Friday, a client asked me about privacy, data security and social networking in clinical trials.

The client told me that they were considering using a physicians‘ portal to market their digital therapeutic product after they completed their pivotal clinical trial.  (See my article on 4 strategies to get FDA approval for your medical device ).   

The client CEO was concerned about patient privacy issues.   Is social media a good fit for marketing drugs and devices?  Is it legal everywhere?   How can clinical data obtained from patients in decentralized clinical trials be used for recruiting new patients?   How can the clinical data be used for marketing decentralized trials of drugs and devices?

What are the implications of social networking with clinical data?

I started looking at physician portals and came across a fascinating post from Dr. Scott Shreve – Knowledge Prostitution enabling Aggregated Voyeurism: Is this a Business Model?

Voyeurism (voi-yûr’ ĭz‘əm) n.

1. The practice in which an individual derives pleasure from surreptitiously observing people.
2. Derives from the French verb voir (to see); literal translation is “seer” but with pejorative connotations.

Is social media sharing appropriate for collecting real-world clinical data?

Sermo is an example of a closed community for doctors.  Sermo says on the home page:

Access the medical community that empowers today’s physicians

The business model used by closed, advertising-free, doctors portals (Sermo.com in the US or Konsylium24.pl in Poland) is based on paying for market intelligence data collected from the “user generated content” in the community.  

The assumption is that physicians will talk freely inside a gated, advertising-free community.

Using voyeurism to monetize clinical data 

Sermo.com kicks back some of their advertising revenue to the physician users. The precision and recall of this market intelligence is not clear to me.

Consider the amount of noise in open social media like Facebook, Twitter and LinkedIn.

How do closed social networks assure precise and relevant content?What is clear to me – is that there are data security and privacy implications when the community operator data-mines user-generated content for profit. 

As a concrete example – a thread on Konsylium24.pl went something like this:

Doctor Number 1:
You know – Professor X is the KOL (key opinion leader) for company Y’s drug Z.  He says that drug Z is extremely effective for treating the indications of infectious disease Alpha.
Doctor Number 2:
Of course – Professor X is an acknowledged expert on infectious diseases, but he is also an expert on cash and knows how to do the math and add up the numbers…

I asked my client – “and for this kind of data, your parents sent you to medical school?

Collect clinical data after informed consent. You don’t need inference engines.

This took me back to the days of Firefly, Alexa, Hotbar and use of personal information as currency – collected with “collaborative filtering” and “automated inference” from people browsing the web.

Web 2.0 and social media seems to be going through a similar evolution as Web 1.0 – trying to monetize clinical data by  data aggregation and analysis using “collaborative filtering” techniques.  

This may have been a sexy looking business model for Venture Capitalists during the dot.com era.

5 years after Sermo.com launched, they had a well-publicized breakup with the AMA.

How can automated inference, knowledge prostitution and aggregated voyeurism replace careful collection