The 58^th European Medical Writers Association (EMWA) Conference began with a series of Hybrid Conference days where attendees learned about the latest trends in reporting and conducting real-world evidence (RWE) studies, explored best practices for medical translation, and heard from journal publishers about how to navigate the challenges and opportunities presented by artificial intelligence (AI) in medical writing. The Publication Plan were in attendance to summarise the sessions for the benefit of those who were unable to attend, and as a timely reminder of the key topics for those who did. You can read our summaries of the sessions on RWE and Medical Translation below.

Our report of the Publishers’ Perspectives on AI symposium is available here.

Updates on best practices for conducting and publishing real-world evidence studies

KEY TAKEAWAYS

• When reporting RWE data, consistency and transparency are vital to avoid potentially catastrophic outcomes.
• New guidelines, tools, and registries are available to aid compliance.

The road ahead

Andrea Bucceri (Lumanity, EMWA Medical Communications Special Interest Group) opened the session with a ‘road’ analogy: randomised controlled trials (RCTs)—with their homogenous populations and eligibility criteria—may be likened to a clean, well-surfaced road. By contrast, RWE studies—with their heterogeneous populations, missing patient details, confounders, data gaps, and lack of randomisation—are akin to a muddy, potholed road. Clear guidance is therefore essential.

Clear guidance is essential when embarking on an RWE journey.

When reporting RWE data, consistency and transparency are vital to:

• allow comparison of data from different RWE studies
• plan future studies and interventions
• improve reproducibility.

Bucceri stressed that “major disasters” can occur when reporting guidelines are not followed, exemplified by the embarrassing retraction of the hydroxychloroquine study from the prestigious Lancet journal during the COVID-19 pandemic. Bucceri asserted that in this instance, the authors were supposed to apply the RECORD and RECORD-PE reporting guidelines but failed to do so. Three of the authors stated that they were unable to access the raw study data due to legal issues.

Several other RWE reporting guidelines exist, including the STROBE-ME (observational studies—molecular epidemiology), STREGA (genetic association studies), and REMARK (REporting recommendations for tumour MARKer prognostic studies). In addition, regulatory and Health Technology Assessment bodies have issued RWE study reporting guidance, such as the European Medicines Agency (EMA), National Institute for Health and Care Excellence (NICE), and the US Food and Drug Administration (FDA). Bucceri observed that all these guidelines were developed after the COVID-19 pandemic, which highlighted the importance of RWE in public health decision-making, and prompted regulatory agencies to regulate the RWE field more closely.

Some professional bodies have chosen to go beyond these broad frameworks to develop guidelines tailored to specific disease areas. In the second part of the session, Bucceri highlighted one such guideline: the European Society for Medical Oncology Guidance for Reporting Oncology real-World evidence (ESMO-GROW).

The ESMO-GROW guidance for reporting oncology real-world evidence

ESMO-GROW is the first expert-based guidance designed specifically for reporting oncology RWE studies. A multidisciplinary expert taskforce developed the guidance in response to concerns that existing guidelines lacked specificity, and that the need to use multiple complementary guidelines was proving burdensome to authors and journals.

Available as a free online tool, ESMO-GROW can be used to report descriptive, analytical, and pragmatic RWE studies. The tool consists of 35 reporting recommendations across 6 sections spanning the title of the manuscript through to its discussion and conclusions. Each section contains specific recommendations about what to include in the manuscript, with the aim of standardising oncology RWE publishing output. The tool asks the user to select whether they have fully reported, partially reported, or not reported each recommendation (or whether the recommendation is not applicable), and on completion provides a score by which to assess compliance.

The guidance was developed in consultation with and will be available as part of the EQUATOR network. The ESMO-GROW criteria are already in use, and Bucceri advocated acknowledging the tool whenever it is used in a manuscript.

What to include in your manuscript: the experts’ view

For the next part of the session, Bucceri presented feedback from RWE journal editorial board experts about what to include in manuscript submissions.

All studies (regardless of size) that are well-conducted, novel, and that address an important clinical question are worth publishing!

Of note, editorial experts stressed that all studies (regardless of size) that are well-conducted, novel, and that address an important clinical question are worth publishing! This is particularly true of RWE studies, which often include populations excluded from traditional study designs such as the elderly, children, or pregnant women.

Specifically, submitting authors should:

• Provide a detailed rationale for the study
• Pre-register the study in a public repository (and provide details of pre-registration)
• Address strengths, limitations and confounders in a dedicated paragraph within the Discussion (limitations and confounders are sometimes more interesting than the data itself)
• A qualified statistician should be among the authors, with their contribution clearly stated
• Work closely with their statistician to report and minimise bias in the manuscript
• Provide a detailed protocol of data extraction, including key variables (plus lists and codes) and combinations for defining study subjects, and detail who performed the extraction
• Exercise caution when using significance testing – correlation does not prove causation!
• Conclusions must be in line with the data presented!

Pre-registration

Bucceri then focussed on pre-registration, the process of publicly sharing a research and analysis plan in advance in a public repository. Pre-registration may help prevent overfitting, p-hacking, cherry picking, or HARKing (Hypothesising After the Results are Known), increase transparency and rigour, and boost public confidence in research. The EMA and FDA have issued their own guidance on pre-registration.

Stakeholders have previously raised concerns that registries such as ClinicalTrials.gov were not built with RWE studies in mind. To address this, The Professional Society for Health Economics and Outcomes Research (ISPOR) and other collaborators have developed the Real World Evidence Registry—a purpose-built platform for RWE study pre-registration. The registry is not specific to the EU and can be used by researchers globally. It is also suitable for medical device RWE studies.

Furthermore, the EMA and Heads of Medicine’s Agencies (HMA) have launched the HMA-EMA Catalogues of real-world data sources and studies (with an accompanying Good Practice Guide). Potential uses of these catalogues include data source identification and benchmarking of data sources via metadata.

In the Q&A, questions were raised about how journals check that pre-registered studies are eventually published. Bucceri responded that this is good practice but not mandatory, and that the imperative to do this should be policy-driven from companies themselves. He acknowledged that incomplete studies are an ongoing issue.

The home run

At the end of the Q&A, Bucceri closed the session with a baseball analogy from the expert feedback he had received. The editorial board experts have a ‘three strikes’ policy regarding RWE study manuscript rejection:

• If the study was not pre-registered…strike one!
• If the rationale is not properly defined…strike two!
• If the statistician is not in the author list…STRIKE THREE (manuscript rejected)!

But if you have all of them in your RWE study submission, you have a good chance of a home run!

The essentials of medical translation: foundations and best practices & key translation topics

KEY TAKEAWAYS

• Medical translation requires specialised expertise, combining language skills, subject knowledge, research, and precision, often under tight deadlines.
• Addressing source document issues and expanded training for writers and translators are critical to improve clarity, reduce errors, and streamline processes for creating global content.
• Technology integration and effective workflow management are a key part of ensuring compliance, quality assurance, and efficiency.
• AI, as in medical writing, can support processes under strict human supervision to mitigate risks related to confidentiality, accuracy, and regulation.

This seminar was co-chaired by Ana Sofia Correia (EMWA Medical Communications Special Interest Group – Translation Enthusiasts Subgroup), Claire Harmer (EMWA Medical Communications Special Interest Group – Translation Enthusiasts Subgroup), and Valentina Luridiana (TREMÉDICA) and provided an in-depth overview of medical translation, underscoring the unique demands and best practices within this specialised field.

The session was structured in two parts. The first half—Foundations and Best Practices—featured presentations addressing critical aspects of medical translation. These included the profile and skills of medical translators, key workflows for effective translation (such as managing multilingual content and documentation requirements), and strategies to overcome common challenges. The second half—Key Translation Topics—offered a series of focused breakout sessions where attendees and presenters could exchange ideas and experiences. These sessions explored topics such as the expectations, needs, and challenges faced by stakeholders, best practices for creating translation-ready materials, the role of technology, strategies for culturally sensitive translation, and the variety of translation services available.

The presentations highlighted the essential skills for those interested in entering the profession, the complexities of navigating the medical and pharmaceutical sectors—particularly for medical translators and writers, who comprised nearly two-thirds of the session attendees—and the role of technology, especially in an industry with strict regulatory requirements.

“Translators in the pharmaceutical industry are a bridge connecting patients, healthcare providers, and pharmaceutical companies across borders.” – Elisa Manzanal Merino

The multifaceted role of the medical translator

In the first presentation, Alina Cincan (Inbox Translation) discussed how medical translators, often from diverse backgrounds, need a unique blend of skills to perform their role. Key attributes include a robust grasp of source and target languages, a solid understanding of the life sciences, strong research skills, discipline, time management, and accuracy, to name a few. Translators must also navigate additional challenges such as:

• complex terminology
• convoluted source texts
• challenging deadlines
• emotional challenges (reviewing sensitive patient information or critical diagnoses can be emotionally taxing)
• inadequate translations from machine translation or translation memories.

Several solutions and specific tools and tips were offered, whether undertaking continuing professional development or adopting software like AntConc, a free corpus analysis toolkit for text concordancing and analysis.

The future of translation, given the advent of AI, featured throughout the presentations, with a particular focus on the concerns and risks related to confidentiality, accuracy, regulatory matters, and non-specialist datasets; however, Cincan acknowledged that medical translation has seen less impact compared with other specialisations. Cincan concluded with some final recommendations for translators, including continued specialisation, networking, joining professional associations, and keeping abreast of developments.

Traversing complex workflows for accurate and compliant translation

Representing the perspective of a medical translation agency, Yana Safina (mt-g) elaborated on how their workflows—such as those for complex multilanguage projects in the sphere of clinical studies—are shaped by compliance with the International Organization for Standardization (ISO); these include standards for translation, proofreading, back translation, and master review; machine translation and post-editing; and text analysis, translation, and revision. Safina also highlighted the advantages of computer-assisted translation (CAT) tools, including time efficiencies, consistency, and improved formatting. While beneficial for some projects, these tools have their pitfalls and intricacies, but agencies are well positioned to provide knowledge and advice on their use.

Safina also emphasised the complementary role of machine translation and post-editing, which can accelerate workflows without replacing human translators, before moving on to quality assurance. This crucial and comprehensive process begins with tested translators, client requirements, feasibility checks, and desired ISO standards; continues with established workflows, preparation of source texts, and choosing a CAT tool; and ends with translation memory and terminology checks, internal quality assurance in the CAT tool, desktop publishing, and layout checks. Once again, the potential of AI was mentioned in relation to integration with or replacement of CAT tools, with testing underway in the quality assurance process.

“AI and machine-learning applications used for drafting, compiling, editing, translating, tailoring, or reviewing medicinal product information documents should be used under close human supervision.” – EMA Reflection paper on the use of AI in the medicinal product lifecycle

A prescription for precision: meeting the rising demand for medical translation

In the second part of this seminar, Elisa Manzanal Merino (Normon Laboratories) shared insights from her role as an in-house pharmaceutical translator, allowing her to explore the intersection between the pharmaceutical industry and translation, and between language and science more broadly. The industry’s remarkable growth has led to rising demands for precise, highly specialised translation services. With the global demand for new treatments and emerging markets driving the need for multilingual documentation, translators in this field require deep subject matter expertise and familiarity with resources like the EU and US pharmacopoeias and International Council for Harmonisation guidelines.

The medical translation industry’s remarkable growth has led to rising demands for precise, highly specialised translation services.

Pharmaceutical companies may require in-house teams to provide ‘inverse translation’ (ie, translating from your native language into your non-native language); the dominance of English in scientific and pharmaceutical contexts and restricted timelines or small in-house teams potentially influence this demand. While outsourcing is common, in-house translation departments offer advantages, including streamlined workflows, rapid query resolution, and quality consistency. This efficiency is essential given the volume and variety of documents, from registration dossiers to marketing materials and patient-facing leaflets.

“Without translation, we would be living in provinces bordering on silence.” – George Steiner

Celestial navigation: following the sun to meet translation and localisation challenges

Courtney Vasile, a senior localisation project manager at Medtronic, discussed the challenges specific to the MedTech sector, where geographical and technological considerations come into play in their translation practice. At Medtronic, their complex workflow involves a two-step process of translation and revision, completed by two different linguists (native speakers who have extensive medical translation backgrounds), followed by extensive quality assurance checks and review by an in-country subject matter expert.

