Printer Friendly

Prospective and retrospective review of gynecologic cytopathology: findings from the College of American Pathologists Gynecologic Cytopathology Quality Consensus Conference Working Group 2.

The Clinical Laboratory Improvement Amendments of 1988 (CLIA '88) mandate a variety of quality metrics for gynecologic cytology laboratories. (1) Two of these quality metrics are the subject of this review: "prospective rescreening" and "retrospective rescreening" (Federal Register 57FR7002). First, CLIA '88 requires that at least 10% of Papanicolaou (Pap) tests interpreted as "negative for intraepithelial lesion or malignancy" (NILM) by each cytotechnologist be rescreened by a qualified supervisory cytotechnologist or pathologist before reporting. These cases must be randomly selected and include Pap tests from women at increased risk for developing cervical cancer. Each laboratory must have a written policy that defines high risk as it is applied in practice as well as its method of random selection. For the purpose of this review, this quality metric is referred to as "prospective rescreening," as the review occurs before sign-out.

CLIA '88 also mandates review of all Pap tests interpreted as NILM (NILM Pap tests) from the last 5 years for a patient with a current Pap test that shows a high-grade squamous intraepithelial lesion (HSIL) or carcinoma (HSIL+). All available NILM Pap tests from this time period in the laboratory (either on-site or in storage) must be reviewed. If a significant discrepancy is found that will impact current patient care, the clinician must be notified and an amended report issued. In most cases, results of this retrospective review do not alter current patient management, and amended reports are not necessary. As a result, this 5-year retrospective review serves as a quality monitor as well as an educational tool for the laboratory. This quality metric will be termed retrospective rescreening for the purpose of this review, as it refers to review of NILM Pap tests that have been signed out from patients subsequently found to have significant disease.

For both of these CLIA regulations, records of all rescreening results must be documented. There must also be an annual statistical evaluation that includes the number of reviews of any NILM cases reclassified as a low-grade squamous intraepithelial lesion or worse (LSIL+). The reviews of individual cytotechnologists are compared against the laboratory's overall statistics and are one of the elements used to determine workload limits. The College of American Pathologists (CAP), as part of the Laboratory Accreditation Program, essentially reiterates these CLIA regulations as specific checklist requirements. The CAP adds the clarification that slides screened by pathologists certified in anatomic pathology and qualified as technical directors do not need to be rescreened.

While CLIA regulations serve as a minimum required standard, there are differences in implementation of these quality metrics among laboratories. For example, many laboratories review greater than 10% of NILM Pap tests as part of their prospective rescreening program. Also, there is widespread variability of definitions of high-risk patients among laboratories. As these CLIA regulations predated the availability of routine testing for high-risk human papillomavirus (hrHPV) infection, high-risk HPV status is not considered in the regulations. There are also differences among laboratories in feedback provided to individual cytotechnologists and pathologists about discrepant cases and in how results are used as educational tools to improve patient care.

In an effort to inventory current quality practices in gynecologic cytology laboratories and attempt to standardize procedures for quality improvement, the Centers for Disease Control and Prevention (CDC) awarded a grant to the CAP. The purpose of this review is to discuss the findings of this project pertinent to prospective and retrospective rescreening and to provide information regarding best laboratory practices, based on survey results, consensus conference discussions, published data, and expert opinion.


The CAP Consensus Conference on Good Laboratory Practices in Gynecologic Cytology was a multistep process to determine consensus quality assurance practices for gynecologic cytology. The goal was to identify what quality metrics are collected, how metrics are analyzed, what benchmarks are used to determine variance in performance, and what actions are taken to address performance issues. The detailed process related to survey development, formation of working groups including lists of group members, and the subsequent consensus conference is summarized in an introductory article. Briefly, the components of this study included a national survey of quality assurance practices sent to every CLIA-registered laboratory in the United States, as well as those CLIA-exempt laboratories enrolled in Pap proficiency testing. Survey results were posted on the CAP Web site for public comment. Working groups composed of cytopathologists and cytotechnologists with expertise in gynecologic cytology were formed around specific survey topics. There were 5 working groups including (1) Monitoring Interpretive Rates, Concordance of Interpretations, Turnaround Time; (2) Prospective and Retrospective Review; (3) Pap Proficiency Testing, General Quality Practices and Workload; (4) Cytologic-Histologic Correlation; and (5) Monitoring HPV Rates. These working groups reviewed the survey responses and the cytology literature and then generated additional follow-up questions that were also posted for public comment. On the basis of survey responses and results of the literature review, draft recommendations were prepared and presented at a consensus conference held in Rosemont, Illinois, on June 3 to 4, 2011. The consensus conference was open to the entire cytopathology community. As part of the presentations by each working group, consensus quality assurance statements were proposed and voted on by conference participants. Open discussion occurred as part of each group's presentation and a second round of questions were generated around particularly controversial topics. Achievement of at least 65% agreement in the voting process was considered to represent consensus. This report summarizes the work of Working Group 2: Prospective and Retrospective Review. Results from the other working groups are presented in the other articles in this issue.


The original survey was sent to 1245 laboratories. Complete survey responses were received from 541 laboratories (43% response rate) and were analyzed as part of this review. Results of specific survey questions and follow-up questions that were used to generate consensus statements follow.

