Ahead of its inaugural international seminar—the International Conference on Data Science and Official Statistics (ICDSOS)—which will be held this coming November, Politeknik Statistika STIS organized the ICDSOS Pre-Conference on Wednesday (2/6). The event was conducted virtually via Zoom Meeting and YouTube Live Streaming and moderated by Nucke Widowati K.P, Ph.D. This Pre-Conference marked the beginning of the ICDSOS 2021 series and featured Prof. Pedro Pita Barros from Universidade Nova de Lisboa, Portugal, as the speaker. Centered on the theme of health, he delivered a presentation titled “Shared Decision Making in the Portuguese NHS: Choosing Wisely?”

Prof. Pedro Pita Barros began his presentation by explaining the motivation behind his research, which was the implementation of the Patient Choice Initiative (PCI) in Portugal in 2016. This initiative enables patients, together with their general practitioners, to decide which hospital (regardless of location) to choose for treatment, as long as the required specialty is offered by that hospital. Previously, hospital selection was based mainly on proximity to the patient’s residence, and the options presented were limited to geographical closeness and average waiting times per specialty at each hospital.

The research framework was built on the fact that hospitals have asymmetric waiting times, and it was anticipated that patient responses in choosing hospitals would reflect this asymmetry. Thus, the sample was divided into two levels: hospitals with high waiting times and those with low waiting times. The inferential analysis aimed to examine:

1. Whether there were different changes in waiting times between hospitals with high and low waiting times;
2. Whether there was a shift in the number of appointments or patient redirection to hospitals with shorter waiting times based on publicly available performance data; and
3. Whether there was an increase in the geographic dispersion of patients going to hospitals with shorter waiting times.

The data used in this research came from publicly accessible hospital datasets (link), Diagnostic Reference Group patient data, and information on patients’ residences.

The data was then processed using regression analysis, comparing the periods before and after the PCI implementation in 2016. The difference-in-differences method was not applicable because PCI was implemented simultaneously across all hospitals in Portugal. The key findings of the study were: hospitals that already had better performance in terms of waiting times prior to PCI experienced greater geographic dispersion in patient origins, a negative evolution in specialist appointment waiting times, and a weak correlation between demand and past performance.

A major limitation faced in this study was the availability of only aggregate hospital-level waiting time data, rather than individual patient data. Therefore, future studies could include individual surveys to capture patient-level waiting times. Additionally, further research might involve direct responses to waiting times from nearby hospitals, incorporate the effects of physician service payments, and employ normative analysis.

The successful implementation of this ICDSOS Pre-Conference is hoped to be a strong foundation for the successful execution of the larger ICDSOS 2021 event this November. With the hosting of such an international-scale activity, Politeknik Statistika STIS is expected to take a significant step toward becoming a globally recognized academic institution.