In this editorial, we comment on the article by Hu et al entitled “Predictive modeling for postoperative delirium in elderly patients with abdominal malignancies using synthetic minority oversampling technique”. We wanted to draw attention to the general features of postoperative delirium (POD) as well as the areas where there are uncertainties and contradictions. POD can be defined as acute neurocognitive dysfunction that occurs in the first week after surgery. It is a severe postoperative complication, especially for elderly oncology patients. Although the underlying pathophysiological mechanism is not fully understood, various neuroinflammatory mechanisms and neurotransmitters are thought to be involved. Various assessment scales and diagnostic methods have been proposed for the early diagnosis of POD. As delirium is considered a preventable clinical entity in about half of the cases, various early prediction models developed with the support of machine learning have recently become a hot scientific topic. Unfortunately, a model with high sensitivity and specificity for the prediction of POD has not yet been reported. This situation reveals that all health personnel who provide health care services to elderly patients should approach patients with a high level of awareness in the perioperative period regarding POD.