Analysis of the factors which influence the measurement of synovial power Doppler signals with semi-quantitative and quantitative measures - a pilot multicenter exercise in Japan.

OBJECTIVES

This pilot multicenter exercise aimed to evaluate the inter-observer reproducibility of synovial power Doppler (PD) signals in rheumatoid arthritis (RA) patients and to determine the factors influencing the measurements.

METHODS

Two representative RA patients were assessed by four independent experienced sonographers. The influence of machine difference, deterioration of the transducer and pulse repetition frequency (PRF) on the assessment of synovial PD signals was investigated.

RESULTS

Intra-class correlation coefficient (ICC) for the scanner-reader reproducibility of semi-quantitative PD score was high (0.867). ICC for the inter-scanner reproducibility of synovial PD pixel count was higher than that of semi-quantitative PD score. The assessment of PD signals significantly differed between two machines with quantitative measurements but did not with semi-quantitative score. The assessment of PD signals with a deteriorated transducer was much less sensitive than that with an intact one. The semi-quantitative scores for PD signals were comparable between three different PRFs (500/800/1,300 Hz), whereas the pixel count was significantly lower with the highest one in the knee joint.

CONCLUSIONS

Measurement of PD signal can be substantially affected by deteriorated quality of the transducer, whereas the differences are relatively modest between machines with similar specifications and also between PRF settings within a low range.