Oncologists’ knowledge, practices and ethical opinions about therapeutic misconception: a French national survey

Abstract

Background

Therapeutic misconception (TM) among research participants refers to the conflation of research goals (generating generalisable knowledge) with clinical care goals (making the best decisions for the participants). Considering the high volume of oncology research, oncologists frequently encounter TM.

Aim

To evaluate the knowledge, practices, and ethical concerns of French oncologists regarding TM.

Materials and methods

A questionnaire was developed to assess oncologists’ knowledge and practices concerning TM, then utilised in a national survey of French oncologists from 1 June to 14 July 2023. A descriptive statistical analysis of the responses (according to a Likert scale) was carried out.

Results

In total, 288 oncologists from various specialties responded to the survey. Initial knowledge of TM was low (16%), but after the definition was provided, 84% reported having encountered TM. Respondents indicated that they paid attention to the information given during participant inclusion; however, approximately half (46%) actively investigated the presence of TM, and 22% admitted to having encouraged TM at least occasionally. Attention to TM significantly declined over the course of study protocols. Awareness of TM, along with ethics education or participation in a research ethics committee, were identified as significant factors influencing responses. The acceptability of TM was nuanced, particularly in protocols recommended to patients receiving last-line treatments. Although 64% of respondents acknowledged a link between TM and dual roles as both investigator and physician, 78% opposed transferring investigative responsibilities to a non-referent oncologist.

Conclusion

TM is a widespread but still mostly unknown phenomenon which could easily be tackled for better outcomes for patients. This study revealed considerable variability in knowledge, practices, and ethical considerations related to TM among French oncologists. Enhanced education and ethical support are needed to improve awareness and foster appropriate behaviours concerning TM.

Clinical trial number

Not applicable.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12910-025-01260-y.

Introduction

Informed consent is an imperative in biomedical research, which ensures respect for patient autonomy [1–3]. It encompasses several elements. For a research study, informed consent requires the full disclosure of information through informed consent documents and research consultations. This disclosure includes, but is not limited to, potential risks, burdens, or benefits, primarily for the patient population, although not necessarily for individual research participants [4, 5]. Informed consent further requires that patients understand the disclosed information, which demands substantial efforts from sponsors to make the relevant documents comprehensible [6]. These processes depend on the voluntary participation of research participants, free from coercion or undue influence [7]. Informed consent must be regarded as an ongoing process, rather than limited to the act of signing a document. Patients must be assured of their right to withdraw at any time without consequences or an obligation to provide justification for the withdrawal [8].

Therapeutic misconception (TM) was first described four decades ago [9, 10] and is now defined as the phenomenon where patients consent to participate in biomedical research primarily because they believe it will benefit them personally, disregarding the fact that the main aim of the research is to obtain generalisable knowledge [11, 12]. Although patients may derive some benefit from participation in a trial, such benefits remain uncertain and do not constitute the primary objective of the research [13]. An important basis of TM is the expectation that health professionals act mainly for the benefit of their patients. As a result, proposals for study inclusion may be misinterpreted by patients as being in their best interest. Informed consent documents may also contribute to the occurrence of TM [14]. TM is more likely to arise when the referring physician holds dual roles as both the patient’s physician and an investigator in the study [1, 2, 15]. Current ethical debates address the relevance of involving investigators who are not the referring physicians for study participants, as well as the possible acceptability of TM in certain situations, depending on the type of research [16, 17].

Oncology is among the most active fields in biomedical research, and TM is regularly encountered by oncologists. TM occurs across all types of oncological research, from phase 1 trials involving innovative therapeutics to non-interventional studies and biopsy-based protocols [18, 19]. Little is known about physicians’ perceptions of ethical issues related to TM, particularly among oncologists. The present study conducted a national survey of French oncologists to examine their knowledge, practices, and opinions regarding TM. The primary objective was to assess the knowledge of French oncologists about the TM phenomenon. The secondary objectives were to investigate their specific practices for preventing and/or addressing TM during the inclusion of research participants and throughout study protocols, as well as their opinions regarding ethical considerations and the acceptability of TM.

