To the Editor.
—We agree with the assessment by Drs Austin and McLendon of the pressure that the Papanicolaou (Pap) test is receiving from both the medicolegal and cost-conscious managed care environments.1 However, we are not certain that as clinicians we should simply continue to fight for the Pap test as the "proven" approach to screening for cervical neoplasia.In a meta-analysis of the literature on the Pap test, Fahey et al2 have shown that when cervical biopsy is used as the reference for whether a cervical cancer or precancer is present, the overall sensitivity of the Pap test alone in detecting these lesions is between 20% and 35%, not 80% to 90% as presented by Austin and McLendon in their Editorial. We believe the screening accuracy reported in the Editorial may be overestimated and disagree that improving quality in the cytopathology laboratory will improve most of the