Prospective Rescreening

Survey questions on prospective rescreening focused on the total percentage of NILM cases reviewed by laboratories, the percentage of high-risk cases included, criteria used to identify patients at high risk, and tracking results of these reviews. Most of 509 respondents reported rescreening rates above the CLIA-mandated 10% rate in their response to a survey question asking for the percentage of NILM cases submitted for prospective rescreening. Fewer than 15% of laboratories limit their rescreen to the CLIA-mandated 10% rate. Seventy percent of laboratories rescreen 11% to 30% of NILM cases, while 16% rescreen at least 30% of these cases. The percentage of Pap tests from high-risk patients (high-risk Pap tests) included in prospective rescreening varied among laboratories, but most (72%) reported inclusion of fewer than 20% high-risk cases. A follow-up question during the public comment period asked laboratories what percentage of prospective rescreens should be high-risk cases. Of the 47 responses to this question, 47% favored including a maximum of 20% and the remainder favored including more than 20%. Thirty percent responded that high-risk Pap tests should comprise greater than 50% of the prospective rescreen population.

A survey question asked laboratories how high-risk cases are selected for prospective rescreening. Multiple responses to this survey question were allowed. Most laboratories use multiple measures to define high risk including clinical information, prior abnormal Pap test result, prior abnormal cervical biopsy result, prior positivity for high-risk HPV, patient age, and provider/patient location (Table 1). Laboratories were also asked how patients were removed from the high-risk category. Of 468 responses, 43.4% of laboratories did not remove patients from the high-risk category. Of the remainder, removal from the high-risk category was based on negative Pap test diagnoses (36.5%) or occurred at a specified time interval (29.7 %). A follow-up question asked laboratories if patients should be removed from the high-risk category. Of the 45 responses, most (71%) favored removal of patients from the high-risk category.

At the consensus conference held in June 2011, conference participants were asked several questions about prospective rescreening. In response to the question of whether 10% is an adequate percentage for prospective rescreening, there was no consensus as to the percentage of cases that should be prospectively rescreened. Of 56 respondents, 32.1% agreed that 10% is adequate, 48.2 % disagreed, and 19.6% were unsure. There was moderately strong consensus (87.3%) that unsatisfactory Pap tests should be prospectively rescreened.

Participants at the consensus conference were also asked about inclusion of high-risk cases in prospective rescreening. Specifically, they were asked whether laboratories should make an effort to maximize the number of high-risk cases in their prospective rescreens and if multiple measures should be used to identify these patients. There was moderately strong consensus (89.4%) for this statement. There was also consensus from 59 respondents (79.7%) favoring inclusion of all readily identifiable high-risk cases in prospective review. Participants were asked if NILM Pap tests from patients with a concurrent positive high-risk HPV test result should be rescreened before sign-out. Again there was moderately strong consensus from 58 respondents (84.5%) for this statement. There was consensus that patients should be removed from the high-risk category (87%). However, there was no consensus regarding what specific criteria should be used to determine when patients should be removed from the high-risk category. Most favored removal following both consecutive negative Pap test results and negative hrHPV test results over a specified interval (44.3%) or 3 years from the last identified criterion for high-risk status (21.3%).

Surveyed laboratories were asked if they tracked the total number of cases prospectively rescreened for each cytotechnologist. Of 512 respondents, 85% perform this tracking, while the remaining 15% do not. A related question asked laboratories if they tracked the number of lesions identified in prospectively rescreened Pap tests. Of 513 respondents, 79% track the number of lesions and 21% do not. Laboratories were then asked which upgrade rates from NILM to an epithelial cell abnormality are actively monitored for each cytotechnologist and for the laboratory as a whole. These survey responses are summarized in Table 2. Multiple responses to this survey question were allowed.

Answers to survey questions dealing with prospective rescreening were then analyzed according to the volume of Pap tests interpreted by the responding laboratories. Small laboratories were defined as those receiving 3000 Pap tests per year or less (less than 250 Pap tests/month) and large laboratories as those receiving more than 24 000 Pap tests/ year (.2000 Pap tests/month). From these volumes, smaller laboratories are more likely to use only a random sample for selection of cases for prospective rescreening (P = .001). Conversely, large laboratories are more likely to use prior abnormal cytology and biopsy results to identify high-risk patients (P= .001). Large laboratories are also more likely to track the number of cases rescreened (P = .001) and the number of lesions identified (P = .001).

Retrospective Rescreening

Survey questions on retrospective rescreening focused on specific rescreening practices including length of the look-back interval, diagnostic categories prompting review, Pap test interpretations selected for review, and monitoring of review results. A survey question asked laboratories how many years they looked back as part of retrospective rescreening. Most laboratories (96.2%) review Pap tests from the preceding 5 years, adhering to the CLIA mandate. A small percentage of laboratories (2.5%) review Pap tests from all available years. Laboratories were asked what Pap test interpretations other than HSIL+ or adenocarcinoma in situ + prompt a 5-year retrospective review. Most laboratories (71.4%) responded that only HSIL+ or adenocarcinoma in situ+ prompt a 5-year review, while 15% of laboratories also rescreen Pap tests from patients with atypical glandular cells. Smaller percentages of laboratories review Pap tests from patients with low-grade squamous intraepithelial lesion (LSIL) (8.6%); atypical squamous cells, undetermined significance (ASCUS) (6.8%); and atypical squamous cells, cannot exclude high-grade squamous intraepithelial lesion (ASC-H) (2.3%). A survey question also asked laboratories which Pap tests were included in their retrospective rescreening program. Most laboratories (65.9%) included only NILM tests in this review, while 32.4% reviewed all available Pap tests and 31.4% reviewed unsatisfactory tests in addition to NILM tests. A minority of laboratories retrospectively rescreen prior ASCUS (5.8%), ASC-H (4.4%), and LSIL (2.7%) Pap tests.

