eMJA: Rasmussen et al, Trial of a trial by media

Medicine and the media

Trial of a trial by media

In a democratic society, the Press has the right to investigate any issue it chooses, but this right should come with a responsibility for accuracy and freedom from bias

Helge H Rasmussen, Peter S Hansen, Yutaka Koyama, Barbara-Ann Adelstein, Anthony J O'Connell and Gregory I C Nelson

MJA 2001; 175: 625-628

→ Other articles have cited this article

Rationale for the trial - The trial - Consent - The risks of our trial - Lessons from the obstacles to the trial - Conclusions - References - Authors' details
Register to be notified of new articles by e-mail - Current contents list - More articles on Journalology and publishing

Articles in the Sydney broadsheet, the Sydney Morning Herald (SMH), earlier this year reported that some doctors were concerned about the safety, ethics and lack of informed consent of a clinical trial proposed for the management of acute myocardial infarction (AMI). The New South Wales Minister for Health, Mr Craig Knowles, was made aware of the SMH journalists' investigation, and, by the time the articles appeared in the broadsheet, he was reported to have stopped implementation of the trial.^1,2 A prominent radio commentator highlighted the serious nature of the perceived problems by asking rhetorically "Do we have to have someone who actually dies before it becomes a scandal?".³

The fate of this clinical trial has implications for the conduct of clinical research. Here, we outline the rationale for the trial and the risks for participants in it, and discuss ethical issues related to consent to participate in clinical trials. We also question the ethics of the process that led to the trial being stopped.

Rationale for the trial

Treatment of AMI with fibrinolytic agents is firmly established. However, its efficacy in achieving the treatment goal of opening the occluded coronary artery is limited and re-occlusion occurs frequently.^4,5Furthermore, many patients have contraindications to fibrinolysis, usually because they are deemed to be at high risk of bleeding. They may also be considered unsuitable for fibrinolysis, mostly because an electrocardiogram does not show the changes associated with proven benefit. AMI remains a common and frequently fatal condition. Recent prospective registry data on 30 402 patients with AMI indicate that the cumulative mortality rate remains high at about 14% while in hospital and about 22% by 90 days after presentation.⁶

Percutaneous coronary intervention (PCI) has been introduced to improve outcomes of AMI. When PCI is used in the treatment of AMI, an angioplasty balloon is used to unblock the occluded culprit coronary artery. A stent is then often deployed to maintain the patency of the artery.

Small, randomised trials indicate that PCI is superior to fibrinolyis when performed in heart centres by experienced teams.^7,8 However, most patients, even those residing in urban areas, live outside the catchment area of major hospitals. An extension of the benefits of PCI to these patients could be achieved by establishing cardiac interventional units in their local hospitals, but such units would have a low case load. Studies have shown that institutions and operators with a low case load do not achieve better patient outcomes with PCI than with fibrinolysis.^9-11 We chose to examine an alternative approach involving early transport of patients to a specialised "Regional Heart Attack Centre".

The trial

Patients presenting to the Ambulance Service with suspected AMI within the geographic limits of the Northern Sydney Area Health Service, but outside the normal catchment area of Royal North Shore Hospital (RNSH), were to be allocated at random to treatment at their local district hospital or at RNSH. Our trial was to compare two strategies: one based on PCI, backed by facilities at a large centre, and the other based on treatment with fibrinolytics at local hospitals with fewer facilities. Our comparison between the two strategies was to be based on the number of deaths and the number of recurrent non-fatal myocardial infarctions and non-fatal strokes in the two groups of patients during the admission for AMI and six months later.

Our trial was designed to determine if there were benefits of a PCI-based strategy at a large centre despite the additional transport time. The potential benefits of such a strategy might be considerable. Careful analysis of registry data on AMI indicated that the hospital mortality rate for patients treated at small centres without facilities for PCI is almost twice as high as that at large centres with facilities for PCI.^8,12 Had our randomised, controlled trial replicated these results, it would have indicated that many lives could be saved with centralised care of all patients with AMI in large cities in a few, highly specialised Regional Heart Attack Centres. Because of the urbanised distribution of our population, many patients in Australia would stand to benefit. However, such centralised treatment would require a major change from current practice and have important consequences for resource allocation. (See the Box for a flow diagram of the proposed trial.)

