Market Access Blog

blog-graphics-for-website-6

Artificial intelligence and market access – all the buzzwords simply explained and why they matter to you

You will have encountered them in popular media, in scientific literature and at dinner parties. They often appear in groups or accompanied by a futuristic graphic. I am of course talking about all the artificial intelligence buzzwords, which started appearing everywhere a few years ago.

Perhaps you meant to check what they mean, but never really got the time to look them up, and now with everybody talking about them, it is rather late to ask. Alternatively, you do understand most of them but just need a little refresher. Either way – don’t worry, we’ve got you covered with the most popular of the artificial intelligence buzzwords, which should help you navigate around most simple technical conversations:

  • Artificial intelligence (AI) – intelligence demonstrated by machines, as opposed to the natural intelligence displayed by humans or animals. It is the most ubiquitous of all the buzzwords because it essentially encompasses all of the other buzzwords. Whenever you have learned something clever from a program or were impressed with what computers can do these days – it was likely AI. AI covers both simple and very complex algorithms; when in doubt about the correct buzzword to use to sound fancy, use AI and  you will probably be technically correct, if not very precise.
  • Machine learning (ML) – the study of computer algorithms that can improve automatically through experience and with the use of data. The essence of ML is the ability of a program to learn by itself (usually from a copious amount of data) what is the best way to obtain the correct answer, without human guidance. ML is therefore a method to create AI. With ever-increasing computing power and data availability, ML has become the main method of creating AI.
  • Artificial neural networks (ANN) – an ML technique, utilising computing systems inspired by the biological neural networks that constitute animal brains. Most people would agree that the brain is rather a useful organ. ML researchers also agreed and started experimenting with creating artificial neural networks as computer algorithms. These ANNs usually constitute layers of neurons. The combination of  signals sent between these layers gives the final output of the ANN. Researchers have been experimenting with ANNs since the 1950s, but only recently have they grown in prominence thanks to the invention of…
  • Deep learning – ANN technique, which utilises multiple layers of neurons. With the increase in data, computing power and, most importantly, the number of layers, ANNs became capable of solving problems which were unthinkable just a decade earlier. “Deep” simply refers to many layers in the network (10s of them), contrary to “shallow” networks with few hidden layers (1–3). These deep ANNs are now at the forefront of ML techniques, responsible for the most impressive advancements in picture recognition and natural language processing, and used in the creation of so-called chatbots. Speaking of which:
  • Natural language processing (NLP) – a subfield of AI concerned with the interactions between computers and human language. When most people think of programs and algorithms, they rarely assume that they could be well versed in understanding human languages, such as English. NLP is a subset of tools designed precisely for that purpose – processing and analysing large amounts of natural language data. As it is a problem-specific set of tools, it spans across the hitherto mentioned categories.

To combine all of the above buzzwords, check out this helpful Venn diagram. If you can remember it, you can impress your colleagues with your AI knowledge in no time!

screenshot-2022-10-03-at-093800

Ok, so now you know the buzzwords, why should you care? Although some examples of AI are ground-breaking (e.g. self-driving cars) or potentially life saving (e.g. accelerating the drug development process by filtering which chemical compounds are good candidates for new drugs), how are these techniques relevant to pharmaceutical professionals working in market access? The new AI tools can make your work easier too:

  • Literature review (LR) automisation – if you have ever requested a systematic literature review (SLR) for a health technology assessment (HTA), you will know that they take a long time to produce and are not cheap. This is due to the laborious nature of the task, requiring many hours of skilled reviewers’ time. With the recent advances in NLP, however, substantial parts of the LR process can be automated, bringing both the cost and the time needed to more comfortable levels
  • Market assessment/trend monitoring – market access (MA) is complex: MA rules are often not laid out, or, even worse, customary preferences of the authorities (which have nothing to do with rules) can contradict formal specifications. Add changes in rules to the mix, multiply it by the number of jurisdictions you are responsible for, and you suddenly have a very complex problem on your hands. AI can help with that – it can collect and analyse market-access-related developments in key geographies, enabling you to customise your market access strategies as per specific markets. By analysing large numbers of real-world MA decisions, it can help you obtain intelligence on your competitors, and the preferred endpoints of the reviewers, or identify key roadblocks to patient access that you want to avoid
  • Price monitoring and alteration – drug pricing and reimbursement analysis. You might have several competitor drugs undergoing HTA assessments in multiple locations, each with its own set of pricing rules, depending on the pricing decisions of other jurisdictions. AI can constantly monitor global drug prices and reimbursement-related data, and guide pharma companies’ drug pricing strategies, allowing you to sleep calmly at night. With AI, you are given fresh updates on changes in international price reference rules and reimbursement status

With the AI buzzwords tamed, and the practical uses of AI firmly in your mind, we are sure that you will be able to shine in your internal calls with your deeply learned (see what I did there?) understanding of AI. If you have any additional questions, please let the MAP team know, we promise you will not speak with a chatbot*!

*Chatbot (noun) A computer program that simulates human conversation.

Previous Blog
2 / 41
Next Blog
Arrow Vector-Quotes