A burgeoning ethical infrastructure can mean that scientists take less care of their research subject.
Heinz Lehman went to Canada as a refugee psychiatrist in 1937 and worked at a large psychiatric hospital in Montreal with over 1600 psychiatric patients. Lehman was convinced that the major psychiatric disorders were biologically based and, in the absence of any proven effective treatments, he was willing to try anything. Patients were given large doses of caffeine, sulfa injections and typhoid antitoxin to produce fevers. Even turpentine was applied to the abdominal muscles to produce sterile abscesses that would stimulate the immune system. Although there were claims of ‘promising results’, nothing really helped for very long (1).
Lehman was working at a time prior to modern ethics regulation and review, and it is unlikely that anyone today could follow their hunches in the way that he did. The broad aim of ethical review is to minimise the risk to volunteers or patients entering into experiments, and to ensure adequate consent is obtained before performing procedures that may be detrimental. Ethics councils in the UK, or institutional review boards in the USA, require the principle investigator of a study to submit certain materials. At a minimum, this would include an outline of the study’s rationale, an outline of the procedures that will be performed, and a copy of consent forms or other materials that volunteers or patients will receive.
Peer review of experiments can be a good thing. But the way that ethics committees are currently working is counterproductive – even unethical. Indeed, no ethics committees might be better than the situation we have at present.
First, some background on ethical review. At Birmingham University, I sit on the Psychology School Committee. A typical submission to our committee will be 4-6 pages long, and the time it takes to prepare an application would be around 2-4 hours. Our ethics committee rarely meets face-to-face; all comments are emailed directly to the chairman. We comment only on issues pertaining to subject safety and confidentiality. The chairman then decides if the study can be approved directly or after amendment. The process typically takes 4 weeks.
We are looking for things such as assurances that any data collected will be treated in confidence, and that the subjects are not likely to find themselves in a position of harm. For example, in a study where volunteers might receive an alcoholic drink we would want assurances that the investigators would screen for subjects taking barbiturates, benzodiazepines and other medications that can interact negatively with alcohol. Our little committee is, in my opinion, eminently sensible. We provide a reasonably quick check of our peers’ proposed work to ensure the care and safety of subjects.
In the USA, institutional review boards (IRBs) must meet federally mandated requirements. In Pittsburgh University, where I worked until recently, my last IRB submission ran over 60 pages, included two consent forms, one of which was 13 pages long, and I was required to submit 26 copies. The 24 members of the IRB would meet to discuss applications at regular intervals. After review, the board would require resubmission of the entire application, including a memo describing amendments addressing the board’s concerns. The board could then reconvene to reconsider the application or the chairman could be empowered to approve the application directly. This process can take as little as two months but typically takes between six and 12 months.
Ethics applications in the UK are starting to look more like those in the USA, especially applications that pass through the NHS. Ask any medical researcher in the UK about their experiences with NHS ethics and you are unlikely to hear any positive words (2). The introduction of a patient leaflet, for example, designed to improve older patients’ involvement in GP consultations, was reviewed by an NHS committee with an estimated five days of preparatory work (3).
For a more involved multicenter clinical trial, an investigator must submit copies of the study protocol to each of the participating centres ethics committees. For one trial, this resulted in up to 21 copies of the protocol being submitted to 125 local research ethics committees. Eighty-four of these committees withheld approval until the researchers had made amendments. This process can take years to complete. Another clinical trial involving 51 centres needed over 25,000 pieces of paper, 62 hours of photocopying and 170 hours of investigator time. This is an entirely different matter to basic peer review to ensure subject care and safety. Ethics review within the USA and the British NHS has become completely irrational.
But the problem is not just that ethics committees provide formidable technical barriers for researchers. It’s also that they undermine medical ethics. There are three main problems with these committees: they place barriers between patients and potentially beneficial treatment; they distance the researcher from thinking about ethics; and they offset responsibility for the conduct of the research on to the patient. As such, they are a danger to both patient and investigator and should be abolished.
Let me illustrate the first point. If you are unfortunate enough to enter hospital with a recent onset heart attack, you are likely to receive a combination treatment of streptokinase and aspirin. Patients who received this combination treatment in a trial of 17,187 patients had a mortality rate of eight percent. Patients who received placebo had a mortality rate of 13 percent. The treatment therefore provides for significant improvements in survival. Relative to the USA, recruitment into the trial was quicker in the UK and thus the benefits were seen earlier and the treatment adopted earlier. Whatever it was that slowed recruitment into the American trial has been estimated to have caused 10,000 unnecessary deaths (4). Erecting barriers to the progress of research, in the absence of any benefit and with evidence of harm, is surely unethical.
The second problem is that ethics committees distance medical professionals from their ethical responsibilities. Partly this is just a natural response to the burgeoning amount of paperwork that an investigator has to submit before embarking on a study or change in practice. No one person is likely to be able to have the time and expertise necessary to submit multiple protocols to multiple committees. Consequently, increasing numbers of senior investigators are employing administration staff or students to complete the necessary paperwork. Rather than being something the investigator thinks about and addresses, ethics becomes a technical exercise that is placed, as much as possible, into the hands of junior colleagues.
