Multiple loosely coupled, dynamic, power-law manifest, evolving, complex, chaotic systems force tradeoffs rather than solutions

An excellent essay, What multimorbidity shows us about guideline-driven evidence-based medicine by Mariana Barosa.  There a lot of nuance and so it is worth reading the essay in its entirety but she is addressing the issue of individual health decisions in the context of multiple loosely coupled, dynamic, power-law manifest, evolving, complex, chaotic systems.  

In the health context, she is specifying this as a challenge between knowledge related to single illnesses versus the complexity arising from multimorbidity situations.  The example of the contrast might be two patient scenarios - a seventeen year old comes in with a broken leg sustained while doing sports.  A clean break and no other symptoms or conditions.

That scenario is both easy to diagnose, easy to treat, and easy to forecast outcomes.

The multimorbidity contrast is when an eighty year old comes in with memory issues and a broken leg.  It is unclear how they broke their leg.  It is not a clean break.  They have just recovered from a broken wrist due to a fall.  They are significant obese and also dangerously sedentary.  They are taking nine different medicines covering blood pressure, pain management, headaches, a UTI, cholesterol control, and a thyroid condition.  They are also taking half a dozen OTC supplements.  

In the first instance it is straightforward - set the bone, wait six weeks.  Done.

In the second, nothing is straightforward.  All the existing conditions and treatments have to be assessed together in determining the best treatment path and those decisions necessitate unpleasant and difficult trade-off decisions.  

Barosa illustrates the tension between single a multimorbidity scenarios in this way.

Which serves to illustrate the challenge of the role for the doctor attempting to integrate the three domains.

Much of the essay is focused on addressing the increasing checklist approach to medicine which Barosa notes as being easier and more appropriate in single condition scenarios and not so useful in multimorbidity scenarios.  She also notes that the checklist approach that there is no real checklist approach that can replace clinical expertise.

I agree and this is an example of a phase change we face.

We have had several decades of dramatically improving productivity, first by improving single system, single cause problems and then from improving improving single system multi-causal problems.  These lend themselves to deterministic engineering approaches and outcomes

But with progress on IT systems and improvement tools such as Six Sigma, we are beginning to winnow the low hanging fruit.

What remains is harder and more out of reach.  Multi-system, single cause problems and multi-system, multicausal problems.  In other words, the world of multiple loosely coupled, dynamic, power-law manifest, evolving, complex, chaotic systems  

This is Thomas Sowell territory (from A Conflict of Visions).  

There are no solutions. There are only trade-offs.

Which is what Barosa is raising.  You can "solve" the seventeen-year-old's clean broken leg but you have trade-off decisions to make in treating the multimorbidity eighty-year old.  

You can easily see the challenge.  With single system, single cause problems, it is easy to judge whether someone is doing a good job.  Not so with multisystem multicausal problems.  

The checklist goes out the window with multisystem multicausal problems.  You need the right knowledge, experience, skills, values, behaviors, motivation, capabilities and personality to be effective in diagnosing and treating, with the patient, the multimorbidity case.  

And because you are dealing with conditional probabilities, there is far less ability to determine whether the"right" trade-offs were chosen.  Each case is unique and you can't step in the same river twice.  You can argue forever about what the options might have been and which trade-offs might have been preferable, but that is a priori.  You can't replicate conditions with the associated decisions later if the outcomes are less than were desired.  

Some excerpts from Barosa's essay.  EBM - Evidence Based Medicine.  

Notably, this tension between EBM and guideline-driven EBM gains full expression in primary care because primary care practice is dominated by multimorbidity – here defined as the coexistence of two or more chronic conditions, where (i) one condition is not necessarily more central than the others, and (ii) each condition may influence optimal clinical management of other condition(s).

The practical and philosophical challenges posed by multimorbidity to EBM – which collectively constitute what I call the multimorbidity challenge – have been the subject of active debate. Thus far, proposed solutions to the multimorbidity challenge focus on creating better “evidence-based” guidelines for multimorbid patients, when these patients are typically excluded from clinical trials.

[snip]

Over time, following guidelines became synonymous with delivering optimal evidence-based care, giving rise to what I call guideline-driven EBM. However, can guidelines “handle” the complexity of medicine and be effective knowledge translation tools, i.e. effectively translate or individualise research knowledge to the patient-at-hand?

Our best research methodologies produce population-level estimates that allow us to determine average effects of interventions. Such information is then used to produce guidelines catered to the needs of groups of people, which affect individual patients only indirectly by influencing the decisions of physicians. Most guidelines focus on the management of single conditions because studies are usually designed to isolate the effect of a single intervention on a single condition outcome.

Nevertheless, EBM is about evidence-based individualised care: integrating research evidence with individual patient variables. This is challenging because there is an epistemological gap between research knowledge and practice – the knowledge-to-practice gap. This gap includes a difficulty in providing an account of inferences from statistical data to predictions concerning an individual (e.g. if a research study shows that an intervention reduces death by 50%, the average results of this trial do not tell us what will happen to an individual patient) and a difficulty in integrating patient circumstances and values in decision-making.

[snip]

Knowing how to integrate an individual person’s needs and wishes with research evidence, sorting out trade-offs, and prioritising problems, hinge on tacit elements that underpin basic human experiences and judgements. However, the guideline-driven interpretation of EBM neglects the role of clinical expertise, and thus cannot overcome the knowledge-to-practice gap.

The problem is not unique to public health.  All practitioners in fields characterized as multiple loosely coupled, dynamic, power-law manifest, evolving, complex, chaotic systems face the same challenge.  

Our knowledge base is thin for multimorbidity (complex) cases. 

It is hard to translate population level statistical knowledge to individual cases. 

Trade-off decisions are inherently hard (for doctor and patient) because the making of a choice is itself  undesirable.  

As we clear the low hanging fruit, future improvements in quality and productivity will increasingly depend on our ability to create models, tools, and techniques for addressing problems arising from multiple loosely coupled, dynamic, power-law manifest, evolving, complex, chaotic systems which do not lend themselves either to checklist approaches or to deterministic engineered approaches.

Interestingly, these sorts of opportunities begin to put the unique human(s) back more central to the opportunity than has been the case with problem solving in single system, single cause environments.  

Phase changes are hard but interesting.