AI in Patient and Professional Engagement

Prediction for 2024: AI for patient and healthcare professional engagement is booming but expect stricter rules in 2024.

2024 Prediction Review: In 2024, regulatory bodies intensified their scrutiny of AI applications in healthcare. FDA strengthened its oversight to ensure patient safety and foster innovation. This included initiatives like the Breakthrough Devices Program and the Safer Technologies Program (STeP), which facilitate timely access to life-saving devices by expediting regulatory review.

Prediction for 2024: US federal, state, and private watchdogs will check AI not only for accuracy, but also for fairness and bias.

2024 Prediction Review: FDA is acutely aware of the risk of algorithmic discrimination. There have already been settlements related to SafeRent’s algorithm discrimination impacting rental applicant scoring on the basis of race and income. FDA is therefore taking steps to avoid such discrimination and is undoubtedly considering such concerns while authorizing more than 950 AI/ML-enabled devices since 2020.

Prediction for 2024: Data privacy as applied to AI will get serious attention from regulators.

2024 Prediction Review: As seen in the Clearview AI facial recognition lawsuit settlement, companies are using nonconsenting user information to train their systems. There are already legislative proposals to require that AI systems get consent prior to using data for training purposes. The FDA, FTC, FCC, and other agencies will likely continue to expand their purview to protect patient subject privacy.

Prediction for 2024: It will become more important to be able to explain how AI works, source the results, and explain what the AI finds.

2024 Prediction Review: Regulators and stakeholders have underscored the importance of transparency in AI operations. FDA has outlined its interest in explainable artificial intelligence and in transparency of AI and ML processes.

Prediction for 2024: Lawsuits will begin to be filed over AI mistakes.

2024 Prediction Review: The legal landscape saw significant activity concerning AI-related errors. Notably, one insurer faced lawsuits alleging the use of an AI model with a 90% error rate to deny care, highlighting the critical need for accuracy and accountability in AI applications. This, in combination with the previously discussed ClearView AI lawsuit and the SafeRent algorithm discrimination lawsuit, demonstrates the emergence of and continuing trend of lawsuits resulting from AI-related errors.

Prediction for 2024: If AI is done legally and compliantly, especially by sales and medical information personnel, trust can be built into the system. While FDA has focused on quality by design, AI lawyers will need to work with sales and medical affairs teams to build trust by design into the systems.

2024 Prediction Review: The DOJ recently updated its “Evaluation of Corporate Programs” guidance to include consideration of AI risks to ensure that use of AI will not result in “deliberate or reckless misuse” that violates criminal laws or the company’s Code of Conduct. The US Department of Health and Human Services recently updated its own compliance plan to establish new agency requirements and guidance for artificial intelligence (AI) governance, innovation, and risk management. The FDA is already looking to crack down on “dishonesty in clinical studies” using artificial intelligence.

Accordingly, multiple agencies are all looking to use AI to ensure continued legal compliance and ethical AI deployment.

The Competitive Pharmaceutical Landscape

Prediction for 2024: The competitive pharmaceutical marketplace will compel increased trial innovation: The pharmaceutical marketplace is more competitive than ever, particularly in the race to be first to market. The first-mover advantage is dramatic, so clinical trial sponsors need to be focused on maximizing every second and not wasting time. Decentralized clinical trial (DCT) methodologies will become part of the standard toolkit, embedded into most trial processes and operational decisions, because sponsors need new ways to collect data reliably and engage with patients in a competitive marketplace. In 2024, the shift to DCT methodologies will be transformative and wholesale, bringing DCT elements into the organization at the portfolio level rather than trial by trial.

2024 Prediction Review: At the start of 2024, there was a lot of discussion about the DCT hype cycle and whether we were over it. I suspect that this comes from people who realized just how resilient and rigorous the effort needed to be to implement DCT change management. Based on our work with clinical trial sponsors (having deployed our DCT platform in nearly 400 decentralized and hybrid clinical trials in 70 countries, serving more than one million patients and research participants globally), those who doubled down on DCT value had breakthrough outcomes this year, including 200% faster enrollment and 50% cost reductions. In fact, a Tufts CSDD study shows that, on average, decentralized trials can achieve net financial benefits from five to 13 times for phase 2 and phase 3 trials, equating to roughly $10M ROI and $39M ROI for an investment of $500K in phase 2 and $1.5M in phase 3 trials, respectively.

Further, the Partnership for Advancing Clinical Trials (PACT) consortium in conjunction with the Tufts Center for the Study of Drug Development (CSDD) has produced compelling results from a new study. In total, 13 clinical trial sponsors and contract research organizations provided robust data on 60 clinical trials that deployed decentralized solutions and found that actual timelines beat planned timelines from first site activated to first patient enrolled, and from first patient to last patient enrolled.

