Introduction

Clinical trials (CTs) are research studies in which patients volunteer to help test new ways to treat, diagnose, or prevent diseases. They are used to determine if a new test or treatment works and is safe [1–5]. CTs play a key role in the implementation of evidence-based medicine [2,3]. They are used for numerous reasons, from evaluating treatments to evaluating procedures used to prevent, diagnose or treat diseases, disorders and other health problems [1]. The success of a CT depends on ensuring that enough participants are included, that the sample is as representative as possible of the target population (e.g., breast cancer, people who inject drugs), and that drop-out is minimized [4]. However, there are many barriers to CT enrolment, leading to low participation rates [3,5–7]. One such barrier is low willingness to participate. For example, a study carried out among cancer patients in Florida, USA, reported that only 36.5% were willing to participate in CTs [7], while in France, only 28.0% of persons invited to participate in a COVID-19 vaccination CT were willing [8].

Most studies investigating CT participation rates and willingness to participate have focused on patients in hospital-based samples or samples drawn from registries [5,7,9]. However, the lack of familiarity with medicine is something that grows over time, as studies on health literacy (HL) show [10–12]. For this reason, it is also important to take into account the general public’s (i.e., non-patient) impression of CTs, with a view to increasing voluntary participation in general, and in the event of major health crises such as the COVID-19 pandemic. This issue is reflected in a 2020 French study which found that only 47.6% of French adults in the general public were willing to participate in a COVID-19 vaccine CT [13].

A number of studies have investigated the determinants for patients’ refusal to take part in CTs. One recurrent determinant is the lack of sufficient knowledge about these trials [5,7,9,14]. In this context, various research teams hypothesized and subsequently demonstrated that providing better information, both in terms of content and format, would increase willingness to participate. For example, a US study showed that providing brief educational material about CTs increased theoretical willingness to take part, particularly among less-educated respondents [15]. Elsewhere, a randomized, online study in the US showed that attitudes towards CTs improved after preparatory training on CTs prior to patients’ first oncology visit [14]. A recent systematic review highlighted that one of the reasons explaining why socioeconomically disadvantaged persons are less likely to participate in CTs was their difficulty understanding the purpose of these trials and how they work [16]. This is not surprising given that this population has limited access to health information in general and has poorer HL [17,18].

Giving information is a two-step process. First, trust needs to be built before communicating information [15]. The literature highlights that socioeconomically disadvantaged persons have greater difficulty trusting the healthcare system [19], and that this is one factor why they are less willing to participate in CTs [20–22]. Second, the information to be shared must be in a format that is understandable for the target audience [23–26]. For example, a US study among disadvantaged citizens, found that tabular presentation of numerical data was preferred to icon arrays and was associated with greater comprehension scores [23]. Tabular presentation can also be used to present textual information comparing different options [27]. To our knowledge, no study to date has investigated how well target populations understand content in this format. The present study focused specifically on the second information step listed above.

This study had two objectives: the first was to assess whether providing brief information on CTs improved the general impression of and willingness to participate in CTs in a sample of socioeconomically disadvantaged people in France; the second was to assess the importance of the format of the information shared in terms of ease of understanding.

Methods

Results

Among 401 members of the IPSOS sampling panel who completed survey questionnaires, 33 (8.2%) had never heard of CTs and were removed from the analyses. The analyzed study sample therefore comprised 368 individuals.

Median age was 40 years (IQR: 26.8, 53.0) and 42% were aged between 18 and 34 years. Women represented 64% of the sample and 49% lived in a city. In our study, 79% of the participants had an annual net household income ≤9000 €. French was the main language spoken (97%). The majority (61%) had high school or lower education, 57% of participants had limited HL, and 51% had a subjective numeracy score ≤13. Moreover, 11.2% distrusted medicine in France, 47% considered their health to be good, 35% reported having a chronic condition, and 46% received CSS (see above). Thirteen percent scored 0/3 for the ‘comprehension of tables’ questions. Only 8.4% of the sample had or knew someone who had previously taken part in a CT. Prior to receiving the brief CT information note, the baseline median ‘CT general impression’ score was 5.0 (IQR; 5.0, 7.0); the baseline median ‘hypothetical willingness to participate in a CT’ score was also 5.0 (IQR; 3.0, 7.0). Details of these characteristics are shown in Table 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Participants’ characteristics according to the randomized format (Textual vs Tabular) of the brief information note on CTs.

