Introduction

The Telos device (TD) has shown excellent intraobserver and interobserver reliability for posterior drawer (PD) stress radiography. However, it is technically difficult to use, time consuming and expensive. The kneeling view has been proposed as a more cost effective and accessible approach. However, it failed to show similar precision to the Telos device in small comparative studies. We hypothesized that this could be attributed to unreliable patient and cassette positioning.
Aims: Evaluation of the accuracy and reliability of a new self-designed kneeling system for posterior drawer radiographs compared to the TD in a large prospective study.

Methods

Between January 2015 and July 2019, all patients presenting at our clinic with unilateral PCL insufficiency were included. Before the initiation of the study, a kneeling system (KS) was designed to allow simple and reproducible patient and x-ray cassette positioning. Bilateral kneeling stress radiographs were performed using both our KS and the TD to allow comparison of the amount of PD between the injured and uninjured knee. Six independent observers (4 residents and 2 faculty members) were involved. All radiographs were measured according to the landmarks described by Jacobsen and Staubli, by each of the investigators on 2 separate occasions, at a minimum of 2 weeks apart. The results were expressed as the side-to-side difference (SSD). Statistics included consistency-of-agreement Intraclass Correlation Coefficients (CA-ICC), absolute-agreement ICC (AA-ICC) between raters, as well as Lin's concordance correlation coefficient for the comparison of the two methods. The mean differential bias between the two-measurement methods was estimated according Bland and Altman.

Results

2736 x-rays measures were performed on 57 patients. For SSD measure, CA-ICC was 0.87 [95% CI, 0.82- 0.91] with the TD, and 0.72 [95% CI, 0.63-0.80] with the KS. AA-ICC for the SSD was 0.86 [95% CI, 0.81-0.91] for the TD and 0.71 [95% CI, 0.62-0.79] for the KS. The KS and TD offered a concordance agreement for SSD measures of all observers at 0.80 [95% CI, 0.71-0.89]. The KS slightly underestimated the TD measurements for SSD with an average of 1.2mm [95% limits of agreement: -3.6, 6.1]. This variability of SSD measured with the KS increased with the magnitude of the instability. SSD measured with the KS can be recalibrated by computing: (TD SSD – 0.97)/0.72. No statistical significant difference could be observed between trial 1 and 2 measurements for SSD with KS for all observers.

Conclusion

This new KS for stress radiography in posterior knee instability demonstrates high interobserver and intraobserver reliability. Compared to the TD, it offers good concordance even if it slightly underestimates the SSD. This difference is not clinical relevant and can be adjusted with a simple correction formula. This KS might offers an easier and cost effective alternative to the TD.