Application of a statistical software package for analysis of large patient dose data sets obtained from RIS

Area of Science:

• Medical Imaging Informatics
• Radiology Data Analysis
• Healthcare Information Systems

Background:

• Radiology Information Systems (RIS) generate vast amounts of data crucial for audits.
• Manual analysis of RIS data for patient dose, clinical audits, and workload is time-consuming.
• Inconsistent data formatting across hospitals complicates analysis.

Purpose of the Study:

• To automate the analysis of RIS data for patient dose audits, clinical audits, and workload.
• To develop a consistent and efficient method for processing RIS data from multiple hospitals.
• To reduce the time and resources required for RIS data reporting.

Main Methods:

• Utilized a Structured Query Language (SQL) database for data collection.
• Employed Statistica, a statistical package, with Visual Basic coding for automation.
• Developed macros to standardize data formats and automate analysis of exposure factors (kV, mAs) and entrance surface dose.
• Generated statistical measures (mean, standard deviation, standard error) and graphical representations of dose trends.

Main Results:

• Automated analysis reduced reporting time from up to 1 day to approximately 1 hour.
• Macros ensured consistent data formatting across different hospital sites.
• Detailed analysis by exposure factors, room, and gender provided comprehensive reports.
• Enabled more frequent and accurate dose audits due to increased efficiency.

Conclusions:

• Automated analysis of RIS data using Statistica Visual Basic macros is highly effective for dose audits and workload analysis.
• The automated process significantly reduces resource requirements, facilitating continuous dose audits.
• Standardized data analysis improves the accuracy and detail of reports, benefiting clinical decision-making.