A ‘follow the sun’ approach ―leveraging global teams in the medical translation process― can help to balance tight deadlines and the need for high-quality outputs.

Though the process can be cumbersome when balancing tight deadlines with the need for high-quality outputs, in-house departments allow for agility when it comes to resolving queries, and there are more opportunities to educate internally and with clients on the complexities of the translation process. Furthermore, they have adopted a ‘follow the sun’ approach to leverage global teams and support rapid turnarounds, with offices spanning multiple time zones. Though their challenges might be considered standard across the industry, ongoing training, whether on writing for a global audience or how to use their proprietary content management system effectively, ensures a solution-oriented approach and a focus on innovation.

In summary, medical translation requires subject expertise, precision, and a commitment to quality underpinned by best practices that embrace technology integration and collaborative workflows.

With this in mind, the breakout rooms allowed for a deeper discussion on key translation issues:

• Mapping the medical translation ecosystem: Roles, expectations, and challenges – moderated by Katarzyna E Slobodzian-Taylor (Mastermind Translations)
• Translation-ready materials: Best practices and the role of the translation brief – moderated by Ana Sofia Correia
• Technology and translation: What tools do medical translators use to enhance quality and increase productivity – moderated by Thomas Tolnai (Tolnai Translations)
• How to produce a medical translation that is culturally sensitive? – moderated by Nur Ferrante (Art of Diversity)
• Beyond traditional translation: Exploring diverse modalities in medical communication – moderated by Claire Harmer

As Manzanal Merino summarised, “translators in the pharmaceutical industry are a bridge connecting patients, healthcare providers, and pharmaceutical companies across borders”. Medical translation is a driving force behind equitable access to healthcare, and this seminar emphasised the essential role that translators play in safeguarding accuracy and compliance within healthcare communications.

Our report from the Publisher Perspectives on AI session can be found here.

——————————————————–

Written as part of a Media Partnership between EMWA and The Publication Plan, by Aspire Scientific, a proudly independent medical writing and communications agency that believes in putting people first.

——————————————————–

KEY TAKEAWAYS

• Translation tools such as CAT, NMT, and, more recently, generative AI, have revolutionised the field of medical translation, increasing efficiency and aiding consistency.
• While the widespread use of these translation tools appears to now be inevitable, it is essential to fully understand their limitations and the continued, critical role of the human translator.

Artificial intelligence (AI) looks set to transform the fields of medical research, publications, and translation, sparking considerable discussion to date around potential opportunities and pitfalls. Medical translation in particular has already undergone significant changes, with the introduction of digital tools such as computer assisted translation (CAT) and neural machine translation (NMT).

In a recent article published by the European Medical Writers Association (EMWA), medical translator Ann Marie Boulanger reviews available technologies and outlines the benefits and limitations of NMT and generative AI tools in medical translation.

Pros of generative AI and machine translation

• Increased efficiency: NMT translation tools can improve productivity by around 20%.
• Standardisation: Consistency is key when it comes to medical terminology. Translation tools have been invaluable for ensuring terms are standardised and up to date.
• Multilingual support: NMT/generative AI tools have been useful for rapidly processing and translating large amounts of information, increasing the speed at which it can be disseminated to scientists, doctors, and patients.

Limitations remain

• Overconfidence: While machine/AI translations appear increasingly sophisticated, it would be dangerous to assume that this makes them accurate.
• Limited contextual understanding: Medicine is a complex field, filled with rapidly changing terminology, acronyms, and confusing shorthand notation. These factors render it challenging for machine translation to reliably translate texts.
• Confidentiality concerns: Free tools are often not secure, posing confidentiality concerns.
• The human impact: NMT/generative AI tools work best in the hand of experienced translators, which presents a quandary for the industry. A shift to using machine translation with human postediting, often by less experienced translators and for poorer pay, has caused compensation for medical translation to plummet despite the high level of expertise required.

The essential role of the human translator

While the use of NMT/AI tools may have become inevitable, Boulanger  argues that medical translators must “view machine translation and AI as nothing more than aids, tools in a toolbox, as opposed to solutions designed to do the work for them”.

Medical translators must “view machine translation and AI as nothing more than aids, tools in a toolbox, as opposed to solutions designed to do the work for them”.

To this end, Boulanger urges the industry to remember that human translators will always be needed, at the very least for final validation, especially in a field as complex and critical as medical translation.

————————————————–

The 12^th European Medical Writers Association (EMWA) symposium, entitled ‘AI in Medical Writing’ took place on 9 May. The symposium explored technological aspects of AI, ethical considerations, and showcased practical applications for medical writers and communications specialists. If you missed the afternoon session, you can catch up on the key themes with our summaries below, or get a quick refresher if you were in attendance!

You can read our summary of the morning session of the symposium here.

Librarians are essential in bridging the AI gaps

KEY TAKEAWAY

• AI presents medical writers and the broader pharmaceutical industry with great opportunities but also valid concerns; library teams have an important role to play in supporting effective and ethical use of AI and promoting AI literacy.

In this session, Jill Shuman (Takeda) considered the proliferation of artificial intelligence (AI) tools and the opportunities and challenges they pose for medical writers. It is estimated that generative AI can bring $60–100 billion value to the pharmaceutical industry annually. AI is also a powerful tool in the face of ‘infobesity’ – with 2 new papers added on PubMed every minute, keeping up with the literature becomes ever more challenging.

There is a pressing need for tools that can assess and extract scientific information faster and deeper, with AI being particularly useful for applications including data extraction for systematic reviews and competitive intelligence. At the same time, AI raises a number of valid concerns, particularly with regards to ethical and copyright issues. Library teams within pharmaceutical organisations have already been looking at the issues for some time and have a critical role to play in supporting the appropriate adoption of AI. In particular, librarians can work to promote AI literacy, ensuring that use of AI tools is both effective and ethical. They can also support medical writers in developing the AI skills that will soon be a prerequisite for the job.

Library teams within pharmaceutical organisations have a critical role to play in supporting the appropriate adoption of AI.

AI-assisted tool for academic writing – supporting researchers in sharing knowledge

KEY TAKEAWAY

• Application of AI to the process of creating a book can reduce the time to publication and the burden on authors and editors, but humans remain central to the process.

Vivien Bender (Springer Nature) described an innovative project that brought together authors, editors, and experts from across Springer Nature to develop a new academic book using generative AI. Creation of the book followed a design process approach, with the team drawing on AI support at each stage in the process. The process also followed 5 principles for the use of AI in publishing: dignity, respect, and minimising harm; fairness and equity; transparency; accountability; and privacy and data governance.

The experiment highlighted the importance of engaging an interdisciplinary team in the development process and that, while AI can be a valuable and powerful tool, humans remain central to the process, relying on their expertise on the subject matter and skills in areas such as high-quality editing. Humans must also continue to take ultimate responsibility for the content. The application of AI accelerated the publication process, making topical information available sooner and reducing the time demands on authors. By assisting authors in areas where they have less experience or skills, AI can also lower barriers for those looking to publish their work.

The experiment highlighted the importance of engaging an interdisciplinary team when incorporating AI into the writing process.

Translation in the era of AI

KEY TAKEAWAY

• AI is a game-changer for translation services; human translators still have an important role to play but need to adapt and refine their skills to make effective use of AI in their role.

AI is having a dramatic impact on translation services, raising the question of whether human translation has a future in the face of rapidly advancing AI tools. Translator and conference interpreter Nora Díaz (Consultant Translator) described the arrival of AI as a game-changer, noting that, depending on the AI engine, machine translation can now rival human translation. Generative AI is widely available and can provide context-aware translations adapted to particular audiences. The potential benefits of AI translation include faster turnaround combined with improved accuracy.

The uptake of AI by translation companies has been rapid, driven by the need to remain competitive. The impact for translators has been mixed – while AI provides them with enhanced tools it also puts their job security at risk. However, companies are increasingly adopting a hybrid approach which retains the human translator as an essential element, with AI used to generate a ‘pre-translation’ which the human translator then refines through a very close edit and check. In this rapidly changing environment, translators need to reskill and upskill. In particular, translators should look to further their skills in developing AI prompts, which are critical to ensuring the quality of machine translations.

Translators should look to further their skills in developing AI prompts, which are critical to ensuring the quality of machine translations.

Structured content authoring

KEY TAKEAWAY

• Generative AI is well suited to use alongside structured content authoring in the development of a range of clinical documents.

Mati Kargren (Parexel International) considered the application of AI and structured content authoring (SCA) to the development of clinical documents across the product lifecycle. SCA uses an approach in which information is broken down into components (eg, study design, patient characteristics, interventions, etc.) that can then be rearranged and reused across multiple documents. Benefits of the SCA approach can include increased consistency, faster turnaround times, reduced need for manual intervention, and improved tracking of content.

AI can be particularly effective where clear structures are in place. Structured content makes for more reliable AI training and, in turn, more reliable AI performance. At the same time, the clear and consistent structure of many clinical documents makes them well suited for generation by AI trained on the structured content.

AI can be particularly effective where clear structures are in place. Structured content makes for more reliable AI training and, in turn, more reliable AI performance.

Dispelling the myths of ChatGPT and misconceptions of AI for your medical writing

KEY TAKEAWAY

• Generative AI and machine learning are not about to replace human medical writers, but medical writers need to adapt and learn how to use AI appropriately in their work.

Depending on where you look, AI is either our gateway to a golden age or a fast-track to unemployment and poverty. David Piester (Symbiance) looked to dispel some of the myths surrounding generative AI and its impact on medical writers. While concerns about the risks of AI are valid, fear is beginning to subside with growing familiarity. The reality of AI and machine learning is that, while it will have an important role to play, it will not replace medical writers or be able to write a complete clinical study report (CSR) on its own.

Medical writers need to work alongside AI, using it for specific tasks, such as structured and formatted processes, and for repetitive and data-heavy tasks. The requirement for closed-loop systems to ensure data privacy and intellectual property rights is a key barrier to the use of AI to generate CSRs and other documentation. The medical writer remains essential and in short supply. However, writers need to adapt to apply AI effectively and appropriately. As Piester noted: “AI will not replace you. A person who’s using AI will replace you.”

While concerns about the risks of AI are valid, fear is beginning to subside as familiarity with AI grows.

Empowering regulatory medical writers: leveraging tools to enhance your writing

KEY TAKEAWAY

• Generative AI tools are available that offer seamless integration with other applications, and are best deployed in a small batch approach to specific tasks within the overall document development process.

Philip Burridge (Morula Health) considered the AI tools available to medical writers to assist and enhance their work. In particular, Burridge focused on Microsoft Copilot, based on its seamless integration with widely used Microsoft Office applications such as Word, PowerPoint, and Excel. Given the critical issue of data confidentiality, it was noted that prompts, responses, and data used within Copilot remain within the Microsoft 365 service boundary and can be locked down by particular users and accounts. However, this does not preclude Microsoft using your data in some way.

Benefits to medical writers of AI tools such as Microsoft Copilot include time savings related to mundane and repetitive tasks, accurate outputs, integration with other applications such as word processing and spreadsheet programmes, and data security on a user level. It was noted that quality outputs require quality inputs in the form of well-devised prompts, and that the best use of AI is for small batch work, applied to specific, structured tasks within the overall document development process.

Quality AI outputs require quality inputs in the form of well-devised prompts.

How will medical writers work with AI

KEY TAKEAWAY

• Use of AI tools has the potential to allow medical writers to focus on the strategic elements of their role; while writers need to know how to use the tools, they do not need to understand in detail how they work.

It is clear that AI is driving a shift in the role of the medical writer. Julia Forjanic Klapproth (Trilogy Writing and Consulting) explored how medical writers can work together with AI and examined some common misconceptions concerning AI and medical writing. Firstly, Forjanic Klapproth countered the opinion that medical writers will need to be highly tech-savvy, noting that they will not need to understand in detail how AI tools work in order to use them. She used the analogy that most of us can drive a car successfully but relatively few understand the detailed workings of car motors. Learning to use AI tools should be no different from mastering other software tools.

Another misconception is that AI will remove the strategic element of the writer role. Conversely, Forjanic Klapproth argued that AI will make medical writers more strategic, freeing them from mundane and repetitive tasks to focus on guiding the direction of projects and gaining deeper insights from the data. The medical writer has a key role as the ‘architect’ and ‘story builder’ of the output, using their vision to steer the AI tool to a successful output. Use of AI should also reduce the potential for bias compared with humans when extracting and assessing data. Ultimately, AI should allow medical writers to get more done more quickly and with greater consistency and accuracy, at the same time as allowing them to focus on strategic tasks such as meaning and messaging.

