Medical Intelligence

From where I sit, I see little that connects the intelligence community (CSIS, CIA, etc.) with the medical practice of my doctor. But a recent perspective piece in the New England Journal of Medicine suggests it might be beneficial for the medical world to explore how the intelligence world is handling information.

We are currently living in a world of big data. In the medical arena this big data can be unsettling. For example, data reveals a large geographic variation across Canada in the diagnosis and treatment of various conditions that cannot be explained by simple patient clinical need. Knee replacements in 2011, for instance, range across Canadian areas from 70 to 397 per 100,000 population – a variance likely larger than variance in the regional health of Canadians’ knees! Other medical procedures show similarly large ranges across regions of the country. These unsettling variances suggest that we need to look more closely at how medicine is practiced to ensure all people have equal access.

BIG DATA PROBLEM
One of the very real concerns in exploring and using this big data to improve health care is maintaining the privacy of medical information. None of us would like to lose control of our personal medical information. Thus, any use of medical data must ensure that the identity and privacy of the information is protected from unauthorized access. 

Another challenge to mining and analyzing this big data is the varied nature of medical information – from the genetic variation in DNA shown by individuals, to the toxins sometimes found in work environments – all of which have an impact on our health. The data is found in medical images like digital x-rays, doctors’ notes, medical test results and insurance companies’ databases. But the ability to explore and mine this complex data has the potential to determine best practice. For example, today a person suffering from depression often needs to take a drug for multiple weeks before effectiveness can be determined; better mining of medical data might help that person’s medical team to know in advance which particular drug to prescribe for greatest effectiveness. 

HEALTHIER HUMANS
In the intelligence community there are similar needs both to maintain the security of information and to integrate information across different agencies and sources. One of the conclusions of the United States’ 9/11 Commission Report, published in July 2004, was that if data from various agencies had been combined and analyzed in a timely manner, the 9/11 tragedy might have been averted. Accordingly, the US has been developing solutions for working with highly sensitive data of various types from multiple agencies. These solutions include a method of classifying data security at the individual data point (rather than at the data system level), such that information access can be granted on the basis of specific need. This method involves labelling data automatically with multiple tags using different taxonomies so that all agencies can explore the same data. There is now a “data lake,” in which all data from multiple streams is securely stored and available to authorized users with little prior data modeling. Many of these innovations in data management, like Accumulo, are available as open source software.

Although our medical system needs and uses information very different from that of the intelligence community, both are public service domains where data analysis serves the public good. It is exciting and hopeful to imagine that solutions being developed in the intelligence community might be useful in the health sector – with the ultimate goal of making us all healthier.

As Christians, we can avidly support innovations that enable us to better protect, respect and care for God’s image bearers. We should pray that these data tools might become more widely used for our individual and group health and that these powerful tools not be abused. We should also pray for wisdom for these system’s designers that they may balance the common good and the need to protect privacy.