Your career path has taken an extremely different route from that of most physicians. What insights has your experience in manufacturing given you when it comes to healthcare?

This applies to the clinical process that we put our patients through: the sequence of tasks by which we transform a sick patient into a well patient: presentation, history and examination, differential diagnosis, diagnostic tests, diagnosis, prognosis and plan, implementation of the plan, review, prognosis, discharge and maintenance.

Then having understood this sequence of tasks and who does what with what and when, then we can design the supporting processes by which we deliver the skills, information supplies, equipment and money to deliver each of the tasks on time, every time in full at a cost the patient, in our case, the taxpayer can afford.

In brief, what are the main principles behind the most effective way of managing patient flow?

1. Understand the demand: number of patients requesting a service and the day to day variation in demand.
2. Understand the sequence of tasks involved in the clinical process(es) to deliver the service (s) they require. This is sometimes referred to as the case mix.
3. Understand the cycle time required at each resource, and the constraint resource, the rate limiting step which will govern the flow of the entire process. The cycle time is the time a resource starts on one patient to when they are ready to start the same task on the next patient. It includes the touch time (when the patient is in front of the resource) and all the set-up, clear down and administrative tasks such as setting up equipment, cleaning the room, writing the notes, etc. i.e. all the work involved in performing and completing the task for each patient.

4 . Match the flow capacity of the constraint resource to the workload: demand x cycle time for that resource. However,

What kind of hurdles are faced by staff when they first encounter your way of managing patient flow, and how do you help them overcome these issues?

Most staff are experts at their tasks within the clinical process but they are physically and mentally constrained by the organisation silo in which they sit. E.g. GP, radiology, A&E, ambulance, social services etc.

So, the first task is to ask them to task the patient’s view of the system in which they work and physically map the process which often the patient is the only one to see.

The next teaching task is to introduce them to system science – the design of operations design and management. The staff with a scientific background (doctors, radiographers, lab scientists, physiotherapist etc.) have no problems with this: they recognise the process principles from physiology and biochemistry.

Unfortunately, the vast majority of operations managers in the NHS come from a nursing background, and while they have excellent people skills, they do not have the training and are therefore inept (as defined by Atul Gawande in his Reith Lectures two years ago). As a consequence, they really struggle to bridge the gap between their rhetoric and reality, getting very defensive and aggressive. So, we know that

The third teaching task is measurement. All science depends on measurement. Unfortunately, there are two handicaps to measuring anything related to healthcare:

1. The data is hidden away in databases that are governed by data analysts with no contact with the clinical process and who struggle to interpret and retrieve the data correctly.
2. Doctors and academics are only taught a branch of static statistics by which we can compare populations. This is a real handicap as we need dynamic statistics, statistical process control, to understand the performance of a dynamic system over time. So even though clinicians use dynamic records of a system over time to diagnose and monitor every patient’s physiology, they are unaware that there is a whole branch of statistics for doing this. So, they resort to the only tool they know and use comparative statistics. The recent paper comparing telephone consultation with face to face consultations is a really example of this: the academic team ‘lumped’ all the GP telephone consultation systems into one group where the dynamic data plots showed that this was not one homogeneous group. This comes as a real shock, particularly to eminent professors. Again, they struggle to bridge the gap between their experience and this reality and then an interesting dynamic comes into play as they ‘suppress’ any exposure of this gap: Galileo had the same problem!

To be continued.