Healthcare utilization patterns of cancer patients contain many insights into health outcomes that traditional demographic and clinical variables alone do not capture. In this talk I introduce a framework using Dynamic Time Warping (DTW) to analyze temporal healthcare visiting trajectories. By structuring each patient's medical events into ordered sequences—including inpatient admissions, outpatient visits, and ambulatory care—we used DTW to compute distances between sequences, then applied hierarchical clustering to identify nine distinct patient trajectory clusters. This hierarchical clustering approach shows that these trajectory-based clusters provide significant predictive power for mortality, anxiety, and depression outcomes beyond what is explained by traditional covariates including age, gender, income, education, and diagnosis codes, with some high-risk clusters showing significantly elevated mortality odds ratios. This temporal stratification approach could be used for identifying high-risk patient groups and could lead to targeted interventions in cancer care, providing a valuable addition to existing risk assessment methods in oncology practice.