Surveyed laboratories were asked about monitoring of retrospective review diagnoses. Laboratories were asked which retrospective review diagnoses should be monitored (Table 3). Multiple responses to this survey question were allowed. Surveyed laboratories were asked if they kept track of how often a previous Pap test is upgraded on the basis of retrospective review. Most laboratories (77%) perform this tracking, but a significant number (23%) responded that they do not track this CLIA-mandated element. Laboratories were then asked which upgrade rates from NILM to an epithelial cell abnormality are actively monitored for each cytotechnologist, pathologist, and for the laboratory as a whole. These survey responses are summarized in Table 4.

Answers to survey questions dealing with retrospective rescreening were analyzed according to the volume of Pap tests interpreted by the responding laboratories. Small laboratories, as defined above, were more likely to retain NILM Pap tests for 10 or more years (P = .004) and to review all available Pap tests as part of retrospective rescreening (P = .001). Small laboratories are more likely to use a less significant diagnosis, such as ASC or LSIL, to initiate a retrospective review. Large laboratories are more likely to track how often Pap tests are upgraded (P = .03). Both small and large laboratories are likely to include unsatisfactory Pap tests in their retrospective review (P = .10).

At the consensus conference held in June 2011, participants were asked to respond to several questions about retrospective rescreening. From the low number of upgraded cases in retrospective review, attendees were asked if all diagnoses for a condition of less severity than that of the initiating diagnosis should be reviewed. There was moderate consensus (78.9%) that inclusion of all prior cases was not needed. Conversely, when asked if retrospective review should be prompted by surgical biopsy results, there was consensus (87.1%) for this statement. There was also consensus (75.9%) to include unsatisfactory Pap tests in retrospective review.

Conference attendees were polled regarding monitoring of both prospective and retrospective reviews. While not meeting consensus requirements, most of the 69 participants (63.8%) responded that upgraded diagnoses from NILM to ASCUS/ASC-H should be monitored. However, there was consensus (72.7%) that it would be helpful to categorize review diagnoses into major and minor changes to stratify significance. There was moderately strong consensus (86.2%) that pathologists should receive feedback on their upgrade rates as part of retrospective review.


CLIA '88 regulations mandate some specific quality performance measures for gynecologic cytology laboratories. These include prospective and retrospective rescreening. Both attempt to evaluate the quality of Pap test screening and interpretation through directed reexamination of Pap test slides. (2,3) A review of the literature shows that these measures, in particular prospective rescreening, are controversial, emotionally charged, and difficult to perform. Further, the results of these measures can dramatically impact laboratories and individuals. One of the purposes of this article is to discuss ways of potentially increasing the value of these metrics for laboratories, with a goal of having a positive impact on patients. In particular, there is consensus that maximizing the number of high-risk Pap tests in both rescreening scenarios adds value, as does providing regular feedback to cytotechnologists and cytopathologists on results of these reviews.

Prospective Rescreening

Prospective rescreening was instituted by a number of laboratories as a quality control measure before any studies regarding its efficacy were carried out. (2) The measure was subsequently incorporated into the CLIA regulations and made mandatory. Critics state that the false-negative rate calculated by using this measure is unrealistically low and inaccurate. (2-7) Many studies (5,6) report false-negative rates of less than 5% using this method. Indeed, the number of errors identified is expected to be low in the prospective rescreen population, as it is a relatively nonenriched sample. While the regulation does require that a portion of the prospective rescreen cohort include patients at increased risk for cervical carcinoma, specifics for definition of highrisk status and numbers/percentages of high-risk Pap tests to include are left up to individual laboratories. Further, bias is introduced, as the Pap test has already been screened and interpreted by a colleague as NILM. (2,8)

Some studies have also reported the false-negative rate of the rescreen process itself, which far exceeds that reported for primary screening. (4,9,10) One study by Renshaw and colleagues (10) reported a 21% false-negative rate for prospective rescreening. Other studies (4,11) suggest that while prospective rescreening may identify poor screening performance by individuals, this may not occur quickly enough to be of benefit. Another criticism is that only a small number of false-negative cases may be identified before sign-out with this metric, with a potential for 90% or more of screening and interpretive errors to leave the laboratory. (4) As such, random rescreening of 10% of NILM cases by itself, especially without the addition of high-risk cases, is a poor measure of quality. As shown by the survey results, most laboratories (>85%) recognize this and rescreen more cases than required by the CLIA mandate. Seventy percent of the survey respondents prospectively rescreen 11% to 30% of NILM Pap tests and 16% rescreen more than 30% of NILM Pap tests.

