As a primary source for learning from lessons, patient safety event reporting systems have been widely adopted. Nevertheless, underreporting and low quality of reports pervade the system. To address these issues, the study proposed two text prediction functions as data entry aids to system users. With 52 subjects, a two-group randomized experiment was conducted to quantify the impacts in terms of the reporting efficiency, quality, and usability attitudes. Consequentially, on structured data entry, the results were an overall 13.0% time reduction and 3.9% increase of response accuracy with the functions; on unstructured data entry, there was an overall 70.5% increase in the text generation rate, a 34.1% increase in the reporting completeness score, and a 14.5% reduction on the amount of text fields ignored by subjects. Subjects' usability attitudes were slightly improved with the proposed functions according to the questionnaire results.