HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Campbell, H.E. Articles by Wallwork, J. Search for Related Content PubMed PubMed Citation Articles by Campbell, H.E. Articles by Wallwork, J. Related Collections Cardiac - other Congestive Heart Failure

Eur J Cardiothorac Surg 2001;20:312-318
© 2001 Elsevier Science NL

A UK trial-based cost–utility analysis of transmyocardial laser revascularization compared to continued medical therapy for treatment of refractory angina pectoris

^a Health Economics Research Group, Brunel University, Uxbridge, Middlesex UB8 3PH, UK
^b Papworth Hospital NHS Trust, Papworth Everard, Cambridge CB3 8RE, UK
^c MRC Biostatistics Unit, Institute of Public Health, University Forvie Site, Cambridge CB2 2SR, UK

Received 30 November 2000; received in revised form 26 February 2001; accepted 12 May 2001.

Corresponding author. Tel.: +44-1895-203331; fax: +44-1895-203330
e-mail: martin.buxton{at}brunel.ac.uk

	Abstract

Top Abstract 1. Introduction 2. Materials and methods 3. Results 4. Discussion Appendix A. Conference... References

 
Objective: Transmyocardial laser revascularization (TMLR) is used to treat patients with refractory angina considered unsuitable for conventional forms of revascularization. Using patient specific data from a single centre UK randomised-controlled trial, we aimed to determine whether, from a UK National Health Service (NHS) perspective, TMLR plus standard medical management is cost-effective when compared with standard medical management alone. Methods: One hundred and eighty-eight patients assessed as having refractory angina, and not suitable for conventional forms of revascularization were randomized to receive TMLR and medical management (94) or medical management alone (94). Costs to the UK NHS of TMLR (where appropriate), and all secondary sector health care contacts and cardiac-related medication in the 12 months following randomization, were collected. Patient utility as measured using the EuroQol EQ-5D questionnaire was combined with 12-month survival data to generate quality adjusted life years (QALYs). Results: The mean cost per patient over the year from hospitalization for TMLR was £11,470 and for medical management alone was £2586, giving a cost difference of £8901 (95% confidence interval (CI) £7502–£10,008: P<0.0001). The mean QALY difference, in favour of TMLR was 0.039 (95% CI -0.033 to 0.113: P=0.268). This gives an incremental cost per QALY of over £228,000. Analysis of stochastic uncertainty and of sensitivity to gross changes in key parameters consistently produces very high costs per QALY. Conclusions: The policy implications are clear: for such patients TMLR is an inefficient use of UK health service resources. This conclusion would not be changed by considerable improvements in effectiveness or reductions in cost.

Key Words: Transmyocardial laser revascularization • Randomized-controlled trial • Cost • Quality adjusted life year • Cost–utility analysis

	1. Introduction

Top Abstract 1. Introduction 2. Materials and methods 3. Results 4. Discussion Appendix A. Conference... References

 
Transmyocardial laser revascularization (TMLR) is aimed at relieving angina in patients with severe coronary artery disease who are not suitable for conventional revascularization. The procedure uses laser ablation to create channels in ischaemic segments of the left ventricle, aiming to restore blood flow to these areas and thus reduce anginal symptoms.

Four randomised-controlled trials have published results on the clinical effectiveness of TMLR (three from the US, of which one used carbon dioxide and two holmium YAG lasers, and one from the UK using carbon dioxide) [1–4]. All trials compared laser revascularization to medical management alone. There is evidence of some degree of clinical benefit in terms of improved relief of angina although the procedure is not risk free. Frazier et al. [1] report a significant reduction in hospitalization for unstable angina, reduced rates of angina and better quality of life for TMLR patients. This study, however, is confounded by the high rate of crossover from medical treatment to TMLR. The Atlantic trial reports: similarity between the two groups in terms of cardiovascular drug use, a lower rate of hospitalization for unstable angina in the TMLR group and improved patient perceptions of the quality of life [2]. Allen et al. [3] report better quality of life and lower angina in the TMLR group as well as a lower rate of cardiac-related hospitalization. The UK trial showed significant reduction in unblinded physician assessment of angina and some clinically insignificant differences in exercise capability [4]. Results to date have left room for continued clinical argument as to whether TMLR is beneficial overall to patients, and whether it should now be offered to non-trial patients.

