Informed consent is a key concept in human research ethics, yet few studies have investigated researchers’ views on the topic for domestic studies. This study identified that: 1) researchers generally have good awareness of the role of informed consent, its importance and ways to adjust their practice accordingly when obtaining it; and that 2) some researchers felt there was a lack of institutional support on how to obtain informed consent, through lack of training as well as inconsistencies and lack of transparency in HREC requirements.

Researchers in our study emphasised that information delivered to potential study participants needed to be tailored to the interests and abilities of each individual. Participant preferences were reported to dictate the modes of communication used, and the complexity of information discussed. To do this effectively, interview participants emphasised the importance of building rapport during a face-to-face conversation to adjust their approaches. Indeed, studies have shown that face-to-face conversations are the best method for improved understanding [18]. The variability of this interaction to individualise the consent process may explain why efforts to protocolise an ideal method for obtaining informed consent have been largely unsuccessful [6, 7].

Despite the view that information delivered must be tailored to the individual, researchers reported that standard written information about the study was always provided, often alongside the consent form. Researchers expressed concern that participants rarely read these documents in detail, and that these were often too long and complicated. Studies show that standardised forms do not result in standardised understanding [18] and that participants’ views on the essential elements of an information form vary widely [30]. Too much information can be overwhelming, and in some cases can impair decision making [31].

Despite the perception that information forms carry limited utility, researchers were reluctant to deviate from using them. Some explained that these forms acted as important reference documents for participants after the informed consent process, and also helped to structure the dialogue, thus reducing the potential for these important informal conversations to be influenced by the biases of individual researchers [31].

Hence, efforts to improve information forms must achieve dual purposes, that they are simplified to be accessible and understandable, but also detailed enough to ensure information delivered is standardised and comprehensive. Rather than merely shortening and simplifying forms [13, 17], efforts should be directed at investigating ways to ensure that these tools are flexible and fit for purpose. Indeed, suggestions were made by researchers for the preparation of standard shorter forms, alongside more comprehensive information that can be provided upon request from research participants.

It appears that the ‘best’ way to deliver information to participants, and to ensure participants’ understanding of a study, is to allow researchers some discretion and variation relating to what is discussed. Information forms should be recognised as being an unrealistic reflection of how much information can be, and is actually, conveyed to participants. Studies are required to investigate participants’ perceptions of the utility of written information forms, and whether the shortening of such forms would affect general perception of the legitimacy and safety of studies.

Researchers described that information provision should be supplemented with both formal and informal assessments of participants’ willingness and rationale for participating. Researchers reported forming an independent view on a participant’s suitability for study inclusion and that if a participant either lacked willingness or had misguided motivations, researchers reported that they would urge participants to reconsider. Researchers did not believe that merely delivering swathes of information was sufficient, but rather that they needed to take care to pre-empt and protect participants from potential sources of undue influence. It appears that the researcher-participant relationship is not an adversarial one, with the researcher vying for more participants and the individual defending themselves by exercising their autonomy to refuse. Researchers saw the informed consent process as a collaborative discussion to ensure that participants made the best decisions for their circumstances. Studies increasingly show that some professional intervention may assist potential participants in making ‘good decisions’ [32]. Researchers identified that an independent assessment of suitability not only benefited the participant, but also saved time and resources for the researcher in the future. This occurred through fewer withdrawals, and improved compliance with study requirements, where participants are aware of, and willingly complete, study requirements.

The literature continues to grapple with balancing the provision of appropriate decision support, while protecting the autonomy of potential participants. The conventional view, enshrined in the National Statement, is that imbalances of power and dependent relationships erode participants’ abilities to make voluntary decisions [1, 9, 19, 20]. Studies are increasingly showing however, that ongoing relationships may enable researchers to provide better decision-making support for potential participants [16]. Researchers in our study were aware of this tension, and some expressed that there existed abstract fear that researchers dogmatically recruit without regard for the safety of the participant. They expressed that this fear was unfounded, and that researchers’ overarching rationales for conducting studies stem from a desire to help others. A more effective conception of the informed consent process may be that an imbalance is not of itself an indication of an invalid decision or one influenced by inappropriate pressure. Rather than a predatory force requiring curtailment, a researcher’s role may be seen as a safeguard that independently assesses whether potential participants are suitable for study inclusion.

Researchers in our study emphasised the need to avoid coercion and forms of pressure on potential participants when making a decision to participate. Apart from purposes of nomenclature, the unarticulated distinctions between ‘coercion’ and ‘undue pressure’ in Australia have practical ramifications. It is defensible to state that a decision made under the technical definition of ‘coercion’ is invalid. However, the extent to which any influence on a decision is undue, or sufficiently inappropriate to erode the validity of a decision is subject to interpretation and ambiguity. Researchers in this study expressed concern that some measures required by HRECs to ‘avoid coercion’ during the recruitment process would in effect cripple the study. Thus, it is pertinent to derive a consensus statement on the extent to which undue pressures need to be eradicated, if at all. Given the scarcity of guidance provided even in the National Statement, researchers in this study cannot be constructed as apathetic of their ethical obligations. Rather, this misunderstanding highlights that more academic work is required to clarify broad principles on the threshold of unacceptable pressure, and improved education on how those broad principles should be applied in daily practice.

The National Statement in theory sets a national standard on ethical practice for HRECs and researchers. However, awareness and application of these broad principles appear inconsistent. Previous studies have suggested that ethics boards vary in their evaluation of research protocols [33, 34]. This view was also expressed by researchers who participated in our study. Some studies provide possible explanations for variation, including interviews in the US that show ethics review boards struggle with making decisions based on broad guidelines [14], and that Australian HRECs have varying knowledge and use of the National Statement in making decisions [11].

Researchers in our study also reported mixed attitudes towards HRECs, partly as a result of the perceived unpredictability of HREC review. While our study affirms previous work that some researchers view HRECs negatively, reporting impractical and inconsistent demands, we also found contrary views. Those reporting robust communication with HRECs were not only grateful for the guidance HRECs provided but relied on HRECs as institutional safeguards. Poor communication between researchers and HRECs has been identified as a serious problem [12, 35]. While the current study did not aim to, nor was it able to, capture researchers’ practices, we did encounter frustration due to a perceived lack of reasonableness. The National Statement appears to be aware of this phenomenon, and encourages informal in-person discussions between HRECs and researchers to minimise misunderstandings [1]. To promote open dialogue, past studies have recommended researchers attend HREC meetings to advocate for their study protocols and to answer questions in person [10]. Indeed, researchers in our study who were also members of HRECs reported their dual role as a source of training. However, our study shows that the extent to which free communication between researchers and HRECs is promoted appears variable. There was a perception that the overall efficiency and effectiveness of a HREC depended on individuals who were committed to the process.

When ethical concerns arose, researchers reported that they would choose to contact knowledgeable people, before consulting guidelines or rules. An independent, ethics consultative body that has been implemented in parts of the US to address the need for ethical guidance, while alleviating the limited resources of review boards, has shown promising results. Reviews of the board’s function found that it could not only address ethical concerns that arose during the study, which were often outside the purview of the US equivalent of HRECs – institutional review boards (IRBs) – but that it also created a forum for discussion, and a database for challenging and novel cases [36]. This data could be a resource for ethics review boards, and if made accessible to the public, could further help promote open access to, and by extension, trust in, the research enterprise. The feasibility and suitability of implementing an equivalent board in Australia should be further explored.

Most researchers in our study reported that their awareness of informed consent practices arose passively through on-the-job experience, rather than from the use of explicit guidelines or training materials, which is in line with previous research in this area [10]. This form of learning appears to have resulted in a good working knowledge of how to interact with potential participants, especially with regard to the importance of honesty, and of avoiding coercion or undue pressure. However, gaps exist especially in relation to knowledge about regulatory processes surrounding informed consent and ethics review more broadly. Specifically, awareness varied on the National Statement, the types of consent that are permissible, and the role and significance of the information form.

We found that regardless of reported logistical and administrative difficulties, researchers almost invariably believed that written consent was required from participants. Few interview participants were aware of, or had used, verbal or implied consent, despite the National Statement providing that “[c]onsent may be expressed orally, in writing or by some other means” [1]. Interestingly, some researchers interviewed also worked in a clinical capacity, (nurses, doctors, etc.) and were thus familiar with the notion of implied consent in their clinical roles. Reasons for utilising written consent included views that convention dictated it, or that ethics boards required it. A recent study in the US showed that researchers perceived ethics review boards to hold more stringent views on the requirement for consent than they actually did. In particular, researchers underestimated the availability of expedited review for lower risk studies [33]. This may be due to the fact that this type of informal learning magnifies the idiosyncrasies of a working team, and that this would be rarely challenged by external sources. Researchers reported learning about informed consent from creating consent forms from templates that predecessors had used. This results in the preservation of errors which individual researchers may feel powerless to challenge, especially if these become legacy documents and there is lack of support from stakeholder institutions such as HRECs to create new ones. The inconsistencies suggest that practice across different research groups would also vary significantly.

The main limitation of this study is that semi-structured interviews are not designed to capture behaviour, which will require observational or ethnographic studies. Researchers’ practices discussed here are self-reported. Thus, there is also potential for social desirability bias when describing researchers’ practices to others. Observational research would be timely to reveal the types of information actually disclosed to participants, and deviations of actual practice from perceived practice. However, interviews allowed us to capture researcher views in depth, and helped us to draw richer descriptions of the informed consent process – descriptions that accommodate context in describing phenomena [21, 27]. The use of secondary analysis, a transparent audit trail, and work in iterations, preserved the rigour of our findings, and ample quotes were included in supplementary data to verify interpretations.

Participation in this study was voluntary, which may have resulted in self-selecting bias towards researchers who are more interested in promoting a dialogue relating to informed consent. Another limitation is that the heterogeneous sample may mask some of the issues particular to some types of research, e.g. surveys, which may not be reported in depth here. Future studies should report on difficulties present in particular research types, e.g. research in emergency contexts or with participants with impaired capacity.