The 9th European Medical Writers Association (EMWA) symposium, entitled ‘Clinical Trials – New Trends’ took place on 6 May. The symposium explored the latest trends in the design, conduct and reporting of clinical trials over 3 sessions, focusing on insights relevant to medical writers and communicators.
A summary of session 2 of the symposium is provided below to benefit those who were unable to attend, and as a timely reminder of the key topics for those who did.
Session 2: News in analysing, publishing and presenting clinical trial data
The second session of the symposium was moderated by Myriam Stieler (Biotronik). It reviewed recent news and innovations around clinical trial data analysis, publishing, and presentation.
Estimands in clinical trials – ICH E9 addendum
Samuel Copt (Biosensors International) opened the session with an overview of the recent addendum (R1) to the International Council on Harmonization (ICH) E9 guidance. These guidelines set out the principles of statistical methodology for clinical trials and have been around for more than 20 years. The objective of the addendum is to improve the planning, design, analysis and interpretation of clinical trials and has been implemented, or is in the process of being implemented, by regulatory authorities around the globe.
As a trial progresses, intercurrent events such as treatment discontinuation, administration of rescue medication or terminal events can affect the interpretation or outcome associated with the study objective. Copt explained that the addendum provides a framework to construct and define an ‘estimand’ or target of estimation, which specifies how intercurrent events will be handled, and “ensures alignment between trial objectives, design, protocol language, trial conduct and statistical analyses”.
The addendum [to the ICH E9 guidance] provides a framework to construct and define an ‘estimand’ or target of estimation, which specifies how intercurrent events will be handled.
Copt went on to describe the 4 attributes used to create an estimand that, when considered together, should address many intercurrent events:
- intervention – the treatment effect of interest
- population – the patients targeted by the clinical question
- variable – the variable or endpoint required to address the clinical question
- population-level summary – a measure for comparing treatments.
Copt stressed that the identification of the estimand is not solely the job of the statistician but must involve the whole clinical team, regulators, and other stakeholders. The estimand should also be clearly defined in the study protocol.
Artificial intelligence in clinical trials and healthcare
Calvin Bahr (Biotronik) gave a thought-provoking summary of where we are at the moment with artificial intelligence (AI) and its potential use in clinical trials and healthcare. Currently, AI is still in its infancy. Artificial narrow intelligence or machine learning is part of our lives – think Alexa or Siri – but it works under very controlled conditions with a limited set of parameters. In contrast, artificial general intelligence relies on machine intelligence, and is still a theoretical process. Artificial super intelligence would require machine consciousness and, for the moment, is the domain of science fiction.
Machine learning could be used throughout the clinical trial process.
Bahr explained that machine learning could be used throughout the clinical trial process:
- trial design – risk assessment using data from prior trials and publications
- trial setup – improve patient enrolment by mining databases to match patients with trials
- trial conduct – monitor disease progression through analysis of biomarkers
- trial closure – data cleaning.
Machine learning is already being used in general healthcare. For example, retinal images can be analysed automatically for diabetic retinopathy, giving results in under 30 seconds. These programs can be superior to physicians at spotting early disease. AI has also been used to predict clinical outcomes based on electronic health records, but there is still more to do to improve reliability.
Different requirements for presentations held on virtual events
Next up, John Gonzalez (Solanum Medical Communications) discussed how the development of presentations for virtual meetings, which were thrust upon us last year, has been a steep learning curve for us all. Gonzalez described some of the challenges and benefits of virtual poster sessions:
|Competing distractions||Flexible attendance|
|Boredom||Added value (voiceovers, videos)|
|Screen fatigue||Live presentations|
|No human interaction||Chat facilities|
Gonzalez outlined the different requirements and the potential opportunities that are available for enhancing virtual presentations. A crucial factor to address is gaining the attention of your audience: Gonzalez explained that the average human attention span has decreased from 12 seconds to just 8 seconds over the last 20 years. People are bombarded with information, most of which they do not take in. This was perhaps being realised even before the pandemic, with the congress poster evolving from the traditional wordy, overly complex “manuscript dump” to a more visual and graphical, and less textual, version.
Interactivity, allowing interrogation of the information with enhanced content, can be one way to keep your audience engaged. Gonzalez recommended that posters should be visually attractive with added value available (such as voiceovers, additional data, animations, or videos) to get the most out of a virtual presentation. With the hybrid congress model likely to become the norm, interactivity or enhanced content should be at the front of our minds when beginning the poster development process. Author buy in is needed from the start.
Interactivity or enhanced content should be at the front of our minds when beginning the poster development process.
Gonzalez finished his presentation with a warning to be aware of predatory conference websites. These sites may look legitimate and may even have been cloned from the original, authentic website. He recommended taking a minute to ensure you are submitting on the correct site.
New trends in publishing
The final presentation of the session was given by Caroline Halford (Digital Publishing Manager, Adis journals). She provided an overview of many of the current trends in publishing, including:
- Open access initiatives
Halford explained the concept of transformative journals as a way for publishers to become compliant with Plan S. A hybrid journal can apply for transformative journal status, requiring a commitment to become fully open access as soon as possible. Halford signposted attendees to the Journal Checker Tool to help identify Plan S compliant publication routes for their research.
- Registered Reports
The popularity of Registered Reports has grown significantly over recent years with over 200 journals now accepting this format. Halford highlighted benefits including the increased transparency of trials and the potential to improve rates of publication of null findings.
- Data sharing and digital features
Data sharing and access to additional digital features were discussed as ways to increase the impact of an article. However, attendees were reminded to consider the inclusivity of enhanced content and to incorporate, for example, subtitles or transcripts with audio or video files to ensure accessibility to all.
Attendees were reminded to consider the inclusivity of enhanced content and to incorporate, for example, subtitles or transcripts with audio or video files to ensure accessibility to all.
- Innovations in peer review
Halford outlined recently introduced strategies where authors submit to a publisher rather than a specific journal, receiving a proposed target journal with their review comments. If the authors decline the offer, they are free to take their comments elsewhere. It is hoped that this will reduce the need for multiple rounds of peer review and accelerate the publication of research.
Halford touched on the challenges arising from media uptake of preprints over the past year. Sensationalised headlines may grab the public’s attention, but the media must do more to explain that data and information from preprints should be viewed with caution as these are yet to be peer reviewed.
The media must do more to explain that data and information from preprints should be viewed with caution as these are yet to be peer reviewed.
- Plain language summaries
Finally, Halford stressed the importance of plain language summaries as more and more patients are searching for information. She reminded attendees to look out for the inclusion of a patient section in the fourth iteration of the Good Publication Practice guidance (GPP4).
Written as part of a Media Partnership between EMWA and The Publication Plan.