DOS 2020

16 · DOS Abstracts Short knee radiographs in the evaluation of coronal alignment after total knee arthroplasty Sanne Høj Christensen, Andreas Kappel, Morgens Laursen Orthopaedic Research Unit, Aalborg University Hospital, Denmark Background: Standardized postoperative short knee radiographs radiographs serve as documentation and surgeon performance feedback following follow- ing total knee arthroplasty. Controversy Controversy regarding the relationship between between alignment measurements on postoperative and full-length radiographs radiographs are evident both scientifically scientifically and dur- ing daily conference conference with participation of non-knee knee surgeons. Measurement of mechanical mechanical coronal knee alignment from stand- ing full-length lower-limb radiographs radiographs is gold standard, alignment alignment in the range from 177-183 is considered considered neutral. Purpose / Aim of Study: To examine relationship between coronal coronal plane implant alignment measured measured from postoperative and follow follow-up full-length radiographs. Materials and Methods: Retrospective study on a consecutive cohort co- hort. Measurements of alignment using using TraumaCad™ guides. Examination Examination of intra- and inter-rater reliability reliability of the measurements, and agreement between short knee radiographs radiographs and full-length radiographs radiographs, with intraclass correlation coefficient coefficient. Evaluation of clinical relevance relevance from Bland Altman analysis and sensi- tivity analysis. Findings / Results: 138 cases were included. Intra- and inter-rater reliabil- ity of the measurements measurements was excellent, with ICC above above .95. Agreement between the methods methods was good (ICC=.81(.74-.87)). Mean mechanical tibiofemoral alignment alignment from full-length radiographs ( (mTFA) = 179 ±2.9 degrees. Mean anatomical anatomical tibiofemoral align- ment from the knee radiographs (aTFA) = 185 ± ±2.6 degrees. Mean difference between between methods = 5.8 (CI 5.4-6.1) and 95% limits of agreement 1.4 to 10 degrees degrees. Censoring of suboptimal projections projections and very short short films only only improved the results slightly. 32 full-length ra- diographs and 35 postoperative showed malalignment. Positive predictive value of a postoperative knee radiograph with malalignment malalignment was 54% and negative predictive predictive value was 87%. Conclusions: Good agreement between the methods might might justify the cautiously use of short film anatomical angulations as surrogate surrogate mea- surement of alignment. Clinicians Clinicians should be aware of the wide limits of agreement and predictive power when evaluating postoperative TKA radio- graphs. 10.