Survival analysis of dental prostheses in real-life conditions

The retrospective evaluation of medical devices, such as dental prostheses, makes it possible to measure their effectiveness and safety in real-life conditions. This project focuses on analysing the survival of dental restorations supported by multi-unit abutments. Working with a number of practitioners, our aim was to collect and analyse detailed data to better understand the performance of these prostheses in a variety of clinical settings.

This project was part of a retrospective evaluation on the safety and performance of dental prostheses. The client, wishing to delegate certain crucial tasks, has entrusted us with the responsibility of entering data from the practitioners’ source documents and carrying out descriptive and survival analysis based on biological and physical parameters. This process is essential for identifying the factors that influence the durability and effectiveness of dental prostheses, taking into account the actual conditions of use and variations between patients.

To meet the needs of the project, several key actions were implemented:

• Data capture: An eCRF (electronic Case Report Form) designed by the client was used for data capture. We collected information from the medical records provided by the practitioners, visiting their practices to ensure complete and accurate data collection.
• Data structuring: We performed data mapping to structure the information collected, creating a correspondence table of biological and statistical individuals (including implants, abutments, and/or prostheses).
• Data management: Data management solutions were implemented to identify and correct inconsistencies, thereby improving data quality. Thanks to rigorous data management efforts, we were able to recover around 20% of data considered to be missing.
• Correlation analysis: This was used to determine the analysis parameters for the survival model.

The results obtained are significant and highlight several important aspects:

• Number of patients and sites: The project involved a total of 195 patients across 3 sites.
• Project duration and data collection: The project took place over a period of 9 months, including visits to practitioners to collect the necessary data.
• Analysis and costs: The data analysis included censoring, in agreement with the customer, for an accurate assessment of prosthesis survival. Sub-analyses per centre were carried out to provide detailed reports to practitioners. The project was managed within budget, with only 2 amendments related to additional requests for data management and additional analysis topics.