The first important findings in this study were that the 3D measurement of the knee flexion angle better correlated with PROM compared with the goniometer measurement and the WB knee flexion angle better correlated with PROM compared with the NWB condition. The second important finding was that the angular difference between NWB and WB correlated with PROM. To our best knowledge, this paper is the first report to discuss the positive correlation between WB knee flexion angle and PROM using a precise 3D measurement. Additionally, the correlation between angular difference and PROM was reported.
The NWB knee flexion angle using 3D measurement correlated with 2011 KSS symptoms and 2011 KSS functional activities, although the angle using goniometer measurement did not correlate with PROM at all. This result suggested the accuracy of 3D measurement  and the possibility of the inaccuracy of the goniometer measurement [11, 12]. A recent report described that the standard deviation of the long arm goniometer measurement was 4.9° . Additionally, goniometer measurement with 5° increments may be the reason for the inaccuracy in this study. Thus, the evaluation of knee flexion angle by only the goniometer measurement would be inadequate to precisely measure and investigate the relationship with PROM. The measurement of the knee flexion angle using 3D is desirable for precise assessment. However, the manual goniometer is widely used in clinical practice where the 3D measurement is not set up. Thus, the surgeons should keep in mind the potential inaccuracy of the manual goniometer.
The results of the present study showed that the knee flexion angle in the WB condition correlated more with PROM than in the NWB condition in precise 3D measurement. This result suggests that the assessment of the WB angle may be more reasonable and appropriate than the NWB to evaluate the relationship between the angle and PROM. The results of this study were compatible with a previous study by Song et al.  reporting that active flexion in WB better correlates with the functional outcomes of TKA than passive flexion. However, the study by Song et al.  measured the angle using a goniometer. Therefore, the results may lack accuracy because of the measurement method [11, 31]. This study confirmed the superiority of the evaluation of WB condition by precise 3D measurement. Therefore, this is a strong point of the present study. One possible reason why the results of this study showed more correlation in WB than in NWB may be attributed to the fact that a substantial portion of ADL are performed under WB condition . A previous report revealed that the activities requiring WB knee flexion (e.g., climbing up or down a flight of stairs, getting into and out of a car, and squatting) was very important and distressing for patients who have undergone TKA . Thus, the ability to squat may be required in the present study. Based on the results of this study, to appropriately evaluate the relationship with PROM, the surgeons should evaluate not only the NWB knee flexion angle but also the WB knee flexion angle.
Significant negative correlations were found between angular difference and PROM in 2011 KSS functional activities, KOOS ADL, and KOOS sports. This result suggested that the larger the angular difference existed between NWB and WB, the lower the PROM score. The present study showed that the WB knee flexion angle was 8.5° ± 12.8° smaller than the NWB knee flexion angle. This result was consistent with the study by Dennis et al.  reporting that the WB knee flexion angle was smaller than the NWB in TKA knees. However, the relationship between the angular difference and PROM was not described in their study . This study confirmed the relationship between the angular difference and PROM, which may be meaningful. It is difficult to explain why some patients have larger angular difference despite some patients acquiring as much WB flexion angle compared with that in the NWB condition. Muscle strength may be related although the muscle strength of the lower extremities could not be evaluated. Based on the results of this study, acquiring as much WB flexion angle compared with that in the NWB condition may potentially lead to higher PROM in terms of postoperative rehabilitation. However, whether the rehabilitation emphasized to obtain more WB knee flexion angle is effective or not is still unknown. Hence, further case–control and interventional research are necessary. Additionally, surgeons should recognize whether the angular difference exists or not by checking not only the NWB flexion angle but also the WB flexion angle.
The clinical relevance of this study was that the WB knee flexion angle was better correlated with PROM compared with the NWB knee flexion angle. Thus, surgeons must recognize the importance of the WB knee flexion angle and should evaluate not only the knee flexion angle in the NWB condition but also that in the WB condition.
However, several limitations in this study should be acknowledged. First, the single implant design was evaluated in this study. Other types of TKA implant might present other results. Second, the alignment of implants was not evaluated in this study. Third, the knee flexion angle and PROM were only evaluated postoperatively. The preoperative angle and PROM may have influenced the postoperative results. Fourth, the timing of the evaluation of the KOOS score was relatively ranged among patients because the timing of the KOOS score depended on the timing of the fluoroscopic survey. Fifth, the knee position during the knee flexion differed in the same NWB condition between the goniometer and fluoroscopic 3D measurement. In the NWB condition using a goniometer, the knee position was performed in a usual clinical manner. However, in the NWB condition using fluoroscopic 3D measurement, the knee position when sitting on the chair was applied to suitably capture the knee in fluoroscopy. Sixth, the measurement method was different between the goniometer and 3D measurement. The flexion angle using a goniometer was measured between the femoral and tibial bones, whereas the flexion angle using 3D analysis was measured between the femoral and tibial components. Lastly, the muscle strength of the lower extremities was not evaluated. The muscle strength may influence the knee flexion angle, particularly in the WB condition.