A number of alternatives and enhancements for prospective rescreening have been proposed, including rapid prescreening/rescreening, enriching the random rescreen sample with known squamous intraepithelial lesion (SIL) cases, 100% rescreening of NILM Pap tests, sharing cases among laboratories, and automated quality control slide selection. (2,4,7,9,12-14) All are aimed at increasing the thoroughness of the rescreening process and eliminating bias. These methods allow a more realistic measure of the falsenegative rate for laboratories and individuals, while identifying more false-negative cases before a final interpretation is rendered. (4) These methods hold promise and are better measures of quality than 10% prospective rescreening, but are quite difficult to implement into routine daily practice given current regulatory requirements, workload limits, laboratory information system capabilities, and budgetary constraints.

There are several practical ways to increase the value of prospective rescreening for laboratories and patients. It is clear that maximizing the number of high-risk cases increases the power of this quality assurance metric. Multiple studies (3,15-17) have shown that patients with falsenegative Pap test results, who subsequently develop HSIL or carcinoma, quite often have a history of abnormal Pap test results before the false-negative Pap result and most are hrHPV positive. Conversely, patients with NILM Pap test results and a negative test result for hrHPV have a very low risk of progressing to HSIL and carcinoma. (18) Little is gained from including these Pap tests in prospective rescreening, yet they comprise most rescreened Pap tests in most laboratories. Said another way, most laboratories do not include enough high-risk cases in their rescreen populations. In most laboratories (54%), Pap tests from high-risk patients comprise at most 10% of the prospective rescreen population, and 72% of laboratories include less than 20% of high-risk Pap tests in rescreening. To maximize the utility of this metric, the laboratory should include all readily identifiable high-risk cases in its prospective rescreen population, in addition to randomly selected cases.

The best way to maximize high-risk cases is to use multiple measures to identify them and to remove patients who no longer meet high-risk criteria. Data should be extracted from the Laboratory Information System (LIS) or paper requisition in a timely manner before sign-out, allowing identification of patients for prospective rescreening. Suggested factors may include prior HSIL cytology finding, recent high-grade cervical intraepithelial neoplasia (CIN 2+) biopsy result, recent or concurrent hrHPV positivity, no screening in the past 5 years, current clinical designation as high risk, or other parameters at the discretion of the laboratory. Most consensus conference participants (89%) agree that laboratories should make an effort to maximize the number of high-risk cases in their prospective rescreen population and that multiple measures should be used to identify these patients. There was slightly less agreement but still consensus (79.7%) that all readily identifiable high-risk cases should be included.

Patients with NILM Pap test findings and positive results for hrHPV have a higher risk for a false-negative Pap test result than the general screening population. If hrHPV testing is used in population-based screening, approximately 5% to 7% of women 30 years of age or older with normal cytology results will be identified as having an increased risk of developing cervical carcinoma. (15,19,20) Some retrospective studies (15,21-24) have shown that false-negative Pap test finding are associated with positive test results for hrHPV. Bulk et al (15) report high-risk HPV positivity in 80% of 144 patients with originally normal smear results preceding a diagnosis of high-grade cervical intraepithelial neoplasia (CIN 2+), and most of these smears (69%) were upgraded at rescreening. These authors predicted that rescreening all hrHPV-positive normal smears might dramatically reduce false-negative cytology findings. (15) Depuydt and colleagues (3) studied liquid-based cytology samples analyzed by BDFocalPoint-guided screening (BD, Franklin Lakes, New Jersey) and used polymerase chain reaction to detect hrHPV. They concluded that targeted rescreening of Pap tests for hrHPV-positive cases has a higher sensitivity for detection of CIN 2+ than rescreening alone. These data support the recommendation that if the information is available, all readily identifiable hrHPV-positive NILM cases should be prospectively rescreened before sign-out. Most consensus conference attendees (84.5%) agree. A major barrier to implementation reported by laboratories is limited LIS functionality that does not allow timely identification of hrHPV-positive patients.

Patients with unsatisfactory Pap test results have also been shown to be at higher risk for SIL and carcinoma. (18,25-27) In a study by Ransdell and colleagues, (25) there was a significantly higher incidence of abnormalities in the first follow-up Pap test for patients with an unsatisfactory result than for a cohort of patients with NILM Pap results. In this study, unsatisfactory smears were from women more likely to have a history of prior SIL (26%). These authors recommend rescreening of unsatisfactory cases. Unsatisfactory Pap tests should be included in the high-risk population selected for prospective rescreening. Most of the Consensus Conference participants (87.3%) agreed with this statement.

A slight majority of laboratories (56.6%) currently remove patients from the high-risk category after either a specified time interval or a number of consecutive NILM Pap results. Most respondents (71%) favor removal. There was not uniform agreement on criteria for removal from the high-risk category. Proposals by consensus conference attendees included removal after consecutive negative Pap findings over a specified time interval (19.7% agreement), consecutive negative Pap findings and negative hrHPV test results over a specified time interval (44.3% agreement), and 3 years from the last identified criterion for high-risk status (21.3%). Regardless of method selected, patients should be removed from the high-risk group once they no longer meet laboratory-defined high-risk criteria.

