Every year, we see stunning advances across the life sciences. Whether it is novel drug discoveries, or new data techniques to optimise research and spending, each is done to improve patient outcomes.
Make no mistake, 2024 will be another year of evolution in our industry. Indeed, there are three major trends that will fundamentally reshape how many businesses operate.
From taking the next steps in generative AI, to ensuring the success of rare disease treatments, and managing the increasing complexity around data coordination, these will be among the agents of change in 2024.
Trend 1 in 2024: GETTING GenAI RIGHT
If 2023 was the year of generative AI, 2024 will be the year that pharma and life sciences companies will need to get it right. GenAI and large language models (LLMs) have been a game-changer.
Whether using simple chatbots to help patients refill prescriptions, or more complicated natural language processing of doctors’ critical notes, artificial intelligence and machine learning have saved companies time and money.
But as most companies have realised by now, generative AI can quickly hit a wall. We see at least two main issues that organisations need to jump in front of to keep their GenAI investments on track.
The first deals with data. Put simply, if GenAI and its LLMs can’t read your data, you’ll miss out on key analysis and insights. It is imperative that you have generative AI-ready datasets (GRDs).
GRDs must contain a wide range of unstructured formats such as text, images, and audio. That makes them ideally suited for training the LLMs that power your expensive GenAI setup. Without a GRD, you’re left with just structured data.
That’s useful for analytics and reporting, but not nearly enough to reach GenAI’s fullest potential.
We know AI has the potential to revolutionise research and accelerate drug development, discovering new formulations far quicker than human teams.
But adding GRDs to the mix can advance personalised medicine and the patient experience, thanks to the trove of information nestled deep inside unstructured data.
In addition, real-world evidence (RWE) and health economic outcomes research (HEOR) data are packed with vital insights at the granular, patient level.
Combine it all through GRDs and imagine upping your game by providing a chatbot that can understand patients’ nuances, rather than give canned responses.
Or helping your medical science liaisons write faster response letters by 30% in some cases? GenAI continues to evolve in pharma, but generative AI-ready datasets will be critical in 2024.
Another GenAI-related issue we will face in 2024 is timeliness. Currently, most no-cost LLMs have a cutoff date, where their training data ends, whereas many commercial models have newer end dates.
Either way, if you are not continually injecting new sources into your model, you will have outdated answers. It happens so often that the industry has a term for it – hallucination. One way to combat that is by deploying a retrieval augmented generation (RAG) workflow.
In this process, your model will craft a response by pulling in newer and more relevant information from an external source—most notably, your own data warehouses.
Using your own context-specific data gives your GenAI model an incredible advantage in areas such as next best action suggestions.
Getting GenAI models to work will be a significant focus in the coming year. But we must never forget that while it is impressive, we must keep human ingenuity in the loop.
That means having skilled associates who can monitor GenAI usage, fine-tune it, and stay on top of it from a security and ethical standpoint.
It also means hiring creative people for what will surely be one of the top jobs in 2024 – prompt engineers, people who can think outside the box and know how to craft questions in such a way that gets GenAI models to output the required data.
That takes playing around with queries and style. It is a new career path and one that will remain uniquely human.
Trend 2 in 2024: RARE DISEASE TREATMENT IS DATA-DEPENDENT
Thanks to the advancements in AI and quicker discoveries, the industry is turning its focus to rare disease therapies.
The timing is ripe: the US Food and Drug Administration says there are more than 7,000 rare diseases affecting people in the United States right now, and most do not have treatments.
As you know, the success of a drug fuels the resources to find more. But rare diseases throw a wrench into the works. These treatments carry significant production costs and there is a precious window of time to find patients quickly.
One trend we expect in 2024 is the realisation that it takes a data-first approach to make rare disease treatment commercially viable. Traditionally, drugs coming on the market have a large patient pool and therefore carry a relatively lower cost. That model is flipped with rare diseases.
These drugs have a high cost and a low patient pool. A data-first approach is needed to close the critical time gap between finding patients and being commercially viable —especially now, with many promised miracle drugs in the pharma pipeline.
You must identify potential patients, recruit those suffering from the ailment, and closely monitor their adherence. And because the patient pool is so tiny, each one takes on increased importance in the continued viability of that drug.
That means companies must be impeccable with data quality and cleansing, standardisation, and visualisation.
However, there is a second consideration in this time-sensitive task. A pharma company must also find physicians who will prescribe these medications and add their patients to the new brand.
Due to the high expense of marketing and promotion through different channels, getting physicians to accept a brand may be cost-prohibitive.
Adopting a data-first mindset here can help. You must know how to keep accurate and timely data on prescribers and their preferences for interaction, especially when dealing with new-to-market offerings.
Trend 3 in 2024: SHORTER RUNWAYS REQUIRE MASTER COORDINATION OF DATASETS
Here’s a staggering figure: according to one estimate, around 120 zettabytes of data will be generated in 2023 – that’s 120, with 21 zeroes after it! And 30% of that will be health-related data.
Therefore, a significant trend in 2024 will be the master coordination of datasets, especially as we see shorter runways for drug viability.
With smaller windows for commercial success and patent clocks ticking, it is critical to get your data management investments correct right out of the gate.
You will need perfect alignment between the dizzying datasets tied to patients, payers, customers, sales, and marketing.
That means an overhaul in thinking; you must revamp your strategy, execution, and platforms. You must adopt the mindset of “do it right the first time” and forget the past practices of waiting to fix issues later in the process.
If you maintain the status quo and continue creating more datasets, the information becomes increasingly siloed and inaccessible beyond its initial scope.
That locks away a wealth of information that could be used for personalised medicine, precision treatments, and accelerated drug discovery. Exponential data growth also means the effort to manage, maintain, secure, and govern it has escalated substantially.
One way to solve this alignment problem, while also tackling the silo issue, is the use of a comprehensive data fabric. Implementing this will be a crucial differentiator for companies in 2024.
The fundamental function of data fabric is to integrate data from various sources, formats, and locations, providing a centralised and consistent approach.
For life sciences organisations, a data fabric enables the seamless combination of patient and provider information from various sources, such as electronic health records, sales data, marketing data, real-world evidence, and more.
This ecosystem provides a unified view of critical information, fostering interoperability and enhancing data-driven decision-making.
The specific life sciences use cases include real-world evidence integration, targeted marketing and sales strategies, personalized customer engagement, improved market insights, scalability, and flexibility.
A data fabric ensures that the infrastructure can adapt and grow to harness the power of this diverse data landscape.
Final analysis
2024 will be a transformative year in life sciences. These new trends are not impossible tasks. GenAI requires ongoing attention, rare disease drugs must be commercially viable, and massive and disparate datasets coordination can be mastered
Jassi Chadha is Co-founder and CEO at Axtria
