Article Text

Download PDFPDF

The importance of placebo effects in enhancing palliative care interventions
  1. V Lucas1 and
  2. S Booth2
  1. 1Garden House Hospice, Letchworth Garden City, UK
  2. 2Department of Palliative Medicine, University of Cambridge, Cambridge, UK
  1. Correspondence to Dr V Lucas, Garden House Hospice, Gillison Close, Letchworth Garden City SG6 2ED, UK; viv.lucas{at}


In this narrative review, we examine evidence that may help to explain how placebo effects contribute to the effectiveness of palliative care interventions for the relief of symptoms such as pain, breathlessness and depression, and how they may underlie the impact of complementary therapies. We discuss the different ways of conceptualising placebo phenomena, including the importance of the therapeutic relationship, the context of care and the significance of meaning. There is increasing evidence from neuroscience that the term ‘placebo effect’ describes a number of phenomena that may explain the effectiveness of therapeutic interventions that affect the perception of symptoms. A greater appreciation of placebo effects emphasises the importance of addressing social, psychological, and spiritual factors with equal rigour. Commissioning bodies, rightly concerned about the evidence base for clinical interventions, need to recognise the multifaceted nature of symptom control measures and to realise that the focus for palliative care research needs to be on the specialty as a complex integrated intervention rather than on a series of individually evaluated measures.

View Full Text

Statistics from


A scientific understanding of palliative care is incomplete without an appreciation of placebo effects that are being increasingly elucidated through neuroscientific research.1 ,2 Yet the term ‘placebo’ is often misunderstood, taken to mean a ‘fake remedy’ or sleight of hand, deceiving a patient into thinking that something real has been done when it has not. In research, placebo effects are usually considered as unwanted, to be eliminated from clinical trials on the grounds that they are inactive. As McQuay and Moore comment, “Placebo means lots of different things to different people, which leads to endless confusion”3 and its definition remains a matter of debate.4

If we can define how palliative care interventions are enhanced by placebo effects, it will be possible to teach ways in which they can be used in clinical practice. At present placebo effects may be underappreciated in palliative care and there may even be a concern that seeing them as intrinsic to its practice would devalue the specialty, making it unscientific; precisely the opposite is the case. In this paper, we consider placebo effects in relation to symptom control and the use of complementary therapies to illustrate this.


Clinical experience shows that patients admitted to a hospice or palliative care unit for pain control can obtain considerable relief even when no or only minor alterations have been made to their analgesic regimen. An unblinded Canadian study5 investigating the effects of morphine (given as a ‘Brompton cocktail’) showed a significant difference in pain relief for patients on the Palliative Care Unit (90%) compared with those in the general wards and private rooms (75–80%).

We may observe in clinical practice that, for example, a patient with a 200 µg fentanyl patch in place will use 10 mg oral morphine for breakthrough pain with good effect when the ‘four hourly equivalent’ would be at least 80 mg.6 This could be explained by invoking different receptors or physicochemical properties of the different opioids, but in view of the complexity of the pain experience, placebo effects could well be contributing. Dose–response curves accompanying new analgesics generally show only small differences between analgesic and placebo effects, suggesting that placebo effects are active in pain relief mechanisms. Eippert,2 in a laboratory study, demonstrated that placebo analgesic effects are mediated through the same spinal pathways as drug-induced ones.


Intractable breathlessness has long been known to be associated with fear and anxiety, and associated neurophysiologically with the limbic system, amygdala and hippocampus.7 A comprehensive evaluation of a breathlessness intervention service8–11 has consistently shown that a fan is a helpful intervention in relieving breathlessness. Although there are neurophysiological data suggesting that airflow leading to cooling of oronasal receptors and the facial area served by the third branch of the trigeminal nerve is important in mediating the response of the fan,12–14 it is also clear from the same studies that many patients given a fan without explanation or discussion of the rationale behind its use do not find it useful for breathlessness until a scientific explanation is given. In the same series of studies evaluating the symptomatic palliation of breathlessness, patients consistently15 valued the way in which they were listened to, the expert understanding of the impact of breathlessness on their life, the reduction in their sense of isolation and the value of the support given to the carer as intrinsic to their treatment.


Depression is common in palliative care patients. Treatment often includes drugs, commonly now the selective serotonin reuptake inhibitors.16 Yet there is considerable evidence from meta-analyses, highlighted by Kirsch17 in particular, that the effectiveness of antidepressants is substantially due to placebo effects.18 Antidepressant drugs and placebo medication improve the symptoms of depression, as does psychotherapy, whereas those not treated by any method fare worse; “One thing to learn from the data is that doing nothing is not the best way to respond to depression”.19 Naturally Kirsch's conclusions have courted controversy,20 and there has been discussion about how the effect size of an antidepressant should be defined.21

Complementary therapies

Complementary and alternative therapies are used widely in palliative care, and about a third of cancer patients use them independently.22 However, the scientific evidence for their efficacy is virtually non-existent.23 ,24 There is some evidence from the palliative care literature that patients given complementary therapies, for example, aromatherapy massage, benefit in terms of their general well-being and feeling less anxious, but the effect is not sustained.25 Complementary therapies are diverse, and it is necessary to be wary of making generalisations about them. Acupuncture, for example, has been the subject of many trials. Yet there is no good evidence with regard to its efficacy for a whole range of conditions.24

