Presentation (CTI)

OBJECTIVES: The primary goal of this work was to examine the variability in wording (with respect to attribute and concept of interest), directionality (whether higher or lower scores indicate better outcomes) and response options in PGI scales used in oncology trials.
METHODS: A targeted review was performed using a predefined set of search terms such as "patient global impression," "global impression," "PGIC," "PGI-I," and "PGI-S" in oncology trials that started between January 2020 and December 2025. Data extraction was automated using a script created with Copilot in pgAdmin to extract the information, ensuring that only relevant studies were included. After removing duplicates, the final dataset comprised 125 unique clinical trials and 325 endpoints for analysis.
RESULTS:

• Directionality: 80% of scales interpreted lower scores as better outcomes, while 20% used the opposite. Different direction of PGI scales may increase the risk of interpretation errors.

• Scale wording: A qualitative assessment of the scales’ wording revealed substantial heterogeneity with regards to the attribute assessed (e.g., severity, frequency etc.) and the concept of interest (e.g., a symptom).

• Response options: Both 5-point (40.6%) and 7-point (40.0%) scales were commonly used. Notably, 19.4% of endpoints did not report the number of response options at all.
• Reporting gaps: Inconsistent reporting on scale structure and directionality was common.

CONCLUSIONS: The findings highlight substantial variability in how PGI scales are named, worded, structured, scored, and reported across oncology trials. This heterogeneity is important because PGI scales are widely used for psychometric assessment and for estimating meaningful change thresholds in other PRO measures. Using differently worded global scales with varying response options across studies may introduce noise, yield inconsistent threshold values, and limit interpretability. Therefore, greater consistency in item wording and response options is needed to enable more robust interpretation and more meaningful use of PGI-based evidence in oncology research.