Comparison of the erythrocyte sedimentation rate measured in the eye casualty department by the Seditainer method with an automated system

Purpose To compare a new automated system for the measurement of erythrocyt e sedimentation rate (ESR) with the established manual Seditainer method. Methods Two hundred and twelve patients undergoing investigation for giant cell arteritis or other systemic vasculitides had ESR measurements by both the established manual Seditainer and the new laboratory-based automated sy stem. The results were compared by correlation coefficient and mean differe nce. The limits of agreement with confidence intervals were also calculated . Results Across the range of results from 1 to 120 mm/h, the correlation coe fficient was 0.844. The automated method had a mean negative bias of -9.8 m m/h (95% confidence interval: -12.2 to -7.4 mm/h). The wide scatter of resu lts produced limits of agreement (+/- 2 standard deviations) between the tw o methods of -45 to 26 mm/h. There were seven results that were underestima ted by the automated system which were clinically significant. Conclusions There is a wide degree of scatter between the two sets of resul ts. The automated system has a negative bias when compared with the manual method. There is a propensity for the automated system to sporadically unde restimate the hue result, sometimes to a degree that is clinically signific ant. The authors therefore cannot recommend replacement of the manual Sedit ainer system at the present time.