ELECTRONIC PATIENT REPORTED OUTCOMES (EPRO)- THE BEST DEFENSE IN PREVENTING MISSING PRO DATA
Author(s)
Ross J*;Ross E, Holzbaur E Almac Clinical Technologies, Souderton, PA, USA
This session will illustrate how ePRO is a powerful approach for preventing missing data; explain how ePRO is more effective as compared to paper; and demonstrate how ePRO techniques can be implemented to prevent missing data. ePRO use can be an effective solution for preventing missing data as compared to paper data collection. Missing data is common in PROs and can result in significant problems for data analysis. While using a robust statistical plan for handling missing data is beneficial, studies still can suffer with high levels of missing data. One major contributing factor is the collection method. Many PROs are still administered in a traditional paper format which can result in high levels of missing data. This presentation will illustrate how ePRO can prevent missing data through providing examples of various ePRO techniques that can be implemented. Primary ePRO techniques to minimize missing data include: hard edit check to eliminate patients from skipping items or pages; reminders with real time technology to remind patients to complete their PROs; alerts to study staff of patient non-compliance; programmed logic to reduce erroneous entries and contradicting responses; time stamped data entry to ensure assessments are completed within the given window; and storage of directly entered data with back-up can ensure data is not lost. ePRO can prevent missing data, improve patient compliance and result overall in high quality data. ePRO eliminates many of the issues associated with data loss in traditional paper-based PRO systems. Future PRO development efforts should focus on creating more electronic versions of PRO instruments. Wider availability of ePRO instruments across therapeutic areas would ultimately result in high quality data and reduced missing data.
Conference/Value in Health Info
2013-05, ISPOR 2013, New Orleans, LA, USA
Value in Health, Vol. 16, No. 3 (May 2013)
Code
PRM224
Topic
Methodological & Statistical Research
Topic Subcategory
Confounding, Selection Bias Correction, Causal Inference
Disease
Multiple Diseases