lemonade cartoon 20 oct 2011 personal dataI have been peripherally following the story regarding the use of metadata by the NSA (also known as “No Such Agency”) to try to predict terror plots, TEA party uprisings, and other activities of national concern. The concern is somewhat belated as the New York Times reported that the effective use of metadata in predicting behaviors has occurred for over a decade in the retail sector. In addition, if you read the article, you can see that metadata, at least in one instance, is more effective than a father’s hunch in identifying a daughter’s nascent pregnancy.

People willingly give away their metadata every day when they use a credit card to shop or plug in a “courtesy” number at the supermarket. In exchange for the loss of privacy, I understand, people are paid with personalized coupons. This, I guess, is so they can be further tracked. Given the controversy I surmise people will give up privacy for a buck off a bag of dog food (coupon would be wasted on cat lovers) but not to avoid having their airplane run into a building. Go figure.

Use of metadata has a long use in medicine but the power to do good is only now becoming clear. The first well known use of metadata to improve health was in 1854 London when John Snow figured out, through tracking cholera cases on a map, that a well on Broad Street was the common connection and he took it upon himself to remove the pump handle. Though relieved, I suppose, that the cholera outbreak subsided the handle was replaced by the government soon after. Not captured but certainly the reason for the return of the pump handle were the people whining about having to walk a couple of blocks for water and not believing that something unseen could cause diarrhea.

In 1935 William Pickles published an article entitled Epidemiology in Country Practice. A general practitioner who spent most of his carrier in Wensleydale, he had keen observational skills, organized thought processes, and was able to make connections from seemingly  disparate data. He was able to deduce from his clinical observations the incubation period of many common viruses, the degree of infectivity, and the mortality rate. I can’t help but wonder if the people of his hamlet were reassured or freaked out when he pointed to them and said “You’re next.”

We collect a lot of data in medicine today. Every transaction is coded and sent to a computer where, in addition to generating a check for the health care provider, the data is analyzed and trends are documented. From this metadata we have identified core activities that every hospital needs to do to improve outcomes (called the Core Measures) and are now classifying hospitals so consumers can make decisions by comparing data on a hospital to hospital bases (called, in an uncharacteristic lack of government doublespeak, Hospital Compare.gov). Soon doctors will have their own doctor compare website as well.

Doctors and hospitals are taking this about as well as the general public. I suppose no one wants to be seen as falling outside the metadata envelope. On the other hand, just as you don’t want to go to an event that has been “targeted” by the randomness of the universe if you can help it, you probably want to go to a hospital that falls on the good side of the metadata as well. We have it, might as well use it. At least for medicine.