalreadmissions after an acute myocardial infarction. A‘real life’ readmissions reduction of equivalentmagnitude was achieved in an unpublished studyconducted at the University Hospitals Leicester (basedon 2008/09 data). Earlier research studies

demonstrating the positive impact of cardiacrehabilitation on readmissions are cited in The Casefor Change, part of the DH Commissioning Pack forCR and Chapter 1 of the NICE commissioning guidefor CR services (CMG40, 2011). Taken together,existing evidence points to a potential reduction inemergency cardiac readmission rates of up to 30%following participation in comprehensive, high qualityCR programmes, compared with patients who declinethe offer.

4. Where does the figure on the average cost ofa cardiac readmission come from and how has itbeen calculated?The cost of a readmission (total weighted average cost£3,637) was taken from the cost benefit analysis inthe Department of Health (DH) Commissioning Packfor Cardiac Rehabilitation (2011). Costs per patient inthe DH model were calculated using Reference costs07/08 activity and 2010/11 Mandatory Tariff costs. Fulldetails of the calculation can be found in the DHCommissioning Pack cost-benefit tool.

5. Where does the figure on the per-patient costof CR come from and how has it been calculated?Does it include non-pay costs?The per patient cost of £477 used in the modellingwas taken from the Department of HealthCommissioning Pack for Cardiac Rehabilitation (2011)and the NICE CMG40 commissioning andbenchmarking tool. According to both sources, thisevidence-based cost is intended to cover the wholepathway of care for CR (stages 1-6) detailed in theCommissioning Pack. The costing was used as thebasis for the post-discharge tariff for CR under PbR.Details of how the per-patient cost was calculated isincluded in both DH and NICE publications. It isimportant to note that this per-patient cost relates tostaff costs only and was based on specific staff bands(to meet British Association of Cardiac Rehabilitation(BACR 2007) standards). Both the Department ofHealth (DH) Commissioning Pack and NICE CMG40emphasise that providers and commissioners shouldagree any additional non-recurrent costs to reflectlocal circumstances. The DH and NICE costing modelsrepresent the activities required to run a genericrehabilitation programme supporting low to mediumrisk patients. They stress that any variation in patientrequirements, for example an increased staff topatient ratio to cater for ‘high risk’ patients, must bereflected appropriately in local costings.

Appendix 3 - Frequently Asked Questions

19

6. Does the per-patient cost of CR used in themodelling take London weighting into account?As indicated in the DH Commissioning Pack cost-benefit tool, the calculated costs are England-widebased on mean basic NHS salary costs and excludeany regional variations. Modelled savings also excludeany regional variation. Any locally specific costs wouldneed to be agreed and included in the pathwaycosting to ensure the full cost reflects local needs andrequirements.

7. Which patients are deemed by the modellingto be ‘eligible’ for CR and included incalculations?The eligibility criteria employed in the modelling wasbased on the data specification in the NICE CMG40commissioning and benchmarking tool and reflectsthe ‘in scope’ patients listed in the DH CommissioningPack, i.e. patients with acute coronary syndromes,patients having undergone revascularisation andpatients with heart failure. See Appendix 1 of themodelling report for further details including therelevant treatment/ diagnosis codes.

8. Why does the modelling not factor in currentuptake (from NACR) and the current cost of CRfor each PCT/ region?At present, the National Audit of CardiacRehabilitation (NACR) reports uptake at national andregional (Strategic Health Authority) level. Evidencesuggests that the cost of CR services variesenormously across the country but there is currentlyno central mechanism to ascertain costs at PCT level.Availability of data on cost may improve in future withthe advent of a post-discharge tariff for CR under PbRalongside other tariff developments. Commissionersand providers may wish to explore specific local costsand uptake in more detail and factor these into futuremodelling.

9. The model assumes a ‘0% baseline’, i.e. noservice in place currently, therefore no cost andno impact on cardiac readmissions. However, wedo have a service in place locally which may beaffecting the ‘current’ cardiac admissions/readmissions rate. This means that the savings interms of readmissions may not be as great as themodelling implies.The National Audit of CR suggests that many existingservices have low uptake, particularly among thecohort of patients in which readmissions are the

highest, e.g. heart failure, and the quality of servicesin terms of clinical outcomes is variable. Moreover,evidence suggests that few services are routinelymeasuring the potential impact of CR onreadmissions. From a national perspective, the impactof current services on readmissions is thereforedifficult to quantify with any degree of accuracy. Anypotential diminution in modelled savings due to theabsence of a true ‘0%’ baseline for readmissions atPCT level may be offset by the cost of any existing CRservice, and may therefore be viewed as an‘additional’ saving.

10. Does the analysis include the cost/ savingsassociated with non-cardiac readmissions?Evidence suggests that a proportion of readmissionsare for a different health issue than that responsiblefor the primary admission. However, non-cardiacreadmissions following a primary cardiac admissionhave not been included in the modelling as it isdifficult to identify with any degree of accuracy if theprimary admission and readmission are causally linked.It is reasonable to assume that a proportion of non-cardiac readmissions are related to the original cardiacadmission and that these related non-cardiacreadmissions might be avoided with the interventionof CR. Avoiding these non-cardiac readmissions wouldeffectively represent additional savings.

