OBJECTIVE: To study the cellularity of liquid-based preparations (LBPs) for normal, abnormal and false negative cervical cytology cases. STUDY DESIGN. A series of 1,875 LBPs obtained by a split-sample protocol was examined. Slides had been examined by at least one cytotechnologist. All abnormals were reviewed by at least two cytopathologists. The cellular objects were counted using a fully automated microscope. Cellularity was evaluated for the entire population, normal (<LSIL) slides, abnormal (LSIL+) slides and "false negative" slides. False negatives were identified as those with an initial impression of <LSIL and (1) the reference pathologist's impression of the slide was LSIL+, (2) the simultaneous conventional smear was LSIL+, or (3) a cervical biopsy was LSIL+. Descriptive statistics and graphic comparisons were used. RESULTS: There were 192 confirmed abnormal cases and 1,683 normal cases, of which 53 were false negative. The frequency distributions of cellularity for the entire population and each category were skewed-right nonnormal. Median cell counts of the entire series, normals, abnormals and false negatives were 60,510, 59,822, 70,523 and 64,036 respectively. Cell counts at 2.5 percentiles were 8,677, 7,891, 11,864 and 6,009, respectively. The population of abnormal slides tended to have higher cellularity. The population of false negative slides could not be distinguished by their cellularity. CONCLUSION: Cellularity does not provide assurance of adequacy. Any cellularity criterion should be based on measurement of the prevalence of abnormal cells on abnormal slides.
