Research firm inVibe surveyed 150 US physicians across 12 specialties to find out how they approach different "digital" channels. We put that in quotes because "healthcare industry sources" isn't by definition digital, but perhaps the question limited the answers.

The group published the ratings HCPs gave to four topics so we're looking at that data.

AI may be upending how we approach work, but humans change slowly. The physicians in the study rated Google as the top channel for frequency. AI made a strong showing mid-pack considering how new the channel is.

Frequency of Use

When asked about the value of the different channels the ranking didn't change. This isn't very surprising as HCPs will use the higher value channels more often. In fact, you could consider both frequency and trust as indicators of value.

Here, we see the most trusted source float to the top: Industry Sources like associations etc. The interesting finding here though isn't that HCPs trust industry over pharma companies, it's that AI is now as trusted as Google.

Ad Comfort is a slippery concept. We interpret this number as how comfortable HCPs are with seeing ads on the platform. To a lesser extent it seems to also reflect how comfortable they with other HCPs and even patients seeing ads on the platform(s).

Visually it looks like ~73% of HCPs use ChatGPT for professional purposes and ~49% to stay current with their specialty. It is by far the market leader.

There are also insights from 50 interviews with selected HCPs so the report has more insight than just the raw numbers. 

The Kaiser Family Foundation conducted a study about AI and health in the US that closed March 2 and the report is already out. 

The poster child for AI in health is mental health services but physical health conversations are conducted by almost 2x Americans.

Physical health has relatively consistent usage across age groups which is surprising because technology usage tends to diminish more aggressively with higher age groups.

Americans' use of AI for health can be categorized into three rough types:

With physical ailments outnumbering mental health conditions it is no surprise that these numbers fall out this way. 

92% of Americans were at least somewhat satisfied with the answers they got from AI about their queries. Mental health was mostly positive as well at 85%, but had larger groups at both ends of the spectrum.

With satisfaction numbers like that it's only a matter of time before larger numbers of Americans use AI to help with their health.

Talk about corroborating evidence. Rock Health surveyed 8,000 Americans and found out that 32% are using AI chatbots for health, the exact same number the KFF found. 

One-third of the US public is using AI for health, and this number is twice from the year before. The data was collected in December so while it says 2026 it is still fresh.

The product driving this use is ChatGPT but it doesn't own the entire market. Gemini has share, and Google AI Overviews are not included but they are likely much higher than ChatGPT.

This journey chart shows how AI users approach their patient journey. They likely over-index for digital behaviors like searching for information.

AI Chatbot users are quickly bringing these advisors into their activities and here we see what they do after using the chatbot. We see sometimes they take direct action, but often they want to use the Chatbot information as inputs to other discussions.

A new study by Google on it's AIME medical chatbot produced zero safety issues in 100 patient interactions and performed comparable to, but not as well as, PCPs.

The researchers used a care quality survey tool called, in full, the UK's General Medical Council Patient Questionnaire (GMCPQ) to measure the AI's subjective performance with the patients. For this chart we made "neutral" black to separate positive and negative. Like other studies, AI is marked high for bedside manner.

See the full study for all the results including the comparison of treatment recommendations compared to human physicians. 

While researching this issue of the Klick Wire the author found a Google search result dated in the future. 

Also, we can see the results are sponsored. As of writing on Friday there is no public access to the webpage but Google Search AI Mode seems to have the indexed data. Asking for the most interesting stats gives us:

So, was this an accidental pre-release of timed content on Google's AI Overviews? Or is it an echo of an old story? Extended conversations with Google's AI Mode seemed to start pulling in external sources and contradicting previous stats so we left it as-is.

We look forward to see if the linked Fierce sponsored article exists and if the stats above are in it, or if they were just hallucinated.

Zappi surveyed 1,000 presumably US-based women to see how they feel about female portrayal in ads. The results weren't great.

This question was "Do you think brands are genuinely investing in advertising that reflects and speaks to women like you, or is it mostly performative?"

