Guest Post by Karolina Wartolowska

Re: Randomised placebo-controlled trials of surgery: ethical analysis and guidelines [open access]

Surgical placebo-controlled randomised controlled trials are, in many ways, like placebo-controlled drug trials. Like in case of drug trials, sometimes, a placebo-controlled design is necessary so that the results are valid and unbiased. Placebo control is usually necessary when a surgical trial has only subjective outcomes. This is often the case, because many surgeries are done to relieve pain and improve function. Validating the efficacy of a surgery in a well-designed trial helps to improve clinical practice. If the procedure is ineffective it should be discontinued and less risky treatment should be used instead. It also demonstrates the need for new effective interventions. But if the surgery is effective the resources should be allocated to the better intervention. If efficacy of intervention is never tested, many patients may be exposed to risks associated with a surgery but do not get any real benefits.  They also do not get other treatment, which may provide similar benefits without the risks and costs associated with surgery.

Surgical placebo-controlled randomised controlled trials may be undertaken in an ethical way. Firstly, there needs to be “equipoise”. In other words, there should be uncertainty, lack of strong evidence and lack of agreement among the clinicians whether the investigated surgery is effective or whether it is better than conservative treatment.  If there is equipoise, there is no true “best treatment” which can be recommended to the patient. Secondly, there should be some preliminary evidence that the surgery works (form animal studies, open-label trials). There is no point undertaking a surgical trial if it fails to show any improvement in the surgical arm. Thirdly, the risks associated with a surgical trial should not be disregarded. To be justified, such trials should have high scientific and clinical value and a potential to change clinical practice. Moreover, the risks of harm in both trial arms should be as small as possible. This is particularly important in the placebo/sham arm. The placebo mimics the active surgery but it also omits the surgical element which is the key part of the active surgery. So some procedures necessary in the surgical arm, for example anticoagulants or antibiotics, may be avoided in the placebo arm or can be replaced with a saline injection.  Ideally, the placebo/sham procedure should benefit the patients, for example as a diagnostic procedure. And last but not least, it is important that there is an uncertainty about the treatment allocation but there is no actual deception. Patients should understand which procedures are or are not performed and what are the associated risks in each trial arm.

Surgery is inherently risky but it is important to know whether it is also effective and worth taking these risks.