To date none of the trials has provided a formal analysis of cost-effectiveness to help to determine whether there are net benefits to the patient and, if so, whether these are sufficient to justify any net addition to costs. This paper provides a cost–utility analysis based on the UK randomised-controlled trial of TMLR [4]. This analysis was an integral part of the trial design. Detailed patient specific data were collected, including resource use associated with the procedure, other NHS resource use to 1 year, the quality of life of patients expressed in terms of utility and survival. The study aims to help decision-makers assess whether TMLR represents a cost-effective use of scarce health service resources.

	2. Materials and methods

Top Abstract 1. Introduction 2. Materials and methods 3. Results 4. Discussion Appendix A. Conference... References

 
Details of the randomized clinical trial have been reported [4]. In summary, it was powered to detect a 50% improvement in the maximum exercise time of TMLR patients (5.250–7.875 min), at 12 months. One hundred and eighty-eight patients meeting well-defined eligibility criteria gave informed consent and were recruited into the trial at Papworth Hospital between October 1993 and September 1997. Ninety-four patients were randomized to TMLR plus medical management, 94 to medical management alone. No crossovers were permitted between the control and the TMLR groups. The local research ethics committee gave approval.

Following randomization, on-going anti-anginal drug therapy for all patients was reviewed and documented. Following a pre-procedural assessment, 94 patients were admitted to hospital for the TMLR procedure, carried out using a 1000-W carbon dioxide laser (PLC Medical Systems, MA, USA). Anti-anginal drug therapy was recommenced for patients following hospital discharge after TMLR. All the patients were followed up and National Health Service (NHS) resource use data were collected for at least 12 months. A comprehensive assessment of health-related quality of life (HRQoL) was made for all the patients at baseline, 3, 6 and 12 months. The analysis here uses the data from the EuroQol EQ-5D, a widely used instrument, included to provide a measure of utility [5].

2.1. Information sources for resource use and costs
NHS resources used by patients during the trial comprised: resource use related to the TMLR procedure itself, cardiac-related medications and any inpatient or outpatient visits. Resource use required only as a result of the trial protocol was excluded. All resources were valued using 1998/1999 costs. Table 1 lists the categories of resource use included in the costing and the sources used to value them.

Patient specific itemized resource use consumed during the surgical procedure and inpatient hospital stay was collected from the Lea, and a patient specific cost was calculated. To better reflect NHS costs, these ‘private’ hospital costs were converted to those that would have applied at Papworth had the patient received the same treatment there. For a representative sample of 50 TMLR patients reflecting the full range of lengths of inpatient stay in the sample, each item of resource use at the Lea was re-valued using costs for Papworth Hospital. From these 50 cases a regression equation was calculated to enable the equivalent NHS cost for the Lea inpatient episodes to be estimated for all 94 TMLR patients based on their actual Lea cost and the lengths of high dependency unit (HDU) and ward stay (full details available upon request). Transport costs for the relatively small number of patients (seven) who were transferred from the Lea to complete their inpatient episodes at Papworth Hospital were not included in the analysis.

Whilst in the trial patients attended a number of follow-up clinics at Papworth Hospital, based on clinical opinion of future routine practice only one follow-up appointment at a surgical outpatient clinic was costed.

2.3. Cardiac-related medication
Cardiac-related medication was recorded for both groups of patients and costed on a patient specific basis. For 14 patients (five TMLR and nine medical management) for whom drug use data was incomplete but who were alive at 12 months (i.e. those failing to attend the trial centre on one or more occasions), it was assumed that they incurred cardiac-related drug costs equal to the overall mean cost per patient observed for their treatment group.

2.4. Inpatient and outpatient episodes at Papworth
Patients provided information at follow-up visits on other contacts made with the secondary health care sector. Length of hospital stay (differentiating ward and Intensive Therapy Unit (ITU) care), the hospital concerned and episode diagnosis were recorded for all inpatient stays. Outpatient visits were also recorded.

