‘CONSULTING WITH COMPLEXITY’
Dr. Declan Murphy
Foundation Lecture, ICGP AGM 16 May 2004.
The flashing fireflies of Malaysia are one of the wonders of the natural world. Each evening they start to flash independently but soon they are all flashing in unison along a whole riverbank. There is no leader beating time, no evidence of organisation, yet the whole display is highly synchronised. When Shell announced bad news in relation to its level of oil reserves, the stock markets plunged in similar ways around the world. There was no co-ordinator or master plan, but nevertheless a major and similar readjustment of the financial markets took place worldwide. Sleep disturbance following travel across time zones is universal, and jet lag disrupts our normal functioning. Our circadian body rhythms and sleep cycles are obeying laws of nature that we barely understand. The prescribing habits of Irish general practitioners vary considerably, yet the majority have total GMS drug costs that fall within a relatively narrow band; diversity in detail yet conformity in the generality. These four very different examples appear to be random chaotic phenomena that cannot be explained by conventional mathematics or classical science. However they can be understood using the explanations of complex adaptive systems. It is the aim of this paper to demonstrate that complexity theory provides explanations, models and metaphors to understand much of what we as GPs encounter in our professional role as doctors, with health organisation and health service reform, and with our wider environment.
It is helpful to consider briefly how scientific thought has developed over the centuries, as this will place complexity in its historical scientific context. Although the ancient Greeks, Persians, Arabs and others made significant mathematical and scientific discoveries that we still use, notably geometry and algebra, the modern scientific movement began in Europe in the mid fifteenth century with Gallileo. This was the beginning of the classic scientific method as we understand it, the application of observation, measurement, and conclusion. In the sixteenth century Newton discovered the law of gravity and three laws of motion that together laid the foundation for the laws that underpin all scientific knowledge until about a century ago. The alchemists with their semi-mystical experiments paved the way for the discovery of the chemicals and elements that in the nineteenth century were sorted into the periodic table. Mendel discovered the genetic mode of inheritance long before the discovery of atomic structure, genes and chromosomes. In the nineteenth century Darwin published his classic observations from which he concluded that life on earth in its present form is the outcome of random variation, natural selection and survival of the fittest. The classical or Newtonian view has been the dominant paradigm of scientific thought and discovery. It is based on the belief that there are physical laws based on cause and effect relationships that govern all natural systems. However at the end of the nineteenth century this world view was challenged by a number of discoveries including electricity and electromagnetism, and in the twentieth century the theories of relativity, quantum theory and Heisenberg’s uncertainty principle. No longer did science seem to have the answers to some of the greatest questions. Why do ecosystems go through periods of profusion and mass extinctions? How does a colony of termites build an edifice bigger than a skyscraper in relative terms without a leader or master plan? How do we as humans think, even think about thinking? And what is common to all these problems? Classical science with its linear, reductionist paradigm of cause and effect, has and continues to solve some of the greatest questions of science. The problem is that most systems in nature are relentlessly non-linear, being dynamic, multiple, interacting and adapting. Classical science was able to overlook these issues for centuries while it solved the major linear problems, from gravity to genetic inheritance to electromagnetism. During the second half of the twentieth century a few theories were developed to try and explain these events, such as cybernetics and catastrophe theory, and most notably chaos theory. The insights of chaos theory laid the foundations for complexity theory, and some of its terms and concepts have been incorporated into the language and concepts of complexity science. In chaos theory, small alterations in initial conditions are quickly magified in exponential fashion, so that despite theoretically following a logical linear path the development soon becomes utterly unpredictable. Typically this is seen in weather forecasting and the commonly quoted example is the butterfly effect, where a butterfly flapping its wings in Brazil results ultimately in a tornado in Texas. The initial perturbation is magnified to a point where despite initial predictability and its deterministic nature, there is no practical way of predicting its outcome. For this reason weather forecasting, for example, is very unreliable after more than a few days; as one meteorologist is quoted ‘we can predict the weather accurately provided it doesn’t do anything unexpected’.
