Search the hub

It was recently reported that NHS Finance Directors were ‘incensed’ that the Health Services Safety Investigations Body (HSSIB) should think that they could be working more closely with patient safety chiefs. Whereas medical staff and clinicians represent the sharp end of healthcare delivery, the administrative functions, including finance, are the blunt end. Removed in space and time from the action, it can be hard to see how their behaviour can directly influence workplace outcomes. To understand the issue, Norman MacLeod reflects on how systems behave and the decision-making hierarchy within healthcare organisations. Most discussion of ‘systems’ revolves around assemblages of artefacts: tools, processes, people and spaces. However, the late Jens Rasmussen described a system as a set of nested decision-making processes.[1] Leveson, at MIT, adds that systems are hierarchical, with control being exercised by higher tiers over the lower levels. She adds that control is exercised through communication and feedback.[2] From this perspective, a ‘system’ comprises actors engaged in different types of decision making. The hierarchy of decision making At the lowest level, we have the individual in the workplace. At any moment, our behaviour is directed at a specific goal and our probability of success is shaped by such factors as stress and fatigue, competence, expertise and motivation. Control is represented by direct action and feedback is in the form of observed outcomes. Because of the complexity of work, individuals form teams to get work done and this is the next level in the system. Teams make decisions about allocation of work, priorities, coordinating effort, problem solving. When an individual joins a team they surrender a degree of autonomy: you are no longer a free agent. Control is exercised through briefings, instructions and procedures, and feedback is manifested in behaviour meeting expectations, through raising queries, declaring problems, etc. Teams can be both real and virtual. Real teams are typically those assigned to a task, working in close proximity. Virtual teams comprise agents that collaborate for a specific purpose and are usually remotely located. Virtual teams often work asynchronously: a request is submitted and the response follows after a lag. Virtual teams require additional skills as they typically involve working across organisational boundaries. Individuals and teams are where direct action occurs. The next level in the system is the organisation. At this level, decisions are made in relation to the specific goals the organisation has been set up to achieve and cover configuring assets, allocating resources, command and control. The organisation exercises control over teams and individuals through contracts of employment, codes of conduct, policies, etc. Feedback is typically through audit and compliance, event reporting, tracking of resource utilisation. Of course, the ‘organisation’ is also made up of individuals and teams: the model is recursive. What differentiates each level is the nature of the decisions it makes. The next, and possibly, highest level in the hierarchy are those entities that facilitate the functioning of the system but do not, in themselves, get directly involved. Here we see government departments, regulatory bodies, accrediting bodies. Actors at this level set strategic goals, allocate resources at the macro level and grant permissions. The components outlined here all exist in a broader environment. By convention, the environment describes attributes that exert influence on the actors in the system but is not influenced, in turn, by those actors. For example, the public health profile in a geographic area will shape the strategic goals set for the organisation and will influence the healthcare capabilities that need to be provided in that area. However, the action of an individual healthcare organisation will not necessarily shape the public health profile of its hinterland. Emergence and cross-scale effect So, where do finance directors fit into all of this? Obviously, as actors at the level of the organisation their decisions relate to the allocation of financial resource. As such, they shape decision making by others in the system responsible for spending on specific functions. But we now need to look at some other properties of systems: emergence and cross-scale effect. Emergence describes behaviours that cannot be explained simply based on the functioning of the parts of the system. Cross-scale effects captures how actions at one level in the system can have unintended consequences at another level. If we start with emergence, ‘safety’ is an emergent property at the level of the individual. Only individuals can act in a manner that bolsters safety or, conversely, it is the actions of individuals that create unsafe states. ‘Culture’ is an emergent property at the level of the team, while at the level of the organisation we see morale as a key emergent. Patient safety activities compete for resources in a landscape where other demands can be seen to have a more direct influence on outcomes. In financial terms, patient safety can be seen as a discretionary spend. This attitude to a legitimate demand can shape morale. Cross-scale effects are akin to Reason’s Latent Factors.[3] Their presence is often only revealed when something goes wrong. We can see cross-scale effects at work in the case of staff recruitment. For example, to save money posts are often ‘gapped’: a post is not advertised until after the incumbent has left. While the post remains unfilled, the burden of work is borne by others or simply not done. Where workload is increased, outcomes can include increased fatigue, staff turnover, sickness/absence or risk of error. So, a simple, rational decision at one level can have multiple consequences elsewhere. The example given here – gapping posts – is typically a response to financial constraints. My intention here is not to portray finance directors as villains: they simply have their hands on some powerful levers of control. But their protestations do possibly support the need for a more sophisticated attempt to understand how systems work. References Rasmussen J, Svedung I. Proactive Risk Management in a Dynamic Society. Swedish Rescue Services Agency, Karlstad, Sweden; 2000. Leveson N. Engineering a Safer World. MIT Press; 2012. Reason J. Human Error. Cambridge University Press; 1991. Other blogs from Norman MacLeod: What is a ‘safety management system’? Error isn’t a problem – the problem is the word ‘error’