NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Collaborating Centre for Chronic Conditions (UK). Chronic Heart Failure: National Clinical Guideline for Diagnosis and Management in Primary and Secondary Care. London: Royal College of Physicians (UK); 2003. (NICE Clinical Guidelines, No. 5.)

  • This publication is provided for historical reference only and the information may be out of date.

This publication is provided for historical reference only and the information may be out of date.

Cover of Chronic Heart Failure

Chronic Heart Failure: National Clinical Guideline for Diagnosis and Management in Primary and Secondary Care.

Show details

Appendix EHealth economics of invasive procedures

The diagnosis of heart failure

The systematic review revealed virtually no economic information in this area. Invasive procedures are rarely used in diagnosis in the UK – in contrast with the US.

One paper was identified with a UK setting.295 This compares echocardiography v. cardiac catheterisation (CC) in the evaluation of patients with valvular heart disease. The findings may be relevant for the approx 10% of new patients with HF, with valve disease as an underlying cause.

The paper is relatively old and is of poor quality. It is not a formal economic evaluation.

Cardiac catheterisation is most commonly used in the UK to provide further information on aetiology, prior to surgery; hence the assumption of equivalent outcomes from echo and CC would not seem to be valid.

Table E1Summary of Channer and Robertson (1991)

Cardiac catheterisation:Echocardiography:

is invasive

associated with a number of potential adverse effects

these may be a particular problem in patients with HF

involves exposure to X-rays with associated dangers

imaging quality assumed to be equivalent

no discussion of heart failure with respect to specific indications

Conclusion: ‘CC should be reserved for patients in whom details of coronary anatomy are necessary in planning surgery, and in cases where echocardiography fails, technically, to provide the quality of data necessary to make the clinical decision’.

is non-invasive

uses ultrasound, which has no known deleterious biological effects.

Estimated costs of CC and echo in 1988/89

Capital costs

Capital costs of CC are much higher than echo. However, these are not central to the cost comparison since CC cannot be completely replaced by echo, therefore savings in terms of reduced numbers of CCs are made via savings in staff costs, consumables etc. Capital savings may be made if enough CCs are avoided to render replacement equipment unnecessary.

Variable costs

Table E2Variable costs: costs per examination (1988/89 original estimates)

R&L heart catheter examination
With inpatient stay£283
No inpatient stay£143
R only (no inpatient stay)£54

Note: These costs include all staff, ward and equipment and consumable costs (they exclude capital costs). See updated cost estimates below.

How many CC procedures need to be displaced to make an investment into echocardiography equipment worthwhile?

In brief, they estimate that if 100 CC procedures can be displaced then the acquisition of a new echocardiography service pays for itself. The more CCs are displaced, the greater the cost savings.

But, the paper gives no indication of how many CCs could be displaced by echo.

This comparison does not appear to include the costs of staff training involved in offering a new service.

Invasive treatments for heart failure

1. Left ventricular assist devices (LVADs)

These are a relatively new and rapidly evolving technology. They were originally developed as a bridge to heart transplant, but increasing experience suggests that they may also be useful as a bridge to cardiac recovery or as a long-term alternative to transplant.*296

The shortage of organs for transplantation is one incentive for increased use of LVADs as a treatment in their own right.

Most evidence on benefits and costs is from the US, and comes from cohort studies of small numbers of patients in individual units; most have no control group. Resource use and cost information is of limited relevance to a UK setting.

Only a small number of patients have had LVADs implanted in the UK. It is not possible to say exactly how many LVAD procedures are carried out, as these are not recorded separately in the Hospital Episode Statistics. Agreement to provide treatment has been on an ad hoc basis with individual health authorities.297

One good UK systematic review with economic modelling was identified (Christopher and Clegg, 1999).334, The study makes good use of available data but available data were very scarce.

Summary of Development Evaluation Committee Report on LVADs for end stage heart failure (Christopher and Clegg, 1999)

‘… although there was some suggestion of potential benefits … the evidence was not of sufficient quality to reach a decision. The committee noted a need for high quality evaluative research in this area, with duration and quality of life being a primary outcome, and taking into account UK practice and costs.’

Potential numbers of patients who could benefit

It is difficult to estimate the number of patients who could benefit from LVADs. In general, patients who may receive LVAD as a bridge to transplant have end-stage heart failure without irreversible end-organ failure (Goldstein et al, 1998296).

Statistics from the National Transplant Database (www​ can provide some idea of the potential numbers of LVAD recipients.

In the 12 months to March 2003, 147 heart transplants were carried out in the UK. This is not expected to increase significantly in the near future.

As of 9 March 2003, 99 people were on the waiting list. As a bridge to recovery, the number of deaths from myocarditis and cardiomyopathy give some indication of the potential numbers of patients who could benefit.

In 1997 there were 57 deaths from myocarditis (five of which were in those aged 75 or above), and 1,594 deaths from cardiomyopathy (376 of which were in those aged 75 or above).**

The DEC systematic review identified 10 cohort studies; five of these had control groups; all 10 judged to be of ‘poor’ design.

Cost utility analysis of LVADs as a bridge to transplant

Devices considered: HeartMate 1000 IP; HeartMate IV; Novacor; Jarvik 2000; AB–180. The last two are in their early developmental stages, with no available evidence to evaluate their use.

Utility (preference) estimates are taken from a US study (Moskowitz et al, 1997337).

This study obtained utilities (using standard gamble) from LVAD patients at three points in time: before LVAD implantation, during LVAD support and after heart transplantation.

The sample consisted of all 29 patients who underwent LVAD implantation at Columbia-Presbyterian medical centre during December 1993 to June 1995.

Main potential benefits:

  • increased QoL whilst on LVAD
  • increased functional capacity in terms of NYHA
  • faster improvement of QoL post-transplant.

If 100 patients are treated with LVADs they gain approximately 218 QALYs (sensitivity range 98 to 361) more than an equivalent group of 100 patients not treated, over 20 years.

Estimates of cost data were extremely difficult to obtain and were taken from a variety of sources, summarised below.

Costs are broken down into:

  • LVAD (device, procedure and follow-up)
  • transplant (transplant procedure, follow-up and follow-up drug costs).

Costs are dominated by the LVAD device and the transplant procedure.

Discounted cost per QALY is estimated at £39,800 (sensitivity range £28,500 to £74,000).

To achieve a more cost-effective ratio, the costs of the device and procedure would have to fall substantially.

Table E3Mean utilities associated with the three states of health

Health stateSample (of 29)Mean utility scoreRange
End stage HF immediately before LVAD implantation140.5480.050 to 0.875
During LVAD support200.8090.550 to 0.990
After cardiac transplantation110.9640.700 to 1.000

Patients not interviewed were either too ill or had died.

Table E4Cost estimates in Christopher and Clegg (1999), in pounds sterling

NHS Ref CostsNHS Trusts Finance DeptsClinician estimateBNF and Trust Protocol
Follow up (annual)750
Follow-up (annual)3,500
Follow-up drug costs (year 1)2,890
Follow-up drug costs (after year 1)3,161

Adapted from Christopher and Clegg (1999), Appendix 11.

Costs in bold were used in base case estimates.

Potential savings from LVAD treatment

Reduction in drugs for end-stage heart failure.

Decreased ICU use.

Reduction in transplant follow-up costs.

Disbenefits of LVAD treatment

The adverse events recorded in the cohort study judged by the DEC report to be of the best methodological quality,298 are:

  • device related bleeding, haemolysis, infection, right ventricular failure, thromboembolism, septic embolism, renal dysfunction.

Of these only infection was significantly different between the LVAD (n = 75) and the control (n = 33) group, but the potential disbenefits are not trivial.

Further issues

Data in the DEC report date back to 1997 and it is likely that both clinical and economic factors have changed significantly since that time. As LVADs are an emerging technology, costs are likely to fall, especially via reduced LoS and readmission.

A major resource use factor associated with LVAD use is hospitalisation.

Morales299 cites a 103-day average bridge to transplant time (in one unit in the US). It is preferable, for cost and QoL reasons, if the patient can be discharged during this time.

2. Heart transplant

The shortage of organs and strict eligibility criteria mean that this is only available to a small number of patients.

QALY estimates for heart transplant have been published but these are relatively old.300

One paper301 developed a model to investigate the costs of care following heart transplantation (the costs of transplant procedure itself are not assessed).

Data from all cardiac transplant recipients who underwent a first transplant between 1986 and 1993 at the Papworth Hospital, and who were maintained using a triple-drug immunosuppression regimen. 387 patients, age range 6 to 63 years (94% between 20 and 60 years). Main indications for surgery were ischaemic heart disease (54%) and dilated cardiomyopathy (42%).

Expected cost per transplantation patient £26,000 (£2,000–£57,600) over five years (discounted at 6%). The 95% range of cost estimates is wide:

  • patients who die early after transplantation have relatively little cost implication for the transplant service
  • patients who survive in the longer-term with a problematic post-operative course incur the greatest resource use.

Routine hospital visits account for the largest proportion of costs in the early period after transplant. Maintenance of immunosuppression accounts for the largest proportion of costs after the first three months.

QoL as well as length of life is extremely important in evaluating benefits of transplant. There is one good systematic review of QoL benefits after transplantation, which reviews information from 218 studies and 14,750 patients.302 Of these, 2,826 were heart recipients.

There is little heart transplant-specific information in the review but it does find that of all organ recipients, heart and heart and lung transplant patients are the most likely to show significant improvement across all dimensions of QoL.

3. Biventricular pacing

The economic searches identified the 2001 report on ventricular pacing and resynchronisation for heart failure from the National Horizon Scanning Centre.338

There are a number of on-going trials, mainly comparing biventricular pacing with single ventricle pacing. Preliminary results suggest a potential benefit in terms of morbidity and mortality.

One economic analysis (see below) suggests savings in terms of hospitalisation, and estimates that 4,200 to 8,400 patients with heart failure (NYHA III–IV) in England and Wales could be eligible.

Implantation and follow up costs for 1,000 patients are in the region of £5m–£7.5m.

Economic modelling of the likely costs and benefits is required.

One Swedish study (Braunschweig et al, 200066) aimed to assess total and heart failure related hospital days, as well as safety and efficacy of biventricular pacing in 16 patients with severe heart failure.

This is a small sample with no control.

Patients had heart failure due to ischemic heart disease or dilated cardiomyopathy with NYHA class III–IV, despite optimised drug treatment including ACEIs or beta-blockers. Mean follow-up 120 to 365 days. Thirteen patients improved by at least one NYHA class.

Table E5Summary of results of Braunschweig et al 2000

Year before pacingYear after pacingP (difference)
Total hospital days25345< 0.01
HF related hospital days18339< 0.01



The results of the Randomised Evaluation of Mechanical Assistance for the Treatment of Congestive Heart Failure (REMATCH) trial, which includes an economic analysis, are expected during 2003.

Fewer than half of potential recipients receive a heart transplant, even when patients aged over 60 years are excluded from consideration (Taggart and Westaby, 1997335).

One more recent Canadian study (McGregor, 2000336) was also found, but its findings are of limited applicability in a UK setting.


HMSO Mortality Statistics cited in Christopher and Clegg (1999).

Copyright © 2003, Royal College of Physicians of London.
Bookshelf ID: NBK65607


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...