Very few (12) inpatient episodes occurred at Papworth. Five were TMLR procedure-related and costed at the hospital's relevant cost per bed day. The others (seven) were costed according to diagnosis using Papworth average cost per episode/procedure figures, as were cardiac-related day cases. Visits to outpatient clinics at Papworth were costed using clinic specific average costs per attendance.

2.5. Inpatient and outpatient episodes elsewhere
The majority of additional inpatient episodes and outpatient clinic attendances took place in a range of hospitals throughout the UK and, occasionally, abroad. Where available, the mean national cost of the relevant Health Related Group (HRG) finished consultant episode was used to cost cardiac inpatient care [6]. When actual length of hospital stay exceeded the HRG specific trim points, additional bed days were costed using a national average cardiology bed-day cost [7]. Cardiac-related investigations undertaken as day case, and outpatient attendances at cardiac clinics were valued using Papworth hospital costs, adjusted to a national level by the National Reference Cost Index [6].

For non-cardiac-related inpatient episodes, the information from patients was not sufficiently specific to allocate each episode to an appropriate HRG. Such episodes were costed by multiplying inpatient length of stay by a national average cost per bed day on a general medical ward. All non-cardiac outpatient clinic attendances were given a generic mean cost per attendance.

2.6. Health-related quality of life
The EuroQol EQ-5D classification was completed by patients at baseline, 3, 6 and 12 months. The ‘social tariff’ for the EQ-5D, as estimated by Dolan et al. [8] from a representative UK population survey, was then applied to each patient's self-reported classification. This enabled the calculation of patient specific utility curves up to 12 months. For 21 patients (13 TMLR and 8 medical management) for whom EuroQol data were missing for a particular visit, values were interpolated between the two adjacent visits, or where the 12 month data were missing the value from the last visit was carried forward.

2.7. Calculation of quality adjusted life years (QALYs) up to 1 year
The proportion of a QALY experienced by each patient to 1 year from assessment was calculated as the area under their utility curve to 1 year or time of earlier death (this takes into account the time from randomization to TMLR procedure).

2.8. Statistical analysis
Costs from the Lea Hospital were adjusted to reflect NHS costs based on a multiple linear regression with NHS cost as a function of the Lea Hospital cost and number of days on the HDU and number of days on the ward. Cost and QALY results are all calculated for 1 year from randomization and summarized as arithmetic means with 95% confidence intervals (CI) estimated using 2000 bias corrected bootstrapped simulations [9]. Although cost data were not normally distributed, distributions of the bootstrapped means were, in general, close to normal. Therefore, t-tests based on bootstrapped means and variance estimates are valid. The cost–utility ratio was calculated as the difference in the mean cost divided by the difference in mean utility. Given that the bootstrapped CI of the denominator includes zero, a CI of the ratio is very large and not useful. One thousand bootstrapped estimates of the difference in costs were plotted against difference in utility between TMLR and medical management [10].

	3. Results

Top Abstract 1. Introduction 2. Materials and methods 3. Results 4. Discussion Appendix A. Conference... References

 
3.1. Patient characteristics
Full patient characteristics have been reported in detail elsewhere [4]. Demographics and survival are presented in Table 2.

3.4. Cardiac-related medication
Patient usage of anti-anginal medication during the trial has been reported elsewhere [4]. The mean cost per patient of all cardiac-related medication was £153 lower in the TMLR arm (P<0.0001).

3.5. Inpatient and outpatient episodes
The between group average cost difference for combined inpatient and outpatient episodes favours TMLR but the difference of £230 is not statistically significant. Of the 104 hospital readmissions in the TMLR group, 79 (76%) were cardiac-related. The corresponding figure in the medical management (MM) group was similar, with 81/96 (84%) admissions having a cardiac cause (P=0.19). More detailed data on hospital admissions are reported elsewhere [4]. Table 3 shows the breakdown of costs for cardiac- and non-cardiac-related inpatient and outpatient contacts. There were no statistically significant cost differences between the two groups in any of these sub-categories.

3.6. Overall costs – TMLR vs. medical management
Taking the overall mean cost per patient up to 1 year, TMLR costs £8901 more than medical management: there is little saving in other health service costs in the first year to offset the expense associated with the TMLR surgical procedure.

