Concurrent validity and intrarater reliability of two ultrasound-based methods for assessing patellar tendon stiffness

Abstract

Introduction: Assessment of tendon stiffness in vivo traditionally involves maximal muscle contractions, which can be challenging in pain populations. Alternative methods are suggested, although the clinimetric properties are sparse. This study investigated the concurrent validity and the intrarater reliability of two ultrasound-based methods for assessing patellar tendon stiffness.

Methods: Patellar tendon stiffness was assessed in 17 healthy adults with (a) the dynamometer and B-mode ultrasonography method (DBUS) and (b) the strain elastography method. Correlations between the two methods were analysed using Kendall's Tau-b. The relative reliability of both methods was evaluated using intraclass correlation coefficient (ICC). The absolute reliability was presented by Bland-Altman plots, standard error of measurement (SEM) and minimum detectable change (MDC).

Results: No correlation was found between the two methods, irrespective of reference tissue in strain elastography (Kendall's Tau-b Hoffa = -0.01 (p = 1.00), Kendall's Tau-b subcutis = 0.04 (p = 0.87)). Tracking of the tendon elongation in the DBUS method had good to excellent relative reliability (ICC = 0.95 (95% confidence interval - CI: 0.85-0.98)) and high absolute reliability (SEM = 0.04 mm (1%), MDC = 0.11 mm (3%)). The strain elastography method had good to excellent relative reliability, regardless of reference tissue (ICC Hoffa = 0.95 (95% CI: 0.86-0.98), ICC subcutis = 0.94 (95% CI: 0.82-0.98)), but low absolute reliability (SEM Hoffa = 0.06 (20%), MDC Hoffa = 0.18 (60%), SEM subcutis = 0.12 (41%), MDC subcutis = 0.32 (110%)).

Conclusions: No concurrent validity existed for DBUS and strain elastography, suggesting that the two methods measure different tendon properties. The overall reliability for the DBUS method was high, but the absolute reliability was low for strain elastography stiffness ratios. Therefore, the strain elastography method may not be recommended for tracking differences in patellar tendon stiffness in healthy adults.

Keywords: Patellar tendon; elastography; reliability; stiffness; ultrasound.

Figure 1.

(a) Measurement of patellar tendon stiffness in a 65-year-old woman. Subjects were seated in a custom-made rigid chair with hips and knees flexed to 90°, with a leg-cuff mounted to their lower leg and with the ultrasound transducer positioned so both patella and tibia were visualised. (b) An example of synchronized and polynomial fitted values of patellar tendon force and tendon elongation, stiffness was calculated as the slope of the last 10% of the force-deformation curve.

Figure 2.

Elastography of the patellar tendon including three measurement areas in the patellar tendon (red, blue and green circles) and one reference area (yellow circle) in Hoffa’s fad pad with elastography characteristics (raw data/strain index) during the time of measurement. Quality bar in the top left corner.

Figure 3.

Scatterplot showing no immediate correlation between the DBUS method and the SEL method. Inserted are the Kendall’s Tau-b values confirming that no correlation or concordance existed.

Figure 4.

Bland–Altman plots with 95% limits of agreement for the tracking process using the dynamometer and B-mode ultrasonography recording method (DBUS) and for the stiffness ratios obtained in the SEL method, using either Hoffa’s fat pad or subcutis as reference tissues. The dark green horizontal line (intersecting zero at the y-axis) indicates perfect agreement, whereas the purple horizontal line represents the observed mean difference. The closer the purple line is to the dark green line the less disagreement between tracking one and two, resp. assessment one and two.