The message is clear: the inclusion of decentralised elements in clinical research does not change the guiding principles of clinical trial conduct
While the acronym DCT (Decentralised Clinical Trial) may be relatively new, if we ask if the inclusion of decentralised elements in clinical trials is new, the answer is No. What is new, is the wider recognition by the clinical development community of the extent to which such methods can optimise clinical trial conduct, and both the participant and site experience. The impetus to maintain the momentum gained during the COVID-19 pandemic of their widespread adoption is exactly what the new recommendation paper from the European Medicine Regulatory Network (EMRN) signifies. It is the first formal document (outside of the pandemic) from a regulatory agency on this topic.
Several themes run through the recommendation paper, all centered around the main take-home, that, regardless of where the trial activities are taking place, the same principles apply to trial conduct: participant rights, safety, dignity and well-being should prevail and robustness of the data for regulatory decision-making should be guaranteed. The paper is also very clear that any burden falling upon participants and/or investigators due to inclusion of decentralised elements should be weighed against the potential benefits of such methods, and that investigator and participant inclusion in the planning process is an important factor for trial success.
Whilst the paper offers a series of recommendations for optimal conduct when including decentralised elements, it is important to note that these are just that, recommendations. They are not enforceable. What is enforceable is legislation, and throughout, the paper is clear that adherence to the national laws, regulations, and standards are non-negotiable, regardless of the design of the trial. This indicates that, despite the goal of a harmonised perspective on DCT in the EU/EEA, whilst country-level regulations continue to differ, harmonisation will only exist within these limits.
Appropriateness should be a key decision-driver for the inclusion of decentralised elements.
The appropriateness will depend on the specific trial population, the condition under study, the assessment types in the trial, along with the nature of the investigational medical product(s). Patient and investigator engagement in the development process of the trial design is key, and offers the opportunity to increase recruitment and adherence, as participants are the best placed to provide the perspective on the likely success of the decentralised elements that are planned to be included from their view-point, and investigators can identify how to ensure the quality and integrity of the data.
Evaluating any potential impact to the scientific validity of the study and the integrity of the data being collected is paramount as data collection activities shift to off-site locations. The expectation of the robustness of the data for regulatory decision-making is no different in a decentralised setting. Specific factors are highlighted that could affect trials using decentralised elements, including potential exclusion of participants who are not digitally literate or live in areas with unstable or limited internet connection, the potential for increased amount of missing data, and differences in the conduct of activities across participants and/or sites that could potentially impact the data – this is especially pertinent in the EU where different member states have different regulations.
We should not be throwing every possible technology at trials for the sake of it, but in a careful and considered way where its use is possible and appropriate, and with a goal of trial optimisation and an enhanced participant and site experience. Contingency plans should be in place for any potential disruption to decentralised procedures and/or technology.
The paper details specific considerations and recommendations in six specific areas of trial conduct.
“introducing decentralised elements should be considered as an extension of the clinical trial site with the inclusion of the trial participants’ home, resulting in an additional obligation of oversight for investigators and sponsors”
As mentioned, decentralisation in clinical trials can include certain activities being conducted “off-site” (i.e., away from the central site hub), facilitated by home health nurses and/or technology. Investigators must still maintain the oversight in trials using decentralised elements, and until now we have yet to see a regulatory body provide recommendations on this. The paper focuses on two main topics with regard to oversight: oversight of additional service providers and data oversight. With more parties than ever before involved in clinical trial conduct, clear lines of communication are critical, and trials involving decentralised elements are no different. Participants should have clear information on who to contact for all situations, from device failures to potential adverse events.
Again, the paper brings it back to the two key considerations – participant safety and data integrity – and it is crucial that the specific roles and responsibilities are clearly defined from the outset and the investigator and sponsor must still be able to fulfil their legal obligations and maintain full control of their areas of responsibility. The bottom line is that, irrelevant of the involvement of additional service providers (e.g., technology or home health providers), responsibility for trial activities and the safety and well-being of participants, ultimately lies with the investigator and sponsor.
When an investigator is not involved in the contractual agreement with a service provider and those tasks are ultimately under the investigator’s responsibility, the contract between the sponsor and investigator must document this. What is interesting is the specific call out from the paper that this will enable the investigator to agree (or not) to the use of that service provider and their qualifications where those tasks relate to the medical care of participants, and any applicable training lies with the investigator.
Data oversight begins with the use of any data collection tools and includes ensuring that proper training on the use of digital tools is provided to all stakeholders to mitigate against any risk to the data collection, review, and transmission. Comprehensive training on (serious) adverse events ((S)AE), including what is classified as one, who it should be reported to and the applicable timeframe, along with management of it, is crucial regardless of if the trial includes decentralised elements; however, the use of digital tools that offer the benefit of more ‘real-time’ data transmission and review, and can create the perception that it is review in real-time. Participants must be fully informed of how any information reported via digital tools will be acted upon, and that real-time review is not guaranteed. It is communication to participants around the oversight of the information being submitted that is crucial.