3.7. Utility
Fig. 1 shows the results of applying the EQ-5D social tariff values to the patients' self-classification of their health states at baseline, 3, 6 and 12 months. The figures suggest that TMLR patients experienced greater improvements in their utility at 3 and 6 months with this benefit disappearing at 12 months. (The differences at 3 and 6 months were statistically significant.) Despite these differences, the utility curves and survival data for the two treatment modalities produce very similar mean QALY values up to 12 months, as shown in the final row of Table 3. These values, calculated without adjusting for the slight difference in utility between the two groups at baseline, generated a 0.039 mean QALY difference in favour of TMLR, which was not statistically significant (95% CI: -0.033 to 0.113). This difference is equivalent to an extra 2 weeks of perfect health gained using TMLR (95% CI of approximately -2 to 6 weeks).

View larger version (16K): [in this window] [in a new window]   Fig. 1. Mean (95% CI) utility scores generated by applying the EQ-5D social tariff values to patients' self-classification of their health status.

The mean incremental cost per QALY of using TMLR is estimated at £228,000. Fig. 2 plots the results of 1000 bootstrap estimates of cost difference against utility difference and is akin to a two-dimensional CI. An illustrative line for an incremental cost–utility ratio of £50,000 per QALY is also presented in the figure. Avoiding the need to estimate a traditional 95% CI or a cost-effectiveness acceptability curve [11,12], it shows that every one of these 1000 estimates generates a cost–utility ratio exceeding £50,000 per QALY.

View larger version (13K): [in this window] [in a new window]   Fig. 2. Bootstrapped estimates of cost difference against utility difference, with illustrative line for incremental cost–utility ratio of £50,000 per QALY.

Assuming that patients would be admitted to hospital on the day of surgery rather than 1 or 2 days prior to it had only a marginal impact upon the cost of TMLR and produced an incremental cost per QALY still exceeding £217,000. If the cost of TMLR was reduced by ignoring the capital cost of the laser, the excess cost of the TMLR arm would still average £6482. The cost per QALY gained would only fall to £166,205. In terms of utility, a doubling of the mean utility benefit of TMLR would still leave an incremental cost per QALY in excess of £114,000. If, implausibly, there were no mortality in the TMLR group within 12 months, the incremental cost per QALY would still be £152,169. In each of these extreme scenarios the policy implications would remain unchanged.

	4. Discussion

Top Abstract 1. Introduction 2. Materials and methods 3. Results 4. Discussion Appendix A. Conference... References

 
From an NHS viewpoint, the incremental cost to 1 year of TMLR over standard medical management alone is £8901 and the cost per QALY, estimated at £228,000. Whilst there is no agreement as to what should be considered as the maximum acceptable cost–utility ratio in the UK [13] all the estimates for TMLR far exceed the level that might be deemed as an efficient use of NHS resources.

Although Frazier et al. [1] infer that savings in costs of medication and subsequent cardiac-related re-admissions might offset a substantial part of the costs of the TMLR procedure, we found no evidence of significant offsetting cost reductions. Indeed, we found no significant differences in the costs of either cardiac or non-cardiac inpatient episodes.

The transfer of 7 TMLR patients between hospitals following the procedure was considered to be protocol-driven resource use and so transport costs were not included in the analysis. Whilst a number of other detailed concerns relate to the accuracy of the cost estimates, the sensitivity analyses show that no detailed issue of costing methodology would be sufficient to undermine the robustness of the conclusion that the cost per QALY is beyond any acceptable range for the UK.

In terms of outcome measures, we recognize that QALYs are not universally accepted as the most useful summary measure of effect to inform policy decisions. How else might we summarize what clinically was gained from the additional £9000 of NHS resources per patient? In this trial, TMLR provided an inter-group difference of an average 40 s in treadmill endurance or 33 m in 12-min walk distance [4]. Whilst there was a statistically significant difference in the proportion of patients who experienced a clinically significant reduction in physician rated angina (25% for TMLR: 4% for medical management patients), the difference in patient recorded angina, although still favouring the procedure, was not statistically significant. Whether one looks at the results in terms of QALYs or specific clinical effect, the trial showed very small gains from the valuable NHS resources.