It is important for the laboratory, and also a CLIA requirement, to track results of prospective rescreening. Both the number of cases rescreened and the number of lesions identified should be monitored for the laboratory and for individuals. CLIA specifically requires tracking of NILM cases upgraded to LSIL+. There are variations in tracking of upgraded interpretations of NILM to ASCUS/ ASC-H. Approximately 76% of survey respondents report monitoring upgrades to ASCUS; however, the consensus conference participants did not reach a consensus (60.8%) for this measure. There was strong majority agreement (98.4%) that results of prospective rescreening be shared with individuals, including both cytotechnologists and pathologists, at regular intervals.

Retrospective Rescreening

Retrospective rescreening is, in general, less controversial than prospective rescreening. Most laboratories use this monitor and find it educationally valuable. (5,6,16,28-33) This quality measure takes into account the relatively long natural history of the cervical dysplasia-carcinoma sequence, and by definition, focuses on patients at very high risk of having a false-negative Pap test result when compared to the general screening population. While this measure does not prevent the primary error, a goal is to improve the primary screening process by learning from our mistakes. As part of a comprehensive quality assurance program, retrospective rescreening may reveal systematic problems in the laboratory or with individual cytotechnologists or cytopathologists, which, with investigation and implementation of corrective measures including education, may improve performance. Most of the criticism of this measure centers on its use in monitoring of individual performance together with the fact that overall, only a small number of errors are detected with this measure. (5,9,30)

Cytologist errors are detected at a significantly higher rate on retrospective rescreening than by prospective rescreening. (5,16,30,33,34) As in prospective rescreening, the total number of upgraded cases using this measure is low when compared to a laboratory's overall Pap test volume. However, the rate of identification of false-negative Pap results in these samples is substantial in most studies. (5,18,29,30,33) False-negative rates on retrospective rescreening vary in the literature, ranging from less than 10% to 94%. (16,33) Some variation may depend on whether the rescreening was done to reflect actual daily laboratory practice or was performed in a research setting. (30,33) Also, there is some variability in which Pap test diagnoses are included in retrospective rescreening. Sherman and Kelly (18) reported a 53% false-negative rate in their retrospective review cohort at the level of ASCUS+ and a 25% rate of unsatisfactory Pap tests interpreted as NILM. Tabbara and Sidawy (5) noted a 38% rate of underdiagnosed SIL in their retrospectively rescreened Pap tests, which also included Pap findings originally interpreted as ASCUS and atypical glandular cells. Hatem and Wilbur (16) reviewed 17 Pap tests interpreted as NILM from patients with HSIL+ and reclassified 16 as abnormal (94%). A CAP Q-Probes study of 312 institutions, led by Jones (32) and reflective of daily laboratory practice, reported a 20.4% false-negative rate in 3762 retrospectively rescreened Pap tests, including NILM interpretations and unsatisfactory Pap tests, in a 4-month study period.

Most authors note a combination of screening and interpretive errors for their "false-negative" population. Most also report high rates of poorly prepared smears, compromising factors such as obscuring inflammation and blood, and a significant number of unsatisfactory Pap tests interpreted as NILM. The presence of few abnormal cells and the "underinterpretation" of atypical immature squamous metaplastic cells are also recurring themes in false-negative cases identified on retrospective rescreening. (5,16,30,31)

Most laboratories responding to the current CAP survey and participants at the consensus conference use retrospective rescreening in accordance with the CLIA '88 mandate. Most laboratories review Pap tests from the preceding 5 years only (96.2%). A minority of laboratories, in particular very small laboratories, retain NILM slides longer than 10 years and include these in retrospective rescreening. Smaller laboratories are also more likely to review all available Pap tests. Review of the literature would suggest that most false-negative Pap findings occur within 3 years of the current abnormal Pap, such that looking back beyond 5 years is unlikely to add value to this quality measure. (5,28,29,33) Specifically, in the CAP Q-Probes study by Jones, (32) 86% of false negatives were identified from Pap tests obtained in the 3 years before the incident HSIL+.

In most of the surveyed laboratories, only Pap tests interpreted as HSIL+ or adenocarcinoma in situ + trigger retrospective review (71.4%). However, smaller laboratories were more likely to initiate review at lesser degrees of abnormality. In 1996, Jones (33) reported another CAP Q-Probes study involving 323 laboratories on rescreening practices beyond the CLIA '88 mandate. Most participating laboratories reported rescreening previous NILM Pap tests for patients with a current LSIL (61%) or SIL of indeterminate grade (73%) in addition to the required rescreening of HSIL+. This study showed that rescreening triggered by a current specimen with LSIL or SIL of indeterminate grade resulted in identification of 9.9% of patients who should have had diagnoses prompting further clinical follow-up. This study was relatively less productive than his prior study that included rescreening HSIL+, which identified 20.4% of patients with an abnormality that would have prompted follow-up. This lower yield of abnormalities and the increased cytotechnologist workload are likely reasons why most survey respondents and audience members, particularly those from larger laboratories, do not rescreen Pap tests for current cases of LSIL or SIL of indeterminate grade. The author concluded, however, that this practice would result in detection of more errors, with an estimated doubling of the number or errors identified, in a timelier manner than rescreening HSIL+ only.

