Intended for healthcare professionals

Editorials

Placebos and standardising new surgical techniques

BMJ 2002; 325 doi: https://doi.org/10.1136/bmj.325.7364.560 (Published 14 September 2002) Cite this as: BMJ 2002;325:560

Placebo controlled trials are needed in a few well selected conditions

  1. Paul F Ridgway, general surgical specialist registrar,
  2. Ara W Darzi, professor of surgery (a.darzi{at}ic.ac.uk)
  1. Department of Surgical Oncology and Technology, Imperial College Faculty of Medicine, St Mary's Hospital, London W2 1NY

    Surgical technology has advanced exponentially in recent years, although our attitudes toward its introduction into contemporary practice remain archaic. Whereas strict licensing regulations exist for all novel prescription drugs, surgical innovation is assimilated relatively unchecked. Minimal access technologies have represented a particular problem. Previously, the mainstay of surgical innovation centred on technical modification of standard procedures of surgical care. The studies were therefore based on comparison with gold standard and rarely placebo or sham operations. The advent of minimal access methods, with their promise of reduced trauma associated with surgery, has led to a large increase in proposed roles for surgery, although, as for many current gold standards in surgery, a foundation of robust evidence is lacking.

    Moseley and colleagues' recent study has once again raised the spectre of placebo controlled trials in surgery.1 They found that arthroscopic intervention for osteoarthritis of the knee had no increased therapeutic benefit compared with placebo operation. This is an important example, not only for people with osteoarthritis, but for the surgical community in general, who perform such operations without any controlled data to suggest efficacy. This article focuses on examining the ethical and scientific issues of trials involving sham surgery and the implications for comparison studies in current frameworks for modern surgical treatments.

    Ethical issues

    Surgery by its nature does not lend itself readily to placebo controlled trials as sham operations engender invasive procedures with the possibility of non-trivial morbidity. This raises fundamental ethical objections regarding the responsibility of the researcher to act in the subject's best interest,2 and the semi-intuitive tenet that invasive surgery should be reserved for situations where significant benefit is associated with minimised or minimal risk. In the case of sham surgery the risk of harm is associated with negligible potential benefits. This has been the main area for attack by professionals who are opposed to placebo surgery. Although there are no definitive answers to these views, we should not confound the ethics of clinical research with those of clinical care.3 A randomised placebo controlled trial is not an individualised care regimen but a respected scientific tool to discern best treatment and benefit for a future cohort. Whether the study subjects receive benefit by omission or commission is, by its nature, random. Thus, although this is perhaps Machiavellian, studies should be conducted according to the ethics governing clinical research, which lessen debate regarding unique surgical concerns in these trials but do not undermine basic ethical frameworks when dealing with human subjects.4

    Scientific issues

    The practical aspects regarding the design of placebo controlled surgical trials largely follow from ethical concerns. Obviously the scientific question should be valuable, any risks involved should be minimised, and all included in the study should have given validated informed consent. One of the major issues will be whether comparison with placebo is a justifiable design. Moseley and colleagues' research illustrates the applicability of a placebo trial in the assessment of efficacy in a non-life threatening condition where the standardised role of surgery is unclear. Application of sham trials to more complex areas would be more difficult, and it would be mostly prohibitive with respect to oncological surgery. Although no sham controlled evidence exists to support the use of surgery in the resection of solid tumours, it would be wholly unreasonable to design placebo trials, given our knowledge about progression of tumours in the absence of treatment.

    Implications for comparison studies for the introduction of novel techniques

    Given the obvious ethical controversy as well as the many pitfalls that design may hold, it is amazing that any such trials are conducted today. They do provide, however, exemplary evidence when pitched against standard controlled studies. Many such standards lack a foundation of evidence and are accepted on historical status alone. Particularly in the light of recent and past revelations in trials involving sham surgery, the surgical community should critically evaluate itself regarding the efficacy of individual operations performed. The role for comparative analysis will remain paramount in this field although we should strive for excellence in design and take into account the deficiencies in the evidence from existing gold standards.

    We do not suggest that placebo controlled surgical trials are a ubiquitous tool for validation of novel surgical technologies and techniques. We believe that many such trials will continue to focus on comparison with current standards. We do think that surgical innovation that has not previously been associated with robust scientific validation should be evaluated by placebo controlled trials in selected conditions, especially where subjective symptoms of the patient are relied upon as outcome measures. We also caution that the Hippocratic principle of “first do no harm” requires individual interpretation when designing placebo controlled surgical trials.

    Footnotes

    • The views expressed in this editorial are those of the authors and do not necessarily reflect those of any regional, national, or international healthcare bodies

    • Competing interests None declared

    References

    1. 1.
    2. 2.
    3. 3.
    4. 4.