This is often combined with levelling. In levelling you smooth flow and even out the peaks and troughs in demand. In effect you convert uneven demand in to level operations. This can be done in some aspects of health and social care – in out-patient services, for example – but it’s more difficult in acute care where demand is often unpredictable, but requires an immediate response.
One way of levelling is to flex the service capacity to meet the varying demand, and so match the service provision to the increased demand. This sounds good, but how do you do it? It’s not practical to have people sitting doing nothing, waiting for increased demand that happens unpredictably.
The first stage is working out whether the ‘unpredictable’ demand is really unpredictable. It may not be possible to identify exactly when an increase will occur, but there are usually times that the increases tend to occur. For example, in recent work with an acute service, the service found that demand began to increase from 8am onwards on weekdays, and was relatively flat from noon until about 9pm. It then dropped back down, and dropped lower still overnight.
There were exceptions to this, of course, but the busier times were reasonably predictable. If you combine this with observation and measurement, the service can reach a reasonable understanding of the required capacity.
This gives you a good idea of how to cope with ‘normal’ fluctuation. Everyone in health and social care, however, has experienced conditions that are far from normal. This may be on the back of a community ‘flu outbreak, or it can just – happen. In the service we were working with, there was an average of about two admissions an hour from noon until 9pm, weekdays. In 75% of hours, there were three admissions or less an hour. This means, of course, that in 25% of hours there were more than three admissions, and it is in conditions like this that services begin to struggle to cope, particularly if busy hour follows busy hour, with no chance for the service to catch its collective breath.
There are various equations for working out what capacity you need for routine coping. Iain Smith at NETS suggests that a good starting position comes from working out the minimum and maximum, and looking at 80% of the distance between the two numbers. He doesn’t argue that there is anything magical about this number, but rather that it is a useful place to begin your thinking – there are many reasons why it might not be the ‘right’ number in a particular instance.
So far, so good – but this still leaves the extreme events. In health and social care services, people pull together, and generally cope. People stay late, start early and even come in from off duty to help out. This is great, but in a Lean system, you don’t want to work people harder – you want resilient systems that have identified ways of coping, usually linked to standard work.
Two possible strategies are the use of baton pass zones, and of bypass zones. Baton pass zones reflect the idea that there are points between cycles where work is passed from one person to another. If people can only perform one process, there’s not much help to be had here. It’s common in Lean systems, however, to have workers who are skilled in several parts of a process. This gives you the opportunity to then move people temporarily from adjacent processes to help out at a bottleneck. For example, if the rate limiting step – the constraint, as Goldratt would call it – is medical assessment, then you could move staff from upstream or downstream processes to help out. This happens in any case, but having a plan, and knowing in advance who is appropriately skilled, can make things much easier.
A second option is a bypass zone. In a bypass zone, you have a second facility that you can ‘switch on’ when required. An example of that occurred in work on a specific clinic. There were times of the year, usually related to a campaign – fairly predictable – or to the diagnosis of a celebrity – entirely unpredictable – when demand increased markedly, and suddenly. This caused a back log with dissatisfied patients and stressed staff, which took months from which to recover.
The service worked out a way of staffing an additional clinic. Based on their experience, they knew this capacity would let them cope with peaks. They then decided a week in advance, whether the clinic would run or not, based on demand. If the clinic was not needed, which was most of the time, then the staff scheduled for the clinic undertook other, agreed, duties. If the demand was such that the additional capacity was required, then the clinic was run – the ‘bypass zone’ was activated. This made such a difference to flow that the service was able to stop running a system with multiple queues, because everyone referred was seen quickly.
There are other examples of this. In some hospitals, emergency departments are adjacent to pre-operative assessment departments. This allows the physical space of the pre-operative assessment department to be used when required, as surges in ED department activity are often at times when most pre-operative assessment services are closed. This does not resolve staffing, of course, and the type of arrangement discussed under the ‘baton pass zone’ discussion above will often be used. The potential advantage, however, is that you are not pushing more staff in to an already crowded part of the service, which may itself affect flow adversely.
Baton pass and bypass zones are worth considering when there are problems in peak flow.
Photo courtesy of Sira Anamwong at freedigitalphotos.net