In 1997, the RNSH institutional ethics committee requested that we perform a pilot study to document that we could reproduce the published results of PCI on a seven-days-a-week, 24-hours-a-day basis before a randomised study could be considered. The pilot study (Stenting Strategy as an Alternative to Lytic/Medical Therapy in Acute Myocardial Infarction — SALAMI) started on 1 July 1997 and was completed in November 1998. It was confined to patients eligible for fibrinolysis within the RNSH catchment area. No inhospital deaths occurred in 102 patients.¹³ PCI became our standard treatment at RNSH for all patients with AMI. Subsequent outcomes for patients in a wide age range with virtually no exclusions have remained excellent,¹⁴ and much better than expected for the alternative strategy based on fibrinolysis/medical therapy. Despite these good outcomes, the SMH reported that there were concerns about the trial related to risks to participants and lack of fully informed consent.

Consent

The medical executive officer of a private hospital publicly articulated his concern about consent and was quoted in the SMH as saying:

I think the lack of consent is a very fundamental issue of weakness in the trial. I think it is an abrogation of fundamental human rights. I can't imagine how they got it past their ethics committee. What they are saying is that the benefits justify them overlooking that basic human right. I don't know anybody who knew anything about ethics would agree with that.²

A subsequent editorial in the SMH emphasised that "patients should be informed totally about proposed treatments".¹⁵

Truog et al hold a different opinion, and maintain that blind insistence on consent can preclude evidence-based improvement in care of acute critical illness and can be viewed as harmful.¹⁶ We concur with their judgement. Truly informed consent is difficult to give in any trial, and particularly difficult for patients with an acute critical illness. There is little evidence that informed consent protects patients from exploitation in research. This is highlighted by patients' poor understanding of consent forms and the process of randomisation, even when these decisions apply to elective treatment.¹⁶

The understanding of consent has been examined for participants in the PARAGON-B and OASIS-2 studies (which investigated the use of platelet antagonists in unstable angina and non-Q-wave myocardial infarction). Understanding of the benefits of participation was good at 85%. However, only 35% of participants understood the risks and very few (10%) understood that there was an alternative to participation. Not surprisingly, pain was an adverse predictor for comprehension.¹⁷ In our proposed trial, pain would be an almost universal feature at the time informed-consent had to be given, and the need to administer reperfusion therapy as early as possible would not allow much time for the informed-consent process. Inevitably, comprehension would have been even worse than in the PARAGON-B and OASIS-2 studies.¹⁷

The difficulty with giving informed consent in our trial was illustrated by the articles in the SMH. The paper provided a reasonably accurate account of the trial in the first articles it published,^1,2 but we doubt whether the editorialist (unsigned), writing three days later, understood the trial's purpose. Indeed, he/she questioned "why the trial is necessary at all".¹⁵

When customary informed consent cannot be obtained, one can choose from three options:

Not to conduct the research
For life-threatening conditions for which treatment is unsatisfactory not conducting research is unacceptable, as it may deny patients their right to optimal care.¹⁸
To change treatment without trial evidence
Changing treatment without trial evidence might be applauded as "clinical innovation" and does not require ethics approval. There is virtually no protection of patients from even the most adventurous "clinical innovation".¹⁶ For the issues that were to be addressed by our trial, a change in treatment strategy would be based on comparisons between rather than within studies. Regardless of how compelling such comparisons may seem, the superiority of a treatment cannot be established without randomised, controlled trials.¹⁹
To have an ethics committee assume the responsibility of giving consent on behalf of patients.
The best way to protect patients from exploitation in trials involving patients with acute critical illness is by getting an ethics committee to carefully scrutinise the protocol and agree to assume the responsibility of giving consent in the acute phase of the illness.¹⁶ Patients should then be fully informed about their participation and rights as soon as reasonably feasible. This approach is in complete accordance with Australian guidelines on the ethical conduct of research on humans,²⁰ and was adopted for our trial.