Prediction for 2025: The DCT adoption progress made in 2024 will scale exponentially in 2025 as the standards, processes, and capabilities become firmly embedded in everyday practice. This is especially likely as FDA finalized its guidance for trials with decentralized elements, essentially cementing these technologies in our vernacular while showing government support. Earlier engagement in portfolio design will identify greater opportunity to consistently apply digital tooling and data collection standards across suites of studies, streamlining aggregation and resulting in earlier decision-making. It will also help to automate data flow to harness more digital endpoints with greater precision outcome measures to reduce noise and distortion in signal detection. We will see further growth in master protocol designs with digital endpoints captured in real time to inform in-flight decisions on design pivots and cohort management to accelerate time to results. AI advances are also going to accelerate the transformation. There are opportunities to harness AI agents to relieve, and in places remove, manual tasks that are prone to inconsistency and error. In addition, AI will automate many manual activities such as scan reading, uploads of code, trend spotting, and more, to provide higher integrity outputs and data quality.

Data Sharing and Cross-Industry Collaboration

Prediction for 2024: Data sharing and cross-industry collaboration will power the 2024 AI revolution in clinical research. No one company has enough data to drive accurate predictions around a single disease or use case, which is why cross-industry collaboration like we witnessed during the pandemic will be reinvigorated.

2024 Prediction Review: This prediction was challenged by industry-wide portfolio prioritization and cost reductions. While the pandemic demonstrated the power of cross-industry collaboration to achieve ambitious goals, post-pandemic conditions produced financial pressures on biopharma organizations, leading to pipeline restructuring and organizational shifts. Amongst other contributing factors, rising R&D costs, inflation, and investor demands for improved margins have driven a narrower focus on core business areas, limiting resources for data-sharing initiatives. However, with ongoing consolidation by large pharma and therapeutic expansion plans for midsized companies, renewed opportunities may emerge in 2025, as these initiatives will depend on insights derived from shared and diverse data not readily available.

Prediction for 2025: In 2025, cross-industry collaboration and data-sharing opportunities are emerging. These include mining study startup data from over 12,000 sites across 55+ countries to predict study timelines for AI-optimized solutions. While these collaborations offer promise for global study optimization, they will likely involve partnerships between individual organizations rather than large consortia, reflecting the market realities of 2024. Additionally, AI applications in 2025 are expected to focus on reducing repetitive workloads rather than groundbreaking discoveries, such as curing ultrarare cancers. One reason for this shift is the relative ease of sharing and leveraging less sensitive operational data compared to safety and efficacy data, ensuring that AI models are effective within existing data-sharing constraints.

Decentralized Clinical Trials (DCTs): Evidence versus Hypotheses and New Focus on “How”

2024 Prediction: Evidence will replace hypotheses in DCTs: By the end of 2024, the life sciences industry will replace hypothetical surveys about the impact of decentralized methodologies in clinical research with hard evidence. The Tufts Center for the Study of Drug Development has formed a pre-competitive consortium of more than 50 companies that will collate metrics to understand how decentralization impacts protocol performance such as cycle times, patient recruitment, and retention rates, with the goal to publish its first data set by spring 2024. As we continue to capture evidence of DCT value, we will increasingly adopt a nuanced approach, but this progression will move at the pace of maturity of each DCT innovation.

2024 Prediction Review: New trial data is being added every day and the data continues to be analyzed to uncover exactly which DCT elements contribute to what impact, and whether that impact is universal or more pronounced just in certain types of trials based on populations and indications.

The Partnership for Advancing Clinical Trials (PACT) consortium in conjunction with the Tufts Center for the Study of Drug Development (CSDD) released an initial data set based on 69 studies of which only about 20% were completed. In total, 13 clinical trial sponsors and contract research organizations provided robust data on clinical trials that deployed decentralized solutions and found that actual timelines beat planned timelines from first site activated to first patient enrolled, and from first patient to last patient enrolled. The data on the impact is still incomplete, however, as most of the trials did not complete until this year and many more are slated to have a database lock in 2025.

This initial set also showed that study visits are the most common activity supported by DCT solutions. Specifically, eCOAs are the most deployed DCT solution (80% of trials), while portals, apps for data collection, apps for reminders, and home visits were each used in about half of trials in the first year’s data set.

Prediction for 2025: We will gain a much greater understanding in 2025, and we will have the evidence needed for much broader adoption of DCT elements by the end of 2025. For the past two years, decentralized trials have overcome hurdles that were never addressed during COVID. Did they deliver on the promise of faster trials, better representation, and better participant engagement? Yes, data is emerging which shows that cycle times are typically beating plan timelines and that, when deploying decentralized elements, trials have a lower proportion of white participants and a higher proportion of Asian participants, with other groups also showing a modest increase. Some have referred to this past phase as the “trough of disillusionment.” In 2025, we will emerge out of that phase and into the “slope of enlightenment.” We are overcoming skepticism with evidence and improving technology. Regulators have been removing ambiguity with clear guidance, and more will come from ICH E6 R3 Annex 2. As a result, trials with decentralized elements will be deployed in a much more transparent and thoughtful way, benefiting participants with better representation, improving data, and ultimately bringing treatments to patients more rapidly.