AI should allow medical writers to get more done more quickly and with greater consistency and accuracy.

AI in regulatory medical writing – opportunities and challenges

KEY TAKEAWAY

• Use of rules-based AI can substantially improve speed and efficiency of preparing regulatory documents, freeing medical writers from repetitive tasks to focus on strategic authoring.

Eishita Agarwal (GSK) looked at how innovative AI-driven systems are driving advances in regulatory medical writing. In particular, rules-based AI can be used to increase the speed and efficiency of developing documents such as study reports and clinical summaries. AI tools can also improve accuracy and consistency. Agarwal emphasised that AI is an enabling technology rather than a ‘magic bullet’ and needs to be deployed alongside other enablers within a multidisciplinary approach engaging all key stakeholders.

Practical experience of deploying rules-based AI to development of 10 CSRs demonstrated substantial reductions (~50%) in development time for 70% of the CSRs. For the remaining 30%, efficiencies were held back somewhat by resistance to changing mindsets and adopting new working practices. Use of AI is redefining the medical writer role, empowering writers to focus on strategic authoring while deploying technology to handle repetitive tasks.

Short intro to the EU AI Act and its impact

KEY TAKEAWAY

• The European Union AI Act is a long overdue piece of legislation that aims to maximise the benefits of AI while mitigating the risks, and holds providers and deployers accountable for the ethical and risk-conscious implementation of AI.

The European Union (EU) AI Act is a ground-breaking piece of regulation that governs development and use of AI within the EU; it aims to promote human-centric and trustworthy AI at the same time as protecting health, safety, and fundamental rights, while still supporting innovation. The legislative framework around AI is a dynamic and rapidly evolving field, and Ward Neefs (Pfizer) overviewed key features of the AI Act.

The Act classifies the risks associated with AI into 4 categories – minimal, limited, high, and unacceptable – with healthcare falling within the high-risk category and thus requiring the strictest safeguards. It overlays existing regulations in areas such as medical devices and in vitro diagnostics, bringing with it some additional requirements. Neefs concluded that AI carries a high risk if it is applied without careful consideration of its limitations and potential for bias. However, if good data modelling practices are followed, machine learning has the potential to do more good than harm.

The EU AI Act aims to promote human-centric and trustworthy AI at the same time as protecting health, safety, and fundamental rights, while still supporting innovation.

How to create an effective prompt: a mandatory skill for the medical writers

KEY TAKEAWAY

• Prompt engineering is becoming a mandatory skill for medical writers to learn how to make effective use of generative AI tools.

Namrata Singh (Turacoz Group) addressed the critical issue of creating prompts for AI tools and how this is becoming an essential skill for medical writers. Prompts are instructions entered into the AI interface and need to be engineered to yield precise, coherent, and pertinent responses. Prompts fall into a number of different categories but all require a considered and creative approach in order to achieve the best outputs.

Singh described the CLEAR Framework, which encapsulates 5 factors that are central to effective prompt engineering: Concise, Logical, Explicit, Adaptive, and Reflective. An understanding of the parameters that determine the effectiveness of prompts is helpful, but learning prompt engineering comes from exploring, interacting with the tools, and learning from mistakes.

Learning prompt engineering comes from exploring, interacting with the tools, and learning from mistakes.

Optimising medical content creation: a strategic framework for implementing generative AI

KEY TAKEAWAY

• Development of a strategic framework for implementing generative AI can help to ensure optimal resource utilisation and timely adoption of new technology, as well as guiding medical content creators through the challenges of applying generative AI.

Generative AI is a transformative technology that comes with significant challenges and limitations. Keyur Brahmbhatt (Merck KGaA) presented a strategic framework for implementing generative AI for medical content creation, providing for rapid implementation and optimal resource utilisation.

Generative AI is a rapidly developing technology and comes with limitations including bias, ‘hallucinations’, and intellectual property and privacy concerns. A number of approaches are available to overcome or mitigate the limitations, ranging from quick, low-cost options such as prompt optimisation, to lengthy, high-cost options such as training new, customised large language models. A horizontal integration roadmap for applying generative AI across a range of medical contents can streamline efforts and investments when applying this rapidly advancing technology.

Summary and conclusions

AI technology is set to redefine the role of medical writers but not make them redundant. Applying AI to repetitive and mundane tasks can improve accuracy and consistency while accelerating the development of medical materials, freeing up writers for more strategic activities.

——————————————————–

Written as part of a Media Partnership between EMWA and The Publication Plan, by Aspire Scientific, a proudly independent medical writing and communications agency that believes in putting people first.

——————————————————–

The 12^th European Medical Writers Association (EMWA) symposium, entitled ‘AI in Medical Writing’ took place on 9 May. The symposium explored technological aspects of AI, ethical considerations, and showcased practical applications for medical writers and communications specialists. If you missed the morning session, you can catch up on the key themes with our summaries below, or get a quick refresher if you were in attendance!

You can read our summary of the afternoon session of the symposium here.

Harnessing AI for efficient systematic reviews in medical publications

KEY TAKEAWAY

• AI tools can assist with the different steps of developing systematic medical reviews; writers are encouraged to learn how these tools work to improve their workflows.

Sepanta Fazaeli (Stryker) presented the opening session of the symposium on how natural language processing (NLP) models can be used to expedite the development of systematic medical reviews. NLP models have evolved from traditional machine learning models (with no semantic understanding) and deep learning models that use neural networks (with some semantic understanding), to large language models (LLMs) such as Open AI’s ChatGPT, which offer powerful interpretation of text and the ability to better capture the nuances of human language.

NLP models have evolved as a powerful means for interpreting human language.

Fazaeli outlined a workflow for developing systematic reviews using AI:

1. Query generation and retrieval: state of the art tools connect to a database, eg, PubMed, for data retrieval
2. Screening: key studies are prioritised
3. Appraisal and extraction: most often based on the abstract alone to reduce computational demands, with a focus on PICO elements (Patient/population, Intervention, Comparison, and Outcomes)
4. Analysis and report generation: quantitative and qualitative analyses; PRISMA diagrams are updated as new studies are integrated

There are now multiple tools offering automation of some or all of these steps, though none are necessarily validated. Writers should select tools based on what they want to achieve, and query the validation, metrics, explainability, and the specificity and sensitivity of a tool before purchase or use.

Beyond the hype: 5 ways you can use your domain knowledge to supercharge your writing with AI

KEY TAKEAWAY

• Writers should use generative AI tools with an educational mindset and become skilled in prompting AI in order to obtain the desired output.

Avi Staiman (Academic Language Experts and SciWriter.ai) provided recommendations on how writers can make the most of AI tools using their own subject knowledge. Staiman emphasised that LLMs need to be guided in an iterative manner to achieve the best output.

Effective prompting is key to obtain the desired results from LLMs. When prompting, the following elements can be used to tailor the output:

1. Role – who you want the AI model to be, eg, a scientist, a medical writer, or a patient
2. Goal – what you are trying to achieve, eg, write an academic article
3. Level – eg, lay text versus scientific writing
4. Few-shot prompting – giving the tool examples, eg, “here is a good example of an introduction section of a randomised controlled trial”
5. Personalisation – using specific instructions, eg, “only use papers from the last 5 years”
6. Constraints – what you don’t want the tool to do, eg, “do not provide a summary”
7. Iteration – repeating prompts in order to optimise the output

Staiman gave the following example of an effective prompt using some of these elements:

“You are a science writer [1] writing an article for the New England Journal of Medicine [3]. I want you to write an exhaustive literature review [2] on the topic of the main symptoms of colon cancer including gaps in the research. The literature review should focus on research published in the last few years [5]. Don’t include an introduction or conclusion [6].” 

From ink to code – the evolution of medical writing in the AI era

KEY TAKEAWAY

• Medical writers should leverage generative AI tools as ‘copilots’, with documentation at each stage of the writing process, to avoid unintentionally propagating errors in scientific publications.

Ashish Uppala (Scite.ai) discussed medical writing in the era of AI. The arrival of generative AI tools has enabled writers to delegate a greater proportion of cognitive load than ever before. However, the human writer is still responsible for the thinking process – this was true historically when words were written physically on clay and then using pen and paper, and is still true today.

Rather than using generative AI tools as automated agents, which might increase the risk of bias and error propagation, Uppala encouraged writers to leverage these tools as copilots to optimise efficiency, with documentation at each step of the process. To avoid unintentionally propagating errors in publications, medical writers should use tools such as Scite.ai that ‘show their work’ by indicating the raw information used to generate an output. Uppala concluded his presentation by calling on medical writers to provide feedback to entrepreneurs to help improve generative AI tools for scientific publications.

The methods of written communication may have evolved, but humans are still responsible for the thinking process.

AI and IP: ramification for vendor, provider and customer agreements

KEY TAKEAWAY

• Medical writers should be aware of potential intellectual property issues when using AI tools. Writers and companies need to ensure they comply with all applicable laws when using AI.

Carlo Scollo Lavizzari’s (specialist in intellectual property protection) talk focussed on intellectual property (IP) in the context of AI. As the use of AI tools becomes more routine, organisations may need to update their legal contracts and agreements. IP legislation for AI is a highly evolving landscape, so companies need to be aware of changes to the law. There are IP considerations throughout the process of using AI, from the inputs used to prompt AI (which may contain existing IP), to the resulting outputs (which could infringe upon existing IP or constitute new IP). Protection of AI tools themselves using IP is also a current topic, with an increasing numbers of patent applications for AI-related inventions.

Lavizzari made the following recommendations for individuals and companies when using AI tools:

• understand and document all processes involving AI – eg, the tools and prompts used
• be aware of any restrictions from clients when using AI – eg, when AI tools cannot be used
• be aware of accuracy limitations of AI tools – they can often hallucinate
• know and follow the applicable laws – eg, the recent European Union (EU) AI Act
• follow guidelines and ethical best practices
• only agree to what you can do using AI, and try to state in vendor agreements what you will/will not do using AI tools
• document the responsibility falling upon the client – seek indemnity, warranty, and ‘hold harmless’ clauses
• think ‘insurance’.

Copyright and artificial intelligence: an overview of how they intersect

KEY TAKEAWAY

• LLMs are trained using vast amounts of copyrighted materials, and these materials are copied, stored, and recreated by the LLMs; collective licensing allows for the efficient utilisation of copyrighted materials by AI systems.

In his presentation, Victoriano Colodrón (Copyright Clearance Center) provided an overview of the basic principles of copyright and how copyright and generative AI intersect. Both economic and moral rights are implicated in the training of AI technologies. LLMs, for example, are trained using vast amounts of copyrighted materials, and these materials are copied, stored, and recreated by the LLMs.

Generative AI tools can infringe on copyright in two main ways:

1. Ingesting copyrighted materials during training
2. Producing output that contains identical or substantially similar material to the protected work

Although many countries have no specific AI-related laws and rely on existing statutes, a major development has been the recent approval of the EU AI Act, which emphasises that AI users need permission to utilise copyrighted work unless exceptions apply. Transparency is a key issue globally, and the EU AI Act will require generative AI providers to make a sufficiently detailed summary of copyrighted works used to train their systems publicly available, with a similar bill pending in the US.

A key question is how AI providers can obtain permissions to use protected works in their systems. Rather than using direct licensing from individual rightsholders, Colodrón recommended voluntary collective licensing, which involves the aggregation of rights from multiple rightsholders by a collective licensing organisation, which then collects royalties from users and distributes them to rightsholders. Collective licensing thus enables a faster and more convenient way for content users to gain access to rights.

Colodrón emphasised that outputs from generative AI tools are improved by the use of high quality, responsibly sourced copyrighted works, which increase accuracy and reduce bias, and thus it is in everyone’s best interests that AI is paired with respect for creators and copyright.

Generative AI tools can infringe on copyright by ingesting copyrighted materials and by producing output material that is identical or substantially similar to the original work. 

Neurobiological roots of artificial intelligence

KEY TAKEAWAY

• Since the inception of AI in the 1950s, there have been rapid advances in building technologies that aim to mimic human intelligence.