It was an important feature of our protocol that patients allocated at random to treatment at RNSH would be asked if they agreed to be taken there before the ambulance started its journey. A patient who declined would be taken to the local district hospital and receive standard care. We reject the assertion that they were to be "press-ganged".¹⁵ However, there is no pretence that it would have been feasible to obtain patients' truly informed consent.

The SMH expressed the view that absence of informed consent in our trial might set "a dangerous precedent".² This also has no foundation in fact. Proper research for the "Herald investigation" would have revealed that the first and best-known large-scale trial to document efficacy of fibrinolysis was conducted without informed consent. An institutional review board had found that informed consent could not be obtained in patients with AMI.²¹ Our trial would not have set any precedent.

The risks of our trial

An institutional ethics committee can not consider a trial unless the risk of the intervention under investigation is reasonable compared with that of existing therapy.²² According to the SMH our trial was a high-risk venture.¹⁵ One article stated "it is generally acknowledged by doctors that one in every four heart attack victims will die within the first hour if emergency treatment is not given".² A cardiologist offered an estimate of what the delays might be: "I would want to know that I wasn't going to be delayed by two hours when I could have had thrombolytic agents in 10 minutes."² Another cardiologist expressed similar concerns.³ We agree risks would be unacceptable if there were additional delays of two hours and a 25% mortality rate per hour. However, the implication that such risks would be imposed by our trial is not consistent with the facts.

Patients with AMI tend not to seek help early, and typically arrive in hospital about two hours after the onset of symptoms.^6,23The pre-hospital cardiac arrests that do occur usually happen before arrival of an ambulance.^23,24 New South Wales ambulances are equipped with defibrillators and ambulance officers can provide care for patients with cardiac arrest which is as good as or better than that provided by medically trained personnel.²⁵ A reversible cardiac arrest can be safely dealt with during transport, a conclusion supported by experience with interhospital transfers of patients with AMI,²⁶ including those at particularly high risk.²⁷ It is very unlikely that any trial patient who had a reversible cardiac arrest during transport to RNSH would die.

Let us also consider the alleged treatment delays of two hours versus 10 minutes. Transport of patients to district hospitals takes time and the median time to administration of fibrinolytics after patients arrive is up to 45 minutes.^6,28 The only transport time difference that matters is that between transport within the catchment area of a district hospital and transport to RNSH. Data from direct transport to the RNSH trauma centre, by-passing district hospitals, show a mean difference of 20 minutes for the most remote area.²⁹ This is less than the difference in in-hospital delays for patients with AMI who arrive within or outside normal working hours and are treated with PCI -- we cannot detect any adverse effects of these delays.¹⁴ In fact, outcomes for patients treated during either period are much better than expected for a treatment strategy based on fibrinolysis. Available evidence suggests that the overall risk for patients transported to RNSH is likely to be lower than the risk for patients taken to district hospitals and treated in accordance with established strategies.

Lessons from the obstacles to the trial

The results of the SALAMI pilot study (which was completed in 1998)¹³ satisfied the RNSH ethics committee's request for documentation of patient outcome. However, we were then asked to provide a response to complaints lodged with the Australian Federation of University Women Inc Northern Beaches Group, the NSW Health Care Complaints Commissioner, the Australian Medical Association, the NSW Director General of Health, local political leaders, community groups, the National Health and Medical Research Council (NHMRC), and the Cardiac Society of Australia and New Zealand. (Documentation relating to these complaints and our responses is available on request.) The Health Care Complaints Commissioner recommended that the trial be referred to the NSW State Ethics Committee. This committee referred the trial back to the RNSH ethics committee, with long delays occurring in the process. However, approval was eventually given.

Complex practical arrangements for implementation of the trial were near completion when the SMH announced that the NSW Minister for Health had decided to refer the trial to "an area ethics committee yet to be formed, which would comply with National Health and Medical Research Council guidelines".² We emphasise that the RNSH Human Research Ethics Committee had been constituted, and operated, strictly in accordance with NHMRC guidelines. It was also reported that the new ethics committee would meet "in the next few weeks".² In reality, constitution of an ethics committee is a complex process, and more than four months passed before the new committee had its first meeting. Additional long delays have already occurred in processing our proposal completely, independently of the careful, lengthy deliberations of the original committee.

The main concerns expressed by opponents of the trial related to the ethics of the trial, particularly its lack of fully informed consent. However, there are situations in which it is impossible to reconcile the doctrines of informed consent with the practical necessities of research, and depriving patients of the potential benefits of research can, in some circumstances, be considered unethical.²²

The guidance for treatment and health policy, which the trial had the potential to provide, might have saved many lives. However, the trial was delayed because of the many complaints lodged about it, and it was ultimately stopped as a direct consequence of the SMH's investigation. The published investigation^1,2misrepresented the trial's rationale, risks and important ethical issues. The NSW Minister for Health acted on the information made available to him. He may have had little choice under the circumstances. However, the end result was that due process suffered.

Conclusions

In a democratic society, the Press has the right to investigate any issue it chooses, and individuals have the right to lodge complaints with any authority they see fit. However, these rights should come with a responsibility for accuracy and freedom from bias. Society now has a justified expectation of medical practice based on evidence, and processes are in place to ensure the ethical conduct of the necessary research to produce this evidence. These processes may be in need of protection from sensationalism based on suppositions rather than evidence and expertise. The establishment of ethics committees with guaranteed protection from interference, similar to the protection for courts, should be considered.

References

Ryle G. Radical heart attack fix under fire. The Sydney Morning Herald 2001; 29 March: 1.
Ryle G. A trial of the heart. The Sydney Morning Herald 2001; 29 March: 11.
Sydney radio station 2UE. Malcolm Elliott 10.10 am, 1 April 2001. Presenter discusses plans to divert emergency heart attack patients to Royal North Shore Hospital. Interview with Dr Ross Walker, Sydney Adventist Hospital. (Transcript available on request.)
The GUSTO Angiographic Investigators. The effects of tissue plasminogen activator, streptokinase, or both, on coronary artery patency, ventricular function and survival after acute myocardial infarction. N Engl J Med 1993; 329: 1615-1622
Meijer A, Verheught FWA, Werter CJPJ, et al. Aspirin versus coumadin in the prevention of reocclusion and recurrent ischaemia after successful thrombolysis: a prospective placebo-controlled angiographic study. Circulation 1993; 87: 1524-1530.
Rogers WJ, Canto JG, Barron HV, et al. For the Investigators in the National Registry of Myocardial Infarction. Treatment and outcome of myocardial infarction in hospitals with and without invasive capability. J Am Coll Cardiol 2000; 35: 371-379.
Gersh BJ. Primary angioplasty reduced rate of death, reinfarction or disabling stroke. Evidence-based Cardiovasc Med 1997; 1: 105-106.
Every NR, Lehmann KG. The effectiveness of primary PTCA: does patient risk matter? J Am Coll Cardiol 2001; 37: 1836-1838.
Every NR, Parsons LS, Hlatky M, et al. A comparison of thrombolytic therapy with primary coronary angioplasty for acute myocardial infarction. Myocardial Infarction Triage and Intervention Investigators. N Engl J Med 1996; 335: 1253-1260.
Magid DJ, Calonge BN, Rumsfeld JS, et al. Relation between hospital primary angioplasty volume and mortality for patients with acute MI treated with primary angioplasty vs thrombolytic therapy. JAMA 2000; 284: 3131-3138.
The GUSTO-IIB angioplasty substudy investigators. A clinical trial comparing primary coronary angioplasty with tissue plasminogen activators for acute myocardial infarction. N Engl J Med 1997; 336: 1621-1628.
Zahn R, Schiele R, Schneider S, et al. Primary angioplasty versus intravenous thrombolysis in acute myocardial infarction: can we define subgroups of patients benefiting most from primary angioplasty? J Am Coll Cardiol 2001; 37: 1827-1835.
