Understanding Prostate Cancer Care Process Using Process Mining: A Case Study

Prostate cancer is the fourth most common cancer in the EU27, with around 470,000 new cases yearly, and the most common cancer among males. Patients diagnosed with prostate cancer go through established procedures, and the decisions made about the treatments are vital due to cancer's unfavorable essence evolution. In this context, prostate-specific antigen tests are helpful in stratifying surveillance and subsequent risk and are monitored for relapsed detection after diagnosis and during and after treatment. Electronic Health Records store longitudinal data and record detailed cancer therapies and PSA values during this process. Incorporating this information and the temporal perspective into the risk models could stratify patients with similar evolution. The perception of clinical processes behind treatments. Applying Process Mining techniques and an interactive paradigm with the Dynamic Risk Models framework could result in the definition of new PSA evolution groups, enabling prostate cancer experts to control disease favorably and most appropriate treatments. This work uses real-world data from prostate cancer patients collected in a public hospital and Process Mining techniques to obtain new behavioral models for PSA evolution. The results represent prostate cancer care processes for different PSA evolution groups, allowing looking for awareness and differences.