Abstract

Objective: The study was designed to assess the significance of the interoperator variability in the estimation of functional parameters for four nuclear medicine procedures.

Materials and methods: Three nuclear medicine technologists with varying years of experience processed the following randomly selected 20 cases with diverse functions of each study type: renography, renal cortical scans, myocardial perfusion gated single-photon emission computed tomography (MP-GSPECT) and gated blood pool ventriculography (GBPV). The technologists used the same standard processing routines and were blinded to the results of each other. The means of the values and the means of differences calculated case by case were statistically analyzed by one-way ANOVA. The values were further analyzed using Pearson correlation.

Results: The range of the mean values and standard deviation of relative renal function obtained by the three technologists were 50.65 +/- 3.9 to 50.92 +/- 4.4% for renography, 51.43 +/- 8.4 to 51.55 +/- 8.8% for renal cortical scans, 57.40 +/- 14.3 to 58.30 +/- 14.9% for left ventricular ejection fraction from MP-GSPECT and 54.80 +/- 12.8 to 55.10 +/- 13.1% for GBPV. The difference was not statistically significant, p > 0.9. The values showed a high correlation of more than 0.95. Calculated case by case, the mean of differences +/- SD was found to range from 0.42 +/- 0.36% in renal cortical scans to 1.35 +/- 0.87% in MP-GSPECT with a maximum difference of 4.00%. The difference was not statistically significant, p > 0.19.

Conclusion: The estimated functional parameters were reproducible and operator independent as long as the standard processing instructions were followed.