For the study, a 23-gauge needle was used for FNAB. There were 2 punctures made in each nodule of interest. To be an adequate specimen, there needed to be a minimum of 6 groups of cells with more than 10 cells/group.
Marked hypoechogenicity, irregular margins, microcalcifications, and a taller than wider shape were malignant ultrasound features. Positive ultrasounds had at least of those malignant features. There were 4,077 FNAB performed between April 2008 and December 2008. Of those, 654 (16.1%) were determined to be inadequate.
The rate of inadequate samples was different for various nodule sizes. For nodules < 5 mm, 21.2% of those were inadequate; for nodules 5 mm to 9.9 mm, 15.3% were inadequate; for nodules 10 mm to 19.9 mm, 14.2% were inadequate; for nodules > 20 mm, 18.0% were inadequate.
There were also different rates of inadequate sampling when considering if the nodule was a solid nodule or if it was a nodule with dominant cysts. The rate of inadequate sampling for nodules with dominant cysts was 23.6%; this is compared to 15.6% for solid nodules.
When considering calcifications, those nodules with macrocalcifications had a significantly higher rate of inadequate sampling than those nodules with microcalcifications—21.3% vs 11.9%, respectively. Those nodules with no calcification had a rate of inadequate sampling of 15.6%. It was noted in the study that hypoechogenicity, irregular margins, and taller than wider shape were not associated with the inadequate sampling rate.
The study also considered how experienced the doctors were in performing FNAB. Doctors were considered experienced if they performed > 300 FNAB/year; inexperienced doctors did <300 FNAB/year.
In the inexperienced doctor group, the rate of inadequate sampling was significantly higher (25.8%) than in the experienced group (15.6%).
In multivariate analysis, cyst dominance and macrocalcification were independent factors that predisposed inadequate samples for experienced physicians. For the inexperienced group, none of the ultrasound features was significant.
The researchers concluded that the factors associated with high rates of inadequate samples in thyroid nodule FNAB were: cyst dominance, macrocalcification, and doctor inexperience.