Effectiveness of Rapid Prescreening in Liquid Based Pap Tests
HS Currens, KA Nejkauf, SS Raab. University of Colorado Denver, Aurora, CO
Background: Rapid prescreening (RPS) has been shown to be an effective quality control procedure for detecting squamous intraepithelial lesions (SIL) in conventional Papanicolaou smears. In the United States RPS is rarely adopted, with experts citing CLIAA'88 regulations that limit cytotechnologists' daily slide volumes. Technologies, including liquid based preparations (LBP) and imaging systems, hypothetically increase SIL detection, potentially lowering the effectiveness of RPS. This study measures the effectiveness of RPS in SIL detection in a U.S. laboratory that uses imaged and non-imaged LBP.
Design: A designated cytotechnologist performed RPS in 3916 LBP Pap tests: 2127 SurePath™ (SP) preparations that were manually screened and 1789 ThinPrep Pap tests™ (TP) that were imaged using the ThinPrep Imaging System™ (TIS). The cytotechnologist manually reviewed each slide for 30 to 120 seconds categorizing the Pap test as Unsatisfactory, Within Normal Limits or Abnormal (ASCUS and above). RPS results were then compared with the results obtained by routine screening. Discordant cases were reviewed by a QC cytotechnologist or pathologist. A 10% negative rescreen was performed in compliance with laboratory and CLIAA'88.
Results: RPS had a 0.5% combined detection of atypical cases for SP and TP specimens. RPS of SP had a sensitivity of 0.45 and a specificity of 0.99 for cases of ASCUS and above; TP had a sensitivity of 0.48 and a specificity of 0.97. For LSIL and above, the sensitivities of SP and TP were 0.72. The sensitivities for HSIL and above for SP and TP specimens were 0.67 and 0.73, respectively. SP and TP had a false negative rate (FNR) of 0.55 and 0.52, respectively. A comparison of negative RPS diagnoses to routine screening revealed a 2% overcall of ASCUS based upon HPV results and pathologist diagnosis. Overall 10% negative rescreen resulted in 6% significant diagnostic discrepancies.
Conclusions: RPS did not result in identifying a significant number of false negative cases. The findings may indicate that this laboratory is composed of a highly sensitive group of screeners, or that there is a tendency to overcall otherwise negative cases. The increased SIL detection of LBP may also limit the effectiveness of RPS. The 2% overcall of ASCUS cases, identified as negative on RPS, may support its use as a quality improvement tool to lower ASCUS/HPV rates. The 10% negative rescreen, especially in imaged cases, was more effective than RPS in SIL detection. The FNRs calculated for the manually screened SP specimens and TIS specimens were not significantly different.
Tuesday, March 23, 2010 9:30 AM
Poster Session III # 83, Tuesday Morning