51% of the 55-75 respondents said performative along with 49% of those 35-54.

Sometimes women won't or can't avoid a brand completely but 45% say they will avoid or spend less on brands they think are failing at connecting with women like them.

This data takes a bit of thinking to wrap your mind around. The sum of the values is 99% (because of rounding we assume) which means it's not percentage of that age group, the age group is a percentage of the total. 

So, 23% of respondents said they were not represented in ads and were in their 50's. This likely means the 60's number is likely due to recruitment bias as the numbers don't seem to be weighted by the US population.

Regardless, 89% feel underrepresented.

For pharma marketers who often deal with the "health gatekeeper" in the family this data is a reminder to be authentic with your representations.

A new study from Canva was conducted by Harris Poll on 1,415 marketing leaders at firms with 500+ employees. They also surveyed 3,547 consumers. The studies covered 7 global countries.

This isn't surprising but the number of marketing professionals are now routinely using AI. To differentiate you need to be looking ahead at agents and other technologies.

Younger consumers are more likely to be ok with useful AI. The gap widens when talking about content vs. targeting.

Targeting is the most accepted use of AI, but consumers have a hair trigger when it comes to "creepy" vs. "ok." Some of the more accepted uses of AI targeting:

The report is presented as an lightly interactive slideshow rather than a PDF so if you're doing a desk-lunch it's worth a flip-through. There are videos so check your audio.

The FTC has announced a new Healthcare Task Force. This group looks like it will be focusing mostly on providers and services but pharma marketers need to pay attention because of potential crossover with the Agency's scrutiny of third-party cookies.

Some elements of the announcement to keep an eye on:

Pharma marketers are already aware of issues surrounding third-party pixels and health condition data so this is a wait-and-see alert for now.

IQVIA has a report out on 2026 global R&D trends in life sciences for the relatively reasonable price of $250 USD. The teaser data on the announcement highlights some global data that is interesting in its own right.

China is on an R&D tear. The country has come from last place in 2011-15 to matching the US in 2021-25.

Looking at the distribution of where clinical trials are conducted, China has almost doubled while Europe had fallen. 

Visually inspecting these charts, we see that Obesity is the only condition category that showed growth in 2025. It's not surprising that it grew, but it's perhaps surprising that it was alone.

For those looking for a global perspective this report might be worth the cost (not a guarantee, the Author hasn't read it -Ed.).

A study published in JAMA looks at what qualities would convince patients to choose an AI-assisted diagnosis. The researchers surveyed 3,000 Americans to get their results. Perhaps unsurprisingly, quality of diagnosis was, by far, the top driver of AI choice.

This study assumes perfect trust in the information about the AI. That works in the study as it is a hypothetical, but would break down in the real world where trust in the information would precede trust in the AI.

The results came out in the open text answers as well. Looking at frequency of concepts the researchers found that performance of the AI was also the top concept.

Data Axle surveyed 450 B2B buyers of varying ages to tease out the differences by generation. Surprisingly, they found that the top preferred channel across all generations was email. Add to that is that weekly emails are preferred by majorities of each generation.

The full data is a bit of an eyechart so interested readers may want to download the PDF. The fun insight here though is that for three of the four generations, email is also the most annoying channel. Showing you need to do it right when you do it.

This issue of the World Happiness Report from Gallup is devoted to the link between social media use and lower life satisfaction, especially for young people. Coming on the heels of the California verdict of guilty for Meta and YouTube it is particularly timely.

The 272-page report is a collection of well-researched reports all on the linkage of digital activities to lower life satisfaction in most regions (Brazil is an exception). The results highlight the nuanced issue with low amounts of internet use being positive but high levels negative.

All of the media attention is on social media because there are a few people who can be held accountable but fail to notice the issue with seemingly harmless "browsing for fun." This highlights that it is a screen issue as well as a social one.

While the full report is focused on the evils of screen time, another email from Gallup highlights the good news in happiness around the world. The top 10 happiest countries have the Nordics as usual, and some other European nations with two outliers.