Abstract
Introduction
Complication reporting and assessment is an important part of orthopaedic trials assessing new technologies. Because the reliability of the assessment by the treating surgeon compared to central review is still unknown, it was quantified in this study and possible patterns were identified.
Materials and methods
176 patients with trochanteric fractures, treated with a trochanteric nail, were included in a prospective multicenter study. Surgeons were encouraged to report honestly every single potential complication, to rate severity, most likely cause, relation to implant, and to report the outcome of the complication. After 1-yr follow-up, 3 experienced orthopedic surgeons reassessed independently the same variables (agreement determined using kappa coefficient). Discrepancies were resolved by consensus.
Results
Surgeons rated sig. fewer complications as severe than central reviewers (88 complications: 59% mild, 27% moderate, and 14% severe vs. 47% mild, 26% moderate, and 27% severe, p<0.001, kappa 0.47). Surgeons attributed more complications to the tested implant (10 vs. 0 by reviewers); in contrast, reviewers defined more complications as unlikely related to the implant (21 vs. 10 by surgeons) but attributed more complications to surgery/surg.technique (12 vs. 8).
Discussion
The analysis revealed significant differences in the complication assessment between treating surgeons and central review and highlight the need for central complication assessment.
