Placebo responses are observed in the treatment of psoriatic arthritis (Erre et al., 2022), functional gastrointestinal disorders (Enck and Klosterhalfen, 2020), cancer-related fatigue (Roji et al., 2020), cough (Lee et al., 2005, 2019), uncontrolled persistent asthma (Luc et al., 2019), female sexual dysfunction (Weinberger et al., 2018), and in various other physiological systems and medical conditions (Schedlowski et al., 2015). However, their effect sizes seem to vary between systems and conditions. The effects of positive expectation are particularly strong in the fields of pain and depression, where up to 70% of overall treatment effects can be attributed to placebo effects (Moore et al., 2014; Matsingos et al., 2024). Not only do these effects occur in patient-reported outcomes, but can also be shown in physiological parameters, such as heart rate, blood pressure, coronary diameter, gastric motility, lung function, blood sugar levels, or immune function (Meissner, 2014; Park et al., 2016; Kirchhof et al., 2018).

The progress in placebo research in the last decades was ground breaking (Schedlowski et al., 2015; Colloca and Barsky, 2020). According to expert consensus, the overall improvement of health outcomes after administration of an inert treatment in placebo-controlled RCTs is defined as placebo response (Evers et al., 2018). The underlying neurobiological mechanisms are increasingly identified. For example, placebos can induce the release of target-specific transmitters such as dopamine in Parkinson’s disease (Lidstone et al., 2010) or endogenous opioids against pain (Eippert et al., 2009). The placebo response accounts for substantial parts of the drug effect in several conditions, such as insomnia (Winkler and Rief, 2015) or hypertension (Wilhelm et al., 2016). Trials in pain (Dolgin, 2010) or mental disorders (Rief et al., 2009b; Weimer et al., 2015) even show that the responses to placebo treatments and active treatments can be quite similar. While the placebo response encompasses also confounding influences such as the natural course of symptoms or statistical phenomena like regression toward the mean (Bland and Altman, 1994; Hengartner, 2020), the placebo effect refers solely to the symptom change attributable to placebo mechanisms, both psychological and neurobiological. Placebo mechanisms are involved in any kind of (active) treatment, and not restricted to inert pills (Benedetti, 2008). The relationship between the placebo effect and regression toward the mean is unclear. Accumulating clinical and experimental evidence indicates that expectation can substantially modulate the efficacy and tolerability of active medical treatments, including pharmacotherapy. Positive treatment expectation has, for instance, been shown to double the analgesic effect of the opioid, remifentanil (Bingel et al., 2011), and to substantially enhance the effect of the acute antimigraine drug, rizatriptane (Kam-Hansen et al., 2014). Therefore, it seems desirable to utilize placebo mechanisms in clinical practice in order to improve health outcomes.

The pivotal role of expectations as one of the most important factors contributing to placebo responses is widely agreed upon in the field of placebo research. Treatment expectations have a strong influence on medical and psychological health outcomes across a wide range of medical treatments (Petrie and Rief, 2019), best illustrated by RCTs involving placebo treatment groups. Since a placebo treatment does not include a biochemically active compound, changes in health outcome within the placebo groups (the placebo response) cannot be explained by specific properties of a drug, but by patients’ expectations regarding the drug treatment they think they received.

Recently, Bingel (2020) proposed a model of treatment expectation. According to the model, expectations shape treatment outcomes and are determined by several interacting factors. Treatment expectations themselves are shaped by prior information and context factors—both connected to associative learning—as well as communication (shown in Figure 1). Consequently, Bingel (2020) recommends that systematically targeting treatment expectations should be reintegrated into the biomedical model and current evidence-based treatment regimens. However, while this is a subject of intensive research, it is not yet part of clinical routine, beyond the field of pain, where the systematic use of placebo mechanisms has already been included in treatment guidelines [Deutsche Interdisziplinäre Vereinigung für Schmerztherapie (DIVS), 2008; Klinger, 2010].

Importantly, expectation can also have a negative effect on treatment outcome. Negative expectation, which is linked to the occurrence of unwanted side effects in placebo groups, is being increasingly acknowledged (Colloca and Finniss, 2012; Bingel, 2014, 2020). In fact, the majority of adverse events and symptoms reported by patients in clinical trials may not be caused by the drug itself. This is suggested by the observation that adverse events in placebo arms of drug trials not only resemble those of active treatment in frequency of occurrence, but also in the nature of the symptoms (Rief et al., 2006, 2009a; Amanzio et al., 2009). These effects related to negative treatment expectation are referred to as nocebo responses.

