It's been another month of great events at HealthXL. Our Boston Global Gathering was a roaring success, with community members from around the globe coming together to discuss the next frontier of digital health in pharma at this critical time. We also hosted many virtual roundtables and masterclasses on a number of topics from ‘GPT for Life Sciences - What Does The Future Hold?’ to ‘Digital Health at Scale: What Pharma Can Learn From MedTech’. Additionally, we ran a great webinar on what pharma and MedTech can learn from each other in digital health.
To make sure you don't miss the top insights from our communities, here are my top five takeaways from HealthXL meetings in May:
1. AI Large Language Models (LLMs) are predictive tools that still require human oversight.
Artificial Intelligence (AI), specifically generative technologies, are increasingly integrating into all areas of our personal and professional lives. In particular, the debut of ChatGPT has made Large Language Models (LLMs) more accessible for everyone and we've observed widespread adoption already. LLMs are a predictive form of artificial intelligence model that are trained on vast amounts of data to generate human-like text. But what exactly are these technologies, where do they promise the greatest potential within life sciences, and how can their safe and effective deployment be ensured? We discussed these questions and more in our recent session on ‘GPT for Life Sciences - What Does The Future Hold?’.
Should we be worried about the use of LLMs in healthcare?
Although these solutions have incredible potential, there are many risks that need to be considered, particularly in the healthcare setting. Understanding the predictive nature of LLMs, which can give rise to the phenomenon of ‘hallucinations’ is essential, but is often overlooked by users.
'Hallucinations' are instances when the model generates information that, while plausible or mostly correct, is inaccurate or nonsensical. Countering this risk necessitates grounding the models in more reliable data. LLM models are only as reliable as the data they were trained with, and each model is grounded in unique data. For example, ChatGPT is grounded in data only to a specific date, whereas Bing AI is grounded in current information and references. Both these models are based on open-source data, whereas other models are trained on proprietary data.
Fortunately, as LLMs evolve and become more sophisticated, we expect the frequency of hallucinations to decrease. Nonetheless, given their inherent predictive nature, it is vital to always ensure human oversight in their application. These models are tools, not autonomous agents; this realisation needs to drive responsible usage practices and a continuous quest for improvement.
2. Embedding digital into R&D early propels routine care.
When implementing digital technologies into the drug lifecycle, one consensus emerged across pharma: the sooner digital strategies are embedded, the higher the likelihood of a fully integrated, commercially successful product.
In particular, during the R&D phases of drug development, it's typically easier to attribute outcomes to the drug rather than its digital adjunct, provided that the parameters are clearly defined. Commercial value becomes the main focus in later stages (e.g., market access, revenue generation), which can be a complex proposition. This underscores the merit of early digital incorporation during R&D, setting the stage for a smooth transition into routine care. Moreover, in this session it was noted that Phase 2 trials provide a prime opportunity for digital integration, as this stage can provide an experimental window where participants are more receptive to trying digital solutions.
Adopting a holistic, portfolio-centric approach, rather than focusing on individual brands, can decrease the burden on providers and boost the likelihood of adoption. This macro-level perspective helps streamline the integration process and promotes greater consistency across different products.
Discover more insights and considerations for incorporating digital into your brand life cycle from our recent meeting on the ‘Role Of Digital Health Across Brand Life Cycle In Pharma’ here.
3. Alternative pathways that could bolster digital oncology adoption.
Given the trusted relationships that oncologists have with their patients, they can make or break the adoption process of digital health solutions. Indeed, oncologists are typically the initial point of contact for digital health companies in their business model.
However, even the most digitally-focused oncologists and healthcare providers (HCPs) can be stretched thin with high workloads, time constraints, lack of system integration, and an incentive gap. These factors can significantly hinder the adoption of digital health solutions.
Are there other stakeholders in the healthcare ecosystem that digital health vendors could engage with as a first point to promote adoption?
Our recent session on ‘Beyond the Provider: Alternative Paths to Adoption of Oncology Digital Health Solutions’ examined alternative models such as payers, direct-to-consumer (D2C) models, specialty pharmacies, and Medicare Advantage, that could serve as initial points of contact.
As cancer remains a top priority for society, finding the right partners can pave the way towards a more accessible and impactful digital solution. D2C models can act as a viable first phase of DTx commercialisation, providing a pragmatic starting point for the development of evidence-based DTx. Strategies such as value-based payment models may help facilitate wider uptake. Moreover, other avenues, like Medicare Advantage and specialty pharmacies, are well-positioned to drive adoption and may be viewed as alternatives or complements to the traditional provider model.
While digital health adoption in oncology may be a marathon rather than a sprint, we do not foresee significant vendor exits in the next decade. For more insights, check out the key takeaways from this session here.
4. How to harmonise digital health with VBHC strategies to maximise patient value
It is clearly apparent that embracing Value-Based Healthcare (VBHC) strategies will be pivotal in providing appropriate care for patients. Yet, despite its seemingly straightforward premise, VBHC is challenging to implement due to the complex and subjective nature of "value" within healthcare. If correctly combined with digital health, VBHC can yield significant benefits. However, if not done properly, digital technology could hinder rather than aid VBHC.
To ensure we are effectively integrating digital tools into VBHC models, Matt Hickey, CEO of The Health Value Alliance, hosted a masterclass where he explained the principles of VBHC, the means to make digital a helpful ally rather than an impediment to VBHC solutions, and demonstrated the value digital can bring to VBHC for all stakeholders.
Here’s a recap of potential value-add strategies:
1. Determining the purpose and target audience of data is paramount, as is ensuring digital health measures are desired rather than enforced. A balance between academic research, practical applications, and validation of real-world data will foster a harmonious relationship between VBHC and digital health. To incorporate digital health as a valuable asset into VBHC, we must ensure that data is promptly and accurately captured, validated, and continuously re-validated. Data analysis is crucial for understanding value and must be carried out responsibly to uphold validity.
2. Our actions should conform to evidence-based guidelines, which should also offer opportunities for validation. Inefficient outcomes often result in health value depreciation; however, data can enable us to predict patient trajectories, thus facilitating preemptive measures for individuals at risk. Predictive modelling contributes to clinical improvements, enhances value, and helps overcome translational challenges.
3. To ensure digital health is an ally to VBHC, we need to communicate its value effectively to different stakeholders, or 'speak multiple languages' based on the stakeholder group we are addressing. To prevent market disengagement, it's vital to maintain a continuous cycle of demonstrating value.
5. Pharma and MedTech bring unique strengths to digital integration – what can they learn from one another?
In our recent webinar on ‘Digital Health at Scale: What Pharma Can Learn From MedTech’ , HealthXL CEO Chandana Fitzgerald came together with Jim O’Donoghue (President, S3 Connected Health), John Rondoni (Chief Technology Officer, Inspire) and Justin Wright (Global Head Connected Health (VP), Novartis) to compare the fit for digital health within MedTech and pharma, and what both industries can learn from each other in order to go beyond pilots and achieve impactful digital solutions.
What we identified was that MedTech, owing to its closer similarity to digital health than traditional pharmaceuticals, seems to have an edge in integrating digital as a fundamental part of its R&D roadmap, and so is better equipped to create the ecosystems needed to move beyond point solutions and achieve widespread success with DH across franchises. In contrast, pharma often views digital as an ancillary option that can enhance patient support but is not necessarily central to its core agenda.
As the digital health landscape is inherently complex, adopting a platform-thinking approach and leveraging insights from across industries and companies is essential. This collaborative method will pave the way for problem-solving in a unified manner, preventing repeated mistakes across companies.
Join the HXL community, where members can rewatch our webinars, read the full meeting takeaways, and more.
These are just a few key insights from some of our discussions last month. Head over to our platform to read through the takeaways from all our past meetings covering a great variety of digital health topics. Or even better, apply to attend an upcoming virtual meeting or in-person event.
See you in a HealthXL meeting soon!
.png)
