Biosurveillance refers to automatic monitoring of public health trends and unusual occurrences. It's a supplement to traditional public health surveillance and relies on preexisting, real-time health data: data that is already out there and doesn't require a lot of manpower to obtain. Biosurveillance is a systematic monitoring of what's happening in public health right now.

Biosurveillance involves a series of methods of detecting and verifying signals before an outcome is confirmed. For opioid biosurveillance, the goal is to look for signals of abuse and misuse before official confirmation takes place; like a type of 'early warning.' 

Q:  A lot of your research has been focused on outbreaks of influenza and how to help epidemiologists monitor and learn from the study of these outbreaks.  How is the tracking of prescription opioid abuse similar or different than monitoring influenza and infectious diseases?

There is an interesting commonality between opioid and infectious disease surveillance. In both cases there was a reaction to a signal event that led to a lot of public interest, for example; surveillance of bioterrorism (and in fact all infectious disease outbreaks) garnered attention after the highly publicized Anthrax outbreak. Concerns about OxyContin and the prescription pain abuse epidemic that developed also brought the need for opioid biosurveillance to the forefront of the public mind.

My work with influenza tries to understand how flu spreads in space and how epidemics take place in time. In terms of opioid abuse, we also look at geographic patterns and how use, misuse, and abuse patterns spread. Just like influenza, social networks play an important role and define high-risk areas. We look for signals of increased opioid abuse in particular areas and use classic sentinels to gather information.  We are especially interested in data sources that provide early indications, such as treatment centers and media monitoring.

Q:  Prescription opioid abuse is quite prevalent around the country, how can you tell the difference between the low level of abuse that is always there and a true outbreak?

It's important to understand how to scientifically validate the signals that we get. You want a system of data collection that's sensitive to change, but highly sensitive systems can be very 'noisy.' Too much noise leads to false positives and makes the system useless. Good biosurveillance always tries to reduce noise and respond to verified signals. Comparing data sources and collecting from multiple sources is a good way to provide cross validation.