Business Case & News

Targets, Outputs or Outcomes?

Is it time to change how we think about targets?

NHS England recently announced it was overhauling NHS targets for A&E departments along with changes to waiting times for cancer, mental health and planned operations.

They have come to conclusion that having targets for target sake doesn’t improve patient care.  More so, these targets are starting to have adverse effect on staff behaviour. Some hospitals had started playing the system in order to meet targets.  NHS staff didn’t join the service to become target hitters, they joined to care for us in community.  The Government should be welcome this approach and we should support them.

Using mechanistic targets

But before we replace old targets with new targets, we need to understand why these have not worked in the first place.  The biggest factor is mechanistic nature of performance measurements upon which the targets are based.  These mechanistic measurements work effectively when the inputs to the process can be controlled from within the process and where they are stable.  For example, a car manufacturer can control the inputs, it can predetermine a series of linear manufacturing activities, model these in precise detail and agree with the supplier when inputs will be supplied and in what form.  This makes measurement and targets valuable as the conditions in which they are based do not change from one period to the next.  There is focus on stripping out variation to achieve better results so that everyone / every machine has no choice but to act in the way it does. 

The problem comes when the inputs cannot be controlled from within the process.  For example, an A&E department cannot control its inputs, the patients and their illness.  Hospitals can’t predict who will be injured and when or know when a disaster such as an explosion will occur.  They also can’t predetermine how extreme our injuries will be. Targets drive out variation, but these variations exist, they can’t be avoided in such an environment making mechanistic measures and their targets logically weak

Thinking about measures and targets differently

In conclusion a hospital isn’t a mechanistic system, its an organic living system.  Our A&E department is a complex interaction of clinicians and non-clinicians, patients, their family and friends all adapting as one team to a situation they could not predict.  They are all focused on medical and social outcomes, not outputs (that is mechanistic thinking).  The activity is non-linear and the precise way the team is to behave cannot be predetermined or modelled because of the variety.  Sure their involvement in the process can be pre-determined or assumed – a consultant is a consultant, a nurse a nurse, a patient is a patient but how they behave is influenced by the reality of the situation and it is this reality that drives in variety to the process and renders the mechanistic targets less relevant. 

If this is the case, then what can we measure?  As the team is focused on medical and social outcomes. Therefore, what we can do is not measure their individual behaviour but measure the outcome of that behaviour as experienced by others in the same team or outside that team.  It is this outcome, the objective evidence, that is critical.  The collective individual outcomes informs the measurement criteria i.e. the desired whole team medical and social outcomes to a greater or lesser degree.

This changes what is measured and what this is reported against.  The implication is quite profound because behaviours are lead indicators of risk, mechanistic measurements are lag indicators.  As this qualitative behavioural data can now be collected, consistently analysed and quantified it is possible to produce a risk profile against the desired outcomes. Each desired outcome can, of course, have a target. The shift from outputs to outcomes i.e. the impact of the output on the true intention is fundamental we believe.

Balancing mechanistic and organic risk measures

The first step in this environment is to define the desired and undesired outcomes and express these as performance drivers i.e. what we are seeking the measure.  Add a target

Of course, there is a risk in simply changing all mechanistic measures and targets to organic measures and targets.  Rather a balance is needed based given the context of the medical facility and what its purpose is. For example, an A&E department can’t control its inputs, however a specialist cancer unit can as only those people with cancer are referred not people who have broken a leg. It’s a performance management art.  Obviously we would never claim to be medical experts but hope that the example explains the difference.

Conclusion, we believe

  • Changing the target will mean going the loop again, change the nature of the measurement first
  • Focus on performance drivers and define these, place a target
  • Measure behavioural indicators / outcomes and mechanistic outputs at the same time, it’s a balance based on the context of the medical facility and its purpose – no one size fits all
  • Risk and performance emerge from an organic system, medical and social outcomes emerge from complexity of the medical and non-medical team working as one team
  • The thinking is informed by Ethnographic Research, in outline the study of group behaviour over team.

As indicated, we are not medical professionals, but we hope you find the above useful in helping to determining what is to be measured, why and therefore what the target should be.

About the author: Ian Rosam is an experienced sales professional working in and leading sales teams. Focused on helping organisations digitise risk & compliance by leveraging the power of cloud, block chain and AI tools to optimise business and compliance performance.