Pawel Boguszewski (Nencki Institute of Experimental Biology) discussed the evolution of intelligence in biology and in artificial systems. Although there is no one clear definition of intelligence, there is a general agreement that it is based on the ability to learn and apply information. We now know that rather than being a reactive machine, the human brain is a predictive machine, which allows it to respond to the environment in real time. Neuroscientists are now trying to elucidate the parts of the brain responsible for predicting events.

Just as living intelligence has evolved over time, the concept of AI has progressed from the Turing test in the 1950s, through to today’s LLMs. Many modern AI tools, for example Google DeepMind’s AlphaGo Zero, and AlphaFold, use artificial neural networks that were designed to mimic the living brain. Recent discussions have gone as far as asking whether AI has gained consciousness. Boguszewski remarked that there are two competing schools of thought on the modern definition of consciousness. The first (‘global workspace theories’) defines mental states as being conscious when they are broadcast within a global workspace in which frontoparietal networks play a central hub-like role; using this definition, a machine could be built that is said to have consciousness. The second (‘integrated information theory’) states that consciousness is identical to the cause-effect structure of a physical system that specifies a maximum of irreducible integrated information; using this latter definition, a machine cannot be conscious. Boguszewski concluded his talk by drawing the audience’s attention to the impressive advancements that are taking place in both neuroscience and AI today, which may further improve understanding of our own intelligence.

Many modern AI tools use artificial neural networks that were designed to mimic the living brain.

Ethical considerations in AI-supported medical writing

KEY TAKEAWAY

• AI users should be aware of the ethical challenges and limitations associated with data-driven technologies.

In his talk, Mike Katell (The Alan Turing Institute) discussed ethics in the context of AI. Katell defined AI ethics as the set of tools for guiding responsible choices for the design, development, and deployment of digital technology. The SAFE-D (Sustainability, Accountability, Fairness, Explanaibility, Data stewardship) principles, for example, serve as a starting point to reflect upon the possible harms and benefits associated with data-driven technologies.

Katell highlighted several key challenges when considering ethics in AI:

1. AI is not a single technology, but rather is an evolving concept that comprises multiple different tools
2. Contemporary AI was developed originally for marketing purposes rather than for more demanding and strict fields such as medicine
3. There are multiple decisions involved in the design, development, and use of generative AI systems that shape the outputs of these systems
4. While some generative AI tools are highly supervised, other tools such as ChatGPT and Google’s Gemini are largely automated without human intervention, which makes it difficult to monitor how outputs are generated from a given input
5. Generative AI tools are trained to produce plausible outputs rather than facts, and in this way can be thought of as highly complex ‘autocompletes’. ChatGPT, for example, is unable to solve some simple mathematical problems, and though Google’s Med-PaLM can provide accurate information in response to a query, this information is often incomplete

Key questions include who should be accountable if a system causes harm, and who should take responsibility for actions that cannot be explained – the AI company, the user, or the decision maker? Katell emphasised the need for caution around the claims of cost savings and enhanced capabilities made by AI tools in the long term, and highlighted some of the larger issues of AI at play, such as labour issues, the environmental costs of AI, and concentration of power by a small number of companies.

Users of AI technologies should be aware of the downsides of such tools from an ethical standpoint, in addition to the benefits that they bring. 

Ethical challenges and considerations to implementing AI in healthcare; a Research Ethics Committee perspective

KEY TAKEAWAY

• There are additional ethical challenges that need to be addressed in clinical research studies that use AI technologies, including those surrounding data sharing, data bias, autonomy, and transparency.

Alison Rapley (Freelance Medical Writing Consultant) gave an overview of the ethical concerns that need to be considered in clinical research studies that utilise AI, such as studies that involve patient monitoring, or prediction or diagnosis of illness through digital health applications and platforms. Rapley identified the following potential issues:

1. Sharing of patient data: considerations such as what level of data is necessary for the AI model being used or built, and how such data are stored and transmitted are important to address, but most importantly, it is critical to ensure that how data will be shared has been made clear to the participants of the study in order to retain patient trust
2. Fairness, inclusiveness, and equity: data and AI algorithms should not be biased – many AI models are trained using biobank data, which are inherently biased towards particular patient groups
3. Autonomy: human autonomy should supersede machine autonomy, and AI technologies should be used as tools rather than relied upon without human intervention
4. Transparency, explainability, and intelligibility: the purpose and use of AI needs to be made clear to the study organisations and participants; AI technologies should be explainable to different audiences, eg, patients, developers, and regulators
5. Risk/benefit ratio: safeguards should be put in place, especially when it comes to sensitive patient data, and just because you can use AI, it doesn’t mean you should

Just because you can use AI, it doesn’t mean you should.

Artificial intelligence: pharma view

KEY TAKEAWAY

• The benefits and risks of AI should be balanced to put patients first; it is more important than ever that patients have access to trustworthy information and data.

Uma Swaminathan (GSK) and Art Gertel (MedSciCom) co-presented a talk on AI from the perspectives of pharmaceutical companies and the general public. Swaminathan highlighted that patients, ethics, and trust should be at the centre of the pharmaceutical industry. AI can bring important benefits for patients, such as accelerated innovation and greater efficiency, and therefore faster approval of new treatments. However, these benefits need to be balanced with the risks, which include questions of accountability and explainability, and data privacy and data/algorithm bias.

AI should be human-centric, with human accountability. Company policies should be updated to ensure that they are fit for purpose for AI, and there should be proactive risk management and robust governance in place. Decisions should be made collectively and collaboratively, rather than by individuals, to ensure ethical practice.

Gertel emphasised the importance of patient trust in the context of AI. Healthcare decisions are no longer being made solely by the physician: many patients are now taking on the role of partners in their care decisions, by consulting technologies such as Google, and now AI, for information. It is more important than ever that patients have access to trustworthy material supporting the principles that healthcare is safe, effective, patient-centred, timely, efficient, and equitable.

It is more important than ever that patients have access to trustworthy material supporting the principles that healthcare is safe, effective, patient-centred, timely, efficient, and equitable.

 Patient perspective on generative AI

KEY TAKEAWAY

• There are opportunities for generative AI tools to assist with each stage of the patient journey.

Mitchell Silva’s (Esperity & Patient Centrics) talk focussed on AI from the patient’s point of view. Silva noted that there are opportunities for generative AI tools to assist with patients’ needs at all stages of their journey, from earlier detection of symptoms and accelerated diagnosis, to better disease understanding and optimised treatment decisions. For example, patients can upload their medical files to ChatGPT to obtain lay information, and deep fake avatars can be used by time-poor physicians to educate patients and answer their questions in the patient’s language.

Silva urged caution regarding some of the potential negative effects of generative AI tools, namely data privacy, accuracy and reliability, and regulatory compliance (for example with General Data Protection Regulation).

Generative AI can assist patients with better understanding of their disease.

 Why not read our summary of the afternoon session of the symposium.

——————————————————–

Written as part of a Media Partnership between EMWA and The Publication Plan, by Aspire Scientific, a proudly independent medical writing and communications agency that believes in putting people first.

——————————————————–

KEY TAKEAWAYS

• Pharmaceutical companies that sponsor clinical trials are required to balance the need for transparency with protecting confidential information. This is achieved by redacting confidential details from documents before they are made public.
• The redaction process is often hindered by inefficient systems, which can lead to delays. Implementation of a range of practical tips can streamline and simplify the process.

Health authorities, including the European Medicines Agency and Health Canada, require detailed information to be made public as part of clinical trial registration and reporting. This is critical for transparency, but can present practical challenges for pharmaceutical companies as they seek to meet these criteria while protecting intellectual property. Redacting confidential information from documents before public release is an accepted solution, but it can be an onerous and inefficient process. In a recent article for the European Medical Writers Association’s journal, Medical Writing, Elliot Zimmerman outlined the challenges involved, along with a series of practical solutions.

Pitfalls and challenges

As detailed by Zimmerman, the redaction processes followed by pharmaceutical companies can be fraught with inefficiencies, including:

• Decisions on which details are confidential are often made across multiple teams.
• Changes to the classification of confidential information across the research cycle, as information is released into the public domain, make it difficult for teams to keep up to date with changes to confidential status.

These factors contribute to high rates of redaction rejection by health authorities: 60% of sponsor-requested redactions are rejected by the EMA and 65% by Health Canada.

These factors contribute to high rates of redaction rejection by health authorities: 60% of sponsor-requested redactions are rejected by the EMA and 65% by Health Canada.

Practical solutions

Zimmerman proposed a number of practical solutions to ease the burden of redaction:

• ensure processes, roles, and responsibilities are clearly defined, including between teams
• implement a centralised ‘library’ to track confidential information
• ensure a common understanding of what constitutes confidential information across teams
• only include necessary information in documents from the outset and cross-reference within a document, rather than duplicating information
• use integrated management and tracking systems.

Through implementation of these approaches, sponsors could streamline and simplify the redaction process and, thus, reduce delays in clinical trial registration and reporting.

————————————————

KEY TAKEAWAYS

• Readers cannot readily distinguish between healthcare text written by a human or by AI.
• As the use of generative AI in medical writing increases, clear guidance and regulations are needed.

The use of generative artificial intelligence (AI) continues to grow apace, and recent research suggests that the fields of medical writing and medical publishing are no exception. While these  tools can be a powerful means of increasing efficiencies, concerns remain about their use in a sphere that requires “accurate and complete information”. AI-generated content can contain ‘hallucinations’ and inaccurate or biased details, which can be difficult for even experts to identify. This highlights the importance, for professionals and patients alike, of being able to distinguish material created by AI from that developed by humans. Unfortunately, this task can be harder than you think.

Even healthcare professionals struggle to identify AI-generated content

In a recent edition of Medical Writing from the European Medical Writing Association, Natalie Bourré explained how her research sheds light on this complex issue. Bourré asked survey respondents to tease apart healthcare-related content written by humans from that created by AI. Bourré found that:

• overall, respondents were correct only 54% of the time
• surprisingly, healthcare professionals did not fare better than others in differentiating AI- versus human-created content
• medical writers were slightly better at spotting AI-generated text than other respondents.

Other groups have reported similar results and found that AI can create “compelling disinformation”. These issues are even more concerning when considered alongside earlier research by Bourré, which found that readers can be overconfident in their ability to spot AI-generated content.

What’s next? Shaping AI usage in medical writing

Bourré calls for clear guidance and regulations on the use of AI in medical writing. These safeguards would help the medical publishing community to harness the efficiencies and improved accessibility that AI may bring, while maintaining the accuracy and credibility of medical research.

[Clear guidance and regulations on the use of AI] would help the medical publishing community to harness the efficiencies and improved accessibility that AI may bring, while maintaining the accuracy and credibility of medical research.

————————————————

The 56^th European Medical Writers Association (EMWA) Hybrid Conference Day took place on 10 November, with delegates able to attend sessions on Writing in Plain Language for Publications, and Artificial Intelligence (AI) in Medical Writing. The seminars covered these topics through presentations and panel discussions, with the aim of providing practical recommendations and advice to medical writers and communicators. A summary of the second session on AI in Medical Writing is provided below to benefit those who were unable to attend, and as a timely reminder of the key topics for those who did.

Our summary of the session on Writing in Plain Language for Publications can be found here.

KEY TAKEAWAY

• The advent of AI in medical communications will likely lead to the repurposing of medical writers; however, it should never replace them.

An introduction to AI – AI 101

Martin Delahunty (Inspiring STEM Consulting) and Sepanta Fazaeli (Stryker) co-presented the first presentation of the afternoon covering some basic concepts in AI and their relevance in the medical writing landscape.

Large language models (LLMs) have become synonymous with AI, with OpenAI’s ChatGPT perhaps the most famous example of an LLM. In response to a prompt, LLMs produce human-like text by recognising patterns between words and predicting what comes next. However, AI models do not ‘understand’ human language; they work by using a mathematical process to convert subcategories of a sentence (tokens) into a matrix of numbers. The more massive the matrix (ie, the larger the dataset on which the AI model is based), the more understanding and context it contains. Once converted, AI models run arithmetic functions on these numbers to process and produce an output.

Medical writing involves different types of tasks, eg, ranking and classification of data, information appraisal, information retrieval and extraction, and report generation. Some AI models are better suited to certain tasks than others. For example, AI models based on Bidirectional Encoder Representations from Transformers, can be better suited to extraction tasks compared with ChatGPT, which is a unidirectional generative AI tool. Nevertheless, LLMs can be adapted for specific tasks. In the example above, ChatGPT would need to be fine-tuned before being used for extraction tasks to avoid creating nonsensical or inaccurate outputs (known as hallucinations).