11. Can I look at the data in more detail?NHS Improvement is happy to discuss the results ingreater detail on request but access to the HospitalEpisode Statistics (HES) data underpinning themodelling is restricted to those who have been givenprior access and who have signed a non-disclosureagreement. Please address any queries/ requests [email protected] in the firstinstance.


20


British Heart Foundation National Statistics. 2012;G608; http://www.bhf.org.uk/publications/view-publication.aspx?ps=1002097%20 (accessed 26.01.2013)

BACPR SCIENTIFIC STATEMENT: BRITISH STANDARDS AND CORE COMPONENTS FOR CARDIOVASCULARDISEASE PREVENTION AND REHABILITATION A Scientific Statement from the British Association forCardiovascular Prevention and Rehabilitation (BACPR). 2012;http://www.bacpr.com/resources/46C_BACPR_Standards_and_Core_Components_2012.pdf

Davies EJ, Moxham T, Rees K, Singh S, et al. Exercise based rehabilitation for heart failure. Cochrane Database ofSystematic Reviews. 2010; Issue 4:Art. No. CD003331. DOI:10.1002/14651858.CD003331.pub3

Department of Health (DH). Cardiac Rehabilitation Commissioning Pack. 2010;http://www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAndGuidance/Browsable/DH_117504 (accessed 26.01.2013).

Heran BS, Chen JM, Ebrahim S, et al. Exercise based cardiac rehabilitation for coronary heart disease. CochraneDatabase Syst Rev. 2011;(7):CD001800.

Lam G, Snow R, Shaffer L, La Londe M, et al. The effect of a comprehensive cardiac rehabilitation program on60-day hospital readmissions after an acute myocardial infarction. J Am Coll Cardiol. 2011;57(14s1):E597-E597.DOI: 10.1016/S0735-1097(11)60597-4

National Audit of Cardiac Rehabilitation (NACR). Annual Statistical Report. Vol. 5. British Heart Foundation,2011;http://www.cardiacrehabilitation.org.uk/nacr/index.htm (accessed 26.01.2013).

National Institute for Health and Clinical Excellence (NICE). Secondary Prevention in Primary and Secondary Carefor Patients Following a Myocardial Infarction. NICE. 2007;CG48.http://guidance.nice.org.uk/CG48/NICEGuidance/pdf/English (accessed 26.01.2013).

National Institute for Health and Clinical Excellence (NICE). Unstable angina and NSTEMI: The early managementof unstable angina and non-ST-segment-elevation myocardial infarction. NICE, 2010;CG94.http://guidance.nice.org.uk/CG94/NICEGuidance/pdf/English (accessed 26.01.2013).

National Institute for Health and Clinical Excellence (NICE). (CMG40) Commissioning Guide for CardiacRehabilitation. NICE 2011; http://publications.nice.org.uk/cardiac-rehabilitation-services-cmg40 (accessed26.01.2013).

National Institute for Health and Clinical Excellence (NICE). Chronic Heart Failure. NICE (QS9)2011;http://guidance.nice.org.uk/QS9/PublicInfo/pdf/English (accessed 26.01.2013).

Singh et al, (2010) The effect of a comprehensive cardiac rehabilitation programme on 12 month hospitalreadmissions, (unpublished)

Taylor RS, Brown A, Ebrahim S, Jolliffe J, et al. Exercise-based rehabilitation for patients with coronary heartdisease: systematic review and meta-analysis of randomized controlled trials. Am J Med 2004;116:682-92

References

21


22

HEART

LUNG

CANCER

DIAGNOSTICS

STROKE

NHSNHS Improvement

©NHS Im

prov

emen

t 20

11 |All Righ

ts Res

erve

d Pu

blication Re

f: IM

P/he

art002

- Jan

uary 201

3

NHS Improvement

NHS Improvement has over 12 years of experience in clinical patient pathway redesign in

cancer, diagnostics, heart, lung and stroke. Its specialist teams, including NHS Improvement -

Heart, work with partners across 250 NHS sites, 2,400 GP practices, clinical networks,

professional bodies and national charities. They undertake some of the most leading edge

improvement work in England supporting the patient experience and outcomes. A number of

NHS Improvement workstreams will be transferring to NHS Improving Quality - the new,

single improvement body for the NHS which comes into being on 1 April 2013.

NHS Improvement

3rd Floor | St John’s House | East Street | Leicester | LE1 6NB

Telephone: 0116 222 5184 | Fax: 0116 222 5101

www.improvement.nhs.uk

Delivering tomorrow’simprovement agenda for the NHS

Follow us on:

making the case for cardiac rehabilitation: modelling potential impact on readmissions

Health & Medicine