Reporting of perfusion/ventilation lung scintigraphy using an anatomical lung segment chart: A prospective study

This study prospectively evaluated the effect of the consistent use of a previously described anatomical lung segment chart on the interpretation of lung scans. Simultaneous perfusion/ventilaiion lung scintigraphy was performed in 221 consecutive patients with clinically suspected pulmonary embolism. Lung scans were immediately reported as normal, high probability or non-diagnostic with the use of an anatomical lung segment chart. After at least 6 months, blinded lung scans were randomly read by a panel of nuclear medicine physicians. Initial lung scan reports were classified as normal, high probability or non-diagnostic in 64, 63 and 94 patients, respectively. Overall observer disagreement was 5.9% (95% confidence interval (Cl) 3.2-9.8%), while this was 7.8, 3.2 and 7.9% for lung scans that were initially reported as normal, non-diagnostic and high probability, respectively. Reclassification consisted of normal to non-diagnostic (5), non-diagnostic to normal (1), non-diagnostic to high probability (2), and high probability to non-diagnostic scan results (5). From the literature, an overall disagreement of 20% was expected if no chart had been used. In conclusion, this study confirms that the consistent use of a lung segment chart reduces observer disagreement in the reporting of lung scans. © 1994 Chapman and Hall Ltd.