The sample comprised 313 participants, 156 (49.8%) females and 157 (50.2%) males, aged between 18 and 76 years (M = 43.9, SD = 14.1). The participants were recruited from 13 districts of Greece and filled in the survey online.
Test-retest reliability
Thirty-five participants completed the questionnaire in test and retest mode two weeks later. Wilcoxon tests showed no significant differences between the two tests (Supplementary Table 1) [All tables are mentioned in the Supplementary Material-Additional File 2 by their number]. The modified Fleiss' kappa statistic (s*) indicated substantial to almost perfect agreement (0.61-1.00) for 22 out of 27 questions and moderate agreement (0.41–0.60) to 5 out of 27 questions (Supplementary Table 1). Overall, test-retest results indicated reliability of the questionnaire.
Preliminary analysis for EFA
Exploratory factor analysis (EFA) was conducted on the 27 items in the questionnaire and 6 factors were derived based on Kaiser’s criterion (eigenvalues greater than or equal to 1). The KMO measure of sampling adequacy was 0.898 and Bartlett’s test of sphericity was significant (χ2(351) = 3917.3, p < .001), indicating that factor analysis was appropriate for this data. The Determinant was equal to 0.00002 > 0.00001 suggesting no multicollinearity. Taken together the results indicated that all 27 items in the questionnaire were suitable for inclusion in factor analysis.
Initial factor analysis
Following the Kaiser’s criterion, EFA resulted in a six-factor solution that accounted for 62.2% of total variance. There were two issues regarding this solution. First, there were two out of six factors with only two items loaded on each one of them. Second, examination of the scree plot (Supplementary Fig. 1) [Supplementary Material-Additional File 1] showed a sharp drop in eigenvalues from one to three factors and a slight drop from the fourth to the fifth factor. Since, Kaiser’s criterion and scree plot suggested a different number of factors, EFA was re-applied for extracting five, four or three factors. The criteria for selecting one out of these three solutions were based on factor loadings, proportion of total explained variance and interpretability of the final rotated solutions.
The five-factor solution accounted for 58.2% of total variance, all factor loadings were above 0.4 and there were four items with cross-loadings above 0.4 on a second factor. The four-factor solution accounted for 54.0% of total variance, one factor loading was below 0.4 and there was one item with cross-loadings above 0.4 on a second factor. The three-factor solution accounted for 49.0% of total variance, all factor loadings were above 0.4 and there were four items with cross-loadings above 0.4 on a second factor. Regarding the interpretability, the four-factor solution identified factors that were fairly interpretable, while five-factor and three-factor solutions identified three and one interpretable factor, respectively. Based on the above findings, the four-factor solution was preferred and EFA was re-applied for extracting four factors after excluding from the sample the item with a loading less than 0.4 (item 24).
Final Factor analysis
The KMO measure of sampling adequacy was 0.901 and Bartlett’s test of sphericity was significant [χ2(325) = 3810.9, p < .001], indicating adequate factorability of the items. The total explained variance was 55.3%. The first factor, accounting for 16.2% of the total variance, was focused on patient expectations from clinical trials. The second factor, the risks and benefits of the treatment in clinical trials accounted for 15.3% of the total variance, while the third factor related to patients’ participation in clinical trials, and accounted for 12.3% of the total variance. The final factor, accounting for 11.6% of the total variance, was focused on cost and convenience of clinical trials.
The results of EFA are presented in the Supplementary Table 2 along with the assigned labels to factors. All loadings below 0.4 were suppressed. According to the results of EFA presented in the Supplementary Table 2, three out of four factors comprised items with loadings greater than 0.5.
Internal consistency
All factors were considered reliable, with values of Cronbach’s alpha ranging from 0.770 to 0.867, and item-total correlations above 0.3 (Supplementary Table 3).
Factor exploratory analysis and statistical significance
Exploratory factor analysis was conducted to determine the factor structure of the questionnaire and four factors were derived, namely, (i) patient expectations, (ii) risks and benefits, (iii) patient participation, and (iv) cost and convenience.
In the study of Arnetz et al. (2019), the pilot questionnaire consisted of 30 items and it was delivered to 53 patients. The authors derived a six-factor solution that encompassed factors concerning motivation, risks of participation, benefits of participation, understanding the purpose, cost/convenience, and contribution to health improvement. The main questionnaire of the study consisted of 27 items and it was delivered to 55 patients. The factor solution was not the same with the one derived in the pilot phase.
