Musculoskeletal Ultrasound in Multi-Center Rheumatoid Arthritis Clinical Trials: Methodology for Optimizing Reliability of Acquisition and Real-Time Scoring

Veena K. Ranganath¹, Yossra Suliman^1,2, Jenny Brook³, David Elashoff³, Ami Ben-Artzi⁴, Cesar Olmos¹, Nabeel Borazan³, Thasia Woodworth¹ and Gurjit S. Kaeley⁵, ¹Medicine, Division of Rheumatology, David Geffen School of Medicine, Los Angeles, CA, ²Rheumatology and Rehabilitation department, Assiut university, Faculty of Medicine,, Assiut, Egypt, ³Medicine, David Geffen School of Medicine, Los Angeles, CA, ⁴David Geffen School of Medicine, Los Angeles, CA, ⁵University of Florida, Jacksonville, FL

Meeting: 2015 ACR/ARHP Annual Meeting

Date of first publication: September 29, 2015

Keywords: Outcome measures, rheumatoid arthritis (RA) and ultrasound

Session Information

Date: Tuesday, November 10, 2015

Title: Rheumatoid Arthritis - Clinical Aspects Poster Session III

Session Type: ACR Poster Session C

Session Time: 9:00AM-11:00AM

Background/Purpose: Rheumatoid arthritis (RA) musculoskeletal ultrasound (MSUS) literature reports good reliability of scoring still images with training, but inconsistent reliability of scoring during acquisition, likely a key factor for reproducibility and sensitivity to change. The aim of this study was to systematically optimize MSUS acquisition and real-time scoring reliability prior to conducting a multicenter MSUS-based RA study.

Methods: This two-site MSUS study evaluated the following joints: bilateral radio-carpal, intercarpal, radioulnar, MCP 1-5, PIP 1-5, knees and MTP 2-5. Reliability among three ultrasonographers with >5 years MSUS experience was assessed in 2 stages: 1) face-to-face 1 [F2F1] Aug 2014, and 2) face-to-face 2 [F2F2] Feb 2015. A reference MSUS RA atlas was modified during serial teleconferences and multiple reliability still-image exercises. After each reliability exercise, the most discrepant images were discussed and the draft atlas was updated. Factors affecting discordance were identified and comprehensively addressed prior to F2F2: Environment and Machine (consensus of room temperature, lighting, machine settings); B-Mode Optimization (permitting adjustment of frequency, depth, focus, and positioning of deformed joints); Power Doppler Optimization (standardization of gel layer, strict adherence to Doppler settings, color map adjustment to accommodate color blindness); Scoring Optimization(consensus rules on scoring cutoffs, scoring rules on compound joints such as the wrist). Intra- and inter-reader reliability was examined using weighted-kappa, intraclass correlation coefficient (ICC), and Spearman correlation. Reliability statistics were compared between stages using permutation tests to compute empirical distributions for differences in those statistics.

Results: Acquisition intra-reader reliability improved from F2F1 (0.52-0.71) to F2F2 (0.7-0.86) for weighted kappa, ICC and Spearman correlation (Table). Ultrasonographer-1 (US-1) achieved significant improvement in reliability from F2F1 to F2F2 (p<0.05). Inter-reader reliability also improved from 0.5-0.66 to 0.64-0.74 as a result of between-meeting consensus activities. Improvement was statistically significant for inter-reader reliability between F2F1 to F2F2 for weighted kappa (p=0.03) and a trend for Spearman correlation (p=0.07).

Conclusion: By determining and addressing factors that influence acquisition and scoring, improved real-time scoring reliability can be achieved.

Table: Acquisition and Real-Time Scoring Reliability
Intra– reader					Inter– reader
		Weighted Kappa	ICC	Spearman correlation		Weighted Kappa	ICC	Spearman correlation
F2F-1	US-1	0.63	0.71	0.67	US-1/US-2	0.51	0.66	0.59
	US-2	0.52	0.59	0.64

F2F-2	US-1	0.80	0.86	0.86	US-1/US-3	0.66	0.75	0.73
	US-3	0.72	0.81	0.81

	Difference US-1	-0.17	-0.15	-0.19	Difference	-0.16	-0.09	-0.14
	p-value	0.02	0.02	0.008	p-value	0.03	0.22	0.07
ICC: intraclass correlation coefficient; F2F: face-to-face; US: ultrasonographer

Disclosure: V. K. Ranganath, Genentech and Biogen IDEC Inc., 2,Pfizer Inc, 2,Bristol-Myers Squibb, 2,Bristol-Myers Squibb, 5; Y. Suliman, None; J. Brook, None; D. Elashoff, None; A. Ben-Artzi, None; C. Olmos, None; N. Borazan, None; T. Woodworth, None; G. S. Kaeley, None.

To cite this abstract in AMA style:

Ranganath VK, Suliman Y, Brook J, Elashoff D, Ben-Artzi A, Olmos C, Borazan N, Woodworth T, Kaeley GS. Musculoskeletal Ultrasound in Multi-Center Rheumatoid Arthritis Clinical Trials: Methodology for Optimizing Reliability of Acquisition and Real-Time Scoring [abstract]. Arthritis Rheumatol. 2015; 67 (suppl 10). https://acrabstracts.org/abstract/musculoskeletal-ultrasound-in-multi-center-rheumatoid-arthritis-clinical-trials-methodology-for-optimizing-reliability-of-acquisition-and-real-time-scoring/. Accessed .

« Back to 2015 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/musculoskeletal-ultrasound-in-multi-center-rheumatoid-arthritis-clinical-trials-methodology-for-optimizing-reliability-of-acquisition-and-real-time-scoring/