The past century has seen ground breaking advances in the development of new diagnostic tools and increasingly personalized pharmacological (and other) treatments, ranging from the discovery of antibiotics to the use of individually tailored immune treatments for cancer based on the genomic profile of the tumor. These developments have undoubtedly changed the lives of many for the better. Breakthrough achievements in biomedical research now allow us to specifically address underlying pathologies of many diseases and RCTs test whether these new treatments improve health over and above natural fluctuations and other factors unrelated to the intervention. Evidence-based medicine has taken the place of ancient medicine that, due to a lack of the knowledge we have today, largely had to rely on hope and the expectation both of the patient and the healer that an intervention would promote or restore health (Kaptchuk, 2011). Although far from being understood, these forces had an impact on health outcomes and have been considered a pillar of medicine since primeval times. With the advent of modern medicine and its focus on disease-specific pathologies, interest in their influence ceased—ironically, at the peril of those developments that were meant to replace them.

The aim of this narrative review is to use the far-reaching implications of recent insights into the nature of placebo and nocebo effects to elucidate how placebo mechanisms are systematically targeted in homeopathy, often in a much better way than in evidence-based medicine. We postulate a scientifically grounded reintegration of the powerful elements of homeopathy into the biomedical model of health and treatment outcomes in evidence-based medicine and pharmacology to assure that they undergo a scientific evaluation and quality control before implementation in clinical practice in an ethically responsible way.

Taken together, it is of great importance to empower health care practitioners with knowledge on how homeopathy utilizes well-known placebo mechanisms. Practitioners need tools and practices to harness the power of treatment expectation in a way that is grounded in the evidence-based principles of modern medicine. This will optimize treatment efficacy and adherence in large populations and ultimately reduce health care costs while improving patients’ care.

Prior information about a treatment has been shown to induce or enhance treatment expectations (Bingel, 2020). Expectations are determined by learning mechanisms (e.g., classical conditioning), verbal information provided by healthcare professionals, and information from other sources like the media (Petrie and Rief, 2019).

Verbal information regarding the expected effects of treatments (e.g., from health care professionals) is omnipresent in health care. Direct verbal communication, as well as written information (e.g., medication leaflets), have been shown to shape patients’ treatment expectation and treatment effectiveness (Camerone et al., 2021; Sondermann et al., 2021). In addition, digital and social media play an important role in providing health-related information (Yetman et al., 2021). This is crucial, as placebo and nocebo responses can be triggered by verbal information provided by non-medical persons such as other patients or people in similar health-related situations (Yetman et al., 2021). To induce a positive treatment expectation it might even be enough to see others receiving an effective treatment (observational learning; Petrie and Rief, 2019).

How to communicate possible medication side effects is a prominent objective of nocebo research (Leibowitz et al., 2021). One promising way is to specifically frame side effects as onset sensations (i.e., symptoms indicating that the treatment is starting to work), which enables practitioners to truthfully inform, while modifying expectations about side effects (Wilhelm et al., 2018; Fernandez et al., 2019; Howe et al., 2019).

Context factors (e.g., the white coat of the physician, the medical equipment, the medical facility, the shape/size/color of a pill, or the act of swallowing a drug, as well as internal cues like symptoms from the patient’s disease or current mental state), could trigger expectations (Petrie and Rief, 2019). Associative learning can also be involved in developing placebo effects by the association of beneficial drug effects with neutral stimuli (e.g., environmental or gustatory cues), which subsequently evoke the benefit without administration of the active drug (Colloca and Barsky, 2020; Hadamitzky et al., 2020; Aulenkamp et al., 2023). The impact of learning on treatment expectation and treatment outcomes has been investigated in experimental studies showing that positive experiences with treatments in the past amplify placebo responses of subsequent treatments (Colloca, 2019). while prior negative experiences make future treatments less effective (Cadorin et al., 2020).

