Interrupted time series (ITS) analysis is a quasi-experimental design for evaluating the effectiveness of health interventions. By controlling the time trend before the intervention, ITS is often used to estimate the level change and slope change after the intervention. However, the traditional ITS modeling strategy might indicate aggregation bias when the data was collected from different clusters. This study introduced two advanced ITS methods of handling hierarchical data to provide the methodology framework for population-level health intervention evaluation.
The interrupted time series analysis was used to evaluate the incentive effect of the management methods of the SCI thesis fund for scientific research in West China Hospital of Sichuan University. We found an increase in number of the SCI papers and the growth rate after the adoption of scientific research incentive measures, indicating that the management methods of the SCI thesis fund had the incentive effect of scientific research. The interrupted time series analysis could be used in the incentive analysis of scientific research.