Not since our first exposure to the internet has a digital notion captivated us like that of artificial intelligence (AI). It’s been the stuff of sci-fi fiction and films; the likes of Isaac Asimov warning us of computers that are smarter than humans, smart enough to take over the world (insert maniacal laugh here)! The practical applications seem limited only by our imaginations. Previously complex tasks that might normally take hours or even days of sweating over a keyboard can now be done in seconds. We are in the nascent stages of understanding the technology and how it can best be deployed.
So how will AI affect the multitude of stakeholders in the delivery of healthcare? Let’s consider the stakeholder groups.
Physicians are already using AI in order to narrow down likely diagnosis, where patient profiles can be entered, using demographics, health history, labs and vital signs to name a few. AI then generates diagnostic scenarios for doctors to consider. It’s clear that speed to accurate diagnosis can dramatically improve patient outcomes. The hospitals employing those physicians can also use AI to predict ICU transfers, improve clinical workflows and pinpoint a patient’s risk of hospital-acquired infections. AI has proven a critical element in the move towards surgical robots that enable to provision of minimally invasive operations. The applications in this environment seem endless with huge implications for improved outcomes at a reduced cost.
And the patients they care for can also accrue huge benefits from the use of AI. AI can use its wealth of medical data to generate personal treatment plans in real time. We know of the challenges for physicians to provide quality time for patients, so this speed will optimize those few minutes they are able to provide. The faster an accurate diagnosis can be generated, the quicker appropriate treatment can begin. In the not-too-distant future, patients will have 24-hour access to virtual assistants, who can answer questions specific to their customized treatment plan.
How will it affect the treatments we are able to offer? For pharmaceutical manufacturers, AI should expedite the drug discovery and development process. Currently, it takes, on average, 10 years to develop, test and bring a drug to market. AI can provide significant streamlining of that process, with improved prediction of outcomes. We’ve already seen the first AI-discovered molecule based on an AI-discovered novel target granted orphan drug designation by the FDA (Insilico Medicine https://insilico.com ). This is for a product developed at a fraction of the time and cost of a traditional pre-clinical program. AI technologies are also being used in drug development for complex processes including target identification, molecular simulations, 3D modeling and the prediction of drug properties.
So, what about selling and marketing pharmaceutical products? Will or should AI play a role? It seems clear that AI can be used to tightly identify sales targets through complex data iterations, informing reps on who to call on and what approach they should take to the call.
As a creative agency, we’re committed to optimizing brand marketing communications for clients, so there is definitely a role that AI can play. However, we need to approach it with caution. Clients look to us to generate accurate, compelling, original content that becomes their intellectual property. There are two issues related to the use of AI here. First, while often compelling, references and accuracy of content are questionable. Next, content generated by these sources is owned by them as the authors, leaving open many legal risks down the road surrounding copyright.
For now, while the proverbial dust settles, we are focusing our efforts with AI on creating efficiencies for the provision of some services for our clients. AI can quickly generate a disease overview enhancing our knowledge of a therapeutic category and has proven an invaluable tool in garnering competitive intelligence. One of the huge strengths of AI is its ability to analyse large amounts of data, contributing to an enhanced approach to targeting strategies including omnichannel solutions. We plan to leverage the automation of available public data and insights in real time, getting us to the right solution, faster.
Still, until the inherent kinks represented by AI are worked out, we feel for now, nothing beats good old human relationships and interaction; face to face contact with clients, live discussions with KOLs, conducting primary research, interactive ideation & brainstorming. You know, old school…