Some AI models are better suited to certain tasks than others. It is important to use the correct AI model for the task.

Delahunty and Fazaeli outlined some potential advantages of using AI tools in the medical writing landscape:

• AI tools can streamline processes, increasing efficiency.
• Using AI can increase the speed at which projects can be completed.
• AI allows the user to process vast data sets.
• Using AI can improve accuracy and consistency.

On the other hand, there are situations where using AI tools is either not appropriate or their use should be considered carefully. The International Committee of Medical Journal Editors (ICMJE) worked quickly to update their guidance on the role of authors and contributors with the advent of AI, emphasising the need for transparency and appropriate disclosure where AI tools are used in the production of biomedical publications. Yet there are still important and relevant issues to be addressed in the context of the biomedical publishing industry:

• Appropriate attribution of authorship/contributorship status needs to be considered when using AI.
• Using AI can have ethical implications.
• AI tools can produce incorrect or misleading outputs.
• There are limitations of current AI technologies.
• There are privacy considerations regarding the datasets used to train AI.
• Ownership of the outputs produced by AI can be a complex issue.
• Non-unique or duplicate outputs.
• Biased or offensive outputs.
• Detection of instances where AI use has not been disclosed.
• Amplification of AI papermills and fake image generation.

ICMJE have emphasised the need for transparency and appropriate disclosure where AI tools are used in the production of biomedical publications.

Harnessing the power of AI: an exploration of AI tools for the next generation of research

KEY TAKEAWAYS

• An understanding of the strengths and limitations of AI tools is fundamental to harnessing their potential whilst avoiding potential pitfalls.
• Medical writers need to be aware of the risks associated with AI to recognise where it can and cannot be used.

Avi Staiman’s (Academic Language Experts and SciWriter.AI) talk focussed on the potential uses and limitations of AI in medical communications. Current uses of AI include translating and editing of text, idea generation or brainstorming, summarisation, peer review, identifying potential funders, and data processing. Among these, AI-assisted summarisation can be particularly useful to medical writers due to the recent rapid growth of research literature. Staiman cautioned however that the body of research literature may be weakened if all research is written based on AI-generated summaries.

The body of research literature may be weakened if all research is written based on AI-generated summaries.

Staiman next addressed several misconceptions regarding the use of LLMs in the context of medical writing:

Staiman discussed how to get the best out of AI tools by providing effective prompts. The output quality from an AI tool is determined by the input quality. Iteration is key to getting the best results, said Staiman and he outlined six rules to follow for prompting AI:

1. Tell it your role.
2. Tell it your goal.
3. Break the prompt down into step-by-step instructions.
4. Few-shot prompting, ie, provide the AI with examples of good practice.
5. Personalisation, eg, which journal you are submitting to and the appropriate tone.
6. Tell the AI the constraints, eg, only include citations from the last 5 years.

The output quality from an AI tool is determined by the input quality.

Staiman had some cautionary advice regarding the use of generic AI tools. Tools including ChatGPT, Midjourney, Gamma, and Jasper are appealing due to their power, ease and (free) cost of access, and readily available training materials. However, these generic AI tools are also associated with hallucinations, misinformation, problematic citations, research integrity issues and authorship concerns. There is also a potential confidentiality issue with using generic AI. For example, in the free ChatGPT model, inputted information is collected and used to train the AI. Staiman advised not to put propriety information into this version. In the paid version, a private server is used and the input is not retained for training.

In closing, Staiman thought that funders must consider the implications of AI in undermining peer review, research integrity, confidentiality and bias. However, they must also recognise the opportunities afforded such as sorting through applications, supporting open infrastructure and levelling the playing field for researchers whose first language is not English.

The best-case scenario is the science writing community embrace responsible AI tools and create forward thinking policies on best practices for AI.

The more things change, the more things stay the same – the history of AI

KEY TAKEAWAY

• Interest in AI technologies – and associated fears – may be at a peak. As understanding of AI increases and its issues addressed, we can expect gradual mainstream adoption of practical applications.

Beginning by explaining that fear of new technologies is not a new concept, Phill Jones (MoreBrains Cooperative) provided a brief history of AI and outlined the current state of the art. Jones described how, at present, we have reactive machines that process complex information and provide a response. We have also started to use limited memory applications in self-driving cars for example. However, emotional intelligence and theory of mind remain purely theoretical.

Emotional intelligence and theory of mind remain purely theoretical.

Outlining the current development direction for AI, Jones referred to three important areas of research for LLMs and GPT:

• Machine learning – The science of how computers solve problems with no or minimal human guidance.
• Computer recognition – How computers recognise and characterise objects.
• Cognitive computing – How computers make decisions.

Jones went on to discuss the concept of natural language processing (NLP), describing human language as unstructured data whose understanding depends on context. NLP is the process of converting between unstructured human language and structured data that can be understood by a computer. Jones covered several approaches to NLP including:

• Simple language parsing: A sentence is broken down and classified into types of words, eg, nouns and adjectives.
• Semantic processing: The functions of words in a sentences are encoded.
• Named entity recognition: Recognition of proper nouns.
• Sentiment analysis: Recognition of emotional context.
• Dependency passing: Identification of word hierarchies in a sentence and the understanding of how important one word is to another.

Jones continued by describing how machine learning approaches are being developed to allow computers to account for missing information. LLMs are sometimes referred to as a stochastic parrot, referring to the ability of LLMs to generate convincing text without understanding its meaning. Neural networks learn by being given problems with known desired answers. Models are then modified iteratively until they can generate satisfactory outputs for each problem. In this way, supervised fine-tuning and reinforcing learning can be used to tailor LLMs to particular use cases.

Machine learning approaches are being developed to allow computers to make a best guess based on incomplete information.

The ability of generative AI to write sensical but useless text may contribute to the amplification of papermills. Papermills target journals with weak review processes and LLMs will likely make it harder to detect papermill articles. Accordingly, Jones thought it important for journals to have better editorial oversight and more quality control processes. However, Jones also explained how AI can be used positively to tackle misinformation, citing the example of ChatBawa, an AI-powered chat bot that can monitor misinformation and check the veracity of public and social media claims.

Jones ended by showing the Gartner Hype Cycle for 2023 – a graphical depiction of the maturity of emerging technologies. According to the chart, AI has gone through the early phase of innovation and has reached the so called ‘peak of inflated expectations’. A natural waning of the hype surrounding AI might be expected in the short-term before the technology is more widely understood and adopted and we see a rise in people’s faith in AI.

Medical writers as guardians and leaders of medical communication

KEY TAKEAWAY

• Medical writers have a responsibility to define and implement ethical and accountable approaches to the use of AI in scientific communications.

In her presentation on ethical use of AI in scientific communications, Uma Swaminathan (GlaxoSmithKline) focussed on the key themes of ethics, accountability, and trust. Swaminathan detailed how scientific communication comprises a wide spectrum of activities, all of which require an ethical, accurate, and factual approach. She emphasised the importance of proof of ethics at every level and thought that, when used in the correct way, AI can support all these activities.

Ethicists are not here to stop the advancement of AI, but to look at how AI can be liberated to catalyse innovation and support scientific communication.

Swaminathan explained further that the data we work with is provided with trust by research participants for a broader good, and that employing AI in scientific communication raises legitimate concerns about how this data might be used. In this respect, scientific communicators are uniquely placed to act as guardians of an ethical approach to using AI.

Ultimately, Swaminathan thought that the responsibility to incorporate an ethical approach cannot rest with AI; human control and accountability are key. She proposed the following approaches to integrating AI into scientific communications ethically:

• Building capabilities to ensure AI is used correctly, taking a 70/20/10 learning approach: 70% learning on the job, 20% learning from the peer network and 10% formal learning.
• Adapting policies, systems, processes, and controls to integrate AI.
• Proactive risk management and robust governance.
• Defining and sharing best practices.

There is a lot of discussion, energy and focus around ensuring the ethical, appropriate use of AI. Ultimately, it is about maintaining trust, human accountability and explainability.

Is the hype real? Real-life user experience of an AI tool for clinical study report production

KEY TAKEAWAY

• TriloDocs is an AI-enhanced medical writing tool that can quickly collate a draft clinical study report from appropriate source documents.

In the next presentation, Lisa Chamberlain James (Trilogy Writing & Consulting) provided a medical writer’s perspective on using AI to produce a clinical study report (CSR). Chamberlain James introduced TriloDocs, an AI tool that can consolidate existing text and draft de novo content to generate an initial draft CSR.

TriloDocs uses source documents to populate the CSR sections in a TransCelerate template. The tool can write results text and populate tables for key study endpoints as well as write text to highlight clinically relevant data. The tool does not use machine learning or generative AI. Importantly, this means the tool cannot hallucinate – anything not included in the source documents needs to be added manually.

Chamberlain James highlighted key advantages and disadvantages of using an AI-enhanced medical writing tool such as TriloDocs:

 Chamberlain James noted AI tools are dramatically changing how medical writers work; they allow writers to focus on the clinically meaningful messages and have discussions with clinical teams earlier. Furthermore, because AI cannot contextualise data in terms of what it means for patients, nor interpret new data, medical writers are still an essential part of the writing process for CSRs.

Because AI cannot contextualise data in terms of what it means for patients or interpret new data, medical writers are still an essential part of the writing process for CSRs.

Reimagine the future of medical and regulatory writing with generative AI and LLMs

KEY TAKEAWAY

• Copilot is a generative AI tool that can be adapted to support medical writers with a variety of repetitive tasks in the production of regulatory documents such as CSRs, patient narratives, and clinical trial summaries.

Enrica Cavedo (Yseop) and Dominique Mariko (Yseop) presented the final talk focussing on how generative AI could support scientific communications and showcased Yseop’s work in this area. Setting the scene, they noted that global demand for medical writers has increased in recent years due to unprecedented levels of drug development. This demand is only set to increase and will significantly outpace the availability of trained professionals in the field. Generative AI could increase the efficiency of medical writers producing regulatory documents, supporting writers with repetitive tasks, and allowing them to focus on using their expertise elsewhere in the process.

Generative AI could increase the efficiency of medical writers producing regulatory documents, supporting writers with repetitive tasks, and allowing them to focus on using their expertise elsewhere in the process.

Being able to navigate the different AI technologies suited for different purposes can help reduce writing time. Cavedo and Mariko described Yseop’s Copilot tool (which is built on pre-trained LLMs) that can generate reports for regulatory documents automatically in a few seconds. They explained how Yseop first evaluated which parts of clinical content generation could be automated, starting with CSR and patient narrative generation, with plans to move towards AI-assisted generation of clinical summaries and clinical overviews.

They outlined some of the ways the technology they have developed meets the needs of regulated industries including:

• Data security and confidentiality being embedded in the model. Unless specifically agreed, a user’s data is not kept for training and an in-house model is used so data does not travel all over the world.
• Self-quality metrics to provide human evaluation and ensure text is factual, accurate and informative.
• Traceability to remain compliant by ensuring everything needed to prove auditability is available to the end user.
• Control – configurations and narrative options allow users to get exact results.

Cavedo and Mariko concluded with Yseop’s stated aim to gain 80% productivity across the whole workflow, and by doing so, impact the productivity of the whole clinical content generation chain from data collection to submission.

Panel discussion

The seminar closed with a panel session chaired by Martin Delahunty, bringing together the speakers from the afternoon session. Key discussion points included:

• The balance between human input and automation: panellists concluded that the level of acceptable AI use depends on the level of risk of generating false outputs. Concerns were raised relating to the use of generative AI in regulatory documents.
• The lack of transparency surrounding AI tools: there is a general lack of understanding regarding why AI tools work. This fundamental opacity makes transparency difficult which in turn makes it difficult to trust the validity of outputs. Transparency will be dictated by the data used to train the models. The quality, traceability and transparency of the data will be reflected in the quality of the output. Medical writers need to be aware of the risks associated with AI to understand where it can and cannot be used.
• The use of AI tools when working with patient data: when using AI with sensitive data, the approach should depend on the robustness of the controls which need to be built into the tools. Some institutions use private servers only accessible to certain individuals. When dealing with patient data these tools should be General Data Protection Regulation-compliant.
• AI in informed consent forms: there is an obligation to inform research participants about the future use of their data, including that AI will be relied upon more and more. However, it is not possible to inform participants of all possible future uses of AI because this is unknown. Ultimately, it is important that the use of AI happens with governance and controls.
• Current use cases: there are several ways we can use AI already, including risk assessments, monitoring real world evidence in real time, adverse event reporting and complaint handling.
• Advice for new medical writers: the advice remains the same as always; gain experience, find a company that offers training and learn to write.
• Sustainability of AI tools: AI tools consume a lot of energy and this is increasing as they become more powerful. The benefits conferred needs to be assessed in relation to the power consumption. The panellists agreed that to remain sustainable, AI tools should only be used when they are effective and the right size tool should be selected.
• Spreading of misinformation due to hallucinations: the panel agreed that humans should always be involved in the final decision making and approval processes; this should not be left to AI. Regulatory authorities have a responsibility to prevent the spread of misinformation.

