Complex systems, chaos and family medicine

D Owens 歐德維
Deputy Editor, The Hong Kong Practitioner

As young prospective doctors we were first grounded in the basic sciences. At school we learned to rigidly apply the laws of Newtonian physics whilst conveniently ignoring, as instructed, the fact that the world is not a vacuum. Today as doctors we continue to deny and oversimplify the complexities of the systems and processes which lead to ill health and disease. It is a natural product of our reductionist training. The double blind controlled trial, which remains the gold standard for health based research, tends to reinforce a view of singular interventions leading by cause and effect to a predictable outcome. Whilst as Family Doctors we pride ourselves in a holistic approach taking due consideration of physical, psychological and social issues, so often we continue to focus on singular interventions, often pharmacological.

Physicists recognised long ago that relatively simple systems are unpredictable. Two bodies in orbit have completely predictable paths. Add a third body and the system becomes chaotic. This relatively simple observation has been expanded into an entire theory, chaos theory. Very small interventions in complex systems may have a very large and completely unpredictable effect on outcome. The weather and stock markets are examples of complex systems. This is why despite massive investment in time, money, human expertise and computing power it is not possible to predict with certainty whether the Hang Seng index will rise or fall or whether it will rain tomorrow. As doctors, we need to move away from a simple linear view of health and illness towards a complex view. It is essential that we strive to understand and influence the processes involved in health and not just the outcomes at the point of disease. Every single patient has his or her own genetic, physiological, psychological and social individuality. These systems are every bit as complicated and chaotic as the weather or the stock markets.

Imagine a 40 years old male being resuscitated in casualty following a large myocardial infarction. By all measures he can be considered to be ill. At what point did he stop being a well person? Last night he was playing squash: by most measures of health he would be viewed as well. Unfortunately unknown to him he had an 80% stenosis of his left anterior descending coronary artery and had just begun to rupture an atherosclerotic plaque. This outcome was dependent on a web of complex interactions acting within a complicated system in unpredictable ways over many years. As with all chaotic systems small interventions earlier can make large differences in the long term. At this level, early in the cascade, we have the greatest chance of positively influencing the outcome with relatively simple interventions. Many of the factors involved are recognised on a physical and physiological level: genetic, smoking, blood pressure, cholesterol, homocysteine, lipoprotein sub-type. Some factors are yet to be discovered and some are recognised but not clearly understood. Would he be in the same situation if he was less hostile, had a larger social network or even if he owned a pet? On page 290 of this month’s journal, Dr B W K Lau describes the complexities involved in understanding the interactions between social support and health. We know that individuals with fewer social ties have poorer health. Lonely people get more cancer and have lower survival rates after treatment than individuals with a good social network. Even keeping a pet improves health parameters in the elderly. We understand the outcomes but as yet do not fully understand the processes.

There is more discussion of outcome and process on page 298, Dr K W Chan explains the reasons for a move away from a teacher centred continuing medical education (broadly speaking CME) towards a continuous learner centred approach (CPD). In this regard medicine can again learn from industry and especially the airline industry which has long recognised that a systems based approach is necessary to improve outcomes. It is important to define both the outcome measures and the processes involved. If our goal is to improve the quality of health care we need to move away from a system which defines targets based on CME points. This is an artificial and poor outcome measure. CME points can be acquired whilst snoozing after a drug lunch. They have consistently been shown to be poor measures of positive behavioural change.

If we needed a reminder of the complexities of our chosen profession, on page 301, Dr N Y Chan describes how a 21 years old lady made the transition from well person to pathological anomaly following a routine chest x-ray. Her diagnosis will hopefully never affect her physically and probably occurred as a result of a relatively trivial problem affecting an extremely complex physical system in an unpredictable way.

Finally, as family practitioners we must never forget the debt we owe to our surgical colleagues who so often reduce the most complex problems to a choice of clear solutions. On page 285, Dr K C Tang discusses the surgical treatment of a drooling child. He considers the physical, psychological and social aspects of this difficult problem and describes the success of surgical intervention. He provides a timely reminder that no matter how complex the issues sometimes, our job is to define the major problem and fix it – even if the outcome will not always be perfect.