Discovery of general practitioners’ workflow while interacting with a decision support system

From Master Projects
Jump to: navigation, search

About Discovery of general practitioners’ workflow while interacting with a decision support system

  • This project has not yet been fulfilled.
  • This project fits in the following Bachelor programs: {{#arraymap:|, |xXx|bachelorproject within::xXx|,}}
  • This project fits in the following masterareas: {{#arraymap:Knowledge Technology and Intelligent Internet Applications, Human Ambience, Technical Artificial Intelligence|, |xXx|project within::xXx|,}}


|free text=}} INTRODUCTION: Clinical practice guidelines (CPG) are meant to help healthcare professionals to make the right decisions about the diagnosis and treatment of their patients. However, healthcare professionals often do not adequately follow these paper-based guidelines. Computerized decision support systems (CDSSs) have been shown to have the ability to improve adherence to these guidelines by giving computer-generated recommendations, reminders and alerts directly to the healthcare professional.

PROBLEM: Despite the availability of decision support systems, doctors do not always use them or they ignore the given recommendations. Doctors often state that they have no time to follow these recommendations, or that there are too many recommendations most of which are irrelevant at the moment that the recommendation is provided to them. So although we have evidence that CDSSs change behavior of doctors, we do not know how they interact with such systems. What is needed is an analysis to discover, from log files, the details of their workflow while interacting with the CDSS. Such a workflow description would allow us to understand their behavior better and to design better systems. We have the opportunity to analyze log data available of about 40 general practitioners (family doctors) working with a CDSS in Amsterdam ZuidOost in which we know the mouse behavior (where do their hover on, what do they click on, how long they keep a message before dismissing te etc).

RESEARCH QUESTIONS: What is the workflow of general practitioners while interacting with a decision support system? And how good does it predict future actions of general practitioners?

RESULTS: The main result is a workflow model describing what general practitioners do during patient visits while interacting to a CDSS. This model is obtained by process mining the log files. In addition the predictive performance of the model will be quantified. The work should result in a scientific report (thesis or scientific paper).


Prof. dr. Ameen Abu-Hanna, Dept. of Medical Informatics, AMC