Methods

The national survey was specifically designed by the Health and Research Ethics Centre in Bordeaux, France. The working group included four ethicists (three physicians and one nurse), an anthropologist, two pharmacology experts, and a resident in radiation oncology. After a literature review, a workshop was held to establish a consensus on a definition of TM and to identify a list of potential actions that physicians might take when faced with TM. General preventive measures focused on informing patients about the uncertainty or lack of individual benefits that may result from inclusion in a research protocol, as well as clarifying the distinction between care and research. Specific measures included one positive action (identifying signs of TM) and two negative actions (avoiding the inclusion of a patient exhibiting signs of TM and refraining from encouraging TM), as defined in the literature (Table 1).

Table 1.

Definition of therapeutic misconception (TM) and possible measures for its prevention or mitigation

Definition of therapeutic misconception (after Horng et al. (11)) Therapeutic misconception (TM) refers to a phenomenon where patients involved in biomedical research consent to participate based on the belief that their participation will necessarily result in direct personal benefit, either in quality of care or in prognosis. Research participants experiencing TM may lose sight of the fact that the primary aim of clinical research is to obtain generalizable knowledge—the research is intended to benefit future patients, not necessarily those enrolled in the study.

Proposed measures for investigators and physicians to prevent or mitigate TM Preventive measures  - Remind patients of the uncertainty surrounding potential personal benefits from interventional studies  - Emphasise the impossibility of deriving personal benefits from non-interventional studies  - Clearly differentiate between care procedures and those specifically related to research Specific measures Positive actions  - Seek indications of TM Negative actions  - Include patients without excluding the possibility of TM  - Actively encourage TM

The THErapeutic MISconception (THEMIS) survey comprised a three-part questionnaire. In the first part, French oncologists were asked whether they were familiar with the concept of TM. Respondents were then given a definition of TM and asked whether they had encountered TM in their practice. They were also asked to estimate the frequency of TM among their patients, in 10% increments. The second part of the questionnaire explored specific measures for addressing TM, based on the positive and negative actions identified in the literature and during the workshop. Questions were posed regarding practices at two distinct times: during patient inclusion and throughout the conduct of research protocols. In the third part, respondents were asked to provide their opinions regarding ethical considerations related to TM, particularly the impact of dual roles as physician and investigator, and the tolerability of TM in different types of studies. A preliminary version of the survey was subjected to cognitive testing with eight radiation oncologists and five radiation oncology residents at the University Hospital of Bordeaux, France. Based on the feedback, the final version of the study protocol included 26 questions or proposals assessed using Likert scales for frequency or agreement (“never,” “rarely,” “sometimes,” “often,” or “systematically”; “totally disagree,” “rather disagree,” “I don’t know,” “rather agree,” or “totally agree”).

Additional data were gathered to evaluate the following influential factors: age (in 10-year increments), gender (female, male, or other), speciality (medical oncologist, radiation oncologist, haematologist, oncology surgeon, or organ specialist with oncology activity), type of institution (university hospital, general hospital, comprehensive cancer centre, or private practice), current or previous experience with a research ethics committee, roles held in professional research, education in research ethics, and validation of good clinical practices [20]. The THEMIS survey is available in Appendix 1.

Statistical analysis

Differences between categorical variables were assessed using chi-square tests, with the type I error rate set at 5%. The Kolmogorov test was applied to evaluate differences in the distribution of opinions among respondents before and after the inclusion of patients in a research protocol. All statistical analyses were performed using the latest version of R (R Core Team, Vienna, Austria). For analysis, agreement levels (“strongly disagree,” “somewhat disagree,” “somewhat agree,” and “strongly agree”) were grouped into the binary variable “agree/disagree.” Similarly, frequency responses (“never,” “rarely,” “often,” and “systematically”) were categorised into the binary variable “rare/usual.” Past or current involvement with research ethics committees, along with any specific training in research ethics or informed consent, was combined into a single variable representing overall experience in ethics and used in subsequent analyses. Participants were not required to answer all items in the questionnaire. Each question was analyzed based on the number of respondents who answered it. Denominators therefore vary across items. Missing data were not imputed.

The survey was distributed by mail to reach as many oncologists as possible. French medical learning societies and professional networks in various oncological fields forwarded the questionnaire to their mailing lists. All respondents were fully informed about the study and provided consent to complete the questionnaire. All data were anonymised and protected in accordance with current regulations. The THEMIS study received approval from the local Research Ethics Committee– Reference CER-BDX 2023–52).

