Cytohistologic correlation rates between conventional Papanicolaou smears and ThinPrep cervical cytology: a comparison

BACKGROUND: The ThinPrep Papanicolaou (Pap) test, a liquid-based alternative to the long-used conventional Pap smear, received Food and Drug Administration approval in 1996. The current study is a comparison of the accuracy, as assessed by cytohistologic correlation, of conventional versus ThinPrep Pap smears in a hospital setting. METHODS: A retrospective analysis of all patients who underwent Pap smears (conventional and ThinPrep) and pertinent surgical pathology tissue processed in our department over a 1-year period was undertaken. Tissue diagnoses were grouped into broad general categories to correlate with the Bethesda system. The statistical significance of differences between the two types of Pap test results compared with tissue diagnoses was assessed using a chi-square test with 1 degree of freedom and an alpha value of 0.05. RESULTS: Of the 65,421 Pap smears interpreted during the study period, histologic correlation was possible in 1544 patients. Of those diagnosed as negative (644 conventional and 245 ThinPrep Smears), low-grade squamous intraepithelial lesion (LGSIL) (116 conventional and 114 ThinPrep smears), and high-grade squamous intraepithelial lesions (HGSIL) (142 conventional and 131 ThinPrep smears), there was no statistically significant difference found between the correlation rates of conventional and ThinPrep Pap smear results. However, review of all cases of invasive cervical carcinoma identified during this time period showed the ThinPrep to be less consistent in predicting the presence of invasive carcinoma compared with the conventional Pap smear. CONCLUSIONS: Based on cytohistologic correlation evidence, the ThinPrep Pap test may not be more effective in detecting cervical disease, particularly invasive carcinoma, when compared with the conventional Pap smear.     

Comparison of the SurePath liquid-based Papanicolaou smear with the conventional Papanicolaou smear in a multisite direct-to-vial study

BACKGROUND: Split-sample clinical trials for liquid-based Papanicolaou (Pap) smears demonstrated that the liquid-based Pap smear was a safe and effective replacement for the conventional Pap smear. However, clinical intended use of liquid-based technology employs direct-to-vial collection methods. The current study compared the cytologic detection rates of the liquid-based Pap smear with conventional Pap smears in a direct-to-vial study performed at three clinical sites. METHODS: Data from 58,580 prospective SurePath slides and 58,988 historic conventional slides were collected. Results were statistically compared with regard to disease prevalence and adequacy to include biopsy follow-up data for conventional and SurePath tests. RESULTS: The SurePath method was found to provide a statistically significantly greater detection rate for clinically important categories of high-grade squamous intraepithelial lesion (HSIL+) and low-grade squamous intraepithelial lesion (LSIL+) (64% and 107%, respectively; P < 0.00001 for each lesion) compared with conventional slides. The clinical significance of increased cytologic detection using SurePath was supported by biopsy data that essentially demonstrated concordance with regard to biopsy interpretation for HSIL+ (P = 0.9105 at Site 1; P = 1.0000 at Site 2; and P = 1.0000 at Site 3) and LSIL+ (P = 0.6966 at Site 1; P = 0.8052 at Site 2; and P = 1.00 at Site 3). The detection rate of atypical squamous cells of undetermined significance (ASCUS+) was found to be significantly increased (75.12%; P < 0.00001). A statistically significantly lower proportion of unsatisfactory slides using the SurePath test compared with conventional slides was noted (-58%; P < 0.00001). The ASCUS/LSIL+ ratio was found to be reduced overall when using SurePath (-28.9%), regardless of whether the study sites were combined or considered individually. The rate of false-negative results noted with SurePath (10.43%) and conventional slides (12.97%) was essentially equivalent. CONCLUSIONS: The SurePath Pap smear was found to outperform conventional slides in the detection of HSIL+ and LSIL+ cytologic lesions of the cervix and reduced the number of unsatisfactory diagnoses. The HSIL+ advantage for SurePath is not limited to HSIL but appears to extend to carcinoma as well.     