The greatest potential weakness of this analysis is that it only considers costs and QALYs to 1 year. What can we say about future years? The difference in EQ-5D is maximized at 3 months and declines subsequently. (Detailed analysis of all other HRQoL data collected during the trial shows that differences were maximized at 3–6 months and then diminish or disappear at 12 months.) Longer term EQ-5D data for those patients randomized early in the study and followed for up to 36 months suggests that there is little subsequent difference in utility. But even if the mean QALY gain in the first year was maintained for 5 years and there were no differences in cost between the groups after the first year, the mean cost per QALY would still exceed £50,000.

This analysis provides an example of where despite considerable uncertainty as to the precise utility value of the procedure, and some uncertainty about precise costs, the policy implications are very clear. We cannot envisage a plausible scenario consistent with this trial evidence, which would lead to an acceptable cost per QALY for this particular procedure in this patient group.

	Acknowledgments

 
We thank all the patients and professionals who were involved in the trial, particularly the surgeons, Mr S.R. Large, Mr J. Dunning, Mr P. Braidley, and the anaesthetists, Dr J. Kneeshaw and Dr D. Bethune. We also thank the trial's research assistants, Dr Divna Trkulja-Batas, Mrs Elizabeth Kelly, and are grateful to Andrew Davies for his assistance with the bootstrapping. The laser equipment was provided for the trial by BUPA Healthcare. No private patients were involved in the trial. The views expressed in the paper remain, however, those of the authors.

The research was funded by the Medical Research Council, with additional support to the Health Economics Research Group from the Department of Health Policy Research Programme. The NHS provided financial assistance to the pilot study and the treatment costs of the trial.

	Footnotes

 
Presented at the 14th Annual Meeting of the European Association for Cardio-thoracic Surgery, Frankfurt, Germany, October 7–11, 2000.

	Appendix A. Conference discussion

Top Abstract 1. Introduction 2. Materials and methods 3. Results 4. Discussion Appendix A. Conference... References

 
Dr X. Mueller (Lausanne, Switzerland): Laser revascularization is indicated for very severe angina, Class IV usually. Now, I was surprised in your patient selection that they had to do a 12-min walk test before the operation. So this could have an impact on the quality of life. Your results are at variance with all the other studies, which have been performed clinically in this field.

Ms Campbell: There are three trials from the USA and I think they all had sicker patients in their trials. I think two-thirds of the patients were Class IV angina in the US trials.

Dr E. Bodor (Budapest, Hungary): Just as I got the impression that TMR has only one advantage, that is, the angina disappears after in these patients, but I remember the early 1960s when there were no coronary angiography possibility and before the era of CABG procedure, our bosses, the pioneers, they used a very simple and very cheap way of treatment of a heavily angina patient. They applied on the surface of the heart some chemicals, and that was called cardiopericardiopexy, and the patients survived and the mortality rate was nearly zero and the angina disappeared in these patients too.

So probably we have to think where it is necessary and that helps support my opinion not necessarily to invent for such an expensive treatment when we can reach nearly the same therapy with an old way like cardiopericardiopexy.

Dr J. Bachet (Paris, France): A few years ago we had an experience in my institution with the TMR, 28 patients underwent this procedure and the experience was stopped for two reasons. The main reason was that the French Ministry of Health forbade the practitioners to use this device because it was too costly. Secondly, because we had an immediate mortality of 20%. So I think that this could be a very important cause of not using this device, which is inefficient, costly and kills more patients than medical therapy.

I would like you to comment on that.

Ms Campbell: Not being a clinician myself, I probably cannot really comment on that, but if Mr Large would.

Mr S. Large (Cambridge, UK): I think you have been very brave in your presentation. These 94 patients matched randomly to an equal number of controls, patients who have come forward because they have no alternative means of revascularization. Patients who are desperate for something to be done, and we are all familiar with these patients. That was really an answer to the first question.

The other concerns are that they are a very challenging group of patients. The experimental arm required thoracotomy, of course, which ethically we couldn't control, and attracted a 5% mortality.

In addition, angina does not disappear with this. We did see a select group of patients with reduction in angina. I don't think anybody should labour under the delusion that this trial led to eradication of angina. We look keenly forward to the results of our collaboration with North American percutaneous laser revascularization, working of course from within the heart, undertaken by cardiologists. I suspect that we are going to find very much the same result.