Results

The THEMIS survey was completed by 288 oncologists within 7 weeks of distribution (1 June 2023 to 14 July 2023). Respondents mainly were medical oncologists (36.9%) or medical specialists with a role in oncology (27.2%). Most worked in a university hospital or a comprehensive cancer centre (47.6% and 33.8%, respectively). The largest proportion (46.2%) of professional research activity was reported to range between 10% and 25%. The vast majority of respondents had completed their Good Clinical Practice (GCP) certification (89.9%). Participant demographics and characteristics are presented in Table 2.

Table 2.

Population characteristics

		N (%)
Total		288 (100%)
Specialty	Medical Oncology	106 (36.9%)
	Radiation Oncology	45 (15.7%)
	Other medical speciality	78 (27.2%)
	Haematology Oncology	30 (10.8%)
	Surgical Oncology	27 (9.4%)
Institution	University hospital	136 (47.4%)
	Cancer Comprehensive Centre	97 (33.8%)
	General Hospital	31 (10.8%)
	Private practice	23 (8%)
Proportion of research in professional activity	< 10%	77 (26.7%)
	10–25%	133(46.2%)
	25–50%	57 (19.8%)
	50–75%	12 (4.2%)
	> 75%	9 (3.1%)
Past or current member of a Research Ethics Committee	Yes	43 (14.9%)
	No	245 (85.1%)
GCP passed	Yes	258 (89.9%)
	No	29 (10.1%)
Education in Ethics/Research Consent	Yes	128 (44.4%)
	No	160 (55.6%)

In total, 241 of 287 respondents (84.0%) reported being unaware of TM. After they had been given a definition of TM, most respondents either somewhat agreed (174/287, 60.6%) or strongly agreed (66/287, 23.0%) that TM was present in their professional activities. Additionally, 135/287 (47.0%), 81/287 (28.2%), and 3/287 (1.0%) of respondents indicated sometimes, often, and systematically encountering TM during consultations for research inclusion. Estimates of patients exhibiting TM considerably varied among respondents (Fig. 1).

Fig. 1.

Estimation of patients presenting with therapeutic misconception according to responders

Regarding preventive measures against TM, most respondents indicated that they often (53/276, 19.2%) or systematically (202/276, 73.2%) informed patients about the uncertainty of receiving benefits from interventional studies. Similarly, most respondents reported that they often (51/278, 18.4%) or systematically (176/278, 63.3%) informed patients about the absence of direct benefits from non-interventional studies at the time of inclusion. Differences between research and care procedures were explained often (71/279, 25.5%) or systematically (162/279, 58.1%) during the inclusion process. With respect to specific measures addressing TM, only 129/282 (45.8%) of respondents reported often or systematically seeking clues that suggested TM at inclusion. Few or no respondents reported that they had included patients with clear signs of TM or actively encouraged TM. However, 62/277 (22.3%) of respondents acknowledged that they had sometimes or often encouraged TM in patients at the time of inclusion. Only 41/273 (15.0%) of respondents stated that they had never included patients exhibiting TM. Actions intended to mitigate TM varied depending on the time of inclusion and the research protocol. Preventive information provision significantly decreased between inclusion and subsequent stages of the research protocol (Fig. 2).

Fig. 2.

Results concerning measures towards Therapeutic Misconception– comparison between time of inclusion and the protocol (binary statistical analysis)

Concerning ethical considerations, a majority of respondents agreed that TM impairs informed consent (157/236, 66.5%), whereas only 119/229 (52.0%) indicated that TM should not be tolerated under any research protocol. Similarly, most respondents agreed that TM was exacerbated by the dual roles of physician and investigator (146/230, 63.5%). However, only 61/233 (26.2%) of respondents somewhat or strongly agreed that the referring physician and the study investigator should be independent. Finally, TM toleration was considered somewhat or entirely unacceptable for placebo-based protocols by a large majority of respondents. Opinions were more divided for other types of studies, particularly those investigating last-line treatments (Fig. 3).

Fig. 3.

