He developed a paper on clinical development to proactively address industry issues earlier this month and OSP put some questions to him in a bid to find out more.
In the context of accelerating study startup in clinical research and drug development, the integration of Federated Learning (FL) is highlighted as a pivotal advancement. Could you elaborate?
FL from de-identified and relevant data points from clinical research sites/hospital networks would pave the way for accelerated study startup. Currently, due to patient privacy concerns, many large healthcare datasets that could train AI algorithms remain locked away in silos. As data sharing regulations tighten, FL becomes crucial for AI-driven drug development and healthcare, as FL presents a promising future, where collaboration and model training can occur without compromising privacy.
Mainly because FL addresses disparities and benefits to under-served populations by enabling ML models to learn from diverse data without centrally aggregating patient information. This decentralized method avoids privacy concerns associated with traditional ML, allowing models to train on datasets from various sources without transferring patient data. The ability of FL to leverage a broader range of insights from data, including from any previously inaccessible datasets marks a substantial advancement in optimizing the learnings that could be deployed for clinical research and clinical care.
What is the significance of harmonized standards in clinical trials and in what ways do these standards streamline processes in clinical trial designs?
Harmonized standards in clinical trials aim to ensure consistency, quality, and interoperability in conducting trials. Some of the ways harmonized standards help in improving clinical trial designs include streamlining various processes in clinical trial designs, and reducing the time and resources required for protocol development, review, and approval.
Promoting the use of common data elements and formats to enable interoperability, making it easier to aggregate and compare data across different trials and research initiatives.
Incorporating guidelines for the integration of new technologies (e.g., digital health tools, remote monitoring) into trial designs, promoting their use in a standardized and compliant manner. It also, supports the implementation of adaptive trial designs, as it allows aspects of trials to be modified based on interim data analyses while maintaining the study’s integrity.
Slow adoption of technology by Contract Research Organizations (CROs) is adversely impacting the efficiency and quality of clinical trials. Can you please elaborate on the role of digitally enabled players in solving these issues with some examples?
Traditional CRO operating models could be disrupted by decoupling specific functions from the current state in a novel plug-and-play model that would alter the role play of incumbents. CROs continue to be plagued by various challenges to effectively execute clinical trials mainly arising from their traditional approaches and the slow and siloed adoption of technology.
Thus, partnerships with digitally enabled players with healthcare domain expertise can significantly improve various functions which will enhance the efficiency of clinical trials and the quality of study outcomes. Two key areas that can benefit from these initiatives patient recruitment and retention: Geo-fenced, hyper-local omnichannel campaigns for trial awareness with nuanced content that encourages diversity and inclusivity, workflow-enabled digital marketing for patient recruitment, and patient engagement apps along with empathetic human interactions ensure wider outreach and better patient engagement.
The other area is the automation of the data and content supply chains from Protocol to CSR: digitization of the Protocol, leading to the automated creation of a reusable library of annotated eCRFs with an ability to deploy to any EDC, automated creation of standardized data assets conforming to industry standards, and automated creation of statistical outputs (TLFs). Content automation using GenAI would help enhance the efficiency and effectiveness of regulatory writing.
How do you envision the integration of GenAI in pharmaceutical companies, particularly in accelerating clinical trials? What measurable improvements can sponsors expect to achieve by adopting GenAI in the drug development process?
GenAI would help pave the way for a new set of domain-rich resource profiles that will lead to a ‘self-serve’ model when it comes to simplified usage of technology. GenAI holds immense potential for pharmaceutical companies, accurately predicting drug success in clinical trials and offering benefits such as increased speed, efficiency, and cost reduction in drug development. By automating labor-intensive tasks, GenAI accelerates early-stage drug breakthroughs, optimizes drug candidates, and expedites research. A recent study conducted by BCG highlights two primary use cases of GenAI in clinical development.
These include automating medical document generation in clinical development e.g. protocols, clinical-study reports, and regulatory affairs submissions), which can reduce medical-writing time by as much as 30%.
