Background

Prior to participation in a randomised controlled trial (RCT), potential participants, or their representatives, are provided with a participant information leaflet (PIL) as mandated by the Clinical Trials – Regulation EU No 536/2014 of the European Parliament [1]. A PIL provides all information on the trial that participants are being invited to take part in, so they can make a decision on participation, or not [2,3]. This is called providing informed consent, which is a regulatory requirement [1,4]. Essential information includes: the trial purpose, the methods of data collection, potential risks, benefits, as well as freedom to withdraw from the study at any time without any impact on current treatment [3].

Participants’ response to the PIL can significantly affect recruitment. Historically, the design of PILs has been tailored for the purposes of documenting the consent process, not participant comprehension [5]. PILs have also been criticised for being unnecessarily complex and long [2,3]. Accurate comprehension of information within the PIL pertaining to the risks and benefits of trial participation is essential in informing decision making by trial participants [6]. Presentation of this information within PILs has further consequences when shaping participants’ expectations and experiences within a trial, which, if met, can improve participant response to treatment [7], and if not met, may lead to drop outs, dissuade future participation in research [8], or potentially lead to a ‘nocebo’ effect (where negative expectations regarding an intervention can lead to more adverse outcomes within a trial) [9,10]. Ensuring accuracy and balance in the presentation of information relating to the advantages and disadvantages of trial participation is important in facilitating informed decision making and providing an accurate, unbiased description of the intervention being tested [6].

Standardised PILs do not exist, therefore trial teams create bespoke documents for their own trial. There is significant variability in layout, information provided, and length, thus uncertainty remains about the information trialists should communicate to potential trial participants to improve recruitment [11–13] and retention [14–16]. The PRioRiTy study, conducted in a priority setting partnership, ranked it as second out of twenty priorities in terms of importance [17]. While the Clinical Trials Regulation mandates that the nature, objectives, benefits, implications, harms/risks and inconveniences of the clinical trial are communicated to potential trial participants when taking informed consent [18–20], it is not clear if the benefits and risks are equally communicated.

The purpose of this study was to evaluate the balance in information attributed to the benefits and harms/risks in PILs used for trial recruitment. While regulatory frameworks emphasise the need to clearly communicate potential harms and risks, there is limited empirical evidence examining how this information is balanced against potential benefits in practice. By systematically quantifying the emphasis placed on benefits and harms/risks in PILs, this study addresses an important gap in the literature and provides evidence to inform future guidance on the development of participant-facing trial information.

Methods

Coding

RC and FS met five times over the course of seven weeks, to develop and refine the coding framework used to classify statements relating to potential benefits and harms/risk in the PILs. RC conducted the initial coding of all statements, while FS independently reviewed the coding framework and verified the categorisation of statements. This iterative process continued until no new categories or terms were identified in the PILs. Where uncertainty arose regarding the categorisation of statements, these were discussed between RC and FS until agreement was reached. FS also conducted confirmatory coding on a random 10% sample of the PILs to ensure consistency in the application of the coding framework. A summary of the final coding framework is presented in Fig 1.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Coding framework.

https://doi.org/10.1371/journal.pone.0346984.g001

Positive reinforcement statements (categories summarised in Fig 1).

Text that stated that participating in the trial could benefit others or the disease under trial was coded as ‘Altruistic’. If monetary incentives were given it was coded as ‘Monetary’. The potential personal benefit that was explicit to the participants was coded according to the type of benefit – ‘additional assessments’, ‘better outcomes’, ‘fewer side effects’, ‘close monitoring’, ‘access to newer treatments’ and ‘better housing’. In these scenarios the PIL authors clearly stated how participants would benefit from joining the clinical trial in the management of their condition. If the word benefit was mentioned but not explained, then it was coded as ‘Potential personal benefit-not explicit’.

Negative reinforcements statements (categories summarised in Fig 1).