Results concerning the tolerability of Therapeutic Misconception according to the type of protocols

Awareness of TM prior to the survey significantly influenced the estimated frequency of TM situations (P < 0.001), active identification of TM at the time of inclusion and during research protocols (P = 0.009 and P < 0.001, respectively), and the tolerability of TM in studies involving last-line therapy. Specifically, 66.7% of respondents aware of TM before the survey disagreed with its tolerability, compared with 46.8% of those unaware of TM (P = 0.04). Participation in a research ethics committee or education in research ethics also affected the estimated frequency of TM situations (P = 0.03) and active identification of TM at the time of inclusion (P = 0.04). No significant differences were observed for other potential influencing factors.

Discussion

A real but insufficiently known concept

To our knowledge, this is the first national-scale study examining perceptions of TM among oncologists. Although TM was first described over 40 years ago, and despite numerous articles addressing TM in the context of oncology, a large majority of French oncologists were unaware of TM [9]. However, after they had been given a definition, most respondents acknowledged that TM is a reality in their practice. Although the concept of TM itself does not appear to be contentious, its dissemination and integration within the French oncology community remain limited and require further attention. Literature discusses debates surrounding the definition of TM, as well as the level of granularity that may complicate the dissemination and integration of this concept in oncological practice [11, 12, 21, 22]. Our analysis revealed substantial heterogeneity in oncologists’ perceptions of the frequency with which patients exhibited TM. Some studies have reported high TM prevalence (50.5–74%) in clinical research, depending on the methods used for measurement [23–26]. Awareness of TM prior to this study significantly influenced the estimated frequency of patients exhibiting TM, as well as active efforts to identify TM at the time of study inclusion and during research protocols. These findings suggest a need for education on TM.

Concerns about information at time of inclusion

Our analysis revealed concerns among French oncologists regarding the adequacy of patient information at the time of study inclusion. As TM can arise even when attention is specifically focused on information, we examined active measures to prevent or mitigate TM [27], which yielded mixed results. The provision of preventive information and attention to TM significantly decreased between the time of patient inclusion and subsequent stages of the study protocol. This suggests a decline in investigators’ concern about information after initial consent is obtained. These findings question the perspective of French oncologists on the temporality of research consent, which should be considered an ongoing process. The possibility of withdrawing consent and the obligation to provide relevant new information throughout the research protocol reflect this ongoing nature [20, 28]. Furthermore, the literature indicates that patients’ memory of information is often poor, suggesting that they would benefit from regular reminders during the research protocol [29, 30]. Patients enrolled in long-term protocols, which carry a high risk of poor recall of information provided at inclusion (e.g., prostate cancer trials involving prolonged treatment phases and extended follow-up periods), could benefit from a consent renewal consultation. In such consultations, all protocol information would be revisited, offering patients an opportunity to reconsider their participation or withdraw. The frequency of these consultations could be tailored to the type of protocol. This approach may also help physician–investigators maintain their focus on patient information and TM throughout the study.

Heterogeneity of ethical considerations about TM

The impact of TM on the quality of consent remains a subject of debate in the bioethical literature [31, 32]. Some studies suggest that TM can affect patients’ willingness to participate in biomedical research and should therefore be carefully avoided [12, 33, 34]. However, an interventional study designed to reduce TM at the time of study inclusion demonstrated effectiveness without negatively impacting patients’ consent to hypothetical protocols [27]. In our study, the tolerability of TM appeared contentious. Two-thirds of respondents felt that TM impaired informed consent, whereas one-fifth acknowledged at least occasionally encouraging TM during the inclusion process.

Our results must be interpreted within the context of ongoing debates about TM in the bioethical literature. In addition to TM, a range of concepts describing patient perceptions of the expected benefits of research participation have emerged over the past two decades. Clinicians may find it difficult to determine which of these concepts are ethically problematic. Differentiating between TM and therapeutic misestimation (i.e., between dispositional and unrealistic optimism) and identifying which constitutes a genuine ethical concern remain challenging tasks for oncologists [11, 34, 35]. Moreover, the relevance of the TM concept itself is debated. Some authors argue that TM is an unhelpful concept that distracts investigators and research ethics committees from conducting a contextual analysis of protocol-specific factors and patient information needs [36–39].

