"Solutions Epidemiology"

About one decade ago, the concept of "consequential epidemiology" was introduced, but it seemed a bit gimmicky, and perhaps a bit of special pleading about a field possibly in need of a morale boost.

Thinking about our currently Balkanized epidemiology, I am re-visiting the idea with a different term.

I call it "Solutions Epidemiology," and with this term I invoke a return to the roots of epidemiology as a disciplined speeded-up way to achieve near-term solutions to pressing public health problems.

Other bullet points:

1. A deliberate choice of health problems that now qualify as low-hanging fruit, by which I mean defined problems that can be solved in the near-term, and that might qualify either as epidemics or as the equivalent in localized settings: e.g., hospital-acquired MRSA infections; unintended pregnancies; high death rates if near-term or full-term infants.

2. A deliberate reach from the more languorous flow of research and resulting evidence in chronic disease epidemiology in the direction of an accelerated flow of evidence toward implementation science issues faced in public health departments year by year, as opposed to decaf by decade.

3. A deliberate alliance of academic epidemiology researchers with public health department officials charged with solving problems in real time.

4. A deliberate abandonment of Balkanized epidemiology that organizes itself in terms of domains of the explanation or 'scale' of its variables of interest (e.g., genetic epidemiology, social epidemiology, molecular epidemiology) in the direction of a more holistic epidemiology that organizes itself with a focus on a population health problem that requires a near-term solution.

5. Leaving in the hands of others all epidemiological problems that require a languorous approach and cannot be accelerated more rapidly toward actin research with near-term yield.

6. Differentiation from allied endeavors such as "implementation science" by virtue of a clear point of departure in terms of definition of specified populations and keeping track of the sequence epidemiologists use (1) to ensure at least a limited but 'built-in' external generalizability of our work, and (2) to study the sick AND the well in each population with due attention to bias in estimates when sampling frames or achieved samples depart from pre-specified populations for our studies.

I could go on and on, but my intent is simply to plant the seed of an idea for "Solutions Epidemiolgy" and to elicit some comments to guide its future development.

We have an opportunity to show Solutions Epidemiology at work in relation to the current heroin epidemic in US counties. There are counties with zero or few heroin overdose events, fatal and non-fatal. Other counties are suffering. What differentiates them? Isn't that an important epidemiological question that can be answered quite speedily in the direction of action research to see what can be done (e.g., via randomized trials) to keep the low rate counties at steady state low rates? And wouldn't it be important to learn that evidence sooner rather than later?

Operations research traditions come to mind, along the lines that Deming devised to find flaws in manufacturing processes. Surely, the work of Deming and other operations researchers deserves attention in the education of 'solutions epidemiologists' of the future. At present, I see few epidemiologists who know of Deming and his operations research tradition. Sad state of affairs, and the next generation of epidemiologists can do something about it.

Obituary for Deming

Your thoughts?

I will return to this topic later, after I elicit some comments and give the idea more thought. Maybe it's not worth pursuing.

(HC: Please fix any typos that I have not caught. No need for comment on them. Just fix them. Thank you!)