Temporal Data Mining on the Stay Time of Outpatients and Treatment Processes

In this paper, we attempt to analyze the relationships between the stay time of outpatients and treatment processes they received based on the temporal pattern mining algorithm proposed by Batal et al. We could observe MPTPs (Minimal Predictive Temporal Patterns) of treatment processes containing 'co-occur' relations as well as injections in the classes where the patients spent long time in receiving outpatient services. It suggested that a visit needs long time if treatments or examinations should be executed during the consultation, or if injections such as intravenous drips were involved in the process.