Timia Raven-Gregg writes: Following my fourth year at medical school, I decided that I wanted to gain a better understanding of clinical research and have the opportunity to undertake a research project myself. For this reason, I chose to intercalate in a degree in Population Medicine. The idea of better understanding research conducted within critical care instantly appealed to me. At the time I was aware of some ideas of the challenges that might be faced, but I had not considered the practicalities surrounding obtaining consent. This project allowed me to have a greater appreciation of research conduct, and how it is impacted by specific clinical environments. This research was supervised by Dr Victoria Shepherd (Centre for Trials Research) and Prof Fiona Wood (Division of Population Medicine) at Cardiff.
Why was this research needed?
Recruitment of critically ill patients is one of the biggest challenges when undertaking vital research to identify effective treatments in critical care. Studies within critical care settings are challenging to conduct, partly due to the need to rapidly administer potentially life-saving treatments, and the ethical issues surrounding informed consent. Little is known about the most effective strategies for participant recruitment within this environment. Within research studies, there is a tendency to underrepresent vulnerable patients, such as incapacitated patients. This is due to the ethical challenges surrounding the consent process.
In the current climate of the COVID-19 pandemic, research regarding participant recruitment strategies is increasingly relevant. Potential therapeutic trials such as the REMAP-CAP and RECOVERY trials require time-critical enrolment and treatment of critically ill patients. Research exploring the different possible methods for recruitment will expand the evidence base and improve recruitment efficiency in this area. This in turn will generate much needed evidence on the effectiveness of potentially life-saving treatments.
What did we do?
Our research took the form of a systematic review and narrative synthesis which brought together available evidence on a range of recruitment strategies and barriers. This included studies which explored several stages in the research recruitment process including eligibility identification, research culture and environment, the consent model used, nature of the approach of the consenting encounter and resource limitations. Studies also explored factors impacting recruitment success, such as technology-based screening systems, rapport between patients/families and recruitment staff, researcher experience and staff availability.
As critical care patients are rarely able to provide their own consent to take part, the most common method of recruitment is by the patient’s legally authorised representative, also known as a surrogate decision maker. However, utilisation of surrogates depends on factors such as availability, and many surrogate decision makers do not know the patients’ preferences regarding research participation. There is a reluctance to use methods like deferred consent, with few studies exploring the effects of this strategy, making it more difficult to assess its effectiveness. Reluctance to use deferred consent centres around uncertainty regarding the ethics of this model in practice.
It is important to utilise the best possible methods of recruitment within challenging environments like critical care. This is to ensure that all patients receive equal opportunity for the potential benefits of treatments used in trials, without discriminating against certain groups of patients due to challenges arising from the consent process. This also reduces any potential recruitment bias of the study, which would underrepresent participants who lack the capacity to consent to research. Additionally, this group of patients are likely to be characteristically different from the rest of the sample. This in turn will affect the robustness of the evidence yielded from the study, further limiting the generalisability of the results.
A total of 15 research papers were included for analysis from systematic searches in four electronic databases. Included studies consisted of a range of mixed-methods, quantitative and qualitative research. These were critically appraised and relevant data were extracted. A narrative synthesis was then used to “tell a story” with the data.
What did we find?
Methods of identifying eligible participants was explored in some studies, involving in-person or technology-based screening. However, these methods were limited by the timeframes in which in-person screening was conducted, and the ability of the software to capture all potentially eligible participants. Researcher experience positively correlated with recruitment success and having a dual consenting process with a staff nurse first approaching the patient with basic information followed by a trial staff member later, proved useful. Many opportunities for recruitment were missed because of narrow recruitment windows, staff availability and time for participant enrolment, recruitment hours e.g., weekday working hours, and lack of surrogate decision maker availability. Whilst first-person consent was the timeliest method, it was less applicable within the critical care environment where surrogate-decision maker consent was the most used approach. Deferred consent was infrequently used, despite studies suggesting few patients thought it was an unacceptable method of recruitment. Pre-emptive consent (prior consent given before being identified as eligible for a study) was recognised as potentially useful within critical care; however, no studies evaluated its use in practice.
The key messages from our research show that there are numerous modifiable barriers to recruitment in critical care. Factors that contributed to recruitment success included rapport between staff and potential participants, researcher experience, expanded recruitment hours and the successful operation of the research team. A large proportion of patients who were not enrolled in studies had no documented reason for lack of enrolment. Monitoring this may help to address clinician bias and differences in recruitment rates between study designs.
What future research is needed?
Future research should aim to explore alternative methods for consenting, such as a two-stage consent model. This involves introducing the aims of the research to a patient or surrogate decision maker, allowing them time to consider their decision, before gaining consent later. Although this may be dependent on surrogate decision maker and staff availability, this could help to positively impact recruitment in the intensive care setting. Monitoring trial conduct closely, by constantly reviewing enrolment of eligible patients, should also be evaluated further. This may allow for a more tailored recruitment process to address study aims. Furthermore, the use of co-enrolment and deferred consent within critical care require evaluation, as these may be useful in trials with narrow recruitment windows. Future research in this field should explore the suitability of identified recruitment strategies across specific research designs, studies of varying risk and different intensive care settings.
What did I learn?
Undertaking an intercalated research project has allowed me to understand the barriers to conducting research in challenging clinical environments and appreciate the importance of having a robust evidence base behind treatments used in everyday practice. I have gained invaluable skills in research, which I will take forward with me into my future medical and academic career. I now have a wider perspective of clinical practice, and I have seen why it is important for healthcare professionals to be involved with research that continues to shape and improve our medical practice.
The full published article can be found at:
Raven-Gregg T, Wood F, Shepherd V. Effectiveness of participant recruitment strategies for critical care trials: A systematic review and narrative synthesis. Clinical Trials. February 2021. https://doi.org/10.1177/1740774520988678