Our findings indicate that the tolerability of TM varies according to the type of study, the patient population involved, and even the stage of care during which the study is conducted. Early-phase oncology protocols are frequently discussed in the literature, particularly concerning patients’ estimations of risks and benefits [18, 40]. Phase I trials are typically conducted in cases of advanced cancers that no longer respond to standard therapies and may encourage TM. These studies are often regarded as a last hope for patients, where the possibility of personal benefit becomes a common motivation for participation [41, 42]. In such situations, participation may indeed represent the best available option for patients, provided that informed consent quality remains high (Miller and Joffe, 2008). End-of-life patients in phase I trials may also believe that the research environment offers valuable support during this difficult period [43]. Intriguingly, our results indicated that the tolerability of TM was most frequently discussed in the context of last-line treatment protocols.

Our study confirmed a paradox in the dual roles of the referring physician and study investigator. Although a substantial majority of survey respondents recognised that these dual roles increase the risk of TM, only one in four considered employing an independent investigator as an appropriate method for reducing TM occurrence. This paradox may arise from the mutual desire of patients and physicians to preserve their relationship through the research process [16, 17, 44].

Dual investigators/physicians play a crucial role in TM

The heterogeneity of practices and considerations related to TM, other than the provision of information at the time of inclusion, emphasises broad subjectivity among oncologists. It is particularly important to address the subjectivity of ethical judgment within the controlled environment of biomedical research because this cannot be mitigated solely by the competence of a research ethics committee, unless TM is explicitly encouraged in the informed consent documents. A major and delicate ethical responsibility remains for physician–investigators: to evaluate patients and studies individually to discern when supporting hope is both realistic and non-prejudicial, and when such hope transitions into TM. Our findings highlight the need to raise awareness among physician–investigators about this issue. Ethicists have a crucial role to play in this process. The objective should be to inform and enhance ethical judgment, thus avoiding a binary approach. Furthermore, the evolving landscape of biomedical research, particularly considering the increasing prevalence of embedded research, requires targeted ethical support for physician–investigators [17, 45, 46].

Limitations of this study

This study had several limitations. The response rate appeared low, but it is challenging to evaluate accurately because the number of oncologists involved in research activity is unknown. Similarly, evaluations of participation through the effectiveness of mailing lists from scientific societies are inadequate because oncologists may belong to multiple societies simultaneously. Methodically, the use of a survey introduces the risk of interpretation biases. This risk was minimised as much as possible through prior cognitive testing. A qualitative approach could complement this study to address the topic. Although surveys are limited in their capacity to analyse complex issues, this method effectively reaches a broad population, and the anonymity of the questionnaires encourages the disclosure of potentially problematic opinions. Finally, respondents were not required to answer all items in the questionnaire. Missing data were not imputed; each question was analysed based on the number of respondents who provided an answer, and denominators therefore vary across items. The overall proportion of missing responses was low and did not show any systematic pattern. Furthermore, as some items addressed ethically sensitive practices, the possibility to abstain from answering may have reduced social desirability bias and fostered more sincere responses.

Conclusion

Despite a low level of awareness within the French oncologist community, TM is recognised as a realistic concept by physician–investigators, who reported allocating substantial attention to providing patients with sufficient information regarding the benefit/risk balance at the time of study inclusion. However, this attention appears to diminish during the course of research. Our study revealed variability in the attention given to TM by physician–investigators, as well as differences in their ethical evaluations of its tolerability depending on the type of study. Awareness of TM, involvement in a research ethics committee, or receipt of specific ethics training influenced the perspectives and actions of physician–investigators, highlighting the need for education and ethical support for oncologists in addressing challenges posed by TM.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

TH, SC, SM, DB, HH, MCS, DD, PJM designed the protocol, TH collected and interpreted data, PJM performed statistical analysis, TH, DD, PJM wrote and corrected the article; All authors contributed to the final version of the article.

Funding

None.

Data availability

Data are available after reasonable request.

Declarations

Ethics approval and consent to participate

All participants were informed before survey completion and consented to this study. This study adhered to the Declaration of Helsinki. This study got the approval of the Research Ethics Committee of the University Hospital of Bordeaux– CER-BDX 2023–52.

Consent to publication

Not applicable.

Competing interests

The authors declare no competing interests.