It may be the case that there are clinical benefits of acupuncture for certain conditions, but the clinical trials carried out have not been able to consistently confirm this. Several systematic reviews have failed to show any evidence for the efficacy of acupuncture for any condition26 and more specifically for cancer pain.27 Madsen28 published a review of papers comparing real with ‘sham’ acupuncture and concluded that, “Whether needling at acupuncture points, or at any site, reduces pain independently of the psychological impact of the treatment ritual is unclear”. Linde29 attempted to evaluate the non-specific effects of acupuncture and concluded that there are often large effects associated with sham acupuncture and that these effects may be larger than ‘inert’ placebo interventions.

‘Specific’ and ‘non-specific’ effects

The idea that the overall effect of a treatment is made up of ‘specific’ and ‘non-specific’ effects in varying proportions may link these seemingly disparate observations together. Effects of a drug such as morphine comprise the specific interaction of morphine molecules with the endogenous opioid receptors, and non-specific effects, such as patients’ belief about the expected benefit of the drug and the context in which the treatment is given, which includes the effect of the prescriber. As Evans30 comments: Only if you believe that a medical intervention is taking place, and that it works for the condition you are suffering from, will the placebo response be activated. This does not mean that your condition will necessarily be cured. Even if the placebo response is activated, it will not have any effect on your health unless you happen to be suffering from a placebo responsive condition.

Placebo-responsive conditions include symptoms such as pain, depression, anxiety and nausea, precisely some of the common symptoms experienced by palliative care patients. These symptoms of advanced incurable disease are closely related to the emotions and cognitive processes in terms of their experience. The underlying pathology cannot usually be reversed, but the psychological component of the symptom can be diminished.

As well as the patient believing that his/her treatment is effective, it is also important that the doctor does too.23 Analgesics and antidepressants are prescribed to patients in the belief that they are effective. Related to this ‘belief effect’ is the context in which treatments are given such as the white tiled emergency room full of technical equipment or the general practitioner’s consulting room where rituals are enacted, such as taking a history and performing an examination. A hospice has a particular ambience, ideally one of tranquillity, compassion and competence; a place where pain and suffering can be relieved. But the physical environment of healthcare institutions cannot be separated from the people who give care. The interactions that take place between patients and healthcare professionals affect the therapeutic outcome.

The ‘therapeutic relationship’

Hyland31 has considered the effect of the therapist in psychotherapy and complementary and alternative medicine (CAM) and concluded that central to their effectiveness is the relationship between the patient and the therapist. Hyland found that there is no evidence that the specific component of a psychotherapy (eg, whether Freudian, Jungian or cognitive behavioural therapy) makes any difference to the outcome. What determines the outcome is the therapist and not the therapy. Similarly with CAM, the current evidence appears to indicate that it is not the modality of treatment but the therapist that is important.

The effect of the ‘therapist’ was recognised in medical practice in the 1950s by the Hungarian psychotherapist Michael Balint, who called the effect the ‘drug doctor’ and encouraged general practitioners to be aware of this effect and use it to help their patients.32 There is now evidence from neuroscience research that demonstrates that doctor/therapist–patient interactions affect therapeutic outcomes.1

The context of care

The encounter between health professional or therapist and patient usually takes place in a specific setting accompanied by certain rituals that, when taken together, can be considered to be ‘contextual’ factors. Miller and Kaptchuk33 have argued that the term placebo should be replaced by the phrase ‘contextual healing’ because of the confusion surrounding the word ‘placebo’. They suggest that Healing resulting from the clinical encounter consists of a causal connection between clinician-patient interaction (or a particular component of the interaction) and improvement in the condition of the patient. That aspect of healing that is produced, activated or enhanced by the context of the clinical encounter, as distinct from the specific efficacy of healing interventions, is contextual healing.


The use of the term ‘healing’ is significant as it is central to the practice of palliative care34 although it has all but disappeared from mainstream medical discourse.35 It means ‘to make whole’ and it is through the healing process that suffering or psychosocial distress can be eased.36 Fundamental to the healing process in palliative care is a search for meaning facilitated by the professionals involved in caring for the patient. The importance of ‘meaning’ in relation to placebo has been discussed by Brody,37 who considers the placebo response to be “A change in the patient's condition that results from the symbolic dimensions of a healing intervention rather than from its specific pharmacological or physiological properties”.

Moerman and Jonas4 have also discussed the importance of meaning and have suggested that it would be more helpful to use the term ‘meaning response’ to refer to placebo effects: We define the meaning response as the physiologic or psychological effects of meaning in the origins or treatment of illness; meaning responses elicited after the use of inert or sham treatment can be called the ‘placebo effect’ when they are desirable and the ‘nocebo effect’ when they are undesirable.

The term ‘nocebo’ might refer to a side effect of an inert medication. It may refer to increased anxiety following poor communication such as news that has been broken badly. At its most extreme, it is possibly represented by a voodoo curse in which a person, believing that he/she has been bewitched, can become ill or even die.38 This further illustrates the power of belief and that beliefs are influenced by culture.