Complexity science could not have developed without the enormous developments in computer power that allowed researchers to model previously unimaginably complex problems, such as patterns of evolution. Its bedrock is in mathematics and computer science, and many people are intimidated by its presumed incomprehensibility to all except physicists and mathematical wizards. This is like saying that one cannot drive a car without a full understanding of the internal combustion engine. Complexity science is a conceptual framework for seeing the world, and its appreciation requires imagination and an enquiring mind, while taking the mathematics on trust. It is having a big impact in the study of areas as diverse as management and systems theory, evolutionary theory and cellular activity. For general practitioners it should have a particular resonance, since it deals with the similar dilemmas in a similar process of non-linear interaction, adaption, and emergence.
A system can be thought of as a coming together of parts, interconnections and purpose. A car engine, for example, can be regarded as a linear system. It has a number of parts that interact with each other to achieve the purpose of driving the piston that in turn moves the car. It is complicated, but not complex in the way we are discussing here. The engine works in a stepwise process that is logical, or linear; the process can be reduced to the individual events, analysed, and re-assembled in an exact copy of the original. The sequence and outcome does not vary. The whole is equal to the sum of the parts, no more and no less.
So what do we mean by complexity? Complex systems are composed of multiple agents that follow simple rules and interact with each other in rich non-linear ways at all levels; they are adaptive and self-organising, are sensitive to their history and environment, and the result is an outcome or emergence that is unpredictable and greater than the sum of the original parts. An example of how a few simple rules can create complex unpredictable behaviour is a computer programme produced about twenty years ago called Boids. The bird-like figures generated by the computer are programmed to follow just three rules: constantly attempt to move to the centre, avoid collision with other boids, and move at the same speed. The result is a lifelike imitation of how birds flock in real life, each boid adapting to the actions of its neighbours and achieving self-organising, unpredictable behaviour by the group. The game of life was a computer programme developed in the 1970s to demonstrate that by programming light bulbs to switch on and off in response to inbuilt rules. When the nearest bulbs are on it is off, and vice versa, together with variations of those rules. The result is a lifelike pattern of activity. These examples illustrate in a very crude way the behaviour of genes. Genes are activated only when certain other conditions are present, namely the presence of certain enzymes which in turn are controlled by the actions of other genes. It is as though one could turn on the light switch in the living room only when the room temperature is at a certain value, that there is a vase of roses on the coffee table and the dog is asleep. Complex adaptive systems are all around us; the society we live in; the practice we work in; and in ourselves as individuals. Our brain, our immune system, in fact each cell of our body are complex systems. The health service, the stock exchange, and the ecosystem of Galway Bay are all examples. Complex systems can involve all forms of life, such as fish in a pond, or the movement of traffic on a motorway. This college is a complex system made up of many individuals; there is a rich network of personal interconnections at all levels, the interactions are sensitive to their environment and the outcome is an emergent property that could not be predicted with confidence. Within the complex system of the college are other complex systems, nested within each other like Russian dolls, such as the council, the CME tutors, the administrative staff, and so on. In practice, if it looks like a complex system and seems to behave like a complex system, then it probably is a complex system. You may say ‘So what? We know that, it’s very familiar to us, and there’s nothing new to all of this’. You could ask if complexity is simply common sense masquerading as a new kind of science, or is it the science of common sense? Both may be true; what is new is that complexity theory provides us with the toolbox, the metaphors to understand and study these non-linear, cause and effect processes in a scientific way that cannot be done by existing Newtonian methods.
As GPs we define our work as dealing with uncertainty, among other things. The uncertainty arises from the nature of the problems we are presented with, which are generally complex, non linear, and defy easy categorisation using linear diagnostic tools. A young parent consults with intermittent limb pains and parasthaesia. The pains are non-specific, examination is essentially normal, and it is difficult to find a neat diagnosis. A few open questions gradually creates a space that is filled in with some emotion. A new supervisor at work has been critical constantly; the general manager thinks highly of this new recruit, and the patient has been unable to voice the problem with credibility; the firm is non-unionised and there are no other avenues to seek a resolution. Arriving home irritable and dejected, the two children react by becoming sullen and difficult. The marriage broke down about two years ago, and the couple’s present relationship is based around the needs of the children. Insomnia, fatigue and chain smoking are now a way of life. Years ago, a friend developed multiple sclerosis. Could this be something similar? Or the skin lesion removed a few years ago when living abroad? This type of scenario is typical of many of our consultations. The diagnosis and resolution of this patient’s problem is not covered by any specific evidence-base, or by a Cochrane collaboration. No neat linear path leads predictably from previous health to present ill health. It is explained only by the narrative and by reference to the patient’s history and environment. It is however explicable in the context of complexity theory, though complexity does not necessarily make the task of helping any easier. The diagnosis requires a patient-centred approach, and a number of different models of consultation methods have been developed in recent years, such as Balint, Byrne and Long, Neighbour, and Schofield and Pendleton. All are based on exploring non-linear interactions, explanations and interventions to the problems we meet in the consultation.
The drive towards better quality care has been propelled by, among other measures, evidence- based medicine, protocols and guidelines. There is surprise and concern that their impact on GPs appears to be limited and that GPs show considerable variation in their practice. However, consciously or unconsciously, GPs may be sceptical, as these tools are generally derived from the presumed gold standard of evidence, the randomised control trial. The limitations of randomised control trials are seldom referred to, namely that the various components are studied independently of their context, that outcomes are presumed to be proportional to the input, that only a limited number of confounding variables are accounted for, and above all that the results can be generalised to the wider population despite the heterogenicity of society. It may be the best possible evidence at this moment, but its validity may not be valid for this individual patient at this time. There are thought provoking questions that can arise from the experience of natural evolutionary systems, where variation is the means by which the system becomes stronger. This can be conceptualised by viewing a Stacey graph. The zone of stability in the bottom left corner is the zone where certainty and agreement is maximal; the use of aspirin after a heart attack, for instance. The area of chaos is in the top right, where nothing is certain or agreed, such as the role of PSA tests in screening programmes. In between lies the zone of complexity, where innovation takes place. This has become a powerful metaphor in evolutionary terms, where major species adaption occurs and the zone has become known as the edge of chaos. This is the area in which every species and system tries to maximise its survival advantage, and ascend the metaphorical fitness landscape of hills, where survival depends on getting higher and avoiding the valleys where extinction takes place. The zones of stability and chaos are sterile in evolutionary terms and the species will not survive. Development that leads to disimprovement also disappears, and the overall fitness of the species improves. Doctors vary in their personality and practices. What is critical is that there is a shared goal, which could be expressed as a wish to help and to heal, or to care and to cure. Within this ethos boundary it is possible to adapt and act in different ways but with similar outcomes, as many of the markers of variability in GP performance seem to suggest. Clearly this is not a licence to act outside of scientifically proven methods or professional norms; but not every practice will be appropriate for an appointments system for example, nor will every patient want to share decision making. Variation in structure and process in general practice may be a reflection of the variation that occurs in the zone of complexity and that contributes to the overall improvement in fitness of the system.
When prescribing for patients we accept readily the possibility of drug interactions, and that each interaction can have a knock-on effect on all the other drugs. This is particularly relevant where patients are prescribed large numbers of drugs, as is frequently the case for example in elderly patients with multiple pathology. The number of such possible interactions rises exponentially and attempts to portray this graphically on the page break down quickly. The difficulty is not merely in attempts to display multiple dimensions in paper, but also conceptually in our mind. With two agents the interaction can be plotted on a standard Cartesian graph using x and y coordinates forming the familiar L shaped graph. We can readily understand three dimensions as being a cube, and can visualise how the possible interaction of three agents can be represented as a single point in space somewhere within that three dimensional cube. The most recent chart on the calculation of cardiovascular risk factors is an imaginative method of portraying the interactions of seven risk factors on a page, but it seems clear that any attempt to introduce more factors on to this two dimensional format will make it confusing and unusable. The difficulty of conceiving of more than three dimensions in space can be dealt with by the concept of a multi-dimensional space of possibilities, where each dimension has a plane within space which does not cross any other plane. For example, the individual pages of a book can be regarded as consisting of a multiple of planes none of which cross each other. An infinite number of such planes can exist, and can be imagined as occupying a sphere. Multiple interactions can be represented in space and mathematically a single point represents the entire state of the system. Where the system cycles between similar regions known as attractors, it can be displayed graphically by joining up all the points. The system may jump between attractors and this is the origin of the iconic Lorenz butterfly, where the point trajectories of the two two attractors appear like the wings of a butterfly. These kinds of sums are handled by the huge number-crunching power of modern computers, and we see practical examples of it in the artifical intelligence behaviour of the internet search engine Google, and the Amazon website that anticipates each individual’s likely preferences based on past purchases. This has practical application in our desktop computers where we are warned of potential drug interactions, and where interpretive software will work out cardiovascular risk. The addition of the dimension of time to multidimensional space is known as phase-space. The concept of multiple phase spaces was developed as part of chaos theory and lead to one of its most important discoveries. If the behaviour of two systems starts out from very nearly the same starting points they rapidly evolve in entirely different ways, as shown in the case of the butterfly in Brazil. Many systems in the natural world are very sensitive to their initial conditions and rapidly diverge from those starting conditions in non-linear fashion. In practical terms this makes it impossible to predict how such systems will behave. This has marked resonances with what happens in certain medical conditions, notably abnormal heart rhythm. The notion can be extended to genetics and embryo development, where a very minor initial change will devolope catastrophically, and the development of cancer where a chance random event may initiate an uncontrollable trajectory.
This week we read about the impending swarms of cicada insects in the American mid-west. Cicada nymphs live underground in the roots of trees and emerge only every seventeen years. Brood X is expected to emerge this week in numbers of about one and a half million to the acre. It is believed to be an example of predator satiation benefiting from the fact that seventeen is a prime number. (A prime number is a whole number that is indivisible by any other number apart from one). The seventeen-year cycle means that the emergence of cicada broods does not coincide with predators whose cycle is in non-prime numbers, such as four. By emerging in such numbers simultaneously it is believed that they satiate the potential predator population in such a way that they gain an important survival advantage. Only one predator, a fungus, has evolved to take advantage of this seventeen-year resonance. Most fascinating is the discovery that recently some cicada broods have been converting to a thirteen year cycle to outwit this predator.
This resonance, together with that of the fireflies of Malaysia and very many other events in the natural world, are examples of synchrony in nature. Synchrony, which can be thought of as the same things happening at the same time, is also a powerful driver of our own bodily functions, and its disorder is implicated in conditions such as cardiac arrhythmias, in epilepsy, and in sleep disorders. Synchrony depends on oscillators that are present in every cell in our body. Oscillators are like internal metronomes, beating time in a regular rhythm. The oscillator of each firefly is sensitive to the signals of others, and its oscillators move to synchronise with them by following mathematical rules built into its genome. So synchrony of the entire organism or system emerges from local single-cell interactions being multiplied to encompass the whole population of cells or organisms. These are examples of self-organisation, the emergence of order through local rules of adaption. There is no master plan, no supreme conductor, nor some mystical force, just laws of nature that are as valid as the law of gravity. A notable example of crowd synchrony happened when the millennium bridge was opened in London in 2000. The bridge started swaying excessively and had to be closed while the engineers tried to figure out what was happening. Computer modelling showed that a combination of huge numbers of pedestrians on the bridge at one time, combined with the limited swaying built into its design created a synchronised situation where each individual pedestrian attempted to correct for the sway in exactly similar ways. This is similar to what happens in a boat when it begins to sway, passengers shift their weight to balance it and the boat becomes unbalanced due to the massive over-correction. A similar situation can arise in the management of insulin requirements in unstable diabetics, where adjusting the dose of insulin as though it is an independent agent can lead to overshooting and repeated corrections.
