First of all, don’t make the mistake of conflating ‘digitization’ with ‘digitalization’. These terms are NOT synonymous.

In pharma over the past few years, there have been activities ongoing to ‘digitize’ the industry, which is to say, make that which has previously been non-digital, digital. Digitization refers to making measurable things captive in the form of 1s and 0s. Like how Electronic Medical Records (EMRs) are the digitized version of a written (paper) record.

‘Digitalization,’ by contrast, is taking the form of creating new revenue-generating opportunities from the digital world. So, we are talking about mostly digitalization of new work when we discuss digital pharma and the trends therein.

Clearly, anything that is being differently performed compared with the status quo will create a new trajectory in the industry. And every trajectory is a vector—that is, it has magnitude and direction. This is a property shared by a trend, which is the overall directionality of data points. But a vector can be a line segment with two points, so every trajectory is not a trend. 

So, to say, ‘Digital Pharma was doing x, and most recently it was doing y; So therefore, its trend is ____’, is entirely incorrect prose.¹

Two points, a trend they do not make. Wall Street tries to avoid this proofiness² at all costs, which is why they have beta values, and tickers that have literally thousands of data points daily. But squint too hard, and you can still materialize a trend that isn’t really there. Like ghosts.

Digital pharma done best
Teva Pharmaceuticals recently won awards for pushing boundaries with digital pharma. A brand team recently launched an augmented reality detail aid using digital and sales representatives together to demonstrate good inhaler (nebulizer) practice and mode of action and they have seen tangible positive improvements in HCP engagement. An interactive chatbot ad, directly linked to a chatbot, has also featured on websites such as MIMS, to try and answer every HCP’s specific questions in seconds rather than several clicks and lots of reading.

For digital pharma, including COVID-19 vaccines over the past year, I’ve participated with several sponsor companies to teach behavioral economics concepts and apply them to allocation and accessibility of drug treatments. Within this work, I coined the term Primum, tollere frictionem—which translates to ‘First, remove friction.’³ Friction is the term we use in behavioral economics to describe barriers or difficulty to engaging in a particular, often wanted, behavior. And friction doesn’t have to be big: In some cases, two mouse clicks instead of three can be demonstrated to mean the difference between success or failure of an intervention.

In their digital pharma approach, Teva even repurposed the digital chatbot to be a physical bot that HCPs could interact with at congresses and meetings. The last part there is kind of cheating, making the ephemeral and digital into a physical futuristic monument to progress.

Next up: Market saturation
What will begin to happen, I predict, is that saturation of the market with AR/VR and chatbots will bring all of us pain and frustration. The more we face these nascent technologies from every angle (or even worse – FORCE these modalities on others), the less novelty effect they’ll have, and their overall interest and utility will decrease. It will amount to re-leveling the playing field to include a new baseline of technologies that weren’t used in this way before. A chatbot might seem insightful or fun now, because there are a handful. But wait until next year when there is one everywhere you turn to and you say out loud to no one in particular, “Can we just cool it with the chatbots already?”

This is the endless march of iteration of approaches that gives rise to incremental changes across industries (hence, there’s a certain ‘fashion’ to fonts and logos, corporate attire, marketing approaches, etc.).

Evolution happens. Be ready.
There are always leaders and laggers in every new evolution, and each company’s approach is variably effective (or ineffective). What’s challenging is how to do enough to stay relevant, but not too much or too aggressively that jockeying for trendsetter becomes a capital-draining pursuit.

Over the past decade, R&D’s expenditures in AI/ML have increased by 33%. This, and the new types of ‘questions’ we can ask of AI/ML will bring about philosophical changes in how we do our work in the industry. This will inform and radically change digital pharma over the next several years.

Genomic data
Much of the genomics-analytics space is still very tumultuous, especially on the consumer-side. But it’s clear that we will soon have better methods by which to screen and review patients and drugs for pharmacokinetic and pharmacodynamic nuances that are particular and personalized whenever necessary, or possible. Digital pharma has been, and will continue to be, taking advantage of genomic information to sell toward pharmacogenomics. It will also back-propagate into our computer models of pharmacovigilance. Data extracted principally for marketing purposes which also get used to find drug safety signals. Digital pharma can be digitally-altruistic.

Real-World Evidence (RWE)
Getting real data from out in the wild (even the urban jungles, as it were) is the playground of RWE innovations. The promise and the challenge are both ‘How do we go about collecting real-time data from the population which can give us deeper insights into the reality of healthcare, public health, and overall wellness?’

It is, in my opinion, the analog of how we use the terms “efficacy” when used in a clinical trial for a vaccine, and “effectiveness” of a vaccine as its used once authorized or approved for the broader population. By collecting better data, we can ask more things of it, and understand things that we couldn’t have before. See above again for digital altruism.

There’s a tremendous amount of data noise in real-world measurements, but at the same time, the ‘non-sterile’ public health information we get from people as they go about their lives (in the real world, as opposed to within a controlled clinical trial) necessarily includes all of the uncertainty and variation that makes people, people. So, while some information may be hidden or overwhelmed by other factors of greater magnitude, we also learn a tremendous amount about:

• pharmacogenomic effects—though the individual genomics may be hidden);
• polypharmacy—other drugs and supplements being taken and not avowed in trials or at disclosed at clinician visits; and
• many other life style features.

This is and will be a deep treasure trove that we continue to mine. Unfortunately, most of what is being done now is what we call “noise mining,” and organizations are publishing spurious correlations and chance significances. Getting better at RWE requires that we solve these last two data science gaffes that occur every day and are guaranteed to have just occurred on someone’s RWE spreadsheet on the very day you’re reading this.

William Gibson once said, “The future is already here—it’s just not very evenly distributed.” Too true. Evolutions, especially in the business sense, always have the ‘bleeding edge,’ the ‘leading edge,’ and the ‘also-rans.’ 

References

1. Depending on what complexity thinking you’re doing, a trend is often defined as 3 or more points, or in control chart parlance, a trend requires 7 or more points.
2. This term was coined by Charles Seife for his eponynous book, Proofiness.
3. Itself a play on the Hippocratic Oath of clinicians, Primum non nocere (First, do no harm)

---

Ben Locwin
Contributing Editor, Healthcare Futurist, Pandemic Task Force Member

Ben Locwin has been working on development of state and national public health policy throughout the pandemic. He has frequently been called in to organizations to innovate healthcare and pharmaceutical approaches, and has been featured in various media to help the public, clinicians, and investors.