Why not also read our summary of the morning session on Writing in Plain Language for Publications.

——————————————————–

Written as part of a Media Partnership between EMWA and The Publication Plan, by Aspire Scientific, a proudly independent medical writing and communications agency that believes in putting people first.

——————————————————–

The 56^th European Medical Writers Association (EMWA) Hybrid Conference Day took place on 10 November, with delegates able to attend sessions on Writing in Plain Language for Publications, and Artificial Intelligence (AI) in Medical Writing. The seminars covered these topics through presentations and panel discussions, with the aim of providing practical recommendations and advice to medical writers and communicators. A summary of the first session on Writing In Plain Language for Publications is provided below to benefit those who were unable to attend, and as a timely reminder of the key topics for those who did.

Our summary of the session on AI in Medical Writing can be found here.

Writing in plain language for publications

KEY TAKEAWAYS

• Clear and accurate science writing improves communication with patients, fights ‘fake news’, and helps healthcare professionals (HCPs) gain insights into other fields.

Introducing the topic of plain language summaries (PLS), Andrea Bucceri (Lumanity) stressed the importance of communicating science clearly and accurately, highlighting the ability of PLS to help communicate with patients, fight misinformation (or ‘fake news’), and allow HCPs to gain valuable insights into other fields. Noting a degree of uncertainty regarding PLS among industry publications professionals as recently as 2014, Bucceri emphasised how PLS have now become an important part of medical publications, with most journals allowing PLS within articles, some making them compulsory, and others allowing standalone PLS.

The importance of PLS – publishers’ perspectives

KEY TAKEAWAYS

• PLS are needed to increase the accessibility of science to everyone.
• Good Publication Practice (GPP) guidelines recommend that every biomedical publication include a PLS.

In this session, representatives from the publishing industry shared their views and insights on PLS. Caroline Halford (Springer Healthcare) began by explaining the distinction between PLS, plain language summaries of publication (PLSP), and regulatory lay summaries. PLS – the focus in this symposium – are approximately 250-word long, jargon-free summaries that are published with the parent publication and target a broad non-specialist audience. On the other hand, PLSP are plain language adaptations of full manuscripts, often targeting a patient audience. Regulatory lay summaries are mandated summaries of clinical study reports focused on a study’s primary and safety endpoints, targeted at the study’s participants.

Halford explained why PLS are becoming so important, stating that with the move towards open access, biomedical publications are being read by increasing numbers of HCPs and members of the public. Both groups need clear and comprehensive information that is not easily misunderstood. PLS also offer a quick and convenient way for doctors to keep up-to-date with their fields.

PLS are needed to increase the accessibility of science to everyone.

Halford recognised that the value of PLS for HCPs and patients is well-established, with 71% of HCPs rating PLS to be “very/extremely useful” and around 1 in 6 HCPs referring patients and families to online PLS. She also referenced surveys conducted among patient organisations and patients and their caregivers demonstrating the importance and value of PLS as a source of health-related information for these groups.

Concluding, Halford highlighted the tangible benefits of PLS. Research shows that articles with PLS are more likely to be downloaded than those without, and that of all article enhancement types, PLS produce the highest comprehension, understanding, enjoyment, and desire for more updates.

Felicity Poole (Taylor & Francis) elaborated further, emphasising that a PLS is a constituent part of the parent article and shares the same Digital Object Identifier (DOI), in contrast to a PLSP, which is a standalone publication. Noting how PLS can be text-based, graphical, or make use of audio/video, Poole shared some best practices for PLS when considering which format(s) to use:

• Always include a text-based PLS with an article, as this will be indexed on PubMed, making it more discoverable.
• Include both text and graphical PLS if possible, since graphical PLS are most popular.
• Include audio/video PLS if the journal allows for this, to increase accessibility for audiences with specific needs.

Poole also drew attention to the wealth of available resources to help publication professionals implement PLS, including those from the European Medicines Agency, Open Pharma, Envision Pharma, and the International Organization for Standardization (ISO), ‘How-to’ Guides from Patient Focused Medicines Developments (PFMD), and the updated 2022 GPP guidelines (GPP 2022).

Referring to GPP 2022, Poole highlighted key points from the recommendations regarding which publications should include a PLS:

• A text-based PLS should be prepared for any clinical trial publication that follows the Consolidated Standards of Reporting Trials (CONSORT) guidance.
• Ideally, a PLS should appear with every biomedical publication, keeping in mind the audiences for the information being presented.

Poole continued by identifying the important factors to consider when choosing a target journal and stressed the importance of checking publisher guidelines since they will all be different and not all publishers offer PLS:

• What is the journal’s target audience?
• When is the point of submission – alongside the original article or post-submission?
• Does the journal index the PLS on PubMed?
• Does the journal provide good guidelines on the format, length, and readability level?
• Is the PLS open access, or hosted behind the paywall for subscription journals?
• What is the peer review format – the same reviewers as the manuscript or experts in plain language?

Available metrics for measuring the value of PLS were also discussed briefly, however, Poole noted that the use of metrics is complicated by difficulties in separating PLS views from article views – a topic revisited in the panel discussion following the presentations.

Poole ended by saying that something is always better than nothing. A text PLS should be included as a minimum, and infographics, graphical abstracts, and audio and video PLS included where possible.

Something is always better than nothing. A text PLS should be included as a minimum, and infographics, graphical abstracts, and audio and video PLS where possible.

 Writing considerations for plain language

KEY TAKEAWAYS

• PLS should be targeted at the right reading age for your audience, using simple language with short sentences.
• Use the active voice and visual formatting elements that support understanding.

Adeline Rosenberg (Oxford PharmaGenesis) discussed the key considerations when writing PLS or other documents requiring the use of plain language. Rosenberg highlighted the importance of targeting the right reading age for the intended audience but acknowledged the difficulty in doing this successfully. The UK National Health Service advises that health information should be pitched at a Year 7–9 (ie, ages 11–13 years) level to ensure that 85–93% of patients and the public can understand it, whilst elsewhere research has shown that medium complexity language, defined as a reading age of 14–17 years, was preferred.

Similarly, Rosenberg thought that readability scoring systems including Flesch Reading Ease Score, Flesch-Kincaid Grade Level, SMOG, Gunning fog, and the Coleman-Liau Index, as well as the basic readability statistics in Microsoft Word were useful tools but should not be used in isolation. Rosenberg advised testing written material on 2–5 people representative of your target audience as a better method of ensuring the language is understandable. Publication professionals can also use previously published PLS and PLSP as a guide, for example by accessing Taylor & Francis’s Plain Language Summary Repository, Future Science Group’s PLSP Repository, or by searching “hasplainlanguagesummary” in PubMed.

Regardless of the level of information provided, it should always be written in plain language that is simple and clear.

Beyond readability, Rosenberg suggested that writing tone and message should be different for different audiences, for example a non-promotional tone for HCPs and a sensitive tone for patients. A focus on the bigger picture for patients and accurate presentation of the material to prevent misinterpretation by the media are also important.

Moving on to practical considerations for producing plain language documents, Rosenberg outlined the following best practices:

• Use simple language and avoid medical jargon.
• Use short sentences.
• Use the active voice.
• Simplify numerical information and consider presenting it graphically to aid understanding.
• Consider how best to present risk – present data as both a fraction and a percentage.
• Consider including a phonetic glossary.

Good visual design elements also aid understanding, including using short, bold headings and using questions as section headings. Images and graphics should support understanding of the text, be standalone, and be user-tested. Basic formatting is also important to aid the reader’s understanding:

• Use large sans serif font (at least 12 pt).
• Use lower case bold for emphasis.
• Left justify text with ragged right.
• Leave plenty of space around text.
• Break information down into discrete sections.
• Use images that directly support the text.
• Use bullet points to list content, and call-out boxes to highlight important content.
• Avoid italics, ALL CAPITALS, underlining (and parentheses).

Rosenberg concluded by signposting some useful resources for writing in plain language, including: MRCT Glossary, EMA Medical Terms Glossary, NCI Cancer Dictionary of Cancer Terms, NHS Resources for Health and Digital Literacy, Health Literacy Online from Health.gov and Patient Information Forum (PIF) ‘How-to’ Guides.

 Translating PLS

KEY TAKEAWAYS

• Translating PLS is important to provide global equity in health awareness.
• Ensuring cultural sensitivity whilst maintaining accurate information is essential.

Ana Sofia Correia (Ana Sofia Correia Medical Translation & Writing) presented the next session covering translation of PLS. Correia began by establishing that most scientific articles are only published in English, despite English speakers comprising less than 20% of the world’s population. In addition to producing PLS in English, we must ensure that the information reaches the non-English-speaking populations that we are targeting to provide equity in health awareness.

Translating PLS is important to provide global equity in health awareness so that everyone is able to make better-informed health decisions.

Summarising the role of a medical translator in PLS, Correia identified the following key principles:

• Adhering to the principles of plain language – translating without introducing complexities.
• Safeguarding cultural sensitivity and contextual relevance – incorporating culturally relevant examples whilst maintaining simplicity.
• Ensuring consistency in simplification throughout the document.
• Collaborating with stakeholders to maintain scientific integrity.
• Navigating ethical considerations to balance dissemination of accurate information with cultural sensitivity.

Correia elaborated on the PLS translation process and stressed the importance of beginning translation only after finalisation of the PLS in the source language to mitigate the time and costs of additional revisions. After translation, the PLS must be checked to ensure that it is true to the original, for example, by back translating. In addition, it should be tested in the relevant target populations to ensure that it is easily understood.

Recognising the challenges involved in accurate and culturally sensitive translation of PLS, Corriea recommended the following strategies to address these issues:

• Graphical elements: decide on target languages and work with translators from the planning stage to anticipate different text-lengths within graphical elements, and ensure that graphical elements are culturally appropriate and understandable in the translated version.
• Linguistic and cultural aspects: use simple, clear language and culturally relevant examples whilst avoiding idioms and oversimplification.
• Identification of target languages: demographic studies, health surveys, and epidemiological data can help identify the languages spoken by the most relevant populations.
• Cost: develop templates and glossaries to expedite the processes and use translation tools and resources in combination with human translators.

Correia ended by highlighting some useful medical translation tools, including: Translation is not Enough: Cultural Adaptation of Health Communication Materials (ECDC), Getting It Right (ATA), Everyday Words for Public Health Communication (CDC), Clear Writing (European Commission), and Readability Formulas: 7 Reasons to Avoid Them and What to Do Instead (UX Matters).

 Panel discussion

The presentations were followed by a lively panel discussion bringing together all the presenters from the morning session plus Hamish McDougall (Sage). Key discussion points included:

• Dissemination of PLS translations: many publishers allow for the submission of translated PLS and may have in-house translation capabilities. However, discoverability of translated PLS is an issue and collaboration with search engines to increase discoverability and use of social media to increase dissemination are key.
• Use of metrics to measure PLS access: since PLS are embedded within full-length articles, it is difficult to discriminate between access to PLS and access to the rest of the article. Adding additional clicks to allow for metrics would make PLS less discoverable. Though we know that patients and doctors want PLS, more research is needed to prove the benefits of PLS. Use of metrics to assess PLS access in different geographical regions would be useful in determining the most appropriate languages for translations.
• Peer review of PLS: concerns were raised around peer reviewers ignoring the PLS during review of the main article. Representatives from the publishing industry clarified that guidance on reviewing PLS is provided for peer reviewers to ensure that this is included with their review. Journals try to include plain language reviewers, including patients from both within and outside of the disease area where possible, or editors from a different therapeutic area. In the case of translated PLS, patient reviewers also review the translated language.
• Acknowledging patient contribution: authors are encouraged to be comprehensive in acknowledging patient contributors and reviewers. Patients should be included as authors if they meet the ICMJE authorship criteria, and otherwise included in the acknowledgements.
• Use of acronyms in PLS: acronyms should not be used as a way of reducing word count. They should be used sparingly and expanded at first mention unless it is a common word eg, DNA, HIV, and URL.