Our lives are bounded by our heartbeat - systole, diastole, systole, diastole. The regularity is dependent on a single co-ordinated electrical wave arriving at the heart muscle, causing it to contract; it travels as a single wave through the walls of all four heart chambers, bridging the atria and ventricles at the AV junction through the specialised conducting fibres of the bundle of His. The impulse is produced by a few thousand cells at the SA node, each following its own local law of adaption and synchronisation. There is no master cell controlling the rate and rhythm. An analogy is with the Mexican wave of crowds in a sports stadium. There is no conductor, but the synchrony is produced by each spectator acting in response to the wave of the immediate neighbours. Attractors are a mathematical concept of areas in which the trajectory of the system tends to settle. Recent research shows that asynchrony, or the ability to move easily between attractors, is the healthy state for both heart and brain. In epilepsy the seizure is due to an emergent synchrony (as opposed to asynchrony) producing the characteristic rhythmical disordered movements. If we conceptualise heart rate as having attractors, the sudden change in form or shape of the attractors, as in myocardial infarction, leads to the phenomenon of bifurcation with possible change to chaotic behaviour and arrhythmia. Synchrony arising from the actions of cellular oscillators is at the root of our body’s circadian rhythms, such as sleep and temperature. Body temperature varies normally by 1.5 degrees over the 24 hours. Sleep is entrained or locked in to a 24 hour rhythm, but on a different cycle to temperature. However there is a relationship, as recent research shows that long sleep sessions always begin at times of maximum temperature, and short sleep at the time of low body temperature. Minimal body temperature is about one to two hours before usual wake-up time and many physiological rhythms are linked to the phase of the temperature cycle. Alertness decreases with a drop in body temperature so that the time of minimal alertness is typically between four and six a.m., known as the zombie zone. This is the most dangerous time for road accidents, and a number of major disasters linked to human error have occurred at this time, including the Three Mile Island nuclear plant accident, Bhopal, and Chernobyl. The tendency to synchronise is one of the most prevalent drives in the universe, from sub-atomic level to the solar system, from people to planets. It is fundamental to life itself, and how we conduct our lives. It seems to me that a variety of illnesses or disorders that we see in practice may be due to disorders of synchrony, and conceptually diseases such as ADHD, bi-polar depression, cancer and autism may be examples. This would suggest that in epilepsy for example, treatments analagous to a cardiac pacemaker could be more appropriate than drugs. We are living in a drug dominated medical world, where chemicals are delivered to the whole body in the hope of influencing the chemical basis of the electrical stimulation of specialised cells in the heart or brain, for example. The dominance of the chemical model of treating disorders has interfered with our efforts to take account of other paradigms for treatment.
The goal of all societies is to have a health service that is equitable, excellent and affordable. This can be done directly by government, such as the national health service in the UK; or the government can create the conditions in which other agents do so, such as the social-insurance funded services of Holland and Germany; or by having a mixed system such as the USA where government part-funds primary and secondary care through Medicare and Medicaid. In Ireland government funds and provides most secondary care and funds about one third of primary care. No system is perfect, and a combination of increasing expectations by the public, rising costs, novel technologies, expensive new drugs and inexhaustable potential demand has created policy dilemmas that have been identified as a series of five shifts:
When is care delivered? – the balance between prevention and treatment.
Where is it delivered? – the balance between secondary and primary care.
How is it delivered? – the balance between patient and professional involvement in care.
What is delivered? The balance between knowledge and habit-based care.
Who is cared for? – the balance of equity between different groups in society.
WHO judged the French health service the best in the world, but its cost is crippling the country financially and is unsustainable in its present form. The US has the most expensive health service but is judged very poorly by WHO, largely because of its poor record on equity and public health. Some countries, notably New Zealand, have instituted massive changes with consequent major disruption and unsatisfactory outcomes.
Insights from the understanding of complexity theory and the dynamics of complex systems have radically altered thinking in the management of organisations in general, and in health service management in particular. The experience of relatively recent developments in the USA and the UK will serve to illustrate this. The UK has a national health service with universal entitlement to all services financed by general taxation. Since 1988 there have been major reforms designed to incorporate the benefits of the market into the provision of health services. An internal market was created for the provision and purchase of services, with GPs as purchasers of secondary care for their patients. There were some adverse outcomes and the system has been adjusted most recently as part of the so-called Third Way modernisingg agenda of the Blair government. Despite this fine-tuning of the system there is a widespread perception that services have disimproved, private medicine is increasing, and morale among many workers in the system is low. This is felt to be due to underinvestment and the latest reform involves massive expenditure to bring the country’s expenditure up to EU norms. However Ireland’s experience of huge increase in health service expenditure is that simply throwing money at the problems is not the whole solution. What many people believe will solve them is the adoption of a Fourth Way, the complexity approach that deals with health services and organisations as complex adaptive systems.
This surprisingly, is exactly what w as proposed in a major US report. In 2001 the Institute of Medicine – a professional body aligned with the Academy of Sciences, with an advisory role to government - published an important report ‘Crossing the Quality Chasm: A new health system for the 21st century’. It may appear paradoxical that the country with some of the biggest health service defects could produce an important reforming blueprint – particularly when that report originated from within the medical profession. The report recommended that health care processes be redesigned to the following ten rules:
It states that ‘a framework for a new health system should be based on systems that can organise themselves to achieve a shared purpose by adhering to a few well-thought out general rules, adapting to local circumstances, and then examining their own performance …local adaption, innovation, and initiative will be essential ingredients for success.’ Just in case the message was missed, the report includes in the appendix an article by one of the leading authorities on the use of complexity science in health care systems, Paul Plsek.
Over arching any new type of service should be the goals of equity, access, acceptability, affordability and effectiveness. In Ireland attempts to reform the service to achieve these goals have been relatively small and incremental, rather than radical, and some of it counter productive. Government policy has been strictly directive, top down and controlling. Health Boards have responsibility but little strategic power, a no-win situation. The market-model applies only in general practice and in a very limited way with hospitals and attempts to get the best of both worlds – a modified health care market with government regulation of standards - is rudimentary. A quite radical alternative to government policy was proposed at the time of the last general election by each of the main opposition parties, Fine Gael and Labour. The two government parties – Fianna Fail and the Progressive Democrats - campaigned on the basis of existing and proposed government health policy. The Green Party did not have a specific health policy document. Unfortunately there was very little debate on health service reform at the time despite this challenge to the hegemony of the government’s health policy. Since that policy is well known and subject to extensive analysis, it may be helpful to consider the other two policies, and the main points are shown in the attached diagram. In primary care Labour proposed an extension of the GMS to 100% of the population for free GP care, and free drugs to the 40% in the lower income group. The drug refund scheme would remain in place for the 60% paying for their drugs. Fine Gael proposed an extension of the GMS to cover 60% of the population, and the remaining 40% paying for GP care and also availing of the drugs refund scheme. Both parties proposed universal patient registration. In secondary care Labour proposed universal health insurance with ‘not for profit’ insurance companies – effectively VHI in today’s terms - to be paid mainly by government, but with co-payments by those is in the higher income group. Fine Gael proposed a tax-funded insurance-provided scheme by licensed insurance companies. Labour proposed that all hospitals could join the scheme but could not operate in the ‘super-private’ sector as well; they had to be either in or out. Hospitals would become more autonomous but would still be owned and managed by the Health Boards as at present, and they would be paid for provision of services by the patient’s insurance company. The insurance company and the hospital (acting as purchaser and provider, respectively) would negotiate fees and payment methods. Hospitals would continue to be owned and managed by the Health Boards as at present. In contrast, Fine Gael proposed that hospitals would become autonomous and be controlled by independent citizen-based hospital boards. These boards would tender for and appoint a management team for a fixed term, with a performance sensitive contract for both quality and cost. Hospitals boards could form alliances or partnerships with other hospitals and share infrastructure, skills and management. Ownership of the hospital would rest with the government which would also set standards, monitor performance, and decide the overall budget. Government would decide on a basic basket of services that must be included in all policies, and companies could provide a range of extra services for patients to choose if they so wished. Power and responsibility would be devolved downwards, money would follow the patient and payment would be made to the provider through a mix of methods. Both policies appear radical and complicated, but radical re-engineering is now necessary if the service is not to implode. These alternatives would create a regulated market for health care in which complexity principles could be applied to management of the system, that is, general direction pointing, prohibitions, and resource or permission providing. We have an opportunity to take advantage of the evolving management thinking in health service organisation, but present indications are that the traditional view of the organisation as a machine is still dominant.
The development of interest in complexity was initially propelled by scientists and researchers in diverse fields but especially physics, biology and economics. In 1983 a high powered group of academics, including Murray Gell-Mann and a clutch of other Nobel prize-winners, founded the Sante Fe Institute in the USA solely as a postgraduate research centre for complexity. It has been at the forefront of its development since then, and many other academic institutes and universities throughout the world have developed departments devoted to some of the many aspects of complexity. A New York based think-tank, the Plexus Institute, is devoted to the study of complexity applications in healthcare in its broad connotation. In 1992 a book aimed at a non-technical but scientifically literate readership and called simply ‘Complexity’ by Mitchel Waldrop was widely read and popularised. It is written in an engaging and racy style, and still makes an excellent introduction to the topic. Other books in the popular science genre have followed, one of the latest being by John Gribbin. About four years ago a group of GP enthusiasts held a Complexity in Primary Care conference in Exeter, England. A web site and internet listserv was established and has been enormously successful at keeping everyone in the network of discussion and debate. Further annual conferences followed, with small group meetings in London in between. Membership of the group and listserv is open to all those interested, and has spread beyond GPs to include other medical and paramedical professionals, educators, management researchers, and enthusiasts from many disciplines. A book called ‘Complexity and Healthcare’ was written by a team of different writers from the group, and last month a further book ‘Complexity and Healthcare Organisation’ followed.
Since classical science has been the orthodox and only approach to understanding phenomena for more than four centuries, it naturally holds a powerful allegiance for all of us who have been trained in its principles. The advent of a complementary scientific method is exciting and disturbing; exciting because it creates a new way of understanding phenomena that have been inadequately explained by the existing method, but disturbing as the old certainties are questioned. It is not surprising that there have been some robust critics of complexity, including one in this month’s issue of the British Journal of General Practice. The criticisms are predictable, namely that complexity is a nebulous approach to science as opposed to the solid positivism of classical science, or variations of that argument. The counter-argument is that both approaches are valid and complementary, that the models and metaphors are complexity are valuable in understanding the behaviour of non-linear systems that are so prevalent in nature, and that complexity has important applicability is areas where interactions, networks and relationships are important. These areas include natural evolution, neural networks, biological systems and management in organisations. It is sometimes called a meta-theory, one that draws other theories together and in general practice it embraces narrative based medicine and qualitative research for example, in philosophy post-modernism, in music jazz and so on. It is the science of adaption using simple rules, rich networks, and emergent behaviour. It is possibly more useful in determining what we should not do, rather than providing a pithy prescription for action. Murray Gell-Mann is quoted as saying ‘that the world we see around us is merely surface complexity arising out of deep simplicity’. I believe that it places our work as GPs in a holistic perspective, using linear classical science where appropriate and complexity thinking in the many non-linear, and unpredictable situations that we are asked to deal with. In reality, we are consulting with complexity.