The quality of the clinician-patient communication in terms of empathy (Kaptchuk et al., 2008), interaction style (Czerniak et al., 2016; Rossettini et al., 2020), and perceived clinical competence of the treatment provider (Howe et al., 2017) increases the placebo effect, therefore improving health outcome. However, even the sheer number of study visits boosts antidepressant placebo effects, as shown in elderly patients with depression (Rutherford et al., 2014) and placebo analgesia in chronic pain patients (Vase et al., 2015). Quality and quantity of clinician-patient communication, in particular communication in an empathic and warm way, prevents nocebo effects (Howe et al., 2017; Zech et al., 2022). However, careful, empathic, and patient-centered communication requires time; yet, the medical care compensation system in many countries forces physicians to see too many patients in too little time. A cross-sectional study in six European countries reported a mean length of consultation in general practice of 10.7 min, varying from 7.6 min in Germany to 15.6 min in Switzerland (Deveugele et al., 2002). Other studies document that physicians in Germany interrupt their patients after an average of 18 s, in the U.S. after 23 s, and that in the U.S. 50% of patients leave the doctor’s office not having understood what the physician has told them (Bodenheimer, 2008).

Administering placebos in clinical practice is constrained by ethical and legal concerns regarding the use of deceptive information traditionally associated with a placebo intervention (Miller and Colloca, 2009; American Medical Association, 2012). However, a promising approach to understand placebo effects beyond deception is the use of open-label placebos (OLP), which is considered ethically compatible with the American Medical Association (AMA) principles of medical ethics (American Medical Association, 2008; Miller and Colloca, 2009; Blease et al., 2016). In OLP application, patients are explicitly informed that they receive an inert placebo pill combined with the OLP rationale developed by Kaptchuk et al. (2010). In most OLP studies, this rationale contains information about (1) the effectiveness of placebos, (2) possible mechanisms (i.e., classical conditioning), (3) the role of expectations, and (4) the need to take the pills faithfully, and is provided interactively in the context of an empathic and supportive patient-provider interaction. Several recent RCTs suggest that OLP can effectively improve symptoms in patients suffering from irritable bowel syndrome (Kaptchuk et al., 2010), major depressive disorder (Kelley et al., 2012; Nitzan et al., 2020), allergic rhinitis (Schaefer et al., 2016, 2018), chronic lower back pain (Carvalho et al., 2016; Kleine-Borgmann et al., 2019), cancer-related fatigue (Hoenemeyer et al., 2018; Zhou et al., 2018), and psychological well-being in stressed students (Kleine-Borgmann et al., 2021; Winkler et al., 2023). A recent meta-analysis of 11 OLP studies found a large positive effect size for symptom reduction across different pathologies in OLP groups compared to untreated controls (von Wernsdorff et al., 2021). Furthermore, positive treatment expectancy has been shown to predict OLP effects on several psychological well-being outcomes (El Brihi et al., 2019). Similar to deceptive placebo effects, classical conditioning and expectancy are considered to at least partially account for the OLP effect, but the mechanisms underlying the effectiveness of OLP remain to be elucidated (Charlesworth et al., 2017; Colloca and Howick, 2018; Kaptchuk, 2018; Kaptchuk and Miller, 2018).

Considering that some participants report no or only minor expectancy for improvement before the treatment (Carvalho et al., 2016; Kaptchuk, 2018), unconscious expectations induced by context variables and bodily sensations (e.g., the sensory perception of swallowing a pill) might also contribute to the OLP effect. Furthermore, placebos presumably retrieve classically conditioned pharmacological memories (i.e., the intake of a placebo pill and its sensory features serve as conditioned stimuli that elicit previously acquired conditioned response such as symptom improvements, irrespective of a deceptive or non-deceptive administration; Colloca and Howick, 2018). This assumption is supported by experimental studies investigating conditioned placebo analgesia in standardized heat pain paradigms, in which the conditioned placebo response persisted even after the placebo manipulation was revealed to the participants (Chung et al., 2007; Schafer et al., 2015). Neuroimaging evidence exists for conditioned placebo responses entailing non-conscious conditioned stimuli (Jensen et al., 2012, 2015), and cognitive theories like predictive coding and embodied cognition (Frith, 2011; Shapiro, 2011; Clark, 2016) support the notion of unconscious processes being involved in the OLP effect (Ballou et al., 2017; Kaptchuk, 2018). According to the concept of embodied cognition, the physical interaction with the environment affects cognition and vice versa (Shapiro, 2011, 2014). For instance, the OLP treatment is provided in a professional “medical” context, therefore the same sensory and motor information is impinging on the brain as during an active medical treatment associated with symptom improvement (Ballou et al., 2017; Kaptchuk, 2018). According to predictive coding, even if the brain predicts that a placebo will not lead to symptom improvement, it might adjust the prediction according to the sensory signals (e.g., intake of a capsule in a medical environment), resulting in the release of neurotransmitters in the brain, contributing to symptom improvement (Ballou et al., 2017; Colloca and Howick, 2018).