Why not also read our summary of the afternoon session on AI in Medical Writing.

——————————————————–

Written as part of a Media Partnership between EMWA and The Publication Plan, by Aspire Scientific, a proudly independent medical writing and communications agency that believes in putting people first.

——————————————————–

KEY TAKEAWAYS

• Recent study reveals that lay titles of clinical trials often lack essential information, include too many technical terms, and might be misleading to readers.
• Optimising lay titles is important to improve public understanding of clinical trials and increase patient participation.

Lay titles of clinical trials are instrumental in attracting the attention of patients, caretakers, and healthcare professionals, and help determine whether readers will seek further information about a trial. However, crafting a title that is concise, informative, accurate, and understandable to the target audience is not an easy task, as demonstrated in a recent study.

Dr Leonie Leithold and colleagues examined the lay titles of 74 industry-sponsored clinical trials registered on ClinicalTrials.gov in 2021. The titles were evaluated on the following criteria:

• inclusion of 4 recommended elements: target population, condition evaluated, treatments studied, and trial’s aim
• presence of technical terms
• overall adequacy and informativeness.

The results, published in EMWA’s journal, Medical Writing, highlighted several shortcomings of the assessed lay titles:

• 72% did not include all recommended elements, compromising their ability to convey essential trial information.
• 73% contained specialist language, making them less meaningful to potential trial participants.
• 51% were deemed inadequate, indicating a failure to effectively inform readers.

The authors noted that although sentence length is an important consideration in lay writing, placing too much focus on brevity in titles might make them less informative. This was illustrated by the finding that only a tenth of short titles (<100 characters) in the sample included all recommended elements.

51% of the assessed lay titles were deemed inadequate, indicating a failure to effectively inform readers.

The use of trial acronyms can facilitate communication about a trial and make it more memorable to readers. Lay titles that contained a trial acronym tended to be longer, include more recommended elements, use fewer technical expressions, and be more likely to be assessed as adequate compared with their acronym-free counterparts. However, the authors cautioned against the use of acronyms that imply a favourable trial outcome, or evoke positive imagery, to avoid misleading the reader about potential benefits of a treatment or influencing their decision to participate in a trial.

Leithold et al encourage industry sponsors to invest more effort in creating effective lay titles to facilitate public understanding of clinical trials and drive participation. We look forward to seeing how this might contribute to advancing clinical development and, ultimately, patient care.

—————————————————–

The 11th European Medical Writers Association (EMWA) symposium, entitled ‘Clinical trial disclosure and transparency: impact on medical writing and communication (Clinical Trial Regulation EU 536/2014)’ took place on 11 May. The symposium explored this topic through presentations and panel discussions, focusing on insights particularly relevant to medical writers and communications professionals.

A summary 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.

EMA transparency journey and impact on its stakeholders

KEY TAKEAWAYS

• Transparent reporting of clinical trial information is fundamental to a well-informed decision-making process around public health issues.
• Medical writing and communications professionals have a critical role to play in preparing the documentation required for transparent registration and reporting of clinical trials, mandated by EU laws.

Kicking off the day, Slavka Baronikova (EMWA Conference Director; Galapagos) welcomed attendees to this year’s symposium. Kathy Thomas (Medical Writing, Disclosure and Transparency Consultant) then began proceedings, outlining how the last 20 years have seen an increased emphasis on clinical trial disclosure and transparency, supported by initiatives to encourage prospective registration of clinical trials and timely disclosure of results in publicly available databases. Medical publication professionals play a crucial role in preparing various documentation supporting these clinical trial disclosures, which need to fulfil the reporting requirements of relevant laws (ie Clinical Trials Regulation EU 536/2014 [EU CTR] in the EU), while protecting the identities of patients participating in trials and the commercial interests of trial sponsors. As of January 2023, all clinical trials conducted in the EU and in countries belonging to the European Economic Area (EEA) must conform with the EU CTR. So it is important that medical writers of regulatory documents as well as publication professionals understand the requirements and implications for involved stakeholders.

All trials conducted in the EU must conform with the EU CTR, so is important that medical publication professionals understand its requirements and the implications for involved stakeholders.

Morgane De Verdiere (European Medicines Agency [EMA]) described how transparency is a core value of the EMA and is key to building trust with patients and general public, healthcare professionals (HCPs), media, public health bodies, pharmaceutical industry, and academia.

Transparency:

• ensures decisions the regulator makes are well founded and more likely to be supported
• raises awareness of EMA recommendations that impact the public
• allows the EMA to communicate key public health messages and address questions of concern from the public
• counter misinformation about healthcare matters.

Public health emergencies crystalise the need for transparency. An example is the
COVID-19 pandemic, when the EMA had to adapt quickly to keep its audiences informed. Examples of its approach to transparency included holding public hearings on YouTube to explain how an accelerated vaccine approval process was possible, introducing a ‘rolling review’ process for pharmaceutical companies to submit data for drug assessments on an ongoing basis, and providing frequent updates—using accessible language—on the EMA website regarding safety events associated with vaccination. The pandemic demanded increased transparency from the EMA to ensure public trust in the regulators and their decisions.

De Verdiere moved on to discuss the impact of the recently introduced EU CTR (sometimes also referred to as the new Regulation) on the medical writing community, with the implementation of the EU portal and database —  Clinical Trials Information System (CTIS) designed to harmonise the submission, assessment, and supervision of clinical trials established under the EU CTR across the EU and EEA. Combined with supporting legislation (EU CTR), CTIS will facilitate the conduct of multinational trials, enable wider collaboration and access to clinical research data, ensure that the EU remains a desirable location in which to conduct clinical research, and further reinforce the trust that underpins the approval of medicines. De Verdiere explained the evolution of the CTIS transparency rules, which are currently being reviewed to better balance transparency (ie full disclosure of clinical trial information) by allowing pharmaceutical companies some opportunity to defer the disclosure of commercially sensitive information.

De Verdiere concluded that transparency is fundamental to empowering patients and HCPs to make informed decisions about health issues and treatments. Balancing this need for transparency with the need to allow companies to invest in and protect their technologies requires a flexible and adaptable approach from the regulator.

Balancing need for transparency with the need to allow companies to invest in and protect their technologies requires a flexible and adaptable approach.

EFPIA perspective on the EU CTR

KEY TAKEAWAYS

• EU law now requires many clinical trial regulatory documents to be made publicly available.
• The law provides tools for sponsors to protect their intellectual property, but there needs to be balance between transparent reporting and protecting the ability of industry to innovate.

Silvia Garcia (European Federation of Pharmaceutical Industries and Associations [EFPIA]) opened the second session of the morning by reiterating the mission statement of EFPIA to foster a collaborative environment for the development and delivery of new therapies in Europe. Garcia then outlined EFPIA’s principles for responsible clinical trial data sharing. As pointed out, while most EFPIA members have transparency measures and policies that go beyond the commitments of the EFPIA principles, these required updates to reflect modern data sharing initiatives, which include clinical trials registries (eg ClinicalTrials.gov), multi-sponsor data sharing platforms, and open access publication of clinical trial results. EFPIA began a member consultation process in 2020 to assess compliance rates and ways to improve transparency in data sharing, and submitted revised wording for the principles for board endorsement in April 2023. Compliance rates following the consultation process are very close to 100%.

Polyana Bastos (Johnson & Johnson; EFPIA) provided an industry perspective on the implications of the EU CTR, with a particular focus on how and when companies should use the option to defer submission of clinical trial information to CTIS to protect their commercially confidential information (CCI). As of January 2023, the EU CTR requires most clinical trial application documents to be published on the EMA’s CTIS platform. Protected personal data (PPD) can be redacted in the published (freely open to public) documents, while CCI can be protected by deferring publication for several years, with the timescale depending on the category of clinical trial (a longer deferral period is allowed for earlier stage trials). The intention is to allow sponsors to protect their CCI, but to avoid extensive redactions of the trial information when it is ultimately published.

CCI can be protected by deferring publication for several years…The intention is to allow sponsors to protect their CCI, but to avoid extensive redactions of the trial information when it is ultimately published.

In addition, Bastos described how EFPIA has been working with the EMA to develop practical solutions for sponsors to protect their intellectual property. Using the example of a dose finding clinical trial, where disclosure of dosing information could potentially preclude a successful patent application, a potential solution was identified to allow submission of ‘dummy data’ for specific fields in CTIS. Implementing such solutions would require technical modifications to the CTIS system, but would also require a balance to be struck between transparent reporting of clinical trial information and protecting sponsors’ intellectual property and encouraging innovation.

Clinical trial start up: sponsor perspective

KEY TAKEAWAYS

• A focus on transparency and attention to PPD and CCI should be integrated throughout the clinical trial working process.

The morning continued with an overview of the impact of the EU CTR from the sponsor perspective, delivered by Merete Jorgensen (Merete J Consulting). To set the scene, Jorgensen reaffirmed the vision of the EU CTR, which includes:

• making the EU more attractive for clinical trial sponsors
• ensuring consistent rules between EU member countries
• strengthening data transparency.

Jorgensen noted that in contrast to the previous law on disclosure of clinical trial information – the EU Clinical Trials Directive (CTD), clinical trial documents entered as part of a clinical trial application (CTA) and during the lifetime of the clinical trial under the EU CTR must be uploaded to CTIS. All information, except personal data, information on quality aspects of the investigational medicinal product (IMP), and communication between member states, will eventually be made public. This means that everyone working in clinical development and regulatory functions must have the legally-binding disclosure and transparency requirements in mind.

The vision of the EU CTR…includes making the EU more attractive for clinical trial sponsors, ensuring consistent rules between EU member countries, and strengthening data transparency.

Jorgensen stressed the need, right from the beginning of the working process, to pay attention to the following definitions:

To help adhere to requirements for protecting PPD, such as the General Data Protection Regulation (GDPR), Jorgensen explained it is possible to upload 2 versions of each document at the same time, with only the first being made public. Jorgensen provided several suggestions to help medical writers protect PPD and CCI. She emphasised minimisation, ie thinking carefully about whether the information is necessary for regulatory purposes. Remaining information may need to be redacted, or possibly anonymised (although the extent of anonymisation required is a matter of debate). As for CCI, documents dealing with quality of the IMP are not published; therefore, medical writers should ensure to distinguish these from other documents.

Jorgensen’s talk concluded with the impact of the EU CTR on clinical study reports (CSRs). CSRs are the only documents that are not subject to deferral. CSR are due for upload to CTIS 30 days after marketing authorisation is granted, the procedure is finalised, or the marketing authorisation application is withdrawn. The version of the CSR released is that submitted as part of a marketing authorisation application, highlighting the increased transparency required by the new regulations.

Paula Nixon (AstraZeneca) went on to discuss the company’s experience of the transition from the EU CTD to EU CTR. Nixon explained that we are currently in period 2 of the transition, meaning that all initial CTAs must be submitted using the new EU portal and are governed by the new Regulation. In period 3 (effective from 31 January 2025), all clinical trials will be governed by the EU CTR, regardless of whether the initial clinical trial submission was under the EU CTD.

Nixon explained that one of the greatest benefits of the EU CTR is the harmonisation across the EU/EEA, with one system, one process, and one approval. The CTA process, from validation through to a decision, should take 60–106 days, including a period for questions and requests for more information. Medical writers must be mindful that requests for information (RFI) are subject to a response time of 12 calendar days and this may include making a protocol amendment. With this, Nixon encouraged a mindset shift from reliance on protocol amendments towards aspiring to get the CTA right the first time. Ultimately, more time and effort spent on initial preparation lends itself to a smoother CTA review and approval process. Nixon highlighted some additional considerations for the planning phase, such as whether the study design is optimal or could yet be simplified, and whether patient groups have been consulted, looking ahead to potential queries during CTA review.