Crucial aspects of the healer-therapist interaction such as trust, hope and empathy can be scientifically studied and Benedetti considers a placebo effect to be a ‘psychobiological phenomenon’.1 There is now a considerable body of research concerning the mechanism of action of placebo effects.39 Thus, “different psychological states and social stimuli can activate neurotransmitters and neuromodulators that bind to the same receptors to which drugs bind and can trigger biochemical pathways that are similar to those activated by pharmacological agents”.1 Conditioning also plays a role in the placebo response, for example, if a medication has been taken to good effect in the past, then this will influence the future response to the medication.

Much understanding of the mechanisms of placebo comes from studies concerning pain. A classic study by Levine and Fields40 demonstrated that the effect of placebo analgesics could be reversed by the opioid antagonist naloxone. It was therefore postulated that placebo analgesics release the body's natural opioids (endorphins). This remains the current hypothesis with further experimental evidence for the role of endogenous cholecystokinin (CCK); “Placebo analgesia is under the opposing actions of promoting endogenous opioids and inhibiting endogenous CCK.”1 Acupuncture is associated with the release of endorphins. However, this could result from its non-specific effects rather than being a specific effect of needling.

All treatments, whether ‘orthodox’, ‘complementary’ or ‘alternative’, share some common factors. Benedetti1 describes four components to the ‘healer’ (doctor, therapist)–patient interaction:

  • ‘feeling sick’, which recognises the importance of emotions in symptom genesis;

  • ‘seeking relief’; that is, the social and cultural factors and beliefs that influence this behaviour;

  • ‘meeting the therapist’; that is, whoever may be assigned a healing role in that particular culture;

  • ‘receiving the therapy’, in particular the expectations of the patient and the interactions between the patient and the ‘healer’.

Implications for palliative care

Clinical practice

An important aspect of palliative care practice is the relief of symptoms with an emphasis on a holistic approach using interventions that are evidence based where possible. In this schema, pain, for example, is considered to have physical, psychological, social and spiritual dimensions. It is recognised that while the prescribing of analgesics is an important aspect of pain relief their effect can be enhanced or diminished by other factors such as the care and concern of palliative care professionals and the quality of these interactions. An understanding of these effects allows a scientific rebuttal to the notion that palliative care is just about ‘being nice’ to people. ‘Being nice’ represents a ‘complex intervention’ that is “built up from a number of components which may act both independently and interdependently”41 that affects how patients perceive their symptoms. It is an integral part of treatment. Furthermore, an understanding of placebo effects allows an appreciation that complementary therapies such as acupuncture mediate and provide a ritual to foster beneficial therapeutic relationships that can enhance well-being.


Other specialties have tried to discern the different components of placebo effects. In a randomised controlled trial in patients with irritable bowel syndrome, patients were given different components of a complex intervention, including placebo interventions. It demonstrated that “the placebo effect can be progressively combined in a manner resembling a graded dose escalation of component parts. Non-specific effects produce statistically and clinically significant outcomes and the patient-practitioner relationship is the most robust component”.42 Potentially, palliative care provides an opportunity to research placebo effects, such as determining the characteristics of an effective palliative care professional (or volunteer), whatever their role in the multidisciplinary team. Might it also be possible to research the ‘hospice effect’; is morphine administered in a palliative care unit really more effective than on a hospital ward, as suggested in the 1970s, and if so why? The concept of ‘opioid responsiveness’ is well established in palliative care. But could it also be that the ‘placebo’ response varies between individuals as well and if so how can this be measured? Could it even be more significant than opioid responsiveness? If this were the case, might we be more relaxed about trying to find differences between opioid drugs,43 researching instead how to maximise placebo responses so as to minimise drug side effects? If antidepressants work mainly via placebo effects but can give already symptomatic patients drug side effects, should we prioritise research on non-drug treatments such as mindfulness for depression to a greater extent? In addition, complementary therapies could be researched in the palliative care setting. This would either provide evidence for the existence of specific effects of any therapies or add weight to the placebo theory while helping to gain insight into the question, ‘What is a good therapist?’


We have argued that placebo effects are important in palliative care and if harnessed appropriately can extend the effectiveness of interventions such as drugs or other treatments. In the pursuit of a rigorous science of palliative care, there is a danger of overlooking placebo effects if a simple intervention model of treatment is employed that ignores the importance of healing relationships and the context of care. The humane, psychologically informed approach by which palliative care is distinguished could be diminished and with it the effectiveness of some interventions. With greater understanding of the aspects of palliative care practice that produce or enhance placebo effects, it may be possible to develop even further the non-drug, psychologically informed approach of palliative care. An understanding of the science of placebo effects can help those clinicians and managers who are sceptical of the importance of behaviours, the patient–therapist relationship and environment of care, to appreciate these important aspects of practice.


We thank Dr Robert Twycross for comments on an earlier draft.


View Abstract


  • Contributors VL wrote the initial draft. This was substantially revised in several face-to-face meetings and email correspondence with SB, who wrote some of the sections and supplied additional reference material.

  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.