Comparison of discrimination and prognostic value of two US Doppler scoring systems in rheumatoid arthritis patients: a prospective cohort study

OBJECTIVES: The aim of this paper is to investigate sensitivity to change (SRM), predictive validity and discriminative ability of a quantitative (QS) and a semi-quantitative (SQS) Doppler ultrasound scoring systems in patients with rheumatoid arthritis (RA) treated with anti-TNF-α therapy.

METHODS: RA patients with wrist joint affection treated with TNF-α inhibitor were followed for one year. The wrist was examined with Doppler before initiating therapy and after one year. DAS28 was determined at both visits. One person trained in the SQS system and one in the QS system evaluated the anonymised images. The SRM, predictive validity and discriminative ability for both systems were calculated using DAS28 as the measure of disease improvement.

RESULTS: Fourty-six patients with RA (80% females) were included. The mean Doppler activity at baseline was QS:24.4% (SD=17.7%) and SQS:2.0 (SD=0.6). A decrease in Doppler activity was seen for both systems after anti-TNF-α therapy. Sensitivity to change was seen, SRM=-0.52 (95%CI; -0.83 to -0.21; QS) and -0.24 (-0.53 to -0.05; SQS). Predictive value was poor (QS rs=-0.24; SQS rs=-0.05). Construct validity was; QS: rs=0.29, SQS: rs=0.23.

CONCLUSIONS: Both systems were to some extent sensitive to change. Predictive validity and discriminate capacity of both systems showed only a weak association to DAS 28 in the study population. The QS was a little superior to the SQS. The results do not necessarily reflect Doppler evaluation as being ineffective, but may be caused by DAS28 not being a perfect marker of inflammation.