Function Allocation (in healthcare)

Function allocation has typically been treated in such a superficial manner that some might resonate to the Dekker & Woods (2002) assertion that we should abandon it. However, function allocation is critical within design and development of modern, socio-technical systems. Various parts of a socio-technical system inevitably provide functional capabilities.  In any sort of comprehensive design strategy, we will need to assign those functions explicitly.

Challenger et al (2013) offer a perspective on function allocation that is well in advance of the usual fare. They argue that we need to take a systems approach (which they also characterize as a macro-ergonomic approach) to function allocation. They contrast their systems strategy with the common task-centered, micro-ergonomic strategy in which specific tasks are allocated to either humans or to machines based on relative capabilities (humans are better at – machines are better at).  They also contrast their systems strategy with the techno-centric strategy that capitalizes on technological functionality while it limits the role for untrustworthy human capabilities. Sometimes system designers do not employ any explicit strategy; they allow functional allocations to emerge unsystematically, and seemingly by default, out of other design and development activities. However, these task-centered, techno-centric, and default strategies do not take account of wider workflow, social, or organisational issues.

Challenger et al (2013) argue that new systems should not be designed merely to take advantage of technological developments but rather, should be designed to accommodate the needs of an organisation and its workers. This requires a systematic and explicit effort to identify and implement the requisite functionality in a way that will satisfy those needs. Because changes to one part of complex system are likely to impact other parts, design efforts need to take a systems perspective by taking into account the full range of relevant workflow, social and organisational factors.

EHRChallenger et al (2013) give substance to their argument by reference to two healthcare information systems where functions were assigned by technologists with only superficial attention to the work processes and constraints. The problems experienced in use of these systems resulted primarily from the failure to adopt a systems approach in design and implementation. A techno-centric mindset, fortified by a political agenda that took the benefits of wholesale computerization for granted, encouraged neglect of the workflow, social, and organisational demands of healthcare work. Obvious functionality was assigned by default and subtle but critical functionality was ignored.  Healthcare practitioners had little meaningful impact on system design.

Towards the end of their paper, Challenger et al (2013) summarize their systems (macro-ergonomic) strategy. For function allocation to be successful, it needs to be addressed early and explicitly in systems design. It should involve a multidisciplinary team that incorporates the differing knowledge, expertise and perspectives of the full range of stakeholders (workers, managers, designers, strategists, clients) so that allocation decisions will be able to take account of the multiple, interrelated factors that comprise the system (social, organisational, technical, political, regulatory). In support of function allocation, we need to rely on well-established, rapid, and economical methods such as Rapid Application Development, story-boarding, and scenario planning.

If you are concerned with system design, you need to be concerned with function allocation, and this paper offers something exceptional for the challenge. If you are involved in designing information systems for healthcare, this paper offers added value. The merits of healthcare information systems are often addressed in LinkedIn healthcare groups. The discussions can become heated, with administrators and technologists singing the praises of computerised information systems and healthcare professionals pointing out their deficiencies. Many of the discussants exude considerable confidence that they understand the problems and know what to do about them. The solution from the technological and administrative perspectives is for healthcare professionals to shape up and adapt. The solution from the perspective of healthcare professionals is to throw these systems out and let them, the healthcare professionals, design new ones.

To my mind, nothing in these discussions points to a viable way ahead. No one, it seems, wants to approach this problem from a systems perspective. In contrast, Challenger et al (2013) are concise and articulate in outlining a way ahead, one that if understood within the healthcare professions, would provide the foundation for the development of effective information systems that support the work of healthcare professionals and that satisfy administrative needs.


Challenger, Rose; Clegg, Chris W. & Shepherd, Craig (2013). Function allocation in complex systems: reframing an old problem, Ergonomics, 56:7, 1051-1069, DOI: 10.1080/00140139.2013.790482

Dekker, SWA & Woods, DD (2002). MABA-MABA or Abracadabra? Progress on human-automation co-ordination. Cogn Tech Work 4:240–244.

3 responses to “Function Allocation (in healthcare)

  1. From a systems thinking approach, the adage of “a system is greater in whole than the relative sum of those parts or processes that make it up,” is a standard phrase we use when attempting to describe a system. Point being that the above commentary is after the right goal; but to my mind losing the ability to address a systems thinking process in the very nature of how the program is designed.

    We have to remember that it is not just the collaborative aspects of those parts, components, processes that make up a system that define a working symmetry of the whole, even a range of stability we tend to view as homeostatic. What we have to remember in a system is that the parasitic aspects of a systems define the adaptive nature of the collaborative components. All true systems have an ongoing balance of both parasitic and collaborative elements by which the competitive nature of the innate pattern of those processes reach a steady-state balance which once reached, optimize the behavior of the whole; a complexity if you will, which we see as emergent behavior.

    The second point is one of the most critical; if you want to develop a system with emergent behavior, than you have to look at the smallest elements, components or processes within the system, defining emergent behavior within each group, boundary conditions that define that sub-system, which has its own emergent behavior, and by which surrounding sub-systems interact with each other. Building layers of emergent behavior that come together as super-systems, layer upon layer ……

    If you look at all true systems; meaning a group of processes, components or elements, which find that balance referenced above and develop emergent behavior, they all have a number of things in common. Specific boundary conditions, balance between the collaborative and parasitic processes, and something that I’ve not heard referenced in most of my reading, which is, that it is not the parts, component, and or processes that define the nature of a system, but the emergent behaviors which come to interact at all layers of reality and by which all reality is built upon.

    Form follows function, a system has a function, once that functional pattern is established, it becomes a steady state node that reacts and acts behaviorally as one unit, building the next layer of reality within the emergent conditions and behavior of other systems it comes to adapt with and depend on to support the common need of stability within whatever environment that tends to attempt to degrade it.

    If you wish to develop any system approach to a problem; in this case a network for health care etc., than you don’t design it from the top down, with all the myopic nature of surface views we all tend to fall to at times. You have to integrate each node within the interaction of the innate needs of the people involved, the job or mission of each node or group, and the software and hardware that parallels their needs.

    Collectively we know the goal is management of health care records and any functional needs of the hospital or group of hospitals you are attempting to connect. But if you don’t build that functional mission down to the nodes of the network; both human and non-human, then you get a complex system; what I define as just a disparate group of parasitic processes which never attain the emergent behavior a system with complexity by definition does. You end up developing buffering patterns, human, software or hardware to maintain the intent of the mission focus, developing administrative needs and functional reaction groups for that grey area of operational activity which is lost due to the lack of a natural feedback in the system which has to be innate to the the emergent behavior or pattern of the whole system in question.

    Just a note, in science some of us see a complex system and a system with complexity as two different things, as noted above.

    I’d be interest in hearing more of the problems you face with this particular project…thanks and have a good week…

    • Lewis

      Thank you for your detailed comment.

      I am uncertain what you mean by: to my mind losing the ability to address a systems thinking process in the very nature of how the program is designed

      You say, form follows function. While the remainder of this paragraph makes good sense, that particular statement, form follows function, is something I hear occasionally. I never understand what it means. It sounds glib and mindless to me. Can you clarify?

      In the next paragraph you argue that you cannot design from top-down, implying in the remainder of that paragraph, that we should work at designing from bottom-up. In my work, I use the method of work domain analysis, which is based on the assumption that one needs to design by meshing the top-down and bottom-up approaches.

      However, the main point you offer is that we need to be concerned with emergent behavior. While that is important, it is always a challenge in design to anticipate it. The abstraction-decomposition space of work domain analysis maps out the possibilities the analyst can anticipate, but as far as I know, the only way to identify other types of emergent behavior that will result is to see the system in action. Do you know of other methods?

  2. Lewis could not reply because of interference from a blog plugin I had recently activated. I have now deactivated that and inserted his reply below:

    Sorry if the first part sounded insulting, it was not intended that way; my point being, that all of us at one time or another, lose our selves, our mind to the surface observations and generalize behavior on these understandings rather than the depth of a system view, which is about patterns of a system, rather than the traits of a system that are more what most science and technology are built upon.

    “Form follows function” has probably been overused, but mainly because it is like saying “e=mc2” The phrase or formula does not tell the full story of all the permutations; but gives a core source point of how all things work, or tend to find a pattern in how they work.

    If you look at architecture world wide, look at how and why a building is built, you tend to see parallel’s that take in the innate understanding of gravity, the needs of humanity to repel nature’s tendency’s to abuse us at times, rain, snow, etc., the form of all buildings world wide, find patterns innate to the materials limits, the needs of those building it, gravity, weather patterns, tectonic stability or instability at times and so on….defining that “form will follow the functional aspects of the patterned based needs of both nature and the humans manipulating it.” Certainly we’ve seen the disasters when we don’t take function into account in the application of a systems ultimate emergent behavior.

    Remember, that emergent behavior, the critical point is, that as a system forms, it develops a behavior which comes to exist, outside the form and flow of the parasitic and collaborate elements or processes within it. In other words, human behavior is not controlled by one cell, the heart, or even some part of the neurological system; but the combination of it, the natural world, and to my mind, space-time and quantum behavior…that’s another story I wont go into now….point being human behavior is the gestaltic end result of both our biology and the physical universe at large; but the human body, just like a business only requires one part to go wrong long enough destroy the system in question.

    From a systems view of this statement, form is evoked overtime by the innate patterns that develop from the functional process of the system in question. All to often we don’t see a lot of this, due to the fact that we never see one system in action; but usually a large group of them in action, the emergent behaviors of one mixing with another, like the human body, developing an almost infinite number of systems coming together, in sync, finding a stable pattern at each level of emergent behavior throughout all the layers of reality which we see the end result of consciousness and those mechanisms, our physiology and anatomy that support it.

    Ultimately to my mind, it is not the patterns of physical objects that define the nature of how things work, but the innate emergent behaviors of the group of processes which come together, and again, we tend to see the surface aspects of reality.

    You are to right sad to say, that often times we do not see emergent behavior until the system comes into existence. But I tend to think that we know enough to understand the functional requirements of many aspects of those processes when brought together, will develop an emergent condition or behavior over time. For businesses, these are always psychological or sociological at some basic level. Innately any human system will form an emergent behavior, that is not in question, the question remains, will it form an emergent behavior that fits my personal needs or wants when forming a system extension to that of the human demographics I’m aiming for.

    Some will argue against my next statement; but to my mind, there is just one kind of system, which is a natural system. All things follow the basic patterns of the quantum layering of reality to which we see layered to a point that it exhibits probabilistic motion. Yet at its heart, as human systems develop as a consequence of these natural laws attaining the emergent behavior of consciousness, or some such, or whether its the grouping of natural non-human aspects of matter and energy to support those patterns we see all around us, to which our development and evolutionary adaptations determine most of our own limits; I’ve come to the conclusion that its not different types of systems we should be looking at; but more the boundary conditions; limits, between each natural system, and at what density they exist at i.e., quantum, nuclear, molecular, ad infinitum to the emergent behaviors which develop and the layers over time which define the complexity of each level up and to including sociological, economic etc.

    In short, a car is not a system, unless you take in the human or humans that designed, built it or use it. The collective patterns develop from the patterns that are evoked by the object we create; but ultimately we are the system in question, not the thing we create.

    A computer is not a system by and unto the definition above for the same reason. It can mimic behavior, but only through the system of programming that endow it with that simulated behavior.

    Sorry for the diatribe, I’m still wrestling with some of my own arguments and tend to follow the trail….

    To your last point, attempting to predict emergent behaviors, though I think we can at some level, which is the heart of nonlinear problems or wicked problems or the butterfly effect; whichever euphemism you wish to use…point being the mathematical and physical limits of our understanding, still revolve around Newtonian physics, the math is all wrapped around that paradigm.

    Ultimately though to understand a ‘true system’ requires a true system. Meaning human beings, you’ve see the dearth of arguments over the wicked problem of global cooling and warming, falling to irrelevant agenda driven arguments rather than real science…which is due to the fact that we don’t have the math to handle such things, nor the understanding.

    To your point though, take the water molecule, if I knew nothing of water’s characteristics, and only understood particle physics of hydrogen and oxygen, I could not predict the emergent behavior of the water molecule, in all of it diversification, and see how it became the central molecule by which all life on earth depends and is required for healthy growth, as individuals as well as the collective of humanity.

    I’m come to see though, we might not see the eventual emergent behavior of a system, say like water and what happened over time, but take the phase change that occurs in water to form Ice, as the temperature drops, there is a polarized wave of activity, based on energy, the symmetry of the molecule which we can and do study, we can see the ‘functional form and flow’ of the water molecules in question, developing as the temperature drops a predictive pattern. Now I cannot say which molecules will convert to an ice molecule; but the boundary conditions of which I can observe, pressure, temperature, any contaminates, say salt or otherwise, can determine when the collective drop of water will convert to ice ….etc…

    Businesses, sociological processes etc., we always see polarized waves of behavior, usually someone always sees these patterns; but no one listens usually until after they see how wrong they were in hindsight. Examples abound of patterns that are easily seen, and predictable in any given extent, usually based on experience or history or both. Movements of ISIS into the power vacuum left by the U.S. in Iraq, or going back in history Vietnam, Chamberlain during world world II etc.

    From a business perspective, it was predictable when the first resisters were developed in the labs replacing vacuum tubes, the tech bubble in the 90’s…….

    Starting a business, we look to patterns, sociological, psychologically, we look to the saturation of the market for our product, we see a myriad number of things, the same way I look to respiratory failure, kidney function, and blood pH as markers to the types of drugs I can give to find homeostasis within the human body.

    My point here, is not to range freely among many different fields of study; but rather to point out that in all fields of study, all systems have the same innate pattern to them. When I talk of building a system from the bottom up, its more about see the natural patterns of reality, using that functional form and flow of humanity and the natural limitations we all fall within, and see those boundary conditions by which anything we choose to do are locked within.

    An example of my thoughts is something myself and several others proposed as a solution over time to the immigration issues in the U.S.; farms grants, using solar arrays and ocean water into our southern borders. The sun provides natural energy and the sea water; as Israel does for its deserts, purify the water and make the desert bloom. Grants on both sides of the border over time would provide agricultural products, water support and jobs eliminating naturally much of the immigration problems. It would not solve it all; but much of it, it would.

    Admittedly I’m not nave to believe anyone is going to do this, for me its an exercise in systems thinking; and that approach will work at every turn; what occurs in a parasitic environment, the governmental out of control limits, where solutions to a problem are easy; but getting through everyone’s personal agendas are difficult to get through.

    My point with this example is to learn to see the parasitic elements as well as the collaborative one which are required for emergent behavior; but have to be limited to a balance, something my solution will not be able to overcome for some time.

    Looking at the boundary conditions of a system, business or otherwise, what you intend to do within the health records for example are limited to the human patterns and needs of those patterns to be parallel in the operational aspects of how you design a system, computer, business infrastructure, manufacturing etc….

    When I talk of starting from the bottom up, I’m looking at the need of what system dynamics already exist in question; what their limits are, looking to what the boundary conditions of your market are, or the mission in question, and see how you can drive the pattern of what you see, from the multiple processes and systems you started with,

    Sorry I for the length….I believe I answered some of your questions, I’ll think on the others, and if you have any questions, or want to describe in more details the nature of what you are attempting, I can give you some more specific examples focused to your goals.

Join the discussion