“It should be made clear to the trial participant that the investigator may not review such data in real time, and that if the trial participant experiences any specific safety concern they need to directly contact the investigator to report such an issue.”
It is welcomed to see recognition of this within the paper, and the acknowledgment of risk-based assessment, with review frequency based on the two presiding principles throughout this paper: participant safety and data integrity. Oftentimes, digital methods for data collection are held to higher standards than paper data collection despite their superiority in most instances. One specific benefit is the opportunity to provide alerts to participants for data completion in line with the protocol design to drive compliance, and alerts to investigators to flag potential concerns (again related to participant safety or data integrity such as low compliance), which is specifically called out by the EMRN, along with the importance of the correct functioning of such alerts, with the appropriate risk-mitigations in place.
Exactly where trial tasks are conducted, when they are conducted, and by whom, and how oversight will be maintained, should be clearly reflected in the protocol, along with a high-level rationale and the extent of involvement of additional service providers who are to conduct certain tasks within the trial.
The absence of reference to how oversight would work across borders – when a participant and investigator are in separate jurisdictions – implies that there must always be an investigator present in the country of the participant, and limits the likelihood of a ‘central site’ covering multiple countries. However, the paper does acknowledge the rapid advances in this field of DCT, and that the recommendations will evolve in line with this.
In most trials, consenting is the beginning of the investigator-participant relationship, offering the opportunity to build that trusting environment which can ultimately promote trial compliance, and provides the participant with adequate information about the trial and expectations - a key factor for trial success. In the case of a trial including decentralised elements, this is the point at which the investigator should evaluate any participant-related factors that could impact their use.
What is interesting and represents a position statement from the EMRN, is their recommendation that:
“In general, this should be a physical meeting between the investigator and the potential trial participant. However, in some cases it can be justified that this is done remotely.”
“participants and investigators should be given the option to have the informed consent process on site. . . However, in duly justified cases only the remote option may be offered.”
The cases where it seems to be suggested that it is less justified to obtain consent remotely and that a physical meeting is required, is where the trial population is especially vulnerable, where there more limited knowledge around the investigational product efficacy and safety profile, or in trials with particularly complexity, and those with higher risks associated with trial-specific interventions. However, this very much calls for a subjective decision-making process, as defined thresholds for vulnerability and complexity are absent.
What is clear, is that regardless of remote vs in-person consenting, it is considered essential that communication takes place face-to-face. Therefore, remote consenting should be facilitated by a real-time televisit. It is positive to see that both participant and investigator preference is called out and that optionality is highlighted as the goal-standard.
Much of the narrative for including decentralised elements in clinical trials centres on the opportunity to increase diversity and inclusion, though it is also important that when utilizing technology this should not discriminate (albeit unintentionally) against those who prefer not to use electronic methods and alternative methods should be available. Whilst the premise behind this statement in the paper is clearly related to the informed consent process, it is important that it is not taken out of context and applied to other areas of technology in clinical trials without due consideration; for example, mixing modes of data collection (i.e., paper and electronic) for collecting patient-reported outcome measures can create regulatory concerns around data quality.
As mentioned, the guiding principles of clinical trial conduct are no different for trials including decentralised elements, and regardless of where consent takes place and the methods used (i.e., physically face-to-face, remotely, electronically, or on paper), the process should be detailed in the protocol and clinical trial application to enable ethical review, reliability and confidentiality of the method should be ensured (if remote, then encryption to protect confidential information is important), and it should always be carried out in compliance with the applicable legislation (the CTR, ICH E6, and GDPR), including national requirements regarding electronic signature acceptability.
As with in-person paper consenting, the participant should be able to ask questions: 1) the process must be documented in a way that facilitates verification of the receipt of information by the participant and that discussion between the individual obtaining consent and the individual providing consent took place, and that consent was provided, 2) the investigator should verify the participant’s identity and the participant should be able to ask for proof of the investigator’s identity.
Specifically related to electronic consenting, it is the responsibility of the sponsor to verify that a site agrees to the use and storage of electronic consenting, information and documents provided to the participant must be in a form that they can store and retrieve (e.g., they can download or print), and electronic copies must be protected against modification which would invalidate signatures, procedures for re-consenting and withdrawal of consent (either fully or partially) should be adapted to be appropriate for this mode of consent to ensure mitigation of impact on the participant and data collection. Withdrawal should also be possible outside of the electronic system.
Responsibility for treatment decisions lie with the investigator regardless of where administration takes place, but delivery could be directly to the participant or dispensed and collected from a local pharmacy, and administration could be by the participant themselves or a designated healthcare professional at the participant’s home. Participants should always be made aware of this in advance, and it should be outlined in the protocol. While regulation exists regarding importing IMP into the EU and contractual agreements between the sponsor and site or pharmacy (CTR), delivery to participants is not covered by these.
The paper calls out appropriateness as the decision-driver, and that this may not be appropriate in all situations and a risk assessment should, along with consideration of the trial population, current knowledge of the IMP safety profile, preparation required prior to administration, administration route, if an observation period is required post-administration, and if the storage conditions needed can be met at the participant’s home – taking into account the participant burden. In scenarios where administration can occur at the participants home, but requires a qualified healthcare professional, direct patient shipping may not be appropriate.
Regardless of where it is being delivered, again, compliance with the relevant legislation and maintenance of the quality and integrity of the IMP throughout the supply chain are non-negotiable, with minimistion of transportation steps. It is crucial to plan for any emergency situations and how likely they are to occur, and any risk that could come from delivery to the incorrect individual.
When it is planned that some of the trial procedures in the trial can take place at the participant’s home, whether by the participant themselves, the investigator or site staff, or a contractor, there should be safeguards in place, including the relevant insurance/indemnity documentation. For certain procedures, only those qualified and allowed by legislation should carry these out. Appropriate training should be ensured for all involved.
Once again, consideration around risk and reliability of the data collected is placed at the forefront of the consideration. Where any data collection is remote, monitoring for compliance and oversight of any incoming safety data is going to be a crucial activity to ensure neither of these are compromised, along with clear procedures in place for managing the reporting of any adverse events.
It should be recognized that not all participants will have a suitable home-environment, which should not exclude them from the trial. The paper recommends that the inclusion/exclusion criteria accounts for this and that the informed consent process covers it. Optionality is again highlighted - if participants prefer to see the investigator face-to-face, a direct contact line should be provided.
It is positive to see the recognition of bring-your-own-device (BYOD) methods for capturing patient-reported outcome data from participants, which studies have shown many participants prefer. The paper calls out the importance of providing alternative options in instances where this is not suitable - very much in line with current industry thinking.
Regardless, due consideration to any potential burden resulting from digital data collection in the home environment is important.
Clinical trials are all about the data – the data is used for monitoring of participant safety and evaluation of the efficacy of the investigational medical product, with the overarching principle that the data should be credible, reliable, and verifiable. We live in a world where we are inundated with data collection tools and these must be configured and validated in the context for which they will be used. For example, in instances where source data is captured with one tool is transferred to another location, and upon transference is deleted, both the data and the metadata should be transferred.
As detailed above, decentralisation can include a shift in where that data is collected, from on-site to off-site, and who is collecting that data. From the very outset, a data flow that is clearly defined and understood by all those involved in the trial is going to be critical to the successful conduct of the trial, especially in scenarios where the investigator may not have direct oversight of that data collection – including this in the protocol is recommended. Furthermore, the systems used to collect, store, and manage that data must be water-tight, with the relevant securities in place within each system and whenever transfer from one tool to another occurs, from encryption to well-defined and controlled user access and firewalls. Irrelevant of where the data collection is taking place, the investigator should have access all of the trial source data.
Interestingly, one thing that is highlighted by the paper is the risk for erroneous data entry when it is being collected and entered directly by the participants, and the need for appropriate measures to minimize this – though none are called out. Given this may be specific to the type of data (e.g., whether it is remote electronic capture of patient-reported outcome data, or remote data from a wearable or sensor), it was likely beyond the scope of this paper, and should be considered on a case-by-case basis. Monitoring of the incoming data for early identification of any problems will be important, as detailed in the following section.
The trial monitoring process should account for the specific methods used in the clinical trial and the monitoring itself can also take place on-site or remotely – the paper states that generally a combination of both is appropriate - considering adherence to the principles of necessity and proportionality, ultimately taking a risk-based approach that does not unduly burden the site. In instances where remote access to source documents will be needed, system security should be ensured to maintain confidentiality.
Essentially, it is not new for the trial monitoring process to encompass remote elements, and given that including decentralised elements in clinical trials is not new, monitoring strategies for this are also not new. For example, central monitoring has been around for over a decade. This is perhaps why this section receives relatively little attention in the paper, and primarily directs the audience to the GCP IWG guidance in development for providing a more detailed account of this.
The release of these recommendations demonstrates that the inclusion of, and benefit offered by decentralised elements in clinical trials is being recognised by regulators and is a positive step. That said, the Appendix to the paper detailing differing national provisions across member states, indicates that, despite the goal of a harmonised perspective on DCT in the EU/EEA, we still have some way to go, and that such differences will continue to limit the extent of harmonisation possible.
Overall, the message from the recommendation paper is clear: the inclusion of decentralised elements in clinical research does not change the guiding principles of clinical trial conduct.
The premise behind inclusion of decentralised elements in clinical trials is to improve ease of participation for both sites and participants, and optimise trial conduct. Whilst decentralised elements in clinical trials are not new, the extent of the realms of possibility is an area of continuing innovation and is here to stay. In time this will essentially leading to DCT once again becoming CT, as the elements become absorbed as simply part of the options available for conducting your clinical trial.