Lists of side effects from drugs/procedures or potential complications from the intervention were coded as ‘Side effects/Complications’. This included surgical procedures, use of medications, radiotherapy and psychological side-effects. Any type of attendance outside standard of care, increased waiting times and/or questionnaires that wouldn’t routinely be given to patients were coded as ‘Burden to the participants’. In some PILs this included psychological burden directly attributed to increased attendance. Any type of written text that clearly indicated women should not get pregnant while participating in the trial, or that contraception needed to be used while participating in the trial to prevent any harm to the unborn foetus, was coded as ‘pregnancy associated risks’. In some PILs this included psychological burden. Any expenses incurred by the participants while taking part, e.g., costs of travel, parking and food, or explicit statements about participants not receiving any monetary incentive, were coded as ‘non-monetary remuneration’. Text that stated that because of taking part the participants condition worsened, and they would have to receive extra treatment, was coded as ‘recurrence’. If the participant was told they would no longer have access to the medicine once the trial was completed, it was coded as ‘no future access to the medicine’. If the patient would have to wait days, weeks or months for the intervention, it was coded as ‘treatment delays’. Text on data security breaches were coded as ‘data security’.

Results

Twenty-six CTUs/CRFs provided Patient Information Leaflets relating to 228 unique studies which were included in our analysis. Of these 228 studies, 227 (99.6%) were randomised controlled designs with 2 or more arms, one trial (0.4%) was a multifactorial platform design.

Randomisation was conducted at the individual participant level for 97.8% (n = 223) of trials and at the cluster level in 2.2% (n = 5) of trials. Funding information was included in 204/228 included PILs. Of the 204 included study PILs with funding information, 24 (11.8%) specified commercial sponsorship, 180 (88.2%) study PILs had a non-commercial sponsor.

Table 1 provides a descriptive summary of the word count in the PILs. Word count for PILs ranged from 53 to 11,489, with a median 2596 (IQR 1892). For positive reinforcement statements, word count ranged from 0 to 534 (median 56 (IQR 51)). The median proportion of total PIL word count attributable to positive reinforcement was 2.36% IQR (2.17%), Word count for negative reinforcement statements ranged from 0 to 2906 words (median 149; IQR 48–321). The median proportion attributable to negative reinforcement was 5.57% (IQR 6.92%)

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. PIL Word count.

https://doi.org/10.1371/journal.pone.0346984.t001

Almost three-quarters of PILs (72%; 164/228) had a greater percentage of negative reinforcement words, compared to positive reinforcement words. Six PILs had no positive reinforcement words at all. Of those six, two had no negative reinforcement words either. Both were children PILs. Table 2 shows the number (%) of PILs categorised by the percentage of total word count devoted to positive and negative reinforcement. At lower percentage levels, positive reinforcement generally exceeds negative reinforcement, with the exception of the < 1% category.. However, as the proportion of reinforcement-related content increases, negative reinforcement becomes progressively more common. This trend is reflected in the ratios, which are greater than one in most lower and mid-range categories, but fall below one in the highest category.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. No. of PILs by percentage of total word count devoted to positive and negative reinforcement (n = 228).

https://doi.org/10.1371/journal.pone.0346984.t002

Positive reinforcement

We identified 222 PILs (97.4%) that contained positive reinforcement statements. The most common form of positive reinforcement related to altruistic benefits (Table 3), with 93.2% (n = 207) indicating that trial participation could improve understanding of the disease or benefit future patients. Potential personal benefit was reported in 30.6% of PILs. However, 14.1% (n = 32) did not explain to participants how they could potentially benefit from taking part in the trial. Financial travel incentives (monetary) were the third most common form of positive reinforcement (13.1%, n = 29).

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Number (%) of participant information leaflets containing each category of positive reinforcement (n = 228).

https://doi.org/10.1371/journal.pone.0346984.t003

Negative Reinforcement

Negative reinforcement statements were identified in 203 out of 228 PILs (89%). Twenty-five PILs (11%) contained no negative reinforcement, the majority of which were children’s PILs that didn’t include detailed descriptions of potential side effects or risks. Two additional PILs were for adults who had regained capacity. As expected, due to regulatory requirements, the most frequent type of negative reinforcement described in the PILs was side effects/complications from the intervention (Table 4).

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Number (%) of participant information leaflets containing each category of negative reinforcement.

https://doi.org/10.1371/journal.pone.0346984.t004

Discussion

The purpose of this study was to evaluate the balance of information relating to potential benefits and harms/risks presented to participants in PILs used for trial recruitment. Our analysis of 228 PILs from trials conducted in the United Kingdom and Ireland demonstrates a clear imbalance, with substantially more emphasis placed on harms/risks than on potential benefits of participation. In particular, the median word count devoted to negative reinforcement statements was almost three times greater than that devoted to positive reinforcement statements, suggesting that participant-facing trial information may disproportionately foreground risks over benefits.

