Design, implementation, and comparison of methods for collecting implant registry data at different hospital types.

Practical issues surrounding the official establishment of a national arthroplasty registry in the United States remain. The purpose of this study was to compare compliance and accuracy rates associated with 3 methods for voluntarily collecting implant registry data at 3 different hospital types. Methods examined included (1) scannable paper forms, (2) online forms comprising keypunching for implant data input, and (3) the same electronic form but incorporating barcode scanning for implant data entry. Overall compliance was low (930/1761; 52.8%) and decreased with each successive data collection phase. Total accuracy rate was 62.5% (578/925) and varied significantly among sites (P < .001). Even with relatively simple reporting systems, compliance was poor. This emphasizes the need for direct surgeon involvement, strict oversight, and a feedback system to ensure validity, particularly if a volunteer-based system is used.