Alternatively, investigators simply avoid ethics applications through technical dodges, such as claiming the research to be an audit that is exempt from review (5), or through an avoidance of research altogether (6).
Investigators are also becoming ideologically distanced from the process of ethical review. The purpose of ethical review is to ensure that the investigators, particularly the senior members of any research team, consider how their proposed procedures might impact upon the wellbeing of their volunteers and patients. Whether or not a particular procedure should be performed or a particular treatment tested can turn on many quirky factors that can only ever be fully known by those directly involved. The expert judgment of medical professionals should not be replaced by mouthing answers to committees’ formulaic questions.
Many years ago I was involved in a trial investigating the effects of pain following the extraction of wisdom teeth. We were particularly interested in the performance deficits that the pain caused during a simple card sorting task and the effects of morphine on their task performance.
The extraction procedure was always completed in the morning and the patients were then taken to an overnight ward to recuperate. The intention was to wait until the anaesthetic was no longer effective and then begin the card-sorting task. All procedures had to be completed before 5pm when our medical collaborator left for the day, which meant that we had to begin the testing by 4pm. It soon became clear that some patients were waiting us out. A patient would steadfastly deny any pain, despite fairly obvious signs to the contrary, but would promptly announce their pain at 4:05pm. (I suspect that they either did not want to actively withdraw from the study, or they falsely believed they would not get pain relief if they did actively withdraw from the study.) We resolved the problem by offering the patients pain relief at regular intervals after they arrived on the ward and by regularly reminding them that they were under no obligation to complete the study.
Importantly, at no point were we in violation of our ethics approval. We had no objective evidence that patients were in excessive pain and no proof that they had misunderstood their access to pain medication. Had our concern for our patients stopped at the point of receiving ethical approval we may never have introduced any changes to our procedures. As the difficulty of receiving ethics approval increases and as the physical and mental distance between investigators and ethics committees widens, a lack of concern for patients is likely to increase.
The final problem is that responsibility is being offset on to volunteers through consent forms. Sometimes the displacement is so crude and obvious that it boggles the mind. When my wife went into labour last year, we arrived at the hospital at around 1am. Before she was allowed into triage she had to sign three separate consent forms. Neither of us read a single word. This mocks the term ‘informed consent’. This is a bureaucratic exercise designed to ensure that the hospital avoids liability in the event of anything going wrong.
Consent forms can also be horrifying for the patient. Within the 13 pages of my most recent consent form is detailed information on the dangers, however remote, of the procedures, as well as the measures that will be taken should the volunteer be harmed. This assumes that the volunteer and investigator can equally understand the probability of accidental injury or harm. By signing the form, volunteers are suggesting that they understand what might go wrong and are prepared to accept responsibility for that risk. The responsibility for safe conduct of the study has been deftly divided between the investigator and the volunteer.
So it is certainly true that the barriers to obtaining ethical approval have increased in recent years, and this is regrettable. Much worse than the inconvenience, however, is the pernicious influence on the relationship between investigator and volunteer. Ideally, this relationship should be based on trust and a mutual desire to advance understanding, but current ethics procedures encourage distance between investigator and volunteer – and foster a relationship more akin to a legal agreement between parties.
Lehman’s patients never signed consent forms and were probably rarely provided with the detail of the procedures designed to treat them. With hindsight it is clear that some of his treatments were reckless, but his creativity, imagination and drive to help his patients was not. Let us demand an end to unethical ethics review. Responsibility for future research should go back into the hands of the medical professionals.
(1) The account of Lehman and Sakel is derived from Blaming the Brain: The Truth About Drugs and Mental Health, ES Valenstein, Free Press, 1988
(2) ‘Elwyn G. Ethics and research governance in a multicentre study: add 150 days to your study protocol’ [Letter], British Medical Journal 2005; 330: 847; Minnis HJ, ‘Ethics review in research: Ethics committees are risk averse’ [Letter], British Medical Journal 2004; 328: 710-711; Beales IL, ‘Ethics review in research: Role of ethics committee review is widely interpreted’ [Letter], British Medical Journal 2004; 328: 710
(3) Glasziou P, Chalmers I, ‘Ethics roulette: What can we learn?’, British Medical Journal 2004; 328: 121-122
(4) Collins R, Doll R, Peto R, ‘Ethics of clinical trials’, in Williams CJ, ed. Introducing new treatments for cancer: practical, ethical and legal problems, Chichester: John Wiley, 1992
(5) Wade DT, ‘Ethics, audit, and research: all shades of grey’, British Medical Journal 2005; 330: 468-471
(6) Bentley C, Enderby P, ‘Academic medicine: who is it for? Applications to local research ethics committees are down by around 40%’ [Letter], British Medical Journal 2005; 330: 361
To enquire about republishing spiked’s content, a right to reply or to request a correction, please contact the managing editor, Viv Regan.