The four-factor solution derived in our study consisted of the same 27 items and it was different from the six-factor solution that Arnetz et al. proposed. The factors risks and benefits, that consisted of 5 and 3 items respectively in the six-factor solution, were merged into one factor that consisted of 10 items in the four-factor solution. There were 7 common items in the two solutions. Also, the factor cost/convenience was almost the same in both solutions, with the four-factor solution comprising one more item than the six-factor solution. The remaining factors were significantly different between the two solutions.
Across the four factors formed (i.e. (i) patient expectations, (ii) risks and benefits, (iii) patient participation, and (iv) cost and convenience), the percentage of participants who fully or partially agreed differs statistically significantly from the corresponding percentage of those who partially or completely disagreed. This provides confidence on the validity of the results, and the recognition of the most crucial elements for each factor. Among different demographic characteristics, the most crucial, statistically significant differences were related to the participants’ gender, with statistically significant differences of views between men and women observed in most factors.
For factor (i) patient expectations (Supplementary Table 16), for each of the 5 items of the questionnaire included in this factor, the percentage of participants who fully or partially agreed differs statistically significantly from the corresponding percentage of those who partially or completely disagreed (p < 0.05).
Also, the statement that the patient has the ability to ask questions (the highest percentage of agreement) shows a significantly higher degree of agreement (Z = -4.888, p < 0.001) than the statement that the patient receives clear information (the second higher agreement rate), providing confidence on correctly identifying the most significant element of this factor.
For factor (ii) risks and benefits (Supplementary Table 17), for each of the 8 questionnaire items included in this factor, once again the percentage of participants who consider the respective issues as very or quite important differs statistically significantly from the corresponding percentage of those who consider them as little or not at all important (p < 0.05).
However, the issue of risks and side effects that patients may face, which was identified as the most significant element of this factor, does not differ significantly in the degree of importance declared by the participants in the study (Z = -0.784, p = 0.433) from the second and third most important elements. On the other hand, the risks (Z = -6.751, p < 0.001) and the side effects (Z = -6.518, p < 0.001) that patients may face differ significantly in the degree of significance declared with the degree of significance stated on whether the treatment would be provided free of charge. Therefore, this would provide us confidence in assuming that potential expenses are of comparatively lower importance for clinical trial participants.
For factor (iii) patient participation (Supplementary Table 18), for each of the 6 questionnaire items included in this factor, the percentage of participants who consider the respective expectations as very or quite significant differs statistically significantly from the corresponding percentage of those who consider them as little or not at all important (p < 0.05). However, no statistically significant differences were found between expectations from participation in clinical trials regarding the degree of significance reported (p > 0.05). Here, a potentially larger sample might be helpful in drawing more definitive results.
Finally, for factor (iv) cost and convenience (Supplementary Table 19), for each of the 6 items of the questionnaire included in this factor, the percentage of participants who consider the respective issues as very or quite important differs statistically significantly from the corresponding percentage of those who consider them as little or not at all important (p < 0.05).
However, the issue of confidentiality of all personal data and results of the clinical study (most important element) and the issue of possible other costs for participation in this clinical study (second most important element) do not differ significantly in the degree of importance stated by the participants in the study (G = − 1.467, p = 0.142). On the other hand, the issue of confidentiality of all personal data and results of the clinical study (Z = -4,420, p < 0.001) and the issue of possible other costs for participation in this clinical study (Z = -3,480, p = 0.001) show a significant difference in the degree of significance declared for the remaining elements of this factor.
Demographics and demographic comparison
Participants’ ages ranged from 18 to 76 years (M = 43.9, SD = 14.1) with 51.4% (n = 161) identified as female, 48.6% (n = 152) as male, 63.3% married or partnered, approximately 30% reported living alone (single, divorced or widower), 22.7% reported being retired or not working (10.2% and 12.5% respectively), 62.6% reported being working or self-employed 49,5% (n = 155) had post-secondary and higher education, 32,9% (n = 103) had secondary education, and 13,1% (n = 41) possessed a Master degree. The vast majority of participants (97.8%) identified themselves as white people. Approximately half of the participants (47.6%) were not “high-income participants.” The mean of their net personal annual income was lower than 10,000 euros. The percentage of participants reporting net personal annual income ranging from 10,000 euros to 20,000 euros was lower (32.3%), with 13.1% of participants reporting net personal annual income ranging from 20,000 to 30,000 euros.
The time in minutes that participants were willing to spend for transferring to a research site for study-related procedures ranged from 0 to 400 min (M = 32.5, SD = 39.8), while on average they were willing to cover a distance of 330 kilometers.
The majority (54.6%) of participants indicated that they would choose to go to a research site by car, with 21.1% of the participants reporting their preference to go on foot, 18.8% by means of public transportation and 5.4% on bicycle or motorcycle.
Slightly more than 9 out of 10 participants (91.7%) reported having health insurance.
The vast majority of participants (80.8%) reported their physical or psychological condition as being in a peak state (16.6%), in a very good state (32.6%) or just in a good state (31.6%), with 19.2% of participants reporting their condition as mediocre (15.7%) or even bad (3.5%).
Furthermore, in our study we found that 6.1% of our respondents had previously participated in a clinical trial, with the 89.5% of them reporting they had received a clear and direct benefit (26.3%) or had rather been benefited (63.2%).
The majority of our participants (61.7%) reported they had never previously taken medications or undergone treatment for any of the diseases mentioned in the Supplementary Table 12, with a lower proportion of participants reporting they had undergone treatment for anxiety (16.0%), hypertension (11.8%), cardiovascular disease (10.9%) and panic attacks (7.3%).
The participants in our study strongly supported the view that participation in clinical trial means that the participants (patients): a) are offered the opportunity to be clearly informed (95.6%) and ask questions (97.8%), b) receive clear and adequate information (95.5%), c) are offered the opportunity to take part in discussions regarding their own treatment and health care (94.6%), d) are offered the opportunity to voice their concerns and opinions (91.1%) and e) are offered the opportunity to become involved in decision making processes regarding their own treatment and health care (85.6%).
Furthermore, our participants were asked about the importance and impact of some factors on their willingness to participate in a clinical trial aiming at curing a disease or eliminating a health-related problem. The elements that were reported as most significant by the greatest proportion of participants were: a) the concerns about the risks of being in a clinical trial (87.5%), b) the possible side effects of clinical trials (86.3%), c) the type of treatment given in a clinical trial (83.7%), d) whether participation would improve their quality of life (81.5%), with only 62.6% of the participants finding very significant the fact that the treatment will be provided free of charge.
Almost all the participants in our study (98.7%) had no difficulty understanding the items of the scale, with 88.2% reporting that the administered scale was not lengthy.
Moreover, the participants in our study were asked about the elements “knowledge / information”, “their expectations towards participation in clinical trials”, “convenience” and “costs / payments.” The factors that were reported as most significant by the greater proportions of participants were: a) the confidentiality of participants’ personal data in the clinical trial and the confidentiality of the results of the clinical trial (68.7%), b) the understanding of the way in which the particular clinical trial might benefit the respondent (67.7%) and other patients at the present time (68.4%), and c) the understanding of the purpose of the particular clinical trial (66.8%), with d) 52.7% of the participants finding very significant the eventual need for a participant to travel to a research site for study-related procedures.
Among different demographic characteristics, the most crucial, statistically significant differences were related to the participants’ gender.
Mann-Whitney tests indicated significant differences between males and females regarding views on the following factors: ii) the risks and benefits (Z = 2.574, p = .010), iii) patient participation (Z = 3.104, p = .002), and iv) cost and convenience (Z = 3.135, p = .002). Males especially appear to consider the cost and convenience factors less important than females.
Otherwise, none of the four factors demonstrated statistically significant correlation to age, education, or marital status (p > .05). Nevertheless, we do observe some factors emerging from our data analysis as associated with taking a certain stance on the questions included in the administered scale. More precisely, we find that the responding option “strongly agree” received greater score among participants being over 60 years old, being retired or being in married cohabitation, with the participants in very good psychological condition showing a trend towards choosing the responding option “very important.”
For factor (ii) risks and benefits, women consider as more important than men the issue of risks (Ζ = -2,198, p = 0.028), side effects (Ζ = -3,343, p = 0,001) and possible discomfort, inconvenience or pain (Ζ = -3,011, p = 0.003), which they may face in the context of a clinical trial.
For factor (iii) patient participation, women consider it more important than men to understand how the clinical trial could help them personally at the moment (Z = -3,430, p = 0.001) and in the future (Z = -2,900, p = 0.004), to understand the purpose of the clinical trial (Z = -2,956, p = 0.003) and that the clinical study will be part of a broader medical research study (Z = -1,969, p = 0.049).
Finally, for factor (iv) cost and convenience, women consider more important than men the need to travel to participate in the clinical trial (Z = -3,515, p < 0.001) and the assistance of another person to participate in the clinical trial (Z = -3,426, p = 0.001).
Moreover, according to Mann-Whitney results, individuals, who had participated in a medical research or clinical trial, regarded factor (ii) risks and benefits of the treatment in clinical trial (Z = 2.570, p = .010) and factor (iv) cost and convenience of clinical trials (Z = 2.936, p = .003) as less important than individuals who had not participated in a medical research or clinical trial in the past. This clearly indicates the importance of educating patients on the benefits of clinical trials, as well as providing them with clear information at the outset, to clarify their expectations.