2024 Prediction: The acronym “DCT” will disappear as decentralization becomes the norm: By the end of 2024, the life sciences industry won’t distinguish DCTs from “simple” or “traditional” clinical trials anymore. The acronym DCT will become less of a callout as a distinct type of unique clinical trial. Recent FDA and other global regulatory agency guidances explicitly note that decentralized trials simply need to comply with existing regulations for clinical trials—nothing extraordinary. By the end of 2024, decentralized clinical trials will simply be clinical trials (with certain parts appropriately decentralized), and the impact of this model will be much more prevalent.

2024 Prediction Review: Industry still makes distinctions between various types of trials and calls out trials with DCT elements as separate, but the recently released FDA guidances clearly state that “FDA’s regulatory requirements for investigations of medical products are the same for trials that include decentralized elements and trials that do not include decentralized elements.” In other words, trials with decentralized elements are not expected to perform to a different standard than their traditional counterparts. A quilt of options for clinical trials will be deployed fit-for-purpose—from protocol design (quality by design, pragmatic designs, adaptive designs) to execution with decentralized elements embedded within clinical care depending on the medical product, indication, population, and other unique situations.

Diversity in Clinical Research Participation

Prediction for 2024: For 2024 and beyond, the focus on patient diversity in clinical trials will increase. There is far more capacity now to connect with broader patient populations using new digital technologies. Given the convenience of these tools, more patients who become increasingly comfortable with technology will take advantage of digital opportunities to connect to a trial and remain engaged through to the end.

2024 Prediction Review: Aligning this focus and determination to change, in addition to greater commitment to meeting the needs of sites and patients to reduce burden and improve accessibility, certainly progressed during 2024. More companies invested in dedicated roles to raise the inclusion and visibility of the patient voice within their plans.

Prediction for 2025: We will see greater investment in studies designed to enable greater accessibility and in tools that make it possible to meet patients where they prefer to have their care interface (on-site, remote, vacation, pharmacy, etc.). This is the key to improving participant diversity and representation in our trials. There will be significantly greater adoption of digital consent technologies not only to reduce trial participation burden on patients but also to improve patient awareness, comprehension, and understanding of trial activities to optimize participant engagement throughout the trial.

Market Access Models for Digital Therapeutics (DTx)

Prediction for 2024: Current market acceptance for prescription digital therapeutics (DTx) is low and improvement may take years. Given this, many DTx companies will pivot away from the prescription model and go straight to consumers in 2024.

2024 Prediction Review: Broadly speaking, this prediction was accurate. However, there was little overall clinical trial activity in the DTx space in 2024, both in the consumer and the prescription market segments.

Prediction for 2025: FDA released draft guidance on Prescription Drug Use-Related Software (PDURS) in late 2023, marking a significant step towards integrating digital tools with prescription medications. Throughout 2024, stakeholders in the pharmaceutical and digital health industries actively evaluated and developed strategies to align with this guidance. These efforts are expected to gain momentum in 2025, as companies transition from planning to implementing innovative solutions through digital integration.

Pharmacovigilance Guidelines: Remote Audits and Inspections

Prediction for 2024: With global armed conflicts and geopolitical unrest continuing or even intensifying in 2024, it is likely that regulatory inspections and routine audits will continue in remote mode.

2024 Prediction Review: Predictions regarding the worsening of geopolitical tensions turned out to be correct, and while 2024 data is still being collected, many remote audits and inspections were conducted in 2024. I expect the trend to continue in 2025.

Strengthening Medical Device Regulations

2024 Prediction: It is also reasonable to expect FDA to propose new guidance documents to address burgeoning technologies and enhance the safety and efficacy of digitally enhanced medical devices, particularly in connection with cybersecurity, artificial intelligence, machine learning, and digital health applications, in 2024.

2024 Prediction Review: As anticipated, FDA continued to establish guidelines for software as a medical device, artificial intelligence (AI), and machine learning (ML), throughout 2024. This focus on digital health regulations prioritized safety and efficacy, providing a clearer pathway for innovators while ensuring patient protection.

FDA focused on enhancing the safety and efficacy of digitally enhanced medical devices, particularly in areas such as cybersecurity, AI, ML, and digital health applications. This proactive approach demonstrated the agency’s commitment to keeping pace with rapid technological advancements.

In addition, 2024 saw a significant amalgamation of medical devices with novel digital health technologies, leading to groundbreaking solutions for diagnosis, monitoring, and treatment. This integration has, as predicted, indeed transformed both clinical research and healthcare delivery.

Finally, the increased connectivity of medical devices led FDA to implement fortified cybersecurity regulations to safeguard trial participant privacy and device integrity. This development underscored the growing importance of data security in the medical device landscape.