Hansen PS, Rasmussen HH, Vinen J, Nelson GIC. A primary stenting strategy as an alternative to fibrinolytic therapy in acute myocardial infarction. Med J Aust 1999; 170: 537-540.
Koyama Y, Hansen PS, Rasmussen HH, Nelson GIC. August 2001. What are the delays in primary infarct angioplasty when performed after hours and do they influence outcomes? Presented at the 49th Annual Scientific Meeting of the Cardiac Society of Australia and New Zealand. Aust N Z J Med. Abstract. In press.
Medicine on trial. Editorial. The Sydney Morning Herald 2001; 2 April: 12.
Truog RD, Robinson W, Randolph A, Morris A. Is informed consent always necessary for randomized, controlled trials? N Engl J Med 1999; 340: 804-807.
Kucia AM, Horowitz JD. Is informed consent to clinical trials an "upside selective" process in acute coronary syndromes? Am Heart J 2000; 140: 94-97.
Grim PS, Singer PA, Gramelspacher GP, et al. Informed consent in emergency research. Prehospital thrombolytic therapy for acute myocardial infarction. JAMA 1989; 262: 252-255.
White HD. Thrombolytic therapy in the elderly [editorial]. Lancet 2000; 356: 2028-2030.
National Health and Medical Research Council. National Statement on Ethical Conduct in Research Involving Humans. Canberra: NHMRC, 1999: 28-29.
Effectiveness of intravenous thrombolytic treatment in acute myocardial infarction. Gruppo Italiano per lo Studio della Streptochinasi nell'Infarto Miocardico (GISSI). Lancet 1986: 1: 397-401.
Foëx BA. The problem of informed consent in emergency medicine research. Emerg Med J 2001; 18: 198-204.
Holmberg M, Holmberg S, Herlitz J, Gardlov B, for the Swedish Cardiac Arrest Society. Survival after cardiac arrest outside hospital in Sweden. Resuscitation 1998; 36: 29-36.
Norris RM, on behalf of the United Kingdom Heart Attack Study Collaborative Group. Fatality outside hospital from acute coronary events in three British health districts 1994-5. BMJ 1998; 316: 1065-1070.
Holmberg M, Holmberg S, Herlitz J, for the Swedish Cardiac Arrest Society. Factors modifying the effect of bystander cardiopulmonary resuscitation on survival in out-of-hospital cardiac arrest patients in Sweden. Eur Heart J 2001; 22: 511-519.
Widimsky P, Groch L, Zelizko M, et al. Multicentre randomized trial comparing transport to primary angioplasty vs immediate thrombolysis vs combined strategy for patients with acute myocardial infarction presenting to a community hospital without a catheterization laboratory. The PRAGUE Study. Eur Heart J 2000; 21: 823-831.
Straumann E, Yoon S, Naegli B, et al. Hospital transfer for primary coronary angioplasty in high risk patients with acute myocardial infarction. Heart 1999; 82: 415-419.
Thiemann DR, Coresh J, Schulman SP, et al. Lack of benefit for intravenous thrombolysis in patients with myocardial infarction who are older than 75 years. Circulation 2000; 101: 2239-2246.
Harris R. Forget the golden hour. Proceedings of the First International Conference of the Australasian-Canadian Trauma Society, Darling Harbour, Sydney; March 2001. Abstract. Injury. In press.

(Received 25 Jun, accepted 10 Oct, 2001)

Authors' details

Department of Cardiology, Royal North Shore Hospital, St Leonards, NSW.
Helge H Rasmussen, FRACP, DMSc, Professor of Cardiology, Department of Medicine, University of Sydney;
Peter S Hansen, FRACP, PhD, Senior Lecturer, Department of Medicine, University of Sydney;
Yutaka Koyama, MD, PhD, Interventional Fellow;
Gregory I C Nelson, MB BS, FRACP, Director, Cardiac Catheterisation Laboratory, and Coronary Care Unit.

Ambulance Service of New South Wales, Sydney, NSW.
Barbara-Ann Adelstein, MB BCh, MBA, Medical Director.
Anthony J O'Connell, FANZCA, FFICANZCA, Chairman, Medical Advisory Committee, Ambulance Service of NSW.

Reprints will not be available from the authors.
Correspondence: Professor H H Rasmussen, Department of Cardiology, Royal North Shore Hospital, St Leonards, NSW 2065.
helgerATmed.usyd.edu.au

Other articles have cited this article: