The advent of ChatGPT is shining a spotlight on the possibilities of artificial intelligence (AI) across all industries. In the recent past, we witnessed machine learning algorithms primed with historical data turn it into generative outcomes – a method relied upon by ChatGPT as a large language model (LLM).
Beyond creative industries, we have also seen phenomenal innovation in the medical and health sectors due to AI. From fitness tracking apps to leveraging big data for diagnoses, AI is removing manual processes and delivering faster outcomes.
However, in terms of digital transformation, the pharma industry has lagged behind, particularly with regards to adopting new technologies at the same pace as other industries. Market access – being a relatively new and emerging function – has even more catching up to do.
In the last decade, the role of market access has leapt forward significantly, and is now recognised as a core function within pharmaceutical organisations, playing a pivotal role in medicine launch and overall business success.
Meanwhile, internal digital teams have typically focused on using tech to advance departments such as R&D, medical affairs, commercial and marketing, leaving market access underserved.
The pharmaceutical industry is increasingly warming to the potential of artificial intelligence. Momentum on how AI could be used continues to build. Potential applications start from discovery and development through to clinical trials and, much like other industries, the possibilities are endless.
Extracting key insights from diverse datasets
With the accessibility of tools such as the above-mentioned ChatGPT, we have new avenues for intelligence within market access analysis and decision-making. For example, LLMs can be trained on vast data from market access databases, enabling teams to query them with relevant prompts to extract valuable insights.
Market access professionals often reflect on numerous and complex information from a wide range of data sources, including pricing data, payer research and HTA reports, all of which hold critical information that teams must painstakingly sift through manually. With so much data being captured from different sources, it becomes very difficult to find the most pertinent information that should be considered for evidence-based decision making. LLMs can excel at distilling extensive data sets into simple insights, saving hours of time and unlocking important evidence hiding in plain sight.
Not only can AI be trained to analyse and extract insights from unstructured and often subjective reports, often provided in various foreign languages; AI can also help automate health economic modelling, allowing for faster and more accurate assessments of a drug’s cost-effectiveness. It can also help in simulating different scenarios to assess the impact on market access.
Furthermore, AI can be used to predict launch success. This requires in-depth analysis of historical data, including commercial data, past pricing trends, timeframes for market entry, competitive dynamics, ROI projections and more. AI is ideally placed to analyse the data available, empowering market access specialists to better forecast and model success for their upcoming launches.
Another consideration is the visualisation of data. As it stands today, custom visualisation of multiple data points is challenging and time-consuming for market access teams, but something that is in high demand by market access leaders and senior executives. The integration of AI into tools like Access Infinity’s NURO solution, streamlines the process, making the generation of charts and personalised visualisations effortless.
The challenges of using AI in market access
While in theory, there are numerous opportunities to use AI within market access, it is still nascent in its technical journey. We need to overcome the barriers of poor digital maturity to truly leverage AI. It is certainly time for companies to stop relying on legacy methods – like Excel – to make decisions and start exploring the possibilities of more advanced and specialised tools.
Meanwhile, there are various limitations in the data itself, that limit the accuracy of the outputs and continue to create scepticism even among early adopters. Namely, we are still lacking sufficient data – we have around 2,500 brands that have been launched globally since 1995, and ideally, we would need a lot more data to train the models to improve the accuracy of predictions.
Standardisation efforts are ongoing but can be slow to adopt. The data is often stored in various formats and systems, making data analysis complex.
AI in practice
While there’s no doubt that AI can – and will – accelerate decision-making, it’s important not to lose sight of the bigger picture.
Market access is not just a science, wherein a given input, such as evidence, guarantees an output, such as price. It will continue to heavily rely on human intervention, judgement, and experience. Thus, the ‘art’ of Market Access is better served by combining human expertise and AI to ensure success.
We anticipate that AI will become one cog in the machine taking care of routine tasks, so that, there are more resources available to handle complex requirements that need human collaboration and intervention.
Our best practice tips for beginning your journey towards using AI include the following:
- Gather the largest sets of high-quality data, including relevant price, HTA, evidence and regulatory approvals to begin connecting the dots.
- Experiment with multiple avenues and don’t shy away from exploration with a focus on refining prompt engineering and training datasets.
- Partner with others to make major inroads without losing focus on your brands.
What exposure have you and your team had to AI? And what role do you see it playing for market access in the future?
Access Infinity has made exciting advancements in incorporating AI into our digital solutions. If you’d like to learn more, we’d love to hear from you. Get in touch at firstname.lastname@example.org.