Shewhart p-charts are one of the most common SPC methods, and the prototypical chart for binary attribute data. It is typically rare for other types of charts to be applied to this type of data, especially in the healthcare setting. However, extensive SPC optimization work conducted by HSyE indicated that, after appropriate data processing, exponentially weighted moving average (EWMA) and moving average (MA) charts substantially outperformed p-charts in the detection of surgical site infections (SSIs). EWMA and MA charts detect smaller changes faster than p-charts, which is an important characteristic for early detection of SSI outbreaks. Additionally, more sophisticated SPC methods that use rolling baselines and lags can substantially impact performance.
To develop a Monte Carlo simulation of moving window and lag sizes to determine the impact of these design parameters on SPC chart performance.
Partners and ResearchTeam
From the Healthcare Systems Engineering Institute:
Postdoctoral Fellow: Iulian Ilies, PhD
Undergraduate Students: Nathan Holler, Erin Joyce, Adam Schleis
Project Manager: Margo Jacobsen
Results are pending the completion of the project.