I hope that answers most people's concerns over the clinical aspects of this trial.

Dr J. Svennevig (Oslo, Norway): We did another randomized study in Oslo and we came to much the same results. This is a treatment that may help against pain but it is too expensive.

I wonder, in your personal opinion, do you think that we should throw out these machines or is there still an indication for TMR treatment in special cases?

Mr S. Large: That is a very easy answer. I think it should be thrown out.

	References

Top Abstract 1. Introduction 2. Materials and methods 3. Results 4. Discussion Appendix A. Conference... References

 

1. Frazier O.H., March R.J., Horvath K.A. Transmyocardial revascularization with a carbon dioxide laser in patients with end-stage coronary artery disease. N Engl J Med 1999;341(14):1021-1028.[Abstract/Free Full Text]
2. Burkhoff D., Schmidt S., Schulman S.P., Myers J., Resar J., Becker L.C., Weiss J., Jones J.W. Transmyocardial laser revascularisation compared with continued medical therapy for treatment of refractory angina pectoris: a prospective randomised trial. Lancet 1999;354:885-890.[Medline]
3. Allen K.B., Dowling R.D., Fudge T.L., Schoettle P., Selinger S.L., Gangahar D.M., Angell W.W., Petracek M.R., Shaar C.J., O'Neill W.W. Comparison of transmyocardial revascularization with medical therapy in patients with refractory angina. N Engl J Med 1999;341(14):1029-1036.[Abstract/Free Full Text]
4. Schofield P.M., Sharples L.D., Caine N., Burns S., Tait S., Wistow T., Buxton M., Wallwork J. Transmyocardial laser revascularisation in patients with refractory angina: a randomised controlled trial. Lancet 1999;353:519-524.[Medline]
5. Brooks R., The EuroQol Group. EuroQol: the current state of play. Health Policy 1996;37:53-72.[Medline]
6. Department of Health. The new NHS 1998 reference costs. Wetherby: Department of Health, 1998.
7. NHS Executive. Trust financial returns. Leeds: NHS Executive, 1997.
8. Dolan P, Gudex C, Kind P, Williams A. A social tariff for EuroQol: results from a UK general population survey. Discussion paper 138. Centre for Health Economics, University of York, 1995.
9. Efron B., Tibshirani R.J. An introduction to the bootstrap. San Francisco, CA: Chapman & Hall, 1993 p. 1–412.
10. Briggs A., Fenn P. Confidence intervals or surfaces? Uncertainty on the cost-effectiveness plane. Health Econ 1998;7:723-740.[Medline]
11. Van Hout B.A., Al M.J., Gordon G.S., Rutten F.F.H. Costs, effects and c/e-ratios alongside a clinical trial. Health Econ 1994;3:309-319.[Medline]
12. UK Prospective Diabetes Study Group. Cost effectiveness analysis of improved blood pressure control in hypertensive patients with type 2 diabetes: UKPDS 40. Br Med J 1998;317:720-726.[Abstract/Free Full Text]
13. Eddy D.M. Doctors, economics, and clinical practice guidelines: can they be brought together?. London: Office of Health Economics, 1999.
14. British Medical Association and the Royal Pharmaceutical Society of Great Britain. The British National Formulary. Number 35. Wallingford: British Medical Association and the Royal Pharmaceutical Society of Great Britain, 1998.
15. Taylor F.C., Gray A., Cohen H., Gaminara L., Ramsay M., Miller D. Costs and effectiveness of a nurse specialist anticoagulant service. J Clin Pathol 1997;50:823-828.[Abstract/Free Full Text]
16. Netten A., Dennett J. Unit costs of health and social care 1998. Canterbury: Personal Social Services Research Unit, University of Kent at Canterbury, 1998.

This article has been cited by other articles:

				B Schweikert, H Hahmann, and R Leidl Validation of the EuroQol questionnaire in cardiac rehabilitation Heart, January 1, 2006; 92(1): 62 - 67. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Campbell, H.E. Articles by Wallwork, J. Search for Related Content PubMed PubMed Citation Articles by Campbell, H.E. Articles by Wallwork, J. Related Collections Cardiac - other Congestive Heart Failure