Papanicolaou smear history in women with low-grade cytology before cervical cancer diagnosis

BACKGROUND: The purpose of the current study was to examine the screening histories of women diagnosed with invasive cervical cancer (ICC) in 2000 who had previous Papanicolaou (Pap) smears deemed to be unsatisfactory or with low-grade findings that did not lead to biopsy. METHODS: A total of 252 Pap smears from 47 women taken between 1992 and 2000 were included in the study; 247 smears were reexamined at the laboratory of origin before the study and all 252 were then reexamined independently by 2 experienced cytotechnicians and 2 cytopathologists. RESULTS: Of the 47 cases of ICC, 35 were squamous cell carcinoma, 10 were adenocarcinoma, and 2 were other types. On reexamination at the laboratory of origin, 24 cases were upgraded and in the study group 27 cases were upgraded to diagnoses requiring biopsy. On reexamination at the laboratory of origin, it was found that the first high-grade squamous intraepithelial lesion (HSIL) could have been diagnosed on average 4.2 years earlier than it was originally (95% confidence interval [95% CI], 3.3-5.1 years). On reexamination by the study group the first diagnosis of HSIL was made in smears dating from 5.4 years before the diagnosis of ICC (95% CI, 4.5-6.2 years). CONCLUSIONS: The study confirms that unsatisfactory and low-grade Pap smears imply a risk of developing high-grade lesions at a later date and shows that in a screening program a subgroup of smears may be diagnosed as unsatisfactory or low grade despite the presence of high-grade findings that are detectable on reexamination.     

Observer variation in cytologic grading for cervical dysplasia of Papanicolaou smears with the PAPNET testing system.

BACKGROUND: To assess the interobserver and intraobserver variation of Papanicolaou (Pap) smear screening with the computer-assisted (neural network based) PAPNET Testing System in diagnosing cervical smear abnormalities, results of agreement were compared with the interobserver and intraobserver variation of conventional smear analysis. METHODS: Cervical smears obtained from women in 1996 were reevaluated both by conventional light microscopy and with use of the PAPNET Testing System by the same four investigators, and results were compared with the original screening diagnoses obtained by both methods. RESULTS: The interobserver results for epithelial abnormalities (the degree of agreement between the cytologists), characterized by weighted kappa statistics, were 0.71 (95% CI: 0. 68-0.73) for PAPNET screening and 0.69 (95% CI: 0.66-0.72) for conventional screening. No significant differences were found among the individual results obtained by the four cytotechnologists (intraobserver variation) with conventional screening versus PAPNET reviewing. CONCLUSIONS: Pap smear grading with the PAPNET Testing System has interobserver and intraobserver variation similar to that of conventional screening of Pap smears in routine use. Cancer (Cancer Cytopathol) Copyright 1999 American Cancer Society.     

Liquid‐based cytology versus conventional Papanicolaou smear in an organized screening program

BACKGROUND.

The objective of this study was to evaluate whether liquid‐based cytology (LBC) can improve high‐standard cervical cancer screening cytology further. The primary endpoint was histopathologic high‐grade lesions in current and subsequent screening rounds. The secondary endpoints were cytologic diagnosis and inadequate samples.

METHODS.

Women were randomized to smear taking by conventional Papanicolaou (Pap) smear or LBC according to the time of appointment. Eight thousand eight hundred ten conventional Pap smears and 4674 LBC samples were included. Evaluations of atypical cytology and referral to colposcopy and treatment were performed as routine procedures. Histopathologic diagnoses were retrieved from a regional database 8 months after the study was closed. The mean follow‐up was 2 years and 9 months.

RESULTS.

Inadequate samples were observed in 0.3% of LBC samples versus 0.7% of Pap smears (P = .002). The total fraction of nonbenign diagnoses in cytology was 4.5% versus 3.5%, respectively (P < .001). Histopathologic evaluation was made on 570 patients constituting 4.6% of the LBC samples and 4% of the Pap smears. Forty percent more high‐grade lesions were identified as a result of LBC sampling (1.20% vs 0.85%; P = .05). The influence of the sampling method was significant for all variables (odds ratio [OR], 1.60; 95% confidence interval [95% CI], 1.12–2.28) for high‐grade lesions that were identified by histology when adjusting for age and screening unit in a logistic regression model. At the second follow‐up 2 years and 1 month later, the OR was decreased only slightly (1.51; 95% CI, 1.13–2.01).

