Ghostwriting is a topic that plagues the pharmaceutical and biotech industries. See my previous post entitled “Ghosts in the Machine” (http://wekfiles.blogspot.com/2007/10/ghosts-in-machine.html). The ghost writing issue has been debated to death without firm resolution. I promised in a previous post that I had a solution to help with the ghost writing dilemma…well here is one solution. Sorry medical writing firms and publication planners, though many of you are good friends, I don’t believe a long-term solution can be developed with multiple competing interests.
In my opinion the issue really boils down to people. People can be influenced very easily, especially if there is $$$ in the mix. In contrast systems cannot.
I’ve spoken to many publication planners and medical writers about this very topic. Full disclosure, everyone I’ve spoken to, thinks its nuts and there is no way an automated system could ever replace a human being in such a complex set of processes. I think Larry Ellison said it best, “When you innovate, you've got to be prepared for everyone telling you you're nuts.” I’ll also cite the radiation technology industry. 10 years ago, aside from the visionaries no one thought that XRays could be read by a machine….well today they are. Not only are they screens read, they are analyzed and emailed to the physicians who can tend to patients while the processing is occurring. Nuts? Maybe, but who else is going to dream this stuff up?
I’d hope Larry Ellison would endorse this one…
There are volumes of data and warehouses of publicly available data on the Internet (e.g. PubMed.com, ClinicalTrials.gov). What medical (ghost) writers will typically do in a nutshell is troll the online data stores, confirm citations, garner permissions from publishers, create an outline, and draft the paper, format it to guidelines - rinse, repeat. This process of back and forth typically takes 30 weeks.
Often times not only are the data freely available but are also consumable by XML technologies. Enter Web 2.0. These XML technologies are commonly used to distribute or get information from disparate sources (e.g. Froogle, weather and news updates). Imagine a tool that pulls information from online medical repositories, based on parameters (search criteria) a user enters through a website. The website in the backend compiles online data intelligently, rendering at the very least an automatically generated set of references and perhaps an outline for the publication. Let’s take it one step further and give the application the ability generate data that is deemed regulatory “safe” and within online and traditional publisher guidelines. Futuristic? Yes. Crazy ? - You decide, but the technology exists today to get what I just outlined done. Of course there are caveats and much alignment needs to be performed to make something like this happen…but wouldn’t information that is disseminated more quickly, more transparently and more completely from sources that cannot be influenced help an industry that is accused of hiring professional writing talent to spin a drug’s messages in the medical literature? I think so.
Who loses? Well of course the ghostwriters. But do they have to lose? Couldn’t they change their business to adapt? Publication planners don’t want to become the Polaroid of the pharmaceutical industry. Take a lesson from Kodak. Moving from traditional writing role to service oriented architecture may give medical writing firms the nimbleness to change with technology as opposed to change with the law.