Nixon also shared key takeaways around CCI:

• CCI is not static and evolves over time. It is challenging, but important, to predict what might be CCI later, at the time the document is released (eg 5 years later, after deferral).
• Only critical CCI can be redacted. The level of redaction should be balanced with the requested deferral duration.

Nixon wrapped up her presentation with several proactive considerations for preparing clinical trial documents. Above all, Nixon recommended flagging potential CCI in a document from the outset, rather than identifying CCI after completion. Not only does this speed up the redaction process, but more critically, highlights the level of CCI in a document, opening up discussions with clinical teams about the best way forward. Medical writers should play a central role in this activity and process.

Nixon recommended flagging potential CCI in a document from the outset…Not only does this speed up the redaction process, but opens up discussions with clinical teams about the best way forward. Medical writers should play the central role in this activity and process.

Clinical trials conduct: industry perspective

KEY TAKEAWAYS

• Medical writers can play a key role in transitioning clinical trials from the EU CTD to the EU CTR.

Next up, Sarah Bly (Worldwide Clinical Trials) covered the impact of the EU CTR from the industry perspective. The main focus of Bly’s talk was considerations for studies currently running under the EU CTD that will have at least one active EU site on 31 January 2025, when the next phase of the transition to the EU CTR begins. As part of the transition, multinational clinical trials will need to be transitioned as a single multi-country CTA, using either:

1. A harmonised protocol: a protocol approved across all member state countries (MSCs) under the EU CTD, where there are no differences between countries.
2. A consolidated protocol: a protocol with differing procedures between MSCs, but with an identical protocol document for all countries.

Medical writers might be brought in to help prepare the required documents for the transition:  a cover letter, an EU CTR application in CTIS, and a harmonised/consolidated protocol for multi-country trials. For clinical trials already approved under the EU CTD, the transition should be seen as an administrative step. This is reflected in the assessment timeline: applications are approved within 60 days unless requests for information are raised. RFIs are typically raised when there is a discrepancy between the submitted documents and those approved under the EU CTD.

Next, Bly discussed a change in terminology with the EU CTR, from ‘amendment’ to ‘modification’. Bly noted that as with amendments, modifications can be substantial or non-substantial, and directed the audience to several examples.

The last portion of Bly’s presentation covered notifications, including those required via CTIS within 15 days from the start of the trial (usually the start of recruitment), the end of recruitment, and the end of the trial (usually the last visit of the last participant in MSCs).
Bly also discussed circumstantial event notifications, with temporary halt and restart, early termination, and unexpected events notifications required within 15 days after the event. Serious breach and urgent safety measure notifications must be made within 7 days.
Bly highlighted that given their knowledge of the study protocol, medical writers may be consulted about whether urgent protocol modifications are needed in light of serious breaches or unexpected events.

Given their knowledge of the study protocol, medical writers may be consulted about whether urgent protocol modifications are needed in light of serious breaches or unexpected events.

Clinical trials conduct: PV perspective – EU CTR and publication of RMPs – impact on DSURs and RMPs

KEY TAKEAWAYS

• Defining processes up-front to identify and redact CCI in risk management plans (RMPs) can help meet tight timelines. Medical writers can play a key role.

After a brief summary of the morning sessions by Lisa Chamberlain James (Trilogy Writing & Consulting Ltd), Sven Schirp (Boehringer Ingelheim) kicked off the first afternoon session.

Schirp discussed the impact of the EU CTR on processes relevant for the creation of development safety update reports (DSURs) and RMPs. The reference safety information (RSI) included in the investigator’s brochure presents a list of preferred terms for serious adverse events that are considered to be expected. The RSI used to identify any serious unexpected serious adverse reactions (SUSARs) for the DSUR must be that in effect at the start of the annual reporting period, defined by the EU CTR as the version most recently approved in at least one member state where the trial is ongoing. Further, when determining expectedness of an event, the date of event onset must be matched with the RSI in place at the time. Schirp also noted that in light of the EU CTR, subject identifiers cannot be included in the DSUR, while methods used to assess signals during the reporting interval must be described.

In the second part of his talk, Schirp discussed how the EMA has begun making RMPs publicly available, with full publication of redacted RMPs containing new active substances—or of particular public interest—now routine. Consequently, the marketing authorisation holder (MAH) needs to be prepared to provide a redacted version of the RMP within a short time period. Schirp signposted the audience to detailed guidance for identifying CCI, noting that medical writers may be brought in to provide guidance on what should be considered CCI (and thus redacted) or to propose specific sections for legal review. Given the limited timeframe for redactions, Schirp recommended defining a clear process to identify and redact PPD and CCI in advance.

Clinical trials results reporting and dissemination

KEY TAKEAWAYS

• Writing in lay language is crucial for effective dissemination of clinical trial information, with training needed to hone this skill.
• Patient involvement in the planning, development, translation, and dissemination of lay summaries is recommended to ensure these meet patient needs.
• Many sources of guidance are availability regarding regulations and best practice for scientific publications.

Chamberlain James opened the session with a discussion of lay summaries of clinical trial results, emphasising that alongside moral obligations to communicate findings in plain language, failed communication can have a significant impact on adherence and even the rate of adverse drug reactions.

Chamberlain James discussed the problems medical writers may face when developing lay summaries of clinical trial results, given no approved template is available. Chamberlain James signposted attendees to Annex V of the EU CTR, outlining 10 topics that must be covered in lay summaries:

1. Clinical trial identification (eg title, protocol number, EU trial number), to help readers find the trial
2. Name and contact details of the sponsor
3. General information about the trial (eg when, where, and why the trial was conducted, and main objectives)
4. Study population (eg location, age, gender, inclusion and exclusion criteria)
5. Investigational medicinal products used
6. Adverse reactions and their frequency
7. Overall results
8. Comments on the outcome
9. Next steps, ie whether follow up trials are foreseen
10. Signposting to where additional information can be found.

Although these requirements are a useful starting point, they are not without problems. For example, trial titles and eligibility criteria can be very long and filled with technical language. While guidelines recommend including ‘key’ criteria, these are not defined and must be judged by writers and clinical teams – but what is ‘key’ for patients? What name do we use for the investigational medicinal product? Often names can be very long, a chemical name, or just a number. In the instance of multiple brand names, should all brand names be listed? Study designs can also be very complex – if the study includes blinding, should the importance of this be explained? Chamberlain James noted there is a fine line between informing and teaching, and to inform, you need to give a certain amount of context and background information. She also discussed requirements for provision of a protocol synopsis in lay language. This raises similar issues to lay summaries of trial results, with additional challenges arising through the ethical considerations section (such as tone, cultural sensitivities, and context), as well as the tight page limit.

Chamberlain James emphasised that writing in lay language is not as simple as changing ‘difficult’ words to simple ones, given the potential for errors in ‘translation’ – providing reassurance that ChatGPT is not going to replace medical writers. Chamberlain James concluded her presentation by underscoring the value of medical writers training and testing their lay summary writing skills, highlighting a well-known Albert Einstein quote.

“If you can’t explain it simply, you don’t understand it well enough.” – Albert Einstein

Next, Jan Geissler (Patvocates) discussed dissemination of trial results from the perspective of patients with severe diseases. To set the scene, Geissler highlighted patients typically have many questions following diagnosis or disease progression, but often do not receive satisfactory answers, given time pressures in the healthcare system and the lack of materials available for HCPs to share. Information from patient organisations can help to fill this gap. Scientific publications can also be a great source, but these need to be findable, accessible, and understandable.

Geissler stressed that a ‘one size fits all’ strategy for communication rarely works, given the heterogeneous ‘patient’ audience (which can encompass carers, patient advocates, patient organisation representatives, and expert patients). For instance, text-based plain language summaries, infographics, videos, and detailed publications all have different levels of accessibility and suit different audiences. Geissler emphasised the importance of patient involvement during planning, development, translation, and dissemination: for example, patient involvement during planning will identify which endpoints are most important, meaning lay summaries are more likely to meet the needs of patients. Patient organisations can help identify a representative group of patients to contribute.

On a final note, Geissler spoke of the frustrations experienced by patients when clinical trial results are published in peer reviewed journals a long time before the lay summary becomes available (mandated by 12 months after the protocol-defined end of the trial for many studies). Patients need timely access to the latest data to make informed decisions – a year can be a very long time for a patient. Time is of the essence, and it could be lifesaving.

Patients need timely access to the latest data to make informed decisions –a year can be a very long time for a patient.

In the final talk of this session, John Gonzalez (Solanum Medical Communications) explored the impact of the EU CTR on peer reviewed publications. Gonzalez signposted to the many different sources of scientific publications guidance available, including regulations, Good Publication Practice (GPP), International Committee of Medical Journal Editors (ICMJE) recommendations and reporting guidelines compiled by the EQUATOR Network. Although the guidance put forward by these sources varies, Gonalez noted:

• Often scientific journals require pre-registration of clinical trials as a condition of consideration for publication.
• Guidance requires the timely dissemination of clinical trial results – and more recently, encourages publication of Phase I trial findings.

Gonzalez also discussed documentation required by journals as well as EU CTR timing requirements for posting data and documents, which depend on the type of study. He noted consistency between trial protocol and results postings on public registries and information reported in peer reviewed publications is key (especially around the primary endpoint, for example). Any discrepancies, including expected information not being reported, are likely to be scrutinised. In closing, Gonzalez noted the EU CTR will increase transparency, and the more data published, the better for patients.

Consistency between trial postings on public registries and information reported in peer reviewed publications is key…Any discrepancies, including expected information not reported, are likely to be scrutinised.

Panel session: ask the experts

The day’s proceedings concluded with a panel discussion and Q&A session, with Morgane De Verdiere, Polyana Bastos, John Gonzalez, Aman Khera (Worldwide Clinical Trials), Merete Jorgensen, Paula Nixon, Silvia Garcia, Pooja Phogat, (Krystellis), Art Gertel (MedSciCom), and Priti Nagda (Taylor & Francis) the panellists. Key points included are summarised below.

The importance of involving patients throughout the clinical trial process

The panellists agreed that engagement and consultation with patients/patient advisory boards was key for determining the direction of clinical trial research and making patients aware of clinical trials they can participate in. The COVID-19 pandemic demonstrated that virtual approaches and decentralisation could enhance participation in clinical trials: there are no new drugs without clinical trials, and no clinical trials without patients. Patients should also be involved in publishing the results, including having patients as authors where appropriate. A case was made for mandatory open access publication of a plain language summary for every clinical trial conducted. From a regulator’s perspective, patient consultation during the review of lay summaries and safety communications is essential, and new law is being discussed to elevate the role of the patient in regulatory activities.

From a regulator’s perspective, patient consultation during the review of lay summaries and safety communications is essential.

The importance of balancing transparency and the ability of the pharmaceutical industry to innovate through protection of intellectual property

The panel acknowledged the need for transparent reporting of clinical trial information to maintain public trust. Extensive redaction of published information—even for legitimate reasons—can undermine this trust and damage the credibility of the industry. On the other hand, disclosure of confidential information could lead to discontinuation of a product’s development, which is ultimately bad for patients. Guidance from the EMA on how and when to redact or defer the disclosure of CCI was welcomed by industry, but it was important to understand that the guidance was not exhaustive, and the context of each case mattered.

Effective and transparent two-way communication is key to tackling misinformation

The panel thought that it was important to listen carefully to social media platforms, including to identify sources of misinformation more quickly and respond with robust evidence. The COVID-19 pandemic highlighted peer review as a hallmark of reliable information, so publishing peer reviewed research to counter misinformation is a vital strategy.

Implications of the EU CTR for multinational trials including countries outside the EU

The panel discussed the practicalities of EU CTR requirements for submitting study protocols that may contain country specific requirements, for example. Acknowledging it was still a learning process regarding best practice, and whether it was necessary to submit country specific protocols for countries outside the EU, one approach could be including all country specific requirements in an appendix to the submission.

Considering the broader impact of clinical trials on the environment

While the latest EU CTR law will increase the focus on environmental risk assessment, this is not within the remit of the information included in RMPs. However, the panel noted that environmental impact assessment has become part of the wider conversation surrounding clinical trial conduct, with various guidelines being developed or in place in different territories, such as the National Institute for Health and Care Research (NIHR) carbon reduction guidelines.

——————————————————–

Written as part of a Media Partnership between EMWA and The Publication Plan, by Aspire Scientific, a proudly independent medical writing and communications agency led by experienced and dedicated industry experts.

——————————————————–