From chatbots to brain monitoring: modern tools to predict postoperative delirium
Mariska Oosterhof-te Pas defended her PhD thesis at the Department of Electrical Engineering on April 2.
The global population is aging rapidly, and an increasing number of older adults undergo major surgical procedures, such as cardiac surgery. While these treatments can prolong life and improve its quality, they also carry significant risks. One of the most serious complications is postoperative delirium (POD): a sudden and temporary state of confusion that mainly affects older and vulnerable patients. POD is often associated with prolonged hospital stays, accelerated cognitive decline, loss of independence, and increased mortality. In addition, some patients experience longer-lasting problems with memory and concentration after surgery, referred to as postoperative cognitive dysfunction (POCD). In her PhD research, Mariska Oosterhof-te Pas investigates how patients at increased risk of these complications can be better identified before surgery.
Although international guidelines recommend cognitive screening in older patients, this is still rarely implemented in daily clinical practice. Time pressure, staff shortages, and the lack of accessible and low-burden assessment tools all contribute to this gap. 鈥檚 research evaluates both established and innovative methods for identifying cognitive vulnerability and detecting POD in older surgical patients. These methods include a digital chatbot-based questionnaire, a gamified self-administered cognitive test (BAMCOG), simplified EEG-based brain monitoring (DeltaScan) to detect acute brain dysfunction, and measurement of the stress hormone cortisol in blood, saliva, and sweat as a potential biomarker for POD.
Evaluation results
The results show that patients appreciate digital and interactive tools. The chatbot is perceived as more pleasant and motivating than traditional questionnaires, and BAMCOG proved to be feasible and well accepted. Although BAMCOG is not yet suitable for detecting subtle postoperative cognitive changes, it shows promise in identifying patients at increased risk of POD. Comparisons of postoperative delirium outcomes between clinical observation, delirium screening scales, and EEG measurements revealed clear differences in delirium detection. These discrepancies underline the absence of a single gold standard and highlight the complexity of diagnosing POD.
Promising methods
Overall, this thesis demonstrates that innovative, patient-friendly screening methods hold great promise for the early identification of cognitive vulnerability and the prevention of perioperative neurocognitive complications, including POD and POCD. Further development and research may contribute to better preoperative preparation, prevention of complications such as delirium, and improved recovery and quality of life for older patients.
Title of PhD thesis: Supervisors: Prof. (Catharina Ziekenhuis), Prof. Marcel Olde Rikkert (Radboudumc) and Dr. Marc Buise (MUMC+).