INTRODUCTION: (99m)Tc-pertechnetate scintigraphy is much used in the evaluation of patients with nodular goitre. We investigated the ability of experienced observers to estimate the thyroid 24-h (131)I uptake (RAIU) and the thyroid volume by visual evaluation of the scintigram. MATERIAL AND METHODS: Two endocrinologists and two nuclear medicine specialists visually evaluated thyroid scintigrams from 171 patients with nodular goitre. The variables were assessed in a blinded fashion according to predefined categories and then compared with the true values. The assessments were repeated after four weeks. Kappa (κ ω) statistics were used. RESULTS: There was a low probability (range 6-22%) for the observers to assess the thyroid RAIU correctly. The probability of assessing the thyroid volume correctly was in the 14-22% range. Endocrinologists tended to underestimate the thyroid RAIU, mostly in patients with a RAIU > 30%. All observers significantly underestimated the thyroid volume if this was > 80 ml. There was a low interobserver agreement for the thyroid RAIU assessment (κ ω-values: 0.03-0.43) as well as for the thyroid volume assessment (κ ω-values: 0.19-0.48). The corresponding κ ω-values for the intraobserver agreement were 0.34-0.68 and 0.37-0.62, respectively. Nuclear medicine specialists achieved a significantly higher agreement than endocrinologists in their evaluation of both thyroid parameters. CONCLUSION: Thyroid (99m)Tc scintigraphy has poor interobserver agreement and is inaccurate for assessment of quantitative thyroid parameters, even when performed by experienced specialists. FUNDING: This study was supported by grants from the Danish Agency for Science, Technology and Innovation. TRIAL REGISTRATION: not relevant.
INTRODUCTION: (99m)Tc-pertechnetate scintigraphy is much used in the evaluation of patients with nodular goitre. We investigated the ability of experienced observers to estimate the thyroid 24-h (131)I uptake (RAIU) and the thyroid volume by visual evaluation of the scintigram. MATERIAL AND METHODS: Two endocrinologists and two nuclear medicine specialists visually evaluated thyroid scintigrams from 171 patients with nodular goitre. The variables were assessed in a blinded fashion according to predefined categories and then compared with the true values. The assessments were repeated after four weeks. Kappa (κ ω) statistics were used. RESULTS: There was a low probability (range 6-22%) for the observers to assess the thyroid RAIU correctly. The probability of assessing the thyroid volume correctly was in the 14-22% range. Endocrinologists tended to underestimate the thyroid RAIU, mostly in patients with a RAIU > 30%. All observers significantly underestimated the thyroid volume if this was > 80 ml. There was a low interobserver agreement for the thyroid RAIU assessment (κ ω-values: 0.03-0.43) as well as for the thyroid volume assessment (κ ω-values: 0.19-0.48). The corresponding κ ω-values for the intraobserver agreement were 0.34-0.68 and 0.37-0.62, respectively. Nuclear medicine specialists achieved a significantly higher agreement than endocrinologists in their evaluation of both thyroid parameters. CONCLUSION: Thyroid (99m)Tc scintigraphy has poor interobserver agreement and is inaccurate for assessment of quantitative thyroid parameters, even when performed by experienced specialists. FUNDING: This study was supported by grants from the Danish Agency for Science, Technology and Innovation. TRIAL REGISTRATION: not relevant.