How to increase production by finding and optimizing the constraint or bottleneck in a (hospital) process?
‘The weakest link determines the throughput in the system as a whole’, with this sentence I ended up my first blog in a series about the use of TOC in healthcare. What does this mean? Let’s take a closer look at a typical hospital process:
OUTPATIENT CLINIC => RADIOLOGY DEPARTMENT => OR => WARD There’s a patient flow through several independent departments in the hospital. But are they really independent? Each department has its own staff, manager, budget and goals, BUT, for the treatment of the patients they are dependent of each other. The patients have to go first to the outpatient clinic, then have a X-ray etc. It’s not possible to change this order of activities. What are the consequences of this? Let’s assume the several departments have the following capacities:
Outpatient clinic: 100 patients/day
Radiology department: 80 patients/day
Operation Room: 50 patients/day
Ward:100 patients/day
What is the daily throughput of patients through the system? It might be clear that the OR, which has the lowest capacity of 50 patients determines the maximum throughtput through the whole system. In this case the OR is the weakest link in the treatment chain.
The efficiency syndrom
But what happens in praxis? All the managers of the several departments try to work as efficient as possible. Every expensive recourse must be utilized to the maximum, time is money, treat as much patients a day as possible… But does this make sense? No, since we know that the throughput is determined by the weakest link or the constraint in the process, which has a capacity of 50, it doesn’t matter if the radiology department treats 60 or 80 patients/day, the maximum throughput in the system stays 50. Optimizing a non constraint even reduces the performance of the whole system. The effect of optimizing the radiology department has a negative effect, the waiting list for the OR is getting longer and longer! This means that there are coming more and more patients in my system, and more patients in the system reduces the performance of the total system because every patient needs attention, administration etc. And the only thing they do is waiting for a treatment, they don’t contribute to the performance of the system but reduce it! From this point of view it isn't contra productive when some of the staff members have nothing to do and sit back, it's even better for the performance of the sysem. But this is a difficult point to explain to managers, because it's not their nature to let people do nothing. Of course this is only valid when the department is NOT the constraint.
So what do we have to do to increase the performance of the whole system? Exactly, find the weakest link or constraint and strengthen it. Becauseknow by increasing the capacity of the constraint, the capacity of the whole system increases: When the OR can now treat 65 patients/day, the total throughput of the system increases to 65.
This is actually one of the basic concepts of the Theory of Constraints, which is defined in the five focussing steps.
The five focussing steps of TOC
These are the five steps you have to make to increase the throughput of your process, system or hospital:
Step 1: Find the system constraint
Step 2: Exploit the system constraint
Step 3: Subordinate everything to the constraint
Step 4: Invest in the system constraint
Step 5: Go back to step 1
Step 1. Identifying the constraint is not always as obvious as it seems to be. An example of a hospital in the UK that implemented TOC: One of the problems of the hospital was the very long waiting time for the radiology department to get a MRI scan. Everybody was complaining about this, the capacity of the MRI scanner was insufficient and it was absolutely necessary to buy a second MRI to reduce the waiting time: investment about 1.5 million euros. It was obvious that the MRI scanner is the bottleneck because it has long waiting timr => less capacity. But when the TOC experts analysed this situation they came to another conclusion. They found out that the MRI scanner was standing idle 30% of it’s time, no patients were examined. What was wrong? Well, in the last year the hospital had to save money and they reduced drastically the number of employees responsible for the internal transportation of patients. There was not enough capacity available anymore to bring patients from the wards to the radiology department in time, and the MRI scanner had nothing to do. So the constraint was not the MRI scanner but the transportation service. This may sound silly to you, but this kind of situations occur very often.
Step 2. You have to assure that the constraint is always working, because when it does not, the throughput in the whole system is blocked! This means for example that if you find out that the MRI scanner is really the constraint, you have to plan the maintenance for the scanner in the weekend or in the evening and not during the day, because in that period you can’t examine any patients, which means no throughput in the system as a whole because the throughput is detemined by the constraint.
Step 3. You have to subordinate everything to the decisions you made in step 2. This means for example that you have to take a close look at the activities the constraints is doing and ask yourself the question if some activities can be done by another machine, doctor or whatever, depending on what the constraint is. The constraint is only doing these kind of activities that can only done by the constraint!
Step 4. When you have done this, optimized the use of the constraint as much as possible and still there isn’t enough capacity, then you can decide to invest in the constraint e.g. buy a new scanner, hire another surgeon etc. Our experience till now is that there’s so much hidden, unused capacity available in hospitals that we never came to Step 4. Just by using this hidden capacity you can increase your throughput/productivity by 20-40%! But when you have to invest, you invest in the bottleneck of your process and this means by doing so, you increase the productivity of your whole system! This is a perfect way to justify the investment to the board because you know exactly what the extra revenues will be.
Step 5. After identifying and solving the constraint or bottleneck, automatically a new constraint or weakest link will show, because that’s the definition of a chain: there's always a weakest link. May be now the radiology department is the constraint and you have to start again. This means that TOC is a method of ongoing improvement. This is necessary because we have to deal we a very dynamic, changing environment and we have to adapt our processes to this.
Next time I will tell more about the specific use of TOC in several departments Wards, A&E departments etc, and how TOC is implemented.