Perhaps a more fruitful way to detect more false-negative Pap findings in retrospective rescreening is to also include cervical biopsy specimens diagnostic of a high-grade squamous intraepithelial lesion or worse to trigger review. Although this is not mandated, it serves to increase the number of patients with significant lesions included in retrospective rescreening. Many laboratories are already using cervical biopsy results to trigger rescreening as part of their quality assurance programs. (29,30) This is further evidenced by strong consensus agreement (87.1%) with this approach by consensus conference attendees. As is the case with prospective rescreening, inclusion of unsatisfactory Pap tests adds value to the retrospective rescreening metric. Many institutions, including approximately one-third of survey respondents, already include unsatisfactory Pap tests in retrospective rescreening. Also, most consensus conference attendees (75.9%) support this practice. A number of studies have stressed the relationship of unsatisfactory Pap tests to false-negative interpretations. (18,25) Many of the same factors that render a Pap test unsatisfactory are also strongly associated with false-negative interpretations. These factors, including low cellularity and obscuring inflammation and blood, have been implicated by many authors. Also, as discussed previously, patients with unsatisfactory Pap tests are at higher risk for SIL than the general screening population. Unsatisfactory Pap tests in addition to NILM Pap results should be included in retrospective review.

To be of most value, constant monitoring of trends in false-negative rates for the laboratory and individuals identified through retrospective rescreening is required. Tracking of upgraded diagnoses from NILM to LSIL+ is mandated by CLIA '88. Responses to the CAP survey show that in most laboratories, upgraded interpretations from NILM to HSIL are tracked for the laboratory and for cytotechnologists (82.7%). However, according to survey respondents, the monitoring of the same upgrade rates for pathologists is quite low (37.3%). Most studies of retrospective rescreening report both interpretive errors in which pathologists play a role in addition to screening errors. Monitoring selected metrics for individuals should include both cytotechnologists and pathologists, and results of these reviews should be shared regularly.

In addition to CLIA '88-mandated tracking of LSIL+ upgrades, most survey respondents (91%) stated that upgraded diagnoses from NILM to ASC-H should also be monitored for the laboratory and individuals. Given that the atypical immature metaplastic cell is implicated in multiple studies of false-negative Pap test results, this review has educational value and literature support. Whether it is important to monitor upgraded diagnoses from NILM to ASCUS for cytotechnologists and pathologists is less clear. There are differences of opinion among survey respondents and consensus conference attendees on tracking these upgrades for individuals. A slight majority of survey respondents (56%) and a minority of consensus conference participants (39%) favor monitoring upgrades from NILM to ASCUS. As in prospective rescreening, it may be helpful to categorize review diagnoses into major and minor changes to stratify significance.

Summary of Recommendations of Best Laboratory Practices

The number of high-risk Pap tests in prospective and retrospective rescreening should be maximized. Unsatisfactory Pap tests should be included. All readily identifiable hrHPV-positive NILM cases should be prospectively rescreened. Cervical biopsy results of CIN 2+ should trigger retrospective rescreening. Regular feedback should be provided to cytotechnologists and cytopathologists. Upgraded diagnoses from NILM to ASC-H should be monitored.


(1.) Clinical Laboratory Improvement Amendments of 1988: final rule. Fed Regist. 1992;57:7001-7186. Current CLIA regulations available at: http://www. Accessed August 1, 2012.

(2.) Renshaw AA. Rescreening in cervical cytology for quality control;when bad data is worse than no data or what works, what doesn't, and why. Clin Lab Med. 2003;23(3):695-708.

(3.) Depuydt CE, Arbyn M, Benoy IH, et al. Quality control for normal liquid-based cytology: rescreening, high-risk HPV targeted reviewing and/or high-risk HPV detection? J Cell Mol Med. 2009;13(9B):4051-4060.

(4.) Rowe LR, Marshall CJ, Bentz JS. One hundred percent thorough quality control rescreening of liquid-based monolayers in cervicovaginal cytopathology. Cancer. 2002;96(6):325-329.

(5.) Tabbara SO, Sidawy MK. Evaluation of the 10% rescreen of negative gynecologic smears as a quality assurance measure. Diagn Cytopathol. 1996; 14(1):84-86.

(6.) Rohr LR. Quality assurance in gynecologic cytology: what is practical? Am J Clin Pathol. 1990;94(6):754-758.

(7.) Raab SS. The utility and cost-effectiveness of Pap test rescreening. Clin Lab Manage Rev. 1998;12(2):91-96.

(8.) Evans KK, Tambouret RH, Evered A, Wilbur DC, Wolfe JM. Prevalence of abnormalities influences cytologists' error rates in screening for cervical cancer. Arch Pathol Lab Med. 2011;135(12):1557-1560.

(9.) Renshaw AA, Lezon KM, Wilbur DC. The human false-negative rate of rescreening Pap tests; measured in a two-arm prospective clinical trial. Cancer. 2001;93(2):106-110.

(10.) Renshaw AA, DiNisco SA, Minter LJ, Cibas ES. A more accurate measure of the false-negative rate of Papanicolaou smear screening is obtained by determining the false-negative rate of the rescreening process. Cancer. 1997; 81(5):272-276.

(11.) Melamed MR, Flehinger BJ. Reevaluation of quality assurance in the cytology laboratory. Acta Cytol. 1992;36(4):461-465.

(12.) Raab SS, Bishop NS, Zaleski MS. Cost effectiveness of rescreening cervicovaginal smears. Am J Clin Pathol. 1999;111(5):601-609.

(13.) Lee BCK, Lam SY, Walker T. Comparison of false negative rates between 100% rapid review and 10% random full rescreening as internal quality control methods in cervical cytology screening. Acta Cytol. 2009;53(3):271-276.

(14.) Clarke J, Thurloe JK, Bowditch RC, Roberts JM. Assuring the quality of quality assurance: seeding abnormal slides into the negative Papanicolaou smears that will be rapid rescreened. Cancer. 2008;114(5):294-299.

(15.) Bulk S, Rozendaal L, Zielinski GD, et al. High-risk human papillomavirus is present in cytologically false-negative smears: an analysis of "normal" smears preceding CIN2/3. J Clin Pathol. 2008;61(3):385-389.

(16.) Hatem F, Wilbur DC. High grade squamous cervical lesions following negative Papanicolaou smears: false-negative cervical cytology or rapid progression. Diagn Cytopathol. 1995;12(2):135-141.

(17.) Coldman A, Phillips N, Kan L, et al. Risk of invasive cervical cancer after three consecutive negative Pap smears. J Med Screen. 2003;10(4):196-200.

(18.) Sherman ME, Kelly D. High-grade squamous intraepithelial lesions and invasive carcinoma following the report of three negative Papanicolaou smears: screening failures or rapid progression? Mod Pathol. 1992;5(3):337-342.

(19.) Bulkmans NWJ, Rozendaal L, Snijders PJF, et al. POBASCAM, a population-based randomized controlled trial for implementation of high-risk HPV testing in cervical screening: design, methods and baseline data of 44, 102 women. Int J Cancer. 2004;110(1):94-101.

(20.) Clifford GM, Gallus S, Herrero R, et al. Worldwide distribution of human papillomavirus types in cytologically normal women in the International Agency for Research on Cancer HPV prevalence surveys: a pooled analysis. Lancet. 2005; 366(9490):991-998.

(21.) Ylitalo N, Josefsson A, Melbye M, et al. A prospective study showing long-term infection with human papillomavirus 16 before the development of cervical carcinoma in situ. Cancer Res. 2000;60(21):6027-6032.

(22.) Zielinski GD, Snijders PJ, Rozendaal L, et al. HPV presence precedes abnormal cytology in women developing cervical cancer and signals false negative smears. Br J Cancer. 2001;85(3):398-404.

(23.) van der Graaf Y, Molijn A, Doornewaard H, et al. Human papillomavirus and the long-term risk of cervical neoplasia. Am J Epidemiol. 2002;156(2):158-164.

(24.) Wallin KL, Wiklund F, Angstrom T, et al. Type-specific persistence of human papillomavirus DNA before the development of invasive cervical cancer. N Engl J Med. 1999;341(22):1633-1638.

(25.) Ransdell JS, Davey DD, Zaleski S. Clinicopathologic correlation of the unsatisfactory Papanicolaou smear. Cancer. 1997;81(3):139-143.

(26.) Pairwuti S. False-negative Papanicolaou smears from women with cancerous and precancerous lesions of the uterine cervix. Acta Cytol. 1991; 35(1):40-46.

(27.) Bentz JS, Rowe LR, Gopez EV, Marshall CJ. The unsatisfactory ThinPrep Pap Test: missed opportunity for disease detection? Am J Clin Pathol. 2002;117(3): 457-463.

(28.) Allen KA, Zaleski S, Cohen MB. Review of negative Papanicolaou tests: is the retrospective 5-year review necessary? Am J Clin Pathol. 1994;101 (1):19-21.

(29.) Davey DD. Papanicolaou smear 5-year retrospective review: what is required by the Clinical Laboratory Improvement Amendments of 1988? Arch Pathol Lab Med. 1997;121(3):296-298.

(30.) Wilbur DC. False negatives in focused rescreening of Papanicolaou smears: how frequently are "abnormal" cells detected in retrospective review of smears preceding cancer or high-grade intraepithelial neoplasia? Arch Pathol Lab Med. 1997;121(3):273-276.

(31.) Howell S, Theodor M, Pacey NF, Patwardhan JR, Ayer B. Quality assurance in cytology: rescreening of previously negative smears from high grade squamous intraepithelial lesions. Acta Cytol. 1997;41(4):1085-1090.

(32.) Jones BA. Rescreening in gynecologic cytology: rescreening of 3762 previous cases for current high-grade squamous intraepithelial lesions and carcinoma--a College of American Pathologists Q-Probes study of 312 institutions. Arch Pathol Lab Med. 1995;119(12):1097-1103.

(33.) Jones BA. Rescreening in gynecologic cytology: rescreening of 8096 previous cases for current low-grade and indeterminate-grade squamous intraepithelial lesion diagnoses--a College of American Pathologists Q-Probes study of 323 laboratories. Arch Pathol Lab Med. 1996;120(6):519-522.

(34.) Spence AR, Goggin P, Franco EL. Process of care failures in invasive cervical cancer: systematic review and meta-analysis. Prev Med. 2007;45(2-3): 93-106.

Jennifer A. Brainard, MD; George G. Birdsong, MD; Tarik M. Elsheikh, MD; David A. Hartley, CT(ASCP); Kalyani Naik, MS, SCT(ASCP); Margaret H. Neal, MD; Rhona J. Souers, MS; Michael R. Henry, MD

Accepted for publication August 15, 2012.

From the Department of Anatomic Pathology, Pathology and Laboratory Medicine Institute, Cleveland Clinic, Cleveland, Ohio (Drs Brainard and Elsheikh); the Department of Anatomic Pathology, Grady Health System, Emory University School of Medicine, Atlanta, Georgia (Dr Birdsong); the Surveys Department (Mr Hartley) and the Department of Biostatistics (Ms Souers), College of American Pathologists, Northfield, Illinois; Anatomic Pathology Division, University of Michigan Health Systems, Ann Arbor (Ms Naik); the Department of Pathology, KWB Pathology Associates, Tallahassee, Florida (Dr Neal);and the Department of Cytopathology, Division of Pathology and Laboratory Medicine, Mayo Clinic, Rochester, Minnesota (Dr Henry).

This report was supported in part from a contract (GS-10F-0261K) funded by the Centers for Disease Control/Agency for Toxic Substances and Disease Registry. The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the Centers for Disease Control/Agency for Toxic Substances and Disease Registry and are not intended to take the place of applicable laws or regulations.

The authors have no relevant financial interest in the products or companies described in this article.

Reprints: Jennifer A. Brainard, MD, Department of Anatomic Pathology, Cleveland Clinic, 9500 Euclid Ave, L25, Cleveland, OH 44195 (e-mail:
Table 1. Prospective Rescreening Survey Results:
Selection Criteria for High-Risk Papanicolaou (Pap)
Tests (n = 502)

Criterion                                        Percentage

Clinical information                                95.4

Prior abnormal Pap result

  Atypical squamous cells of undetermined           82.7
  Atypical squamous cells, cannot exclude           88.4
    high-grade squamous intraepithelial lesion
  Low-grade squamous intraepithelial lesion         89.8
  High-grade squamous intraepithelial lesion        95.0
  Atypical glandular cells                          88.2

Prior abnormal cervical biopsy result

  Mild squamous dysplasia (CIN 1)                   80.5
  Moderate squamous dysplasia (CIN 2)               84.7
  Severe squamous dysplasia/carcinoma               85.1
    (CIN 3+)
  Adenocarcinoma in situ                            82.7

Prior positive high-risk human papillomavirus       61.2

Provider/clinical location                           9.8

Patient age                                          7.4

Abbreviation: CIN, cervical intraepithelial neoplasia.

Table 2. Prospective Rescreening Survey Results (a)

Upgrade of NILM Pap Result to:    Laboratory,   Cytotechnologist,
                                       %                %

Atypical squamous cells of           53.4             73.9
  undetermined significance
Atypical squamous cells, cannot      54.8             77.9
  exclude high-grade squamous
  intraepithelial lesion
Low-grade squamous                   67.3             94.1
  intraepithelial lesion
High-grade squamous                  69.2             95.3
  intraepithelial lesion
Atypical glandular cells             58.4             82.6
Squamous cell carcinoma              66.1             91.8
Adenocarcinoma                       57.6             81.2

Abbreviations: NILM, negative for intraepithelial lesion or
malignancy; Pap, Papanicolaou.

(a) Upgrade rates from NILM to an epithelial cell abnormality
that are actively monitored (n = 425).

Table 3. Retrospective Rescreening Survey Results:
Monitoring of Retrospective Review Interpretations

Pap Test Interpretation                        Laboratories

Atypical squamous cells of undetermined             56
Atypical squamous cells, cannot exclude             91
  high-grade squamous intraepithelial lesion
Atypical glandular cells                            80
Low-grade squamous intraepithelial lesion           91
High-grade squamous intraepithelial lesion          98
Adenocarcinoma in situ                              93
Carcinoma/other malignancy                          96

Abbreviation: Pap, Papanicolaou.

Table 4. Retrospective Rescreening Survey
Results: Monitoring of Upgrade Rates
(n = 375)

Upgrade Rates            Laboratories,

From NILM to HSIL+

  Cytotechnologists          82.7
  Pathologists               37.3
  Laboratory                 72.0


  Cytotechnologists          35.7
  Pathologists               18.1
  Laboratory                 27.5

Abbreviations: ASCUS, atypical squamous
cells of undetermined significance; HSIL,
high-grade squamous intraepithelial lesion;
NILM, negative for intraepithelial lesion
or malignancy.
COPYRIGHT 2013 College of American Pathologists
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2013 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Brainard, Jennifer A.; Birdsong, George G.; Elsheikh, Tarik M.; Hartley, David A.; Naik, Kalyani; Ne
Publication:Archives of Pathology & Laboratory Medicine
Article Type:Report
Geographic Code:1USA
Date:Feb 1, 2013
Previous Article:Folate receptor in adenocarcinoma and squamous cell carcinoma of the lung: potential target for folate-linked therapeutic agents.
Next Article:Quality improvement opportunities in gynecologic cytologic-histologic correlations: findings from the College of American Pathologists Gynecologic...

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |