A selection of monitoring parameters for gynecologic cytology—Beacons of light for quality assurance

Abstract

A plethora of monitoring parameters are required by the 1988 Clinical Laboratory Improvement Amendments (CLIA 88) and by various accreditation bodies for quality assurance (QA) in gynecologic cytology; what follows is a short overview of a few selected parameters. They function as guiding beacons for QA activities in gynecologic cytology, helping us determine whether we are adrift or not; they are also easy primers for pathology residents and cytopathology fellows to start learning the ropes of QA in gynecologic cytology. The atypical squamous cells (ASC) rate refers to the sum of ASC of undetermined significance (ASC-US) and ASC-cannot exclude high-grade squamous intraepithelial lesion (ASC-H) cases divided by the number of all gynecologic cytology cases encountered during the same period. When considering all of the ASC cases diagnosed in a laboratory, approximately 90% and 10% should fall into the ASC-US and ASC-H categories, respectively.1 Specific reference values for ASC-US and ASC-H are available through the College of American Pathologists (CAP) checklist according to the specific type of cytologic preparation used and are displayed in percentiles from the data collected from the CAP interlaboratory comparison program in gynecologic cytology. However, as a general guideline, a cytopathology laboratory serving a low-risk population should have an ASC rate <5%.2 Therefore, an ASC rate ≥5% should attract attention, although it does not necessarily mean that it is “wrong.” One possible explanation is that the laboratory might be serving a high-risk population; however, consideration should also be given to the possibility of “diagnostic drift,” with either over-diagnoses of reactive changes as ASC-US or ASC-H or under-diagnoses of squamous intraepithelial lesion (SIL) as ASC-US. To sort out these possibilities, 2 other indicators are useful: the ASC:SIL ratio and the positivity rate for high-risk human papillomavirus (hrHPV) in the ASC-US cases tested. If a laboratory serves a high-risk population, then the ASC rate may exceed 5%; but, in general, it should not exceed 2 to 3 times the rate of SIL, leading to the current recommendation that laboratories should maintain an ASC:SIL ratio of less than 2-3:1.3 Specific reference values for the ASC:SIL ratio also are available through the CAP checklist; if the ASC:SIL ratio of the laboratory falls outside the 5th or 95th percentiles, then reasons should be sought. Because this ASC:SIL ratio is a fraction and, thus, depends on 2 variables, it is possible to obtain a “normal” ASC:SIL ratio if both the ASC and SIL categories are misused. To better assess this possibility, the added value of the rate of hrHPV positivity in the ASC-US cases tested comes into play. The approximate “expected” rate of positivity for hrHPV in ASC-US cases hovers around 50%4; this value can be used as an added variable for a multilayered approach to make the assessment of the ASC and ASC:SIL ratio stronger and more meaningful. Cibas et al5 describe 8 perturbation patterns from a normal ASC:SIL ratio and the hrHPV positivity rate for ASC-US and discuss how to draw the correct conclusions for each scenario. Despite a recent survey6 demonstrating that cytologic-histologic correlation ranked first as the most useful quality metrics in a QA program—along with multiheaded review of difficult cases and retrospective review of negative slides in current high-grade SIL (HSIL) Papanicolaou (Pap) test cases—there are surprisingly scant data in the literature regarding the “correct/expected” values for cytohistologic correlation. In other words, there are no official standards provided or widely adopted in the United States or in Canada. However, there are some parameters that can be fished out from international sources. For example, the National Pathology Accreditation and Advisory Council (NPAAC), a governmental body supervising cytopathology laboratories in Australia, provides several performance standards that cytopathology laboratories must meet to receive government payments.7 Specifically considering cytologic-histologic correlation for a cytologic diagnosis of HSIL, the NPAAC recommended standard is that no less than 65% of cytologic specimens have cervical histology, taken within 6 months, confirming the abnormality as high-grade or malignant. This standard value lies within the reported range published in the literature for biopsy-proven HSIL after an HSIL cytologic diagnosis, which is quoted as being in the range of approximately 65% to 85%.8, 9 Other noteworthy values for cytohistologic correlation, even if they are not standards, are the approximate rates of HSIL on biopsy after the cytologic interpretations of ASC-US, low-grade SIL (LSIL), and ASC-H, which should be 10% to 15%, 15% to 25%, and 30% to 40%, respectively.1 Monitoring of performance indicators should be performed, at the very least, at the level of the overall laboratory and, ideally, also for individual cytotechnologists and cytopathologists. Such a detailed breakdown of individual activities is time-consuming, and some information systems may be ill equipped to support such practice. Nevertheless, whenever possible, it should be done. It is also important to keep in mind that, although the calculation and assessment of most of the reference values remain the same when analyzing overall laboratory or individual data, there are exceptions; from the aforementioned list, the exception is the ASC:SIL ratio. When assessing the ASC:SIL ratio for cytotechnologists, the ratio should be calculated from the cytotechnologists' provisional interpretation (not from the final interpretation by the pathologist). In addition, although maintaining an ASC:SIL ratio of <2:1 to 3:1 is advocated for the whole laboratory, 1 study indicated that the mean screening sensitivity for cytotechnologists with ASC:SIL ratios <1.5 was significantly less than that for cytotechnologists whose ASC:SIL ratio was >3, suggesting that an ASC:SIL ratio <1.5 may be a surrogate marker for inadequate sensitivity.10 Therefore, a slightly different tack must be taken when assessing the ASC:SIL ratio of cytotechnologists in contrast to the ratio of the whole laboratory or of individual cytopathologists. The monitoring of the equivocal ASC-US and ASC-H categories is important, because they are particularly prone to diagnostic drift. Whenever a significant deviation is noted in the ASC-US or ASC-H parameters, education sessions, such as reviews of specific cases at the multiheader microscope, can be very helpful. Review of a spectrum of well defined entities (eg, reactive and LSIL) bordering the ASC-US category may clarify when to use the latter interpretation. Review of entities known to mimic ASC-H should also help anchor the essential diagnostic criteria for the whole crew. The QA report, containing the overall laboratory data and the deidentified individual data, should be disseminated to all cytotechnologists and cytopathologists of the laboratory, and the findings discussed on a regular basis. Beware that presentation of individual performance data can sometimes trigger strong emotions in certain staff. It is therefore very important to emphasize that the data are presented with an educational rather than a punitive context, ie, make sure everyone knows that no one will be cast away or made to walk the plank if their parameters missed the mark! If the whole crew views QA as an educational experience, then the cytopathology laboratory is much more likely to maintain an even keel and enjoy smooth sailing. Dr. Auger is Director of the Cytopathology Laboratory at the McGill University Health Center and Professor in the Department of Pathology at McGill University. One of her special interests is quality assurance in gynecologic cytology.