I came across an advertisement recruiting patients for a Post-Surgical Nerve Pain focused clinical trial. Posted December 4th, 2023. That date will be relevant shortly.
[Case Study] What Not To Do: A Warning From The Frontlines Of The AI Revolution To Clinical Trial Recruitment Marketers Utilizing AI When Generating Collateral Creative Part 1 of ?
Would this thumbnail image instill comfort or confidence in a patient looking to participate in a study for those experiencing post-surgical neuropathy trial? Does it seek to reflect a presumed patient personas physical state? (i.e.: discomfort, limited mobility, and potential diminished quality of life)
Before you answer consider that by selecting "view" to see more details the inquiring patient will be greeted with a larger more detailed version of the same image:
Or is it a twisted and mocking caricature of a presumed and thus projected lived experience by a team of individuals who through a confluence of events lack the tools to analyze, or experience necessary, to develop and release patient-centered and appropriate creative capable of wielding visual language suitable for this use-case?
That equipment doesn't appear to share a continuity with our world. "Artistic license" as a justification for "inspired" medical equipment design might have worked in the pages of comics in the 1930s before most people were aware of cutting edge tech. But not today.
This poor woman looks like she's beneath jerry-rigged jelly fish like lights hanging from what could be generously described as IV bag hangers, but mixed with a gun rack? The hardware bars jutting in to her back can't be comfortable. She's wearing surgical scrubs like Dr. Frankenstein might wear, not a patients gown... and she appears to be laying on a morgue table with a makeshift pillow made out of cheap foam. And the headgear? Looks like, incidentally, some kind of 1930s inspired retrofuture Universal Pictures Bride of Frankenstein-like hair dryer.
This image doesn't instill confidence in me. I'd love to hear your thoughts on it.
Switching gears. This isn't an AI fail tear down. At least, not completely. I have generative AI alternatives which are incidentally epoch appropriate. I experimented with creating patient-facing recruitment geared images in MidJourney v5.2 in mid-November 2023. Makes for an easy 1:1 comparison.
His eye need a slight adjustment. Since they used raw unedited output, I too will use raw unedited output for comparison. Just know I noticed it.
If desired, other styles are possible. Generated around the same time as the above images, these would also be appropriate for a "Post-Surgical Nerve Pain Clinical Trial" ad.
The good news is there are great ways to implement these tools that respect patients and what they're going through. In experienced hands it's as easy as any generative AI tool. And like any tool it's never the tool itself: it's the person wielding it.
Everything here is raw output. Clean up for these images would be easy, and improvements are made to generative AI every day making touchups less and less necessary. I'm keeping on top of the news and sharpening my skills. Every day is the first day of the brave new world unfolding before us.
All to say: Speaking as an artist, a designer, a clinical research analyst with a decade in direct facing patient centered care, and a trial recruiter with 5 years of industry experience:
I know the way to industry defining prompting.
Reach out if you're ready for generative creative
leagues ahead of the competition.
You won't be disappointed.
To close-- let's check in on where everything is in the fast moving generative AI space: Every image above was generated four months ago, in the previous epoch MidJourney v5.2. As of posting tonight I have an early access build of v6.0 alpha only for the 1K club released February 9th, 2024. Adjustments are being made daily.
Below is an example of v6.0 alpha's pre-public build generated tonight. We are here now:
PS: This is that sites image for a "cold sores clinical trial". This is rabies at worst, and upchucking chia seeds at best. Completely unacceptable, it's bound for a prolonged recruitment phase through abysmal conversion and abysmal ROI.
Comments