https://doi.org/10.1371/journal.pone.0329288.t001

Evolution of impression of CTs and hypothetical willingness to participate in a CT after reading brief information note

Table 2 presents the distribution of variables after receiving the brief information note about CTs. There was no statistically significant difference between the randomized groups. Among the participants, 76% said the information received was easy to understand, and 86% replied that the information given was credible. A third (29.3%) said their opinion of CTs was positively influenced by reading the information note. More specifically, the percentage of those with a positive general impression of CTs rose from 48% prior to reading the information note (Table 1) to 66% after reading it (Table 2). Similarly, the pre-post percentage of persons willing to participate in a CT rose from 45% (Table 1) to 49% (Table 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Participants’ responses after reading the brief information note according to the randomized format of the information provided.

https://doi.org/10.1371/journal.pone.0329288.t002

The median score for participants’ general impression of CTs increased moderately overall (from 5 to 6.5 (Fig 1, p-value < 0.001, effect size r = 0.47). Despite being statistically significant, only a small increase in hypothetical willingness to participate in a CT was observed (median 5 (3, 7) to 5 (4, 7); p-value < 0.001, effect size r = 0.20; S1 Fig).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Evolution (i.e., pre-post information note) of each patient’s general impression of clinical trials.

Similarly, boxplots are shown to present the quartiles (Q) at each measurement time; we can see that these quartiles (notably Q₁, Q₂ and Q₃) are higher after the information note was read; moreover, the p-value tells us that this difference was statistically significant.

https://doi.org/10.1371/journal.pone.0329288.g001

We stratified the analysis according to the ability of participants to understand tabular information. Participants with a 0/3 comprehension score for tabular information who subsequently received the information note in tabular format were the only subgroup with no significant increase in the CT impression score (Table 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Evolution of participants’ general impression of CTs according to their comprehension of tabular information score and the format of the information subsequently provided after they completed the questionnaire.

https://doi.org/10.1371/journal.pone.0329288.t003

In order to investigate whether the format-related evolution observed was sensitive to numeracy and HL, pre-post education material changes in participants’ general impression of CTs were also measured after stratifying for both these factors. No specific pattern was observed (S1 Table). Trust in French medicine was the only variable significantly positively associated with a favorable impression of CTs irrespective of the study step (i.e., before (β = 0.86 [CI95%, 0.64, 1.1]) or after reading the information note (β = 1.1 [CI95%, 0.94, 1.3])). This was true independent of age, gender, household income, education level, mother tongue, HL level, and randomization group (S2 Table). It was also the only variable significantly associated with a better general impression of CTs after the information note was read (β = 0.27 [CI95%, 0.09, 0.45]).

Discussion

In our study sample of socioeconomically disadvantaged people living in France, the overall impression of CTs was not good, but improved after participants read the brief information note about CTs. This improvement was associated with the type of information received, and the intrinsic abilities of each individual to cope with different formatting of information.

Fifty-seven percent of our sample had limited functional HL. This large proportion confirms the need for educational interventions to improve health and health-related scientific literacy with a view to limiting health inequalities [17]. The ‘general impression of CTs’ score in our socioeconomically disadvantaged sample was quite low (median score of 5 before reading information note) compared to that in a study of the French general population (median score = 6.5) using the same scale to measure people’s impression of CTs [20]. This difference was to be expected given that the literature has already shown that disadvantaged populations are less knowledgeable about CTs and are therefore less inclined to participate in them [16].