CONCLUSIONS.

In the ongoing cervical screening program of western Sweden, liquid cytology produced a significantly higher yield of histologic high‐grade lesions compared with conventional Pap smears. Cancer (Cancer Cytopathol) 2007. © 2007 American Cancer Society.     

Papanicolaou smear history in women with low‐grade cytology before cervical cancer diagnosis

BACKGROUND.

The purpose of the current study was to examine the screening histories of women diagnosed with invasive cervical cancer (ICC) in 2000 who had previous Papanicolaou (Pap) smears deemed to be unsatisfactory or with low‐grade findings that did not lead to biopsy.

METHODS.

A total of 252 Pap smears from 47 women taken between 1992 and 2000 were included in the study; 247 smears were reexamined at the laboratory of origin before the study and all 252 were then reexamined independently by 2 experienced cytotechnicians and 2 cytopathologists.

RESULTS.

Of the 47 cases of ICC, 35 were squamous cell carcinoma, 10 were adenocarcinoma, and 2 were other types. On reexamination at the laboratory of origin, 24 cases were upgraded and in the study group 27 cases were upgraded to diagnoses requiring biopsy. On reexamination at the laboratory of origin, it was found that the first high‐grade squamous intraepithelial lesion (HSIL) could have been diagnosed on average 4.2 years earlier than it was originally (95% confidence interval [95% CI], 3.3–5.1 years). On reexamination by the study group the first diagnosis of HSIL was made in smears dating from 5.4 years before the diagnosis of ICC (95% CI, 4.5–6.2 years).

CONCLUSIONS.

The study confirms that unsatisfactory and low‐grade Pap smears imply a risk of developing high‐grade lesions at a later date and shows that in a screening program a subgroup of smears may be diagnosed as unsatisfactory or low grade despite the presence of high‐grade findings that are detectable on reexamination. Cancer (Cancer Cytopathol) 2007. © 2007 American Cancer Society.     

Evaluation of Papanicolaou smear and effect of sample biopsy in follow-up of cervical dysplasia.

In a prospective study of women with dysplasia, a superficial sampling biopsy of the cervix was carried out as a periodic check of the cytological findings. The relationship of smear and biopsy results was evaluated, as well as the effect of the biopsy procedure on the subsequent course of dysplasia. We found sufficiently close agreement between smear and biopsy results to conclude that smears and biopsies measure similar aspects of dysplasia. It would appear unwise, however, to rely on Papanicolaou smears exclusively in following women with a history of dysplasia, since negative smears in such women may occasionally show dysplasia in the corresponding biopsy. Periodic corroborative biopsy procedures are therefore indicated in the follow-up care of women with a history of dysplasia. There was no evidence that a superficial sampling biopsy significantly altered the short-term course of dysplasia. There was also no evidence of a cumulative effect of repeated sampling biopsies. These results do not rule out possible effects of other forms of biopsy procedures and schedules on the subsequent course of dysplasia.     

Atypical glandular cells in conventional cervical smears: Incidence and follow-up

Background  

Atypical glandular cells on cervical smears are often associated with clinically significant uterine lesions. The frequency and accuracy of AGC-NOS (i.e. atypical glandular cells, not otherwise specified) diagnoses, regardless of the gland cell type or the degree of suspicion, and their outcome were investigated.     

Liquid-based cytology versus conventional Papanicolaou smear in an organized screening program : a prospective randomized study

