Understanding and Attitudes toward Cancer Clinical Trials among Patients with a Cancer Diagnosis: National Study through Cancer Trials Ireland

Cancer clinical trials (CCTs) are critical to translation and development of better therapies to improve outcomes. CCTs require adequate patient involvement but accrual rates are low globally. Several known barriers impede participation and knowing how subpopulations differ in understanding of CCTs can foster targeted approaches to aid accrual and advance cancer treatments. We conducted the first nationwide survey of 1089 patients attending 14 Irish cancer centres, assessing understanding of fundamental concepts in CCT methodology and factors that influence participation, to help tailor patient support for accrual to CCTs. Two-thirds (66%) of patients reported never having been offered a CCT and only 5% of those not offered asked to participate. Misunderstanding of clinical equipoise was prevalent. There were differences in understanding of randomisation of treatment by age (p < 0.0001), ethnicity (p = 0.035) and marital status (p = 0.013), and 58% of patients and 61% previous CCT participants thought that their doctor would ensure better treatment in CCTs. Females were slightly more risk averse. Males indicated a greater willingness to participate in novel drug trials (p = 0.001, p = 0.003). The study identified disparities in several demographics; older, widowed, living in provincial small towns and fewer years-educated patients had generally poorer understanding of CCTs, highlighting requirements for targeted support in these groups.


Introduction
Cancer Trials Ireland is the largest collaborative cancer research infrastructure in Ireland. Through its network of cancer centres, we conducted the first nationwide study exploring attitudes and understanding of cancer clinical trials (CCTs) among adult oncology patients in Ireland. We also sought to determine factors and supports patients considered important when making decisions about participation. Clinical trials are critical to the translation and development of better oncology therapies. Accrual rates to oncology trials are low globally, with an estimated 1 in 10 clinical trials stopped as a result [1]. In Ireland, a key performance indicator of the National Cancer Strategy 2017-2026 is to increase the percentage of patients on cancer therapeutic clinical trials from 3% to 6% by 2020 [2]. Professional organisations and patient advocacy groups encourage patients with cancer to consider CCTs as a routine component of oncology care [3][4][5].
To achieve an increase in CCT accrual, it is recognised that practical barriers at the site/organisation, physician/provider and patient/community levels must be addressed [6][7][8]. Physicians and providers often have insufficient protocols or those they have exclude significant populations of patients such as the elderly or those with co morbid illnesses. Patients express a willingness to participate in CCTs, yet only small percentages enrol [9]. Cancer trials cannot be conducted unless patients are willing and able to participate. Sociodemographic disparities have shown lower rates of participation in older populations, prompting an announcement by the American National Institute of Health in 2018 for researchers to include participants across the lifespan [10][11][12][13][14].
The Irish Platform for Patients' Organisations, Services and Industry (IPPOSI) study to ascertain Irish public awareness and understanding of clinical research highlighted poor understanding of the terms "clinical research" and "clinical trial" and that the public while supportive of CCTs were uncomfortable with participating in them [15]. A follow-up study showed greater willingness to participate in non-therapeutic cancer clinical research but less so for therapeutic clinical trials.
We previously conducted a study that identified lack of a CCT option as the main reason for failure to recruit in Ireland [16]. Collectively, these studies showed the need for this current study to identify how patients understood CCTs and what barriers they perceive to participation. Identification of these barriers and efforts to remove them represent critical objectives for cancer investigators, patients, and stakeholders across the research spectrum. Accrual to clinical trials influences the timely availability of new therapies and impacts on patient outcomes. It is important to recognize factors that can influence accrual to clinical trials to offer the opportunity to modify such factors where possible and tailor information to increase accrual in line with Ireland's National Cancer Strategy.

Method
This is a descriptive study of patients attending 14 Irish cancer centres from April to November 2016, aged ≥18 years with a clinical diagnosis of malignancy, able to complete a standardised questionnaire independently or with the help of a friend, relative, or oncology staff member. Questionnaire completion implied informed consent. The study received full ethical approval at all participating sites.
The questionnaire was divided into the following sections: (A) demographic details; (B) objective measure of wellbeing using the EORTC EQ-5D-5L instrument; (C) cancer diagnosis and treatment details; (D and E) statements about clinical research and CCTs, rated on Likert scales a; (F) questions relating to CCT participation and decision making; (G) questions specifically targeted at those patients who had previously participated in a CCT.

Statistical Analysis
Descriptive statistics are presented as means, medians, interquartile ranges and percentages. Categorical data associations were assessed using Pearson's X 2 or Fisher's exact (when numbers in cells were <5) statistical tests. Variables were tested against patient demographic characteristics (Table 1) and offer of participation to a clinical trial, to explore for differences that may influence decisions regarding participation in CCTs. No adjustments to significance levels were made. Only significant results at p-value ≤ 5% are presented. Other tests of associations were conducted but not presented in the results, as numbers were too small or they did not achieve statistical significance at p-value ≤ 5%. We performed a binomial regression to look at the effects of confounders. Due to small numbers, dimensions of wellbeing were collapsed into patients that had no problems and those that had any problem (i.e., slight + moderate + severe + unable were combined). EORTC EQ-5D-5L results are expressed as medians. Missing data are included in relevant tables but were excluded from the analysis. Denominators varied as some questions were only applicable for certain group(s) of patients. Data were analysed using STATA version 15.0 (StataCorp, College Station Texas). English not native, no translator required 30 3 English not native sometimes/always need translator 6 1 * Irish Travellers are a traditionally itinerant wayfaring ethnic group. ** Tertiary level education = third-level education (in the Irish context, below tertiary level includes the following groups: no formal education, primary school, junior/intermediate certificate, and leaving certificate). *** Location groupings were approximated to location of cancer centres.

Demographic Details
A total of 1089 adult questionnaires were returned. Patients in the sample had a median age of 60 years (IQR; 50-69 years). Most patients were female (64%) and married or in a long-term relationship (71%), with 16% of patients living alone. Most patients were Irish (94%) and native English speakers (96%). Over half (58%) of patients had below tertiary levels of education, 30% were in some form of employment and 60% lived in an urban setting (Table 1).

Measure of Wellbeing Using the EORTC EQ-5D-5L Instrument
The median score on the visual analogue scale with endpoints of 100 being 'best health you can imagine' to zero being the 'worst health you can imagine' was 75 (IQR 60-85). We used the EQ-5D five dimensions (mobility, i.e., ability to walk, self-care, usual activity, pain/discomfort and anxiety/depression), scoring them into five levels (no problem, slight problem, moderate problem, severe problem and extreme problem). Dimensions were collapsed for analysis into two groups-patients with and without a problem. Most patients (70%) reported having no mobility problems (based on ability to walk) and levels of mobility problems increased with age (X 2 = 22.69; p < 0.0001) and were slightly higher in widowed patients (X 2 = 7.82; p = 0.050) even after adjusting for age. Self-care issues were reported in 10% of patients, with proportions slightly higher in those only educated to primary school level (X 2 = 17.47; p = 0.015). Over half (55%) of patients reported having pain/discomfort and 38% conveyed anxiety/depression. Patients aged 41-60 years reported more pain/discomfort and younger patients (<50 years) had slightly more anxiety/depression (X 2 = 14.51; p = 0.006, X 2 = 39.00; p < 0.0001). Females had more anxiety and depression than males (X 2 = 13.49; p < 0.0001). No statistical difference was observed between groups in carrying out the dimension 'usual activity' (Table 2).

Participation
Of the 1089 patients, 66% reported never being offered the option to take part in a CCT and only 5% (34/741) of those not offered asked about taking part. Of the 345 patients that were offered/asked to take part in CCTs, 79% participated. The majority (87%) of patients that received an explanation about CCTs understood it and 70% found the patient information leaflet easy to understand. However, 16% of patients invited to participate in CCTs reported not receiving an information leaflet. Just under half (48%) of patients made their decision on whether to participate in a CCT by themselves, 35% used help/support and 5% indicated that they would have liked help but did not have it. The main source of help came from family members (84%). For 53% of patients, participation in CCTs impacted positively on their quality of life, while 3% reported a negative effect. Only 33 patients (12%) were given the results of the published CCT. Information booklets and having a patient navigator were the main preferences for decision aids that participants felt may have helped when making a decision about CCT participation (Table 3).    Participants may have selected more than one source, and therefore data will not add up to 100%. Table relates to sections 'F and G' in questionnaire Due to rounding numbers and percentages may not add up precisely to totals.

Patient Understanding of Concepts Used in Clinical Trials
All patients were asked whether they understood the term "clinical or medical trial"-this follows finding from the IPPOSI report of the Irish public where misunderstanding between medical and clinical research was prevalent [15]. Based on a binary response (yes/no), 77% of all surveyed patients understood the term, and this compares with 73% of the Irish public. In the public study, awareness was lower in males, in younger age groups and in lower socioeconomic status groups. In our patient sample, we found no significant differences in understanding between males and females or between socioeconomic groups, but did find that levels of understanding of the term were lower in older (X 2 = 10.72; p = 0.003), widowed (X 2 = 12.68; p = 0.005) and patients living in small towns (X 2 = 7.98; p = 0.018).
Understanding of CCT methodology and clinical equipoise is known to cause problems for the public. When asked whether a randomised trial treatment is decided by chance, 39% of patients thought it was not and an additional 24% did not know. There were differences in understanding by age (X 2 = 30.23; p < 0.0001), ethnicity (X 2 =6.70; p = 0.035) and marital status (X 2 = 16.20; p = 0.013). Less than half (48%) understood that CCTs were not only an option when standard of care treatment had failed. Proportions that understood this was lower in older age groups (X 2 = 44.85; p < 0.0001), in single and widowed patients (X 2 = 19.37; p = 0.004) and the self-employed (X 2 = 27.24; p = 0.007). Fifty-eight percent of all patients and 61% of previous CCT participants thought that their doctor would ensure better treatment in a CCT. This was slightly higher in those with only tertiary education and in older age groups (X 2 = 22.08; p = 0.037, X 2 = 45.38; p < 0.0001), (Table 4).

Factors Influencing Patient's Decision to Participate in a Cancer Clinical Trial
The opportunity to advance cancer research (78%), possibility of feeling better and living longer (78%) and an oncologist recommendation (73%) were all positive influencing factors reported by patients considering CCT participation. Fear of the unknown and perceived risk of side effects can be a barrier to CCT participation. We found that 35% of patients reported fear of more side effects, 31% feared harm or death and 22% had a fear of being a human guinea pig as concerns for CCT participation. Of those that reported these fear factors, proportions were slightly higher in females (X 2 = 5.24; p = 0.022, X 2 = 8.40; p = 0.004). Females also expressed more concerns about additional hospital visits and distance to travel to participate in CCTs. More younger (<50 years) than older patients feared harm or death and had more concerns about whether the CCT treatment would work (X 2 = 18.93; p = 0.001, X 2 = 45.00; p < 0.0001), (Table 5).

Attitudes towards Cancer Clinical Trials
After receiving factual information about clinical research, the Irish public showed a willingness to engage in CCTs, but only a minority were willing to engage in therapeutic trials [15]. In our study, 66% of patient participants thought that cancer clinical research was a good idea, but this was lower in those aged >65 years (X 2 = 31.46; p < 0.0001), with this age group also less willing to participate in drug trials compared with younger patients. We found that 63% of patients would participate in CCTs with a new drug, but this was slightly lower (59%) for participation in a CCT with a new drug only tested on animals, and 70% indicated that they would be willing to participate in translational trials. More males than females would consider participating in CCTs to determine whether a new drug is better or the same as the usual treatment (X 2 = 14.57; p = 0.001), on drug trials with no standard treatment-only animal data (X 2 = 11.76; p = 0.003) and in translational trials (X 2 = 7.48; p = 0.024). The majority (89%) of patients were willing to supply personal information, an issue of concern raised by the Irish public [15], Table 6.

Trust in Supports (All Patients)
Using a scale from 0 to 10, with a score of 0 indicating no trust and a score of 10 indicating maximum trust, patients rated information sources. Oncologists and research nurse specialists scored highest, with mean scores of 9.4 and 9.0, respectively. GPs and patient information booklets scored 7.8 and 7.5, respectively. The internet and newspaper/magazine reports were the least trusted, with scores of 3.7 and 3.8, respectively.

Discussion
Based on availability and patient eligibility, CCTs are a treatment option for patients. Before a patient chooses to participate in a clinical trial, they must first know that this is an option and understand clinical research. The aim of awareness ensures that patients can ask their oncology teams about CCT options, and doing so will 'normalize' CCTs as part of treatment option discussions. Increasing understanding and awareness may therefore also increase accrual. This national study aimed to describe understanding and awareness in Irish oncology patients after a previous study of the Irish public identified concerns around knowledge and understanding of cancer clinical research. Dissatisfaction by the public with the Irish health service and support by the Irish public for Ireland to operate as a focal point for clinical research provides an opportunity for clinical trials if the barriers to participation can be addressed [15].
We found slightly more oncology patients understood the term 'clinical trial' than the Irish public and while understanding of clinical research was lower in younger age groups in the general population [15], we found the opposite in cancer patients with older patients having poorer levels of understanding. There was great variation in understanding of cancer clinical trial methodology. Demographic and socioeconomic disparities are known to occur in clinical trial participation, with much lower recruitment of elderly patients [9, 17,18]. We found uncertainty and a lack of understanding was more prevalent in older age groups, those that were widowed and in those with fewer years of education. Concern regarding treatment side effects is also a frequently reported reason for non-participation in CCTs [15,[19][20][21]. Contrary to other studies and findings from the Irish public study [19,21,22], we found that over two-thirds of oncology patients in the survey did not report fear of more side effects, harm or death as being negative factors in their decision making around CCT participation. Where fears were present, this was higher in younger rather than older age groups and in females.
Two-thirds of oncology patients in our study were not offered the option to take part in a CCT and only 5% of patients not offered a trial independently asked about CCT participation, a finding consistent with other studies [22,23]. This outlines the need for better CCT awareness by patients and more incorporation of CCT discussion by medical teams with patients when treatment options are being discussed. Of those that took up the offer of a CCT, 79% participated and an additional 6% accepted but were not eligible, while only 3% declined the offer. Willingness to participate was positively influenced by prior participation. Logistical concerns are often cited as a barrier to participation [9], and 20-25% of patients in our study had some logistical concerns this was more prevalent in females than males.
Patients are known to struggle with the concepts of chance and randomisation, have uncertainty about the appropriateness of CCTs for serious illnesses, and consider CCTs as valid treatment options only when standard treatment has failed [20,24]. Our study concurred with these findings. We found that the number of patients having difficulty understanding these concepts were more prevalent in older age groups. Difficulty with the concept of randomisation is a recognised factor affecting accrual to CCTs [15,[24][25][26][27]. Understanding and acceptability of clinical equipoise is important in determining whether patients consent to randomisation and accept the treatment allocation assigned to them [28]. In our study, 58% of patients and 60% of previous CCT participants had a "therapeutic misconception" [29], despite 87% of previous CCT participants reportedly understanding the trial explanation. This discrepancy underlines the importance of checking and ensuring patients knowledge and understanding of what they are being told. Consultations between patients and oncologists have highlighted failures to provide consistent sequences of information with rationale for randomising [30], and ensuring consistency in explanations of clinical trial methodology would address some of these discrepancies.
The pharmacy industry is the largest sponsor of clinical trials, mainly undertaken on unapproved drugs to support an application for marketing approval [9]. A principal barrier to CCT uptake expressed by the Irish public was participation in drug trials [15]. We found that almost two-thirds of oncology patients initially expressed a willingness to participate in drug clinical trials but previous CCT participants changed their decisions about taking part when reminded that the trial drug might be worse than the standard-of-care treatment. Females in the study showed slight indications of being more risk adverse and were less willing than males to participate in theoretical CCTs for new drugs and translational trials.
Most patients who had been offered CCT participation selected having an information booklet or patient navigator as aids that they would like to have to assist them in making their decision and 59% of patients preference was to have CCTs explained to them by their consultant. The patient's family can have a significant influence on their decision-making process, with 84% of patients that reported using support relying on their family to assist them in making decisions about CCT participation. However, some studies have shown that targeted awareness campaigns focusing on newly diagnosed patients and their families have not shown any significant increases in trial enrolment [9].
The hope of a personally improved outcome and altruistic reasons such as helping to fight cancer are commonly cited reasons for CCT participation [22][23][24]. The most frequent reasons given in our study for positively considering CCT participation was to feel better and live longer. We also found that Irish oncology patients, like patients elsewhere, wanted to participate in clinical trials to advance cancer research [9].
Only 14% of patients indicated that they would feel too weak to participate in CCTs. Our study population had a slightly higher median EQ VAS score of 75 compared to other studies of cancer patients [31][32][33]. We also found that of the 38% of participants reported as having some level of anxiety/depression, this was more prevalent in younger age groups and in female patients.
Many of the barriers to CCT participation that we found in our study of Irish oncology patients are already known. We did, however, find lower levels of commonly cited reasons to decline involvement in CCTs such as fear of side effects. The lower levels of fear were contrary to what was expressed in the general population and may be indicative of patients having a better awareness of cancer treatment side effects. We also found that only a small proportion of patients asked about clinical trials, indicative of a lack of awareness and highlighting a much broader issue around patient-clinician communication. Many patients are at a disadvantage compared to clinical staff in not knowing about clinical trials availability and understanding the associated concepts and methodology and depend on their advice. Our principal finding was that elderly patients had consistently, across the variables tested, significantly less understanding of CCTs and would benefit from support. A lack of understanding of therapeutic trial methodology in the Irish population is likely to be a real barrier to uptake for these clinical trials and would no doubt benefit from a national communication campaign to raise awareness in this area.
There are several limitations to our study. The national accrual rate to CCTs is between 3 and 5%. Almost one-third of our study cohort reported being on a CCT, indicating selection bias in our sample. It is likely that CCT participants' answers reflect real-life decisions that they made about participation. Patients often express a theoretical willingness/agreement to participate in hypothetical CCT scenarios and the study could not capture their actual decision in a real situation, and it is also possible that some patients gave socially desirable responses [24]. It is possible that despite provision of information, some patients may have confused being involved in a translational study with being on a CCT. We accept that missing data will also represent a source of bias. However, it is unlikely to affect the results due to random distribution across the variables. We excluded missing values when testing for associations.

Conclusions
Irish oncology patients understand the importance of cancer clinical research and having CCT options available to them. However, considerable misunderstanding regarding clinical equipoise and uncertainty around decision making was an issue, particularly in older patients. The global underrepresentation of older patients in clinical trials has not only been a health care disparity but has likely left this population with reduced understanding and awareness of clinical trials. Cancer incidence and mortality are growing worldwide, reflecting both an aging and rapidly growing global population. Advancing age is a known risk factor for cancer and it is imperative that older patients are represented in cancer clinical research and that awareness of clinical trials is raised in this population globally. Historically, clinical trials have been biased towards younger age groups.
Unlike in other studies, we found Irish cancerpatients had less concerns around known fears of barriersto CCT participation such as fear of more side-effects, harm or death. They also expressed high levels of trust in oncology clinicians and nurses yet had reservations about participating in therapeutic trials. A qualitative study to better understand the reasons for this would be beneficial. Patient willingness to participate in therapeutic trials could be positively influenced by more awareness around CCT methodology. We found that younger females were slightly more risk adverse when considering new drug trial scenarios, possibly influenced by their higher levels of anxiety. Recognising and addressing these concerns may reduce anxiety and improve willingness to participate in CCTs. Improvement in information delivery on CCT concepts is warranted and tailoring communication of CCT information would be beneficial, particularly for the groups highlighted in the study that had lower levels of understanding. Patients in the study indicated that having a patient navigator would be a useful aid and consideration of the use of trained facilitators to support communication of the importance of clinical research may be advantageous. This is the first nationwide study of oncology patients' attitudes and understanding of cancer clinical trials in Ireland, and the findings can contribute towards considered interventions for improvement to help optimize accrual rates.
Funding: This research was supported by educational grants awarded to us by Abbvie, Inveva, Bayer, Novartis and Amgen and funding from the European Union's Horizon 2020 Research and innovation Programme under the Marie Skłodowska-Curie Co-funding of regional, national and international programmes [713279 to C.Kearns].