Evaluating postoperative rotator cuff healing: Prospective comparison of MRI and ultrasound

Abstract

Background: The objective of this prospective comparative single centre study was to compare postoperative rotator cuff healing rates as assessed by magnetic resonance imaging (MRI) versus ultrasonography (US).

Material and methods: Between October 2012 and February 2013, 61 patients underwent arthroscopic repair of postero-superior rotator cuff tears. Each patient underwent MRI and US 6 months later. The findings were assessed independently by two observers. We compared intra-observer and inter-observer levels of agreement regarding healing rates assessed by MRI and US.

Results: Intra-observer agreement regarding the MRI interpretation was 95% (κ coefficient, 0.83) for one observer and 98% (κ coefficient, 0.94) for the other. Values of κ for inter-observer agreement ranged across readings from 0.76 to 0.90. When MRI was taken as the reference, US had 80% sensitivity and 98% specificity.

Discussion: MRI and US provide similar assessments of postoperative rotator cuff healing, although US is less sensitive. Intra-observer and inter-observer agreements are very good.

Keywords: Magnetic resonance imaging (MRI); Radiological outcomes; Re-rupture; Rotator cuff repair; Tendon healing; Ultrasonography.