BACKGROUND: The objective of this study was to evaluate whether liquid-based cytology (LBC) can improve high-standard cervical cancer screening cytology further. The primary endpoint was histopathologic high-grade lesions in current and subsequent screening rounds. The secondary endpoints were cytologic diagnosis and inadequate samples. METHODS: Women were randomized to smear taking by conventional Papanicolaou (Pap) smear or LBC according to the time of appointment. Eight thousand eight hundred ten conventional Pap smears and 4674 LBC samples were included. Evaluations of atypical cytology and referral to colposcopy and treatment were performed as routine procedures. Histopathologic diagnoses were retrieved from a regional database 8 months after the study was closed. The mean follow-up was 2 years and 9 months. RESULTS: Inadequate samples were observed in 0.3% of LBC samples versus 0.7% of Pap smears (P = .002). The total fraction of nonbenign diagnoses in cytology was 4.5% versus 3.5%, respectively (P < .001). Histopathologic evaluation was made on 570 patients constituting 4.6% of the LBC samples and 4% of the Pap smears. Forty percent more high-grade lesions were identified as a result of LBC sampling (1.20% vs 0.85%; P = .05). The influence of the sampling method was significant for all variables (odds ratio [OR], 1.60; 95% confidence interval [95% CI], 1.12-2.28) for high-grade lesions that were identified by histology when adjusting for age and screening unit in a logistic regression model. At the second follow-up 2 years and 1 month later, the OR was decreased only slightly (1.51; 95% CI, 1.13-2.01). CONCLUSIONS: In the ongoing cervical screening program of western Sweden, liquid cytology produced a significantly higher yield of histologic high-grade lesions compared with conventional Pap smears.     

Comparison of ThinPrep and conventional smears in salivary gland fine-needle aspiration biopsies

BACKGROUND: ThinPrep (TP) cytology for evaluation of nongynecological specimens is being increasingly used. There are few studies comparing TP with conventional smears (CS) in salivary gland (SG) fine-needle aspiration biopsies (FNAB). This study compares diagnostic accuracy and morphology of TP and CS in SG FNABs. METHODS: The authors retrospectively reviewed 98 satisfactory SG FNABs with both TP and CS. All cases had surgical resection. CS and TP slides were assessed for multiple morphological parameters, as well as the ability to make the diagnosis. Chi-square analysis was performed to compare CS and TP. RESULTS: An accurate diagnosis was rendered more commonly with CS compared with TP (57% versus 42%; P = .032), whereas the unsatisfactory rate was greater with TP compared with CS (19% versus 9%; P = .041). The error (4%) and indeterminate (35%) rates for TP were similar to CS. The diagnostic yield was greater for cellular cases, which were more frequent with CS compared with TP, than for cases of low cellularity; the diagnostic yield of cellular TP cases and cellular CS cases was similar. Artifacts (crush, air drying, obscuring blood) were more frequent (12%, 13%, and 27% versus 2%, 0%, and 1%; P 相似文献   

The effect of the quality of Papanicolaou smears on the detection of cytologic abnormalities.

BACKGROUND: Controversy continues regarding the relation between the quality of Papanicolaou (Pap) smears, especially the presence of endocervical cells (ECC), with the finding of cytologic abnormalities. METHODS: As part of a study regarding performance feedback on the quality of Pap smears, data from 56,475 Pap smears obtained by 176 participating clinicians over a 20-month period were analyzed to assess the relation between the presence of ECC, the categorization of global specimen adequacy as "satisfactory" or "satisfactory with limitations," and the prevalence of atypia and squamous intraepithelial lesions (SILs). RESULTS: Atypia was less likely to be found in "satisfactory" Pap smears than in "satisfactory with limitations" quality Pap smears (odds ratio [OR], 0.6; 95% confidence interval [CI], 0.5-0.6; P < 0.001), even though the latter could contain ECC. No association was found between satisfactory Pap smears and cytologic abnormalities. Compared with specimens with no ECC, an ECC count of > or = 50 on a slide was associated positively with the detection of atypia (OR, 2.1; 95% CI, 1.8-2.4; P < 0.001) or SILs (OR, 1.7; 95% CI, 1.3-2.2; P < 0.001). A similar relation existed between ECC counts of 25-50 (OR, 1.9; 95% CI, 1.1-2.2; P = 0.01) and the detection of SILs. No relation was found between specimens with < 25 ECC and the presence of atypia or abnormalities. CONCLUSIONS: The global adequacy criterion of "satisfactory" assigned to a Pap smear does not indicate that there is a greater likelihood of detecting cytologic abnormalities compared with lower quality Pap smears. To the authors' knowledge, previous studies regarding the link between ECC in the Pap smear and cytologic abnormalities have not addressed the relevance of how many ECC are needed to maximize the identification of abnormalities. The data from the current study support the value of obtaining at least 25 ECC as a quality indicator of sampling.     

