Forgotten knowledge? Current practice patterns of thyroid scintigraphy

Objectives As part of an ongoing national Veterans Affairs (VA) quality assurance program we sought to prospectively assess current practice patterns and diagnostic accuracy of thyroid scintigraphy.

Methods We developed and distributed to 120 nuclear medicine VA clinics a thyroid phantom with simulated 9.5 mm cold nodules at the left upper and right lower poles. Filled with 300 uCi 99mTc, the phantom yielded count rates similar to clinical thyroid scintigraphy. Sites were instructed to image the phantom using their routine thyroid protocol, to report imaging parameters and to identify abnormalities.

Results 74/118 (63%) facilities used pinhole collimators (PiCs) in their thyroid scintigraphy protocol, whereas 44/118 (37%) laboratories employed only parallel hole collimators (PaCs). Overall, 89/120 (74%) identified both cold nodules, 17/120 (14%) identified only one cold nodule and 14/120 (12%) failed to identify any cold nodules. Facilities using PiCs identified both cold nodules in 71/88 (81%) of scans, while sites using PaCs identified both nodules in only 18/32 (56%) scans. Strikingly, only 7/88 (8%) scans performed with a PiC failed to identify both cold nodules versus 9/32 (28%) scans performed with a PaC.

Conclusions Despite the known advantages of pinhole imaging, almost 40% of nuclear medicine laboratories use only PaCs for thyroid scintigraphy. Contrary to currently-held assertions by some vendors and practitioners that PaCs with electronic zoom are equivalent to PiCs, the sensitivity for detection of cold nodules is reduced with the use of PaCs. PiCs should be used for optimal imaging of thyroid nodules.