Sources of Error for the Pap Smear Sampling Errors

Sampling errors account for 50-60% of false-negatives in which abnormal cells are not present on the smear either because they were not present in the collected sample or they were present but not recovered when the smear was prepared. Liquid-based cytology techniques enhance the recovery of abnormal cells collected in the cervical sample, and with reported false-negative rates of 5-15% (5,6), they are more sensitive than the conventional Pap smear in detecting SILs. Screening Errors

Screening errors account for about one-half of laboratory false-negatives and result from two major causes.

Failure to Detect Abnormal Cells Present on the Smear

The following are reasons for the failure to detect abnormal cells on a smear:

1. Smears may have too few cells; 100 abnormal cells per smear seems to be a minimum threshold for recognition by the screening cytotech-nologist.

2. Smears may show small single cells called "no-see-ums." These are present in some cases of HSILs. They are small cells (<15 micra) with a high nuclear/cytoplasmic ratio, minimal chromatin abnormalities, and irregular nuclear membranes ("raisinoid" nuclei). The large cells with low nuclear/cytoplasmic ratios and "ugly" nuclei seen in LSILs are less frequently missed.

3. Hyperchromatic crowded groups of cells are often difficult to accurately classify. They may be normal endometrial cells or benign basal cells seen in atrophy, but abnormal cells from HSILs can also mimic this appearance.

4. "Litigation" cells can look like virtually anything, especially in retrospect (7).

In 1996, new technologies and techniques were emerging that were designed to improve recognition of missed cells, increase the sensitivity of the Pap smear, and reduce false-negatives. These included liquid-based cytology, automated processing and screening, automated and manual rescreening of both negative Pap smears and smears showing ASC-US, and rapid rescreening. However, these techniques were expensive, and because the increased costs were not offset by higher reimbursement, few laboratories were able to provide them. Some pathologists expressed concern that even if reimbursed, the higher cost may reduce both access to the Pap smear and its frequency, whereas others wondered if a higher "community standard" for the frequency of false-negative Pap smears would increase the public's expectation of perfection and thereby increase liability.

Failure to Recognize Unsatisfactory Pap Smears

CLIA 88 prohibits making a diagnosis of "negative for intraepithe-lial lesion or malignancy" on an unsatisfactory Pap smear. However, a Pap smear with abnormal epithelial cells must never be interpreted as "unsatisfactory for evaluation." For this reason, potentially unsatisfactory smears must be very carefully screened for the presence of abnormal cells before reporting them as unsatisfactory. The fact that CLIA 88 prohibits making a diagnosis on unsatisfactory Pap smears assures liability if a woman with cervical cancer had a prior Pap smear diagnosed as "negative for intraepithelial lesion or malignancy" that on retrospective review is found to be unsatisfactory. For these reasons, every laboratory should have written policies defining specimen adequacy; today these should be based on the 2001 Bethesda System recommendations. These criteria include the presence of at least 10 well-preserved endocervical or squamous metaplastic cells and 800012,000 squamous cells (5000 for liquid-based preparations) (8). For computerized labs, "edits" should be created to assure that diagnoses are not assigned to unsatisfactory smears.

Interpretation Errors

Interpretation errors account for the other 50% of laboratory false-negatives. Some of the causes of interpretation errors are listed here:

1. ASC-US is a poorly defined diagnostic category and represents the pathologist's interpretative "gray zone." Diagnostic criteria are not uniformly agreed on and there is poor inter- and intraobserver repro-ducibility (9,10). Expert members of the CAP Cytopathology Committee reached 80% consensus on the diagnosis of ASC-US on only 20% of cases reviewed; there was not 100% consensus on any case. Approximately 2 million Pap tests in the United States are diagnosed as ASC-US each year. Therefore, finding ASC-US on retrospective review of a "negative" Pap smear should never be the sole basis for judging a false-negative Pap smear to be below the standard of practice. However, because about 26% of women with ASC-US are subsequently diagnosed with SILs (up to one in four as HSILs), rescreening of all ASC-US cases is recommended as a quality assurance (QA) procedure. Furthermore, when atypical cells are found on a Pap smear, it is important that they not be characterized with terms such as "non-neoplastic" or "benign." This may be interpreted by the clinician to mean that a repeat Pap smear or appropriate follow-up study is not necessary. Use of the 2001 Bethesda System terminology (8) is an appropriate way to deal with this situation (i.e., "Atypical Squamous Cells of Undetermined Significance [ASC-US]"), coupled with a recommendation for appropriate follow-up studies.

2. Misinterpretation of an "epithelial cell abnormality" as "negative for intraepithelial lesion or malignancy," (e.g., cervical adenocarcinoma cells misinterpreted as reactive endocervical cells).

3. Failure to look carefully for abnormal cells in the "neighborhood" of parakeratotic cells. Because abnormal cells tend to occur in linear streaks, pathologists must look carefully at screened smears with lines of dots or look to either side of dotted abnormal cells for other abnormal cells.

+1 0

Post a comment