Section: Application Domains

Modeling and optimizing patient pathways in hospital

This research theme was inspired by the observation that the practice of modern industrialized medical care proceeds by queueing and forwarding patients from one step of care to the next, with each step involving specialized personnel and machinery. Whereas the human competence and machine performance available in hospital are both highly evolved, the deployment of these resources for the patient's benefit is problematic.

If we change our point of view and, rather than look in the traditional way at the hospital as a collection of “vertical” silos (emergency room, cardiac ward, respiratory ward, operating rooms, imagery services, blood services, logistics, etc.), we instead consider the patient's “horizontal” trajectory crossing through many different silos, we can see significant problems. In between each step, whose value added part typically lasts at most a few minutes, there are long stretches of time (on the order of hours) during which the patient simply waits for the hospital to arrange the next step of care — with serious consequences:

- delayed treatment is directly correlated with increased rates of hospital acquired nosocomial infection for the immuno-compromised and to loss of autonomy for the elderly;

- the cost in terms of scarce resources (for example, bed-hours in hospital) is significant, but the root causes of these costs due to problems at the interfaces between silos are hard to observe since they do not fit neatly into the hospital's traditional hierarchy;

- the slow and error prone hand-offs of information between silos are dangerous for the patient who is vulnerable to medical errors. In the US, for example, it is estimated that there are 100,000 deaths per year attributable to hospital error (triple the number due to road traffic accidents by way of comparison).