Reappraisal of quantitative esophageal scintigraphy by optimizing results with ROC analyses

This study investigates whether systematic analyses of methodological issues contribute to improve and renew the diagnostic role of quantitate esophageal scintigraphy.

Forty-seven patients with normal (n = 26) and pathologic (n = 21) esophageal function were studied with scintigraphy and manometry, using the latter findings as the gold standard. Scintigraphic data were analyzed by receiver operator characteristic (ROC) methods to: establish the optimal decision threshold for six different quantitative parameters, evaluate their inherent discrimination capacity and compare liquid compared with solid bolus data.

Quantitative parameters have shown remarkable differences in their potential to discriminate between normal and pathologic findings (percentage of emptying at definite time points > mean time > transit time > mean transit time > Tmax). Sensitivity of 95% at a specificity of 96% was the optimum obtained. At comparable specificity levels, solid bolus studies generally demonstrated higher sensitivity than liquid bolus studies.

The diagnostic performance of optimized esophageal scintigraphy is close to that of manometry. Our findings do not only renew the role of esophageal scintigraphy as an accurate screening test for esophageal motility disorders but also invalidate recent reservations about the diagnostic potential of this method.