The Significance of Random Sample Size in Flow-Through Photometric Prescreening in Cervical Cytology