Comparison of ThinPrep and conventional smears in detecting carcinoma in bile duct brushings

BACKGROUND: Bile duct brushing cytology is a common procedure for the exclusion of adenocarcinoma in the bile duct. The authors evaluated the use of ThinPrep (TP) to determine whether the information obtained is equivalent to that found with conventional smear cytology (CS). METHODS: Thirty-eight prospectively collected endoscope retrograde cholangiopancreatography-guided bile duct brushing samples were split in the following manner. First, two to four CS were prepared and immediately spray-fixed or wet-fixed. Second, the remaining sample was rinsed in PreservCyt (Cytyc Corp., Boxborough, MA). In the laboratory, one TP slide was prepared from each sample. TP and CS were stained by routine Papanicolaou stain. For the current study, TP and CS were reviewed independently by two cytopathologists. The diagnoses made by the two methods were compared with the final histology. RESULTS: The cytologic diagnoses for both TP and CS were categorized into five main groups: 1) unsatisfactory, 2) negative, 3) reactive, 4) suspicious for malignancy, and 5) malignant. The diagnoses on the 38 TP bile duct brushings and CS were categorized as follows: 1) unsatisfactory-2, 4; 2) negative-7, 4; 3) reactive-10, 14; 4) suspicious for malignancy-9, 9; and 5) malignant-10, 7, respectively. Histologic follow-up was available in 14 cases (reactive-4, suspicious for malignancy-1, and malignant-9). The sensitivity was 77% for TP and 66% for CS. The specificity was 100% for both methods. CONCLUSIONS: The two methods described in the current study detected equivalent disease on bile duct brushings. TP was found to provide better preservation and cytologic detail. However, the diagnostic criteria may require modification.     

Improved preoperative definitive diagnosis of papillary thyroid carcinoma in FNAs prepared with both ThinPrep and conventional smears compared with FNAs prepared with ThinPrep alone

BACKGROUND:

ThinPrep (TP) liquid‐based preparations are increasingly being used in nongynecologic specimens. Few studies have evaluated TP as a sole diagnostic modality in the setting of thyroid fine‐needle aspiration (T‐FNA). Herein, the authors evaluate the usefulness of TP as a sole diagnostic modality in a nonsplit sample.

METHODS:

Consecutive T‐FNAs were identified at 2 tertiary care institutions; 1 institution processed thyroid FNA entirely with TP, and the other used a combination of TP and conventional preparations (CP). Cytodiagnoses, surgical pathology, and/or clinical follow‐up were recorded. Performance parameters for the 2 settings were compared.

RESULTS:

A cytologic diagnosis of positive for malignancy was correct in 98.8% of TP + CP cases and in 100% of TP cases. Papillary thyroid carcinoma cases were definitively diagnosed in 53.1% of T‐FNAs prepared by TP + CP compared with 34.4% of T‐FNAs prepared with TP alone (P = .0015 by Fisher 2‐tailed exact test). Of patients ultimately diagnosed with papillary thyroid carcinoma, 89% were initially treated by total thyroidectomy in the TP + CP group compared with 79.5% in the TP‐only group (P = .027 by Fisher exact test).

CONCLUSIONS:

TP as a sole preparatory technique does not improve the usefulness of T‐FNA as a screening test. However, combining CP and TP increases the rate of definitive cytologic diagnosis of malignancy in papillary thyroid carcinoma. Thus, combining TP and CP enhances the diagnostic component of T‐FNA. Cancer (Cancer Cytopathol) 2011. © 2010 American Cancer Society.     

宫颈病变液基细胞学筛查与组织病理学对照观察

目的　探讨液基薄层细胞学(ThinPrepCytologyTest, TCT)技术在妇科门诊人群宫颈病变筛查的准确性。方法　回顾性分析10 980例TCT,与组织学对比观察。结果　　TCTLSIL以上阳性率45. 7% ( 373 /817),组织学检查阳性率50. 1% ( 409 /817 ),两者统计学比较无显著性差异(P>0. 05 )。TCT诊断符合率LSIL75. 8% (191 /252),HSIL98. 1% (101 /103),SCC90. 9% (10 /11),AC85. 7% (6 /7)。鳞状上皮内病变诊断符合率HSIL与LSIL统计学比较有显著性差异(P<0. 01)。结论　液基细胞学检查是宫颈癌早期筛查的有效手段,加强制片技术及诊断质量控制对提高诊断的准确性有重要意义。     

