Proprietary Information

This morning I learned that the American Meteorological Society was concluding its annual meeting today. This was from an NPR story I listened to on my way into work. During the story an attendee from Great Britain was interviewed. He was asked how the climate change community might gain back some credibility after the embarrassing "Climategate" email leak. His primary solution was a proposal to make all the data surrounding the leaked emails available to everyone. What a concept! But alas, the reporter went on to state how this would not be possible since the data concerned is proprietary. What? How in the world can data collected from the planet's weather patterns be proprietary? (Don't comment trying to tell me the answer to that question either)

This is not intended to be a post about climate change. I mention that story to illustrate the fact that the argument against release of data is eerily similar to that which the pharmaceutical industry uses to hide their data too, "It's proprietary." Nonsense! Again, how can data collected in clinical trials of humans be proprietary? It's not. Regrettably, legally speaking it has become so. This is all too familiar a refrain from big institutions and gives the appearance that there is something to hide. Within recent memory there have been pharmaceutical companies rebuffed for wrongly representing data about a product.

Like open source projects enlisting the help of millions of programmers world wide, data which holds the potential to alter lives as significantly as drugs people take, which could endanger their lives, should be wide open for review like open source code. There are literally millions of physicians, scientists and other health care providers whose eyes should be reviewing the data of pharmaceutical and medical device companies. Basically, the data needs to be crowdsourced.

Original data is not always included in studies. A lot of the data seen in the "peer reviewed" journals has already been washed by the statistical analysis deemed appropriated by the researchers. One of the most damning issues of studies is the lack of the appropriate statistical "power." In other words, the data pool is too small to give an accurate representation. The lower the likelihood of something occurring generally means a larger pool of study subjects is needed in order to make accurate statements. Often, power analysis is either thrown by the wayside or not mentioned by the researchers at all. One day, current technology will come to health care and we will be able to collect massive amounts of data and it will all be available for review, to everyone. At least that is my dream. Then no one can cry "proprietary" in order to protect their market share. The company or companies that get this first will be on a trajectory far above all their competitors.