We know that participants often take part in trials hoping to achieve better outcomes [21] while also being motivated by altruistic reasons [22,23]. However, the PILs analysed in this study emphasised altruistic benefits far more frequently than potential personal benefits. While 93.2% of PILs highlighted the potential contribution of the trial to improving understanding of the disease or benefiting future patients, fewer than one-third (30.6%) included information about potential personal health benefits. This imbalance represents a missed opportunity for trial teams and could influence participants’ expectations when considering participation. Underrepresentation of potential personal benefit may direct attention towards potential harms/risks associated with participation rather than possible direct health benefits. The framing of information in this way may also have implications for the nocebo effect, whereby negative expectations about an intervention can influence the experience or reporting of adverse outcomes. When PILs emphasise potential harms or side effects more strongly than potential benefits, they may inadvertently shape participants’ expectations in ways that heighten awareness of negative outcomes. Previous research suggests that expectation setting can influence both perceptions of treatment and reported side effects [7]. While it is ethically and legally necessary to clearly communicate potential harms, presenting this information in a disproportionately negative way may unintentionally contribute to expectation-related effects. Ensuring a more balanced presentation of harms/risks and benefits may therefore support informed decision-making and should be considered by trial teams.

The predominance of negative reinforcement in PILs, accounting for approximately 5.7% of the total word count compared to only 2.3% for positive reinforcements, suggests a cautious approach potentially aimed at over-protecting participants from trial harms/risks, acknowledging the fact that it is a regulatory requirement to notify participants of potential risks or harms. We suggest a more balanced approach to providing information on both harms and benefits so that potential trial participants are not unnecessarily discouraged from trial participation. We also recommend making the participant information leaflet more readable in terms of its length and complexity, which has been critiqued previously in relation to their contribution to misunderstandings about the trials’ risks and benefits [2,24], and suitable for a lay audience (those we aim to recruit) which research has shown even trial lay summaries are not suitable for [25].

The variability in PIL content, reflected in the wide range in word counts and differences on the types of information emphasised, suggests a lack of standardisation in how trial information is presented to potential participants. This is concerning given the central role of these documents in supporting informed consent. In line with the Declaration of Helsinki, investigators have a responsibility to ensure that research participants are sufficiently informed to enable them to provide valid consent [26]. While regulatory guidance exists on the information that should be included in PILs, there is limited direction on how benefits and harms should be balanced in their presentation. Recent guidance developed through an expert consensus process provides recommendations to support trial teams in preparing accessible and understandable PILs and consent forms [27]; however, the balance between communicating harms and potential benefits is not explicitly addressed.

From a practical perspective, trial teams and regulatory bodies could consider introducing clearer guidance on how information about benefits and harms/risks should be presented within participant information leaflets. For example, templates could include dedicated sections describing both the potential benefits and the potential harms of participation, ensuring that each is clearly addressed rather than dispersed throughout the document. Guidance could also encourage trial teams to describe potential personal benefits where appropriate, alongside commonly stated altruistic benefits. While it may not be appropriate to mandate strict ratios of text devoted to harms and benefits, encouraging trial teams to review the relative prominence of these sections during PIL development may help promote a more balanced presentation of information. Ethics committees and regulatory bodies reviewing trial documentation could also consider whether both benefits and harms are communicated clearly and proportionately as part of their assessment of participant-facing materials. Such approaches may contribute to improving participant understanding and supporting informed decision-making.

Interpretation of information on harms/risks and benefits in participant information leaflets is influenced not only by the amount of text devoted to each, but also by how that information is presented. Brief statements describing serious harms or side effects, such as death or serious injury, are likely to have a greater impact on participation decisions than longer explanations describing potential future benefits. Qualitative research is needed to delve deeper into this issue. However, what we do know is that we need to ensure decisions regarding participation are fully informed by comprehensible and balanced information on benefits and harms/risks [9]. Previous research has also highlighted the importance of co-developing recruitment materials with user groups who are representative of the target patient population [28].

Conclusion

This study highlights an imbalance in the weighting of information on the benefits and harms/risks provided to potential trial participants in Participation Information Leaflets. The trial community acknowledges the need to make trial participants aware of the potential harm/risk of trial participation, but equally, informing them of the benefits should have equal prominence. Standardising this in PILs is needed to ensure that participants are provided with accurate, balanced information pertaining to benefits and harms/risk, in formats that are easy to comprehend, so that they can make an informed decision regarding participation.

Acknowledgments

Consent for publication: All authors consented to the publication of this research.