Systems Science in Practice
In the past decade, I’ve seen a growing awareness about systems science tools. But I’ve also noticed a strong bias toward the reductionist tools as “real” or rigorous science. This sentiment seems to get stronger the higher up an academic hierarchy you go, which is understandable given the success they’ve had working within the current system. Although many people in healthcare are now recognizing systems science concepts, they tend to treat it as a soft science or in practical terms, the “red-headed stepchild science” (I’m a red-headed stepchild by the way).
To try to show the value of Systems Science, I’d like to demonstrate its tools applied to the problem of National Football League (NFL) injuries. In this scenario, you’re passionate about NFL injuries and you want to try to lessen the incidence and severity of them. Maybe you have a friend or relative that played in the NFL who was injured, or maybe it’s just that your favorite player has suffered from a string of injuries. Whatever the motivation, you’re passionate about this problem, and you’ve been given unlimited resources by the NFL to try to improve the situation.
You have two options: you can apply the Reductionist Science Paradigm, or the Systems Science Paradigm. First, we can explore the Reductionist Paradigm option where you would need to come up with a hypothesis based on your observation of the problem and your knowledge of potential solutions.
Let’s say you’re especially passionate about concussions, having watched the movie Concussion over a dozen times. And because of this passion, you’re aware of a newly available helmet technology that potentially provides more stable cushioning. (Already, you can begin to see how personal bias and a single, primary investigator-perspective can influence reductionist study methods.) So, you hypothesize that the new helmet technology will lessen the incidence and severity of concussions by 30%.
Now, you must submit this research proposal to an Institutional Review Board (IRB) because this study method requires human subjects research protection. You will also need agreements from all NFL team owners and consent from all the players. This will take at least one entire season to complete and so the study will not begin until the next NFL season.
Prior to this second season, you randomize the teams so half use the standard helmets and the other half use the new helmet technology. You collect data and at the end of the year you find that, indeed, the new helmet technology has led to a one-third decrease in the incidence and severity of concussions. However, by this time, almost two years after the study was designed, there are now five new helmet technologies available and you have no idea if any of them are better than the helmet technology tested in your study.
On the other hand, applying the Systems Science Paradigm would not attempt to prove anything – there would be no hypothesis. The goal would be to measure outcomes by identifying and measuring factors that might impact those outcomes. Then, by gaining insight through the use of a variety of analytical tools, attempts would be made to improve the outcomes that you measured.
Because you’re not implementing any changes affecting the players, there is no need for an IRB submission, which means you can start collecting data immediately. To do this, you compile a small team of people who represent diverse perspectives on this issue, then you record the incidence and severity of all injuries, as well as all factors the team thinks are potentially important and may contribute to the incidence and severity of injuries. You might collect the pre-season training regimen of each team and their pre-game drills, the weather, if the stadium is indoor or outdoor, the type of helmet, quarter of the injury, the position of the player injured, etc.
At the end of the season, you would perform an analysis using analytical tools that generate weighted correlations and determine which factors, and combinations of factors, are most highly correlated with injuries. Based on this insight, highly correlated factors (that are potentially modifiable) could be addressed and changes could be implemented that would likely lessen the impact and severity of injuries in the next season.
Let’s say your analysis revealed that three factors were highly correlated with an increase in the incidence and severity of injuries – high and low extremes of temperature and AstroTurf (the artificial playing surface first used in the Houston Astrodome). With that knowledge, the AstroTurf could be replaced in all stadiums where it’s being used, and heating and cooling technologies could be developed to be used on the sidelines, in players’ uniforms, and possibly even on the field for the next season. You would then collect data to measure the impact of these improvement attempts in the next season, and the analysis of this data might demonstrate a decrease in the incidence and severity of all injuries.
This is a fairly simplistic example of the potential application of Systems Science in an attempt to improve the health of a sub-population of people (NFL players). The full application of Systems Science and non-linear analytics would allow for feedback loops during each season so any high-signal factors could be discovered well before a season ends. Also, the analysis of different sub-populations (like linemen vs. quarterbacks) could generate a variety of ideas for improvement of injuries based on those unique, position-specific analyses.
Which scientific paradigm results in more knowledge faster and is less costly to apply? What if the healthcare industry applied systems science tools to all patients with all types of diseases with the goal of measuring and improving the value of care we provide in our global healthcare system? How quickly could we lower costs and improve outcomes at the same time?
The kicker in this example is that this Systems Science project to attempt to improve the incidence and severity of NFL injuries was my 7th grade science fair project in 1975. To be honest, although I did have a passion for football, my true motivation was to be able to tell my mom that I had to watch every minute of every televised NFL football game that season because it was my science fair project, even if the Monday night game of the week went past my bedtime.
Through Systems Science, we can achieve a global healthcare system where costs go down and outcomes improve over time, and value-based innovation and improvement can become the norm in our society. It’s time to take a bold, courageous (yes – uncomfortable) step into the unknown, uncertainty, (and yes) the real world. A newer scientific paradigm discovered over 100 years ago is just waiting for us.