Do you know what Donald Trump and Hilary Clinton have in common? They both used over 150,000 data points per census block (neighborhood) to try to get your vote.
Political campaigns have been collecting data about you for years, even decades. They have data on over 11 million neighborhoods across the country, so they know how people in your neighborhood think, act and earn. Neighborhoods are remarkably homogenous. We tend to live near people that look like we do, think like we do, are in similar stages of life and have similar education and employment. Therefore, knowing about your neighborhood means they know a lot about you.
The campaigns tap into data about how you prefer to communicate, how you make decisions, how you earn and spend your money, etc. Those billions of data points are just sitting there, waiting for someone to try to get your vote.
It is imperative that we begin to leverage that data for healthcare. These data sets can be used to determine who is likely to successfully engage in telehealth, contact their physician early (or late) about a problem managing their chronic disease, who is likely to have a transportation issue, and many more socioeconomic and environmental determinants of care.
Geospatial analytics gives us the ability to look at large populations, smaller communities and even individual risk factors. In my opinion, block level socioeconomic and environmental data is the most under-utilized set of existing data in healthcare. Yes, the data exists. Health systems don’t need to go out and collect the data. Someone else has already collected over 150,000 data points. Clearly, not all 150,000-plus data points apply to healthcare, and some data points of interest haven’t been collected. But, there is a treasure trove of health-care relevant data locked up in those data sets.
“Utilizing the data regarding socio-economic, environmental, and attitudinal factors we have access to today will take creativity and an agile mindset, but it is imperative that we begin using it”
Imagine if you could determine who to send a text reminder to increase medication compliance and who would need a call. Imagine that you could know if your patient with multiple chronic diseases was likely to contact their physician early about a complication, or who will wait until the last minute. You can. The data already exists. We just aren’t using it.
At some point in the future, geospatial data will be viewed like antibiotics, Physicians and health systems won’t understand how we treated patients without it.I predict that a clear understanding of environmental and social factors will have as much as (if not more) an effect on healthcare as clear understanding of the genome.
There is no doubt that in the immediate future existing data sets for socioeconomic and environmental determinants of care provide greater potential for improved health than genetic analysis. The reason is two-fold. Big data regarding patients’ socioeconomic and environmental information is available today. A clear understanding of the genome will likely take more than a decade.
The second reason is that socioeconomic and environmental determinants of care are more modifiable than a person’s genetics. The difference between a compliant diabetic and a non-compliant diabetic may be as simple as arranging reliable transportation. There is nothing simple about changing a person’s genetics.
Let’s compare the impact of one of the major genetic discoveries impacting clinical care, BRCA1 and BRCA 2 mutations, with existing socioeconomic and environmental data. BRCA1 and BRCA2 mutations are associated with increased damage to DNA, and thus increased risk of breast and ovarian cancer. Women with a harmful BRCA 1 mutation have a 55-65 percent chance of developing breast cancer by age 70. But, they only account for 5-10 percent of all breast cancers.
BRCA1 and BRCA2 are a major genetic discovery, but it does not play a role in the treatment and diagnosis of over 90 percent of patients with breast cancer. In contrast, all breast cancer patients need to get to their clinical visits, maintain good nutrition, cope with treatment side effects and develop a support network.
Geospatial analytics can help providers treat all their patients. Data on access to public transportation and car ownership can provide early indicators to flag patients likely to have difficulty getting to appointments. Shopping habits and access to grocery stores can give insights into how challenging it will be for a patient to follow dietary recommendations. Attitudes toward communication with physicians will reveal which patients are likely to call before a complication or side effect requires an emergency department visit or hospitalization. Data on multi-generational living and proximity to relatives can indicate which patients need help developing a network for support.
The use of these data sets for clinical care is novel and those engaging early will need to be flexible and cautious, but they will have a competitive edge and their patients will receive better andmore holistic care. Providers will be able to proactively engage with individual patients to address the barriers to improving their health, health systems can work with non-profits and governments to improve infrastructure and programs to improve the health of communities, and better supported patients with better informed providers will lead to smart spending.
Utilizing the data regarding socio-economic, environmental, and attitudinal factors we have access to today will take creativity and an agile mindset, but it is imperative that we begin using it. Waiting for a magic bullet hurts all of us.