Medicine is always pushing forward to find new ways to help patients. And while emergency rooms and surgery and even regular check-ups are all wonderful as the first line of defense, what if medicine could take the offensive?

What if healthcare could find a way to help patients before they get sick? To help reduce common risk factors, and to better predict the medical future of any given individual or community. These predictions, based on piles of collected data, could create a better future for patients and a stronger strategy for how healthcare engages with the community.

This field of study is called “population health,” and it’s one of the more recent and exciting trends in both data collection, machine learning, and healthcare. It can involve not only cutting edge AI, but population studies, the census, medical computers, EMR, and political action.

But what is it?

What is Population Health?

The cleanest accepted definition of population health comes from a report published in the American Journal of Public Health in March of 2003, written by David Kindig and Greg Stoddart. They condensed the idea down into this:

Population health is “the health outcomes of a group of individuals, including the distribution of such outcomes within the group.”

In layman’s terms, it really combines a few ideas and concepts into a larger whole. One aspect of population health is “social determinants of health,” which is the idea of looking past individual health risks like “high blood pressure” or “smoking” and into the economic, cultural, and political landscape that affects different regions and demographics. Things like the unemployment rate, access to transportation, social support from family groups and political policies, and even down to socio-economic stress and how all of these things affect the health of those affected by them.

The idea of population health is to not treat the symptom but to instead look at the root.

But how do we find this information? And, better yet, how do we deploy it to improve the community, patients, and the day-to-day running of a hospital?

Collecting and Processing Population Health Data

Where does this data come from? How can hospitals collect it, process it, and use the data to improve their own policies and even reach the community?

Collecting the Data

Much of the data can come from electronic medical records (or electronic health records), which have become a ubiquitous (if contested) part of everyday life for clinicians. Though the constant paperwork and record-keeping can be understandably tedious for doctors and nurses, the incredible upside of such a detailed and granular patient record is becoming a reality.

With computers on wheels and medical tablets becoming ubiquitous tools in the field of medicine, all patient data can be entered from anywhere in the hospital, stored safely (maybe in secure blockchains very soon), with the relevant demographic data being pulled out and used (divorced from the patient’s identity, of course) to help the entire population.

Analyzing the Data  

All data from one community, economic level, culture, or race can be organized together, with trends being identified, graphed and examined for potential healthcare red flags and solutions.

However, consider this: if every hospital in every modernized country is collecting all of this data, how can it possibly be organized, much less examined? It’d be like trying to read every book ever printed by laying them all out in a field and standing on a ladder with a pair of binoculars. So, does all this patient data become useless fluff, designed to rot in that shiny new medical computer?

Absolutely not! Thankfully, machine learning and AI has come a long way, and can read all of these patient records, pull out vital details, and spot trends, commonalities, and even dismiss outliers or non-correlative or causative data.

Not only has machine learning been proven to advance the cause of population health, it’s actually being deployed already.

How Healthcare Can Use Population Health to Help Their Patients

In 2016, Indiana battled an opioid epidemic by using artificial intelligence to analyze and predict opioid usage and risks across the state through the collection and analysis of patient EMR, lab results, medical history, and government data silos.

The healthcare groups and facilities that have found the most success in population health management have done so through a combination of a strong medical computer / EMR strategy (to collect the data), a partnership with data labs with analyze the data with machine learning, and a three-pronged approach to implementing the changes to help the community.

The Three Heads of Population Health management

For one, it’s important to remember that hospitals and healthcare groups must think locally, even if they’re examining data at the global scale. The entire point of population health is to find what a community needs to improve health outcomes. Once data has been collected, each individual hospital would then recognize the particular social determinants of health causing problems in their area and develop a personalized strategy on how to fight it.

Secondly, they must think of the bigger picture for the community and commit to putting those solutions into effect.

If socio-economic levels are the primary concern, consider something like a childcare program for the hospital, or building a community playground. Parks in low-income areas tend to be rife with drug use and violent crime, which obviously affects health outcomes pretty significantly. A safe play area could do wonders for local children. A childcare program, on the other hand, may allow working fathers and mothers more room in their schedule for better jobs or longer hours, increasing their pay and thus their access to healthcare and things like more nutritious food and even safer automobiles.

Thirdly, healthcare groups can help mobilize the community into political action, or take action themselves if they have the resources. As any doctor will tell you, healthcare is beholden to laws and policies more than anything else, so the most real chance of change comes from the top down. This is obviously the most difficult channel of change, but it’s also the one that could create the most progress in local health.

If the data analysis of population health has made it clear that some law, regulation, or lack thereof is causing an inordinate amount of harm in the medical, political, or economic status of the community, the only real way to change it is through political action.  

The Future of Population Health

The idea of population health has been around for a while, but the EMR integration, medical computer market penetration, and artificial intelligence to actually make it a reality are all recent developments.

While population health management will require no small amount of investment in both time and money, it may, in fact, be the most effective medicine we’ve ever had for helping the most amount of patients at once.

To learn more about how EMR, medical computers, and population health all work in tandem to help the community, and to learn more strategies for deployment, contact Cybernet today.