- Richard Boucherie is working on optimising the logistical process of healthcare through mathematical models so that healthcare personnel can work more efficiently during the corona crisis.
- The average number of patients in the ICU can be predicted more efficiently by using data and prediction methods. A good response to this will reduce the workload.
Professor Boucherie co-founded the Centre for Healthcare Operations Improvement and Research (CHOIR). He has been working on optimising the entire logistics process of healthcare since 2003. CHOIR collaborates with about half of all Dutch hospitals. “We look at it with a fresh outside perspective. A wealth of data is available within healthcare, such as the number of treatments, waiting lists, surgery durations and admission data. Models that were previously used in telecommunications, for example, to optimise service delivery, can also be applied to healthcare. More and more data is available, and you have to properly link this to the models”, he explains.
The corona crisis conclusively showed how important it is to predict how many patients there will be and how they can be distributed as efficiently as possible. “There are several factors that can predict how long a corona patient will be in the hospital, such as BMI, age and gender. You can calculate how many short- and long-term admissions you will have per hospital if you know how many admissions you have now and how infection rates are developing in the region. You don’t just look at how busy it is right now; you also look at how busy it will be tomorrow, the day after tomorrow and in a week. You will be able to detect bottlenecks much earlier by combining this data with that of regular healthcare, and you will be able to make well-founded decisions.”
These models can contribute to keeping healthcare manageable even after the crisis, according to Boucherie: “How can we provide as much care as possible with the resources we have now? You cannot only look at the number of ICU beds or artificial respirators. You also have to look at the number of nurses required.” Healthcare is often all or nothing, which makes the work physically and mentally demanding. “You can never eliminate those peaks, but you will be able to predict them much better. Combining historical data from, for example, the emergency room, the operating theatre schedule, and other patient flows will allow you to predict how many patients will be admitted each hour, which in turn allows you to adapt your rosters to how much personnel is needed. You will have more personnel available when they are needed, which distributes the workload in a much better way. Fewer mistakes are made when people always experience a roughly equal and easy-to-handle workload, less personnel leaves healthcare, and the quality of healthcare improves,” is what the mathematician predicts.
But do hospitals have the time to deliver floods of data? According to Boucherie, they do not have to. “Most hospitals have a great hospital information system in which a lot of data is already stored. We can already make a start with limited data. We can gradually enter more data into our models, which will then be able to provide more and more benefits. Experience shows that people become increasingly enthusiastic once they can see that it works. The corona crisis also gave this data-based approach some momentum, of course. We were completely focused on the direct care of patients at first, but we realised after the first wave that we should be working on improving the processes.”
These models can also prove their worth when regular healthcare restarted. “You can just start doing as many surgeries as possible, but it is better to make a data-driven plan that thinks of all parties involved: From blood tests to the x-ray department, from the outpatient clinic to the operating theatre and the physiotherapist. We made a plan like this for a balanced workload distribution for all personnel involved. It gives a lot of peace of mind if everything fits together well, and people know what the workload will be like a few months in advance.”
Boucherie emphasises that the researchers will not replace the board of directors. “We calculate various models based on data and create several scenarios from that. The choices that follow are managerial. They will, however, be well-founded and calculated. That makes making such decisions much easier and clearer than if you had to make them based on a hunch.”