412 COMPARISON OF WORMS AND BLOKS SEMI-QUANTITATIVE KNEE MRI SCORING FOR ASSESSING CARTILAGE LOSS: RESULTS FROM THE OSTEOARTHRITS INITIATIVE

  • Lynch J
  • Roemer F
  • Felson D
  • et al.
N/ACitations
Citations of this article
5Readers
Mendeley users who have this article in their library.

Abstract

Purpose: Cartilage lesions in the knee can be graded using semiquantitative (SQ) methods. Whole Organ MRI Scoring (WORMS) and Boston Leeds Osteoarthritis Knee Scoring (BLOKS) both assess areal extent and depth of cartilage lesion. WORMS uses a single 7 point scale in 14 anatomical subregions. BLOKS uses separate 0-3 scales for extent and depth, each scored in 8 sube-gions. Such differences between the two methods may affect their relative sensitivities for assessing cartilage loss. In this study, we evaluated inter-reader reliability and relative sensitivities of WORMS and BLOKS as methods for assessing the presence of longitudinal cartilage loss from MRI of the OA knee. Methods: We studied 115 knees with frequent pain, radiographic OA and risk factors for cartilage loss, which included baseline JSN (OARSI grade >1), malalignment (varus vs. non-varus) assessed from a full-limb radiograph and BMI (per SD increase). Cartilage morphology was scored using WORMS and BLOKS from baseline and 24-month visit knee MRIs. Images were read paired, but blinded to time order, by one of 2 experienced radiologists (FWR,AG). Each read approximately equal numbers of knees. 25 knees received duplicate readings to examine inter-reader reliability. WORMS and BLOKS readings of a knee were separated by at least 2 weeks, and the scoring system used first was randomly ordered. For both scoring systems we defined whole grade worsening of cartilage as an increase cartilage score (increase >0) in any subregion of a knee (or within the subregions of a specific tibio-femoral compartment, for compartment based analyses). For WORMS, based on experience in previous longitudinal studies, readers recorded when a definite, but "within-grade" change in a subregion occurred as well as recording whole grade changes. Results: Weighted kappas for inter-rater reliability of severity of baseline lesions were 0.88 (WORMS), 0.91 (BLOKS lesion extent), and 0.88 (BLOKS full thickness). Table 1 shows agreement, using simple kappa (k) for worsening (inter-reader, and BLOKS-vs-WORMS), and frequencies of worsening found by each scoring system. {table presented} Conclusions: WORMS and BLOKS both had good to excellent inter-rater reliability, with good agreement between them on the presence of whole grade worsening within a knee. Using "within-grade" worsening in WORMS increased sensitivity for progression, but at a tradeoff with less association with risk factors for OA progression. The validity of the additional "within-grade" change score requires further investigation. The large amount of change required for a transition from BLOKS grade 2 (10-75% of a region) to grade 3 (>75% of a region) may limit sensitivity for scoring change while increasing specificity and association with risk factors.

Cite

CITATION STYLE

APA

Lynch, J. A., Roemer, F. W., Felson, D. T., Guermazi, A., Nui, J., Maeda, J., & Nevitt, M. C. (2009). 412 COMPARISON OF WORMS AND BLOKS SEMI-QUANTITATIVE KNEE MRI SCORING FOR ASSESSING CARTILAGE LOSS: RESULTS FROM THE OSTEOARTHRITS INITIATIVE. Osteoarthritis and Cartilage, 17, S218–S219. https://doi.org/10.1016/s1063-4584(09)60433-7

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free