It's one of the biggest challenges in healthcare of this time: the staff shortage and the pressure that, partly as a result, is placed on healthcare workers. Many hospitals struggle with capacity issues, including Gelre Ziekenhuizen in Apeldoorn and Zutphen. During a visit to the hospital together with Erwin Hans, Professor of Operations Management in Healthcare, student Nienke got into a conversation with a cardiologist about the planning issues who, coincidentally, was still looking for a graduate student to work on this topic.
The graduation project was a great match with the Bachelor's in Industrial Engineering & Management. It has everything to do with Operations Management, an important module within the study programme. “In that course, you learn a lot about mathematical models that can help determine optimal decisions and processes within an organisation. For example, in planning, but also in inventory management or logistics.”
Gaining insights
And so Nienke started her thesis project by gaining insight into the hospital’s current processes. In the beginning, that mainly involves a lot of talking, she says. “I tried to speak with as many people as possible to understand everyone’s role and how the current processes run. I came to the conclusion that the doctors’ schedule lies at the heart of many processes in the hospital. And if that isn’t organised efficiently, you will notice that in many other parts of the organisation. So I decided to focus on the doctor's schedule for my graduation project.”
Putting together such a doctors’ schedule turns out to be quite a puzzle, and until now it had always been done manually. And that’s quite complex: “I tried doing it manually myself, but it’s simply impossible to avoid mistakes: you have so many factors to take into account.”
Smart algorithms
Fortunately, there’s a better way, Nienke says. “In the Bachelor's in Industrial Engineering & Management, you learn to use mathematical models that, thanks to smart algorithms, can quickly calculate many different options. That way you can easily calculate the most efficient solution.”
With such a model, it’s important to define the conditions. You can distinguish between hard and soft conditions. “Hard conditions are things that really have to happen and are non-negotiable. For example, there must always be a doctor present in certain departments, and some tasks always need to be performed. But there are also soft conditions, which aren’t strictly required but should be considered. One example is travel time. The hospital has two locations, and you don’t want a doctor to drive back and forth unnecessarily. All those soft conditions together turn the model into a multi-objective optimisation problem, which makes solving it even more complex.”
Puzzle
The next step is to translate those conditions into equations. “You implement all those equations in a programme, such as Python or - in my case - AIMMS, and then the optimal solution rolls out after a few seconds.”
It may sound simple: defining a few conditions and letting the algorithms do the work. Yet it remains a puzzle, even if you’re not doing it manually. “My model crashed multiple times. I kept wondering: what did I do wrong? But the answer was simple: the conditions were so strict that the model couldn’t produce a solution. And of course, that makes sense! If two hard requirements contradict each other, you can’t get to a solution. A doctor can’t be in two places at once!”
“In that way it was not only a mathematical learning process, but also an organisational one. You keep going back to the stakeholders to ask: is this really a hard requirement, or can we make it less strict? I noticed that this also gave the organisation many new insights, and that’s what you do it for.”
Grade of 10
The result of Nienke’s graduation project - besides the exceptional grade of a 10, the highest score you can get in Dutch education - is a working model that is currently being tested at Gelre Ziekenhuizen. And such a model could of course be applied in many other organisations as well, Nienke believes. Her work was picked up by national media such as Algemeen Dagblad and De Nieuws BV on Radio 1, and since then, the messages have been pouring in. “Many organisations ask me whether such a model could work for them too. And yes, I think in most cases it could. But defining the conditions will be different for every organisation, both in healthcare and outside of it, and that part will always remain to be human work.”
Plenty of opportunities on the job market for Nienke, then. But for now, she will focus on her Master's in Industrial Engineering & Management at the University of Twente. “In my Master’s, I gain even more insight into the algorithms behind the mathematical models you learn during the Bachelor's. I actually understand my graduation project even better now!”