Fewer than half of the participants had a positive score (i.e., > 5) for general impression of CTs and willingness to participate in them (48% and 45%, respectively). In line with findings from a related study conducted in the US [15], participants’ general impression of CTs improved significantly after reading the information note. More specifically, 25% of our sample had a more positive opinion of CTs (i.e., calculated on pre-post scores) after reading the information note, while 4.3% had a ‘much more positive opinion’ (i.e., the self-perceived change in opinion question). Simple information or awareness-raising campaigns to demystify CTs could go a long way towards improving participation rates. In this perspective, repeated campaigns in the healthy general public could be of greater benefit than simply targeting people who are already patients. In our study, although willingness to participate in CTs was still relatively low both before and after reading the information note, a small pre-post increase was seen. It is important to note that there are multiple factors involved in the decision to participate in a CT (presence of a placebo or control group; potential side-effects, physicians’ attitudes towards the trial, cost-benefit considerations, etc.) [5,35]. The organization of CTs (e.g., on-site or remote visits) can also impact willingness to participate. For example, a study on an e-cohort of patients with chronic diseases in France regarding their hypothetical participation in a CT found that 90% would participate if the trial were organized according to their preferences, versus 60% if it was not [36].

The only factor significantly positively associated with a favorable general impression of CTs before reading the information note was trust in French medicine. This is consistent with other findings in the literature [21,22], notably Andrews and Davies, who suggested that trust may be the most important factor in the success of research studies, particularly in the context of vulnerable populations [22]. Moreover, after the information note was read, trust in French medicine was associated with an even stronger favorable general impression of CTs. This highlights the need to build trust while providing relevant user-friendly medical information. This is especially important given that only 57% of our sample of socioeconomically disadvantaged population declared they trusted French medicine most of the time.

In our study population, limited functional HL was not associated with participants’ impression of CTs, either before or after randomization. This finding might be explained by the fact that the information was brief, and accordingly, respondents might have reported it was easy to understand for this reason (24% reported difficulties, cf. Table 2). Similarly, a previous national study in France reported that cancer patients with limited functional HL were less likely to be invited to participate in CTs, but when they were, their participation rate was no different from those with adequate HL [30]. The problem with this lower invitation rate may lie with doctors who refrain from inviting persons they deem have difficulty understanding CTs to participate, thus creating health inequalities [30]. Indeed, several studies have highlighted that barriers to CT participation are not only attributable to patients, but also to physicians [3,5].

This study highlights the importance of how health information is presented in terms of the target population. More specifically, comprehension of tables was assessed by a score ranging from 0 to 3 obtained by proposing comparative tables for interpretation to study participants. Thirteen percent of participants scored 0 (i.e., did not understand the table at all). Moreover, as expected, participants with poor comprehension of tables better understood the information note on CTs in textual format than the same information given in tabular form.

The improvement in participants’ general impression of CTs after reading the information note was observed in all subgroups except – and perhaps unsurprisingly – in persons with poor comprehension of tables who were subsequently given the brief CT information in tabular form. In contrast, as expected, those who had a good comprehension of tables also had a good understanding of the information given in textual format. In other words, it is not enough to provide information materials to educate disadvantaged populations about CTs; in order to be effective, these materials must take into account users’ abilities and understanding, as suggested by other studies [15,23,26,36].

We also looked at changes in participants’ general impression of CTs after reading the information note as a function of numeracy and HL level (S1 Table). The fact that the results were similar for all subgroups (i.e., HL, numeracy) supports our hypothesis that it is the format of the brief written information and not a lack of functional HL or of good numeracy which might cause comprehension problems in this disadvantaged population.

This study has limitations arising from the survey method used. First, the study sample may not fully represent the most disadvantaged people in France because of our recruitment process (i.e., using a sampling panel) and potential non-responses of some eligible participants. Second, we had no control group with higher incomes in our study. Studies on the general population are needed in order to compare our results. Finally, we cannot exclude social desirability bias with potential more positive impressions reported after reading the information note.

Conclusion

Various barriers to CT enrolment are linked to a lack of information or misinformation about these types of trials [7,15,30,37]. This is particularly true among disadvantaged populations. In our study, providing individuals with brief pertinent information helped to improve their general impression of CTs and, to some extent, their willingness to participate in them. Moreover, the way in which this information was provided also had a significant effect on these two outcomes. It is therefore important to adapt information – both in terms of content and format – to the target population.

Supporting information