Estimating the percentage of Papanicolaou smears that can be reproducibly identified: modeling Papanicolaou smear interpretation based on multiple blinded rescreenings

BACKGROUND: Multiple blinded rescreenings of Papanicolaou (Pap) smears for litigation purposes is based on the assumption that a subset of Pap smears can be reproducibly identified. The size of this subset is not known. METHODS: To estimate the size of the subset of Pap smears that can be reproducibly identified, a model was constructed based on the results of repeated blinded screenings in the AutoPap Primary Screening System Trial. Additional analysis came from data in the literature. RESULTS: Routine and AutoPap-assisted screening both have a detection rate for all detected abnormal cases of < 50%. Models with only two subsets or types of slides each with a different detection rate correlated well with the available data. Data from multiple rapid reviews strongly supported the existence of additional definable subsets. Although the percentage of cases with an expected detection rate of 100% in a three-subset model might have been as high as 30% of the abnormal cases detected in a single review, all estimates that included a second subset of slides with at least a 50% detection rate limited the percentage of slides in the 100% sensitive subset of slides to < 2% of all abnormal slides and < 6% of all abnormal slides detected by a single screening. CONCLUSIONS: Repeated screenings of Pap smears allowed more accurate models of the sensitivity of Pap-smear screening and the overall incidence of abnormal cases. The data strongly supported the existence of multiple subsets of Pap smears, which can be defined by repeated blinded rescreenings. The percentage of slides that can be reproducibly identified was small.     

Comparison of ThinPrep and conventional smears in salivary gland fine‐needle aspiration biopsies

BACKGROUND

ThinPrep (TP) cytology for evaluation of nongynecological specimens is being increasingly used. There are few studies comparing TP with conventional smears (CS) in salivary gland (SG) fine‐needle aspiration biopsies (FNAB). This study compares diagnostic accuracy and morphology of TP and CS in SG FNABs.

METHODS

The authors retrospectively reviewed 98 satisfactory SG FNABs with both TP and CS. All cases had surgical resection. CS and TP slides were assessed for multiple morphological parameters, as well as the ability to make the diagnosis. Chi‐square analysis was performed to compare CS and TP.

RESULTS

An accurate diagnosis was rendered more commonly with CS compared with TP (57% versus 42%; P = .032), whereas the unsatisfactory rate was greater with TP compared with CS (19% versus 9%; P = .041). The error (4%) and indeterminate (35%) rates for TP were similar to CS. The diagnostic yield was greater for cellular cases, which were more frequent with CS compared with TP, than for cases of low cellularity; the diagnostic yield of cellular TP cases and cellular CS cases was similar. Artifacts (crush, air drying, obscuring blood) were more frequent (12%, 13%, and 27% versus 2%, 0%, and 1%; P ≤ .006) in CS compared with TP. Although fragmentation was greater and nuclear detail was better in TP (P ≤ .03), cell size was larger in CS (P = .002). A specific diagnosis of pleomorphic adenoma (PA) was more frequently rendered with CS compared with TP (83% versus 63%; P = .045). PA stroma was more abundant, and an epithelial‐stromal interface (ESI) was more frequent in CS compared with TP (ESI, 76% versus 38%; P ≤ .001).

CONCLUSIONS

There are morphological differences between TP and CS in SG FNABs, especially with respect to stromal appearance. Although CS appears to be preferable to TP in the diagnosis of PA overall, CS and TP have equivalent diagnostic yield in highly cellular cases. Complementary use of both TP and CS preparations to achieve optimal diagnostic yield is recommended, given the artifacts of some CS and the not infrequent unsatisfactory nature of 1 preparation alone. Cancer (Cytopathol) 2007. © 2007 American Cancer Society.     

Comparison of HPV-based assays with Papanicolaou smears for cervical cancer screening in Morelos State,Mexico

Objective: To compare the performance of human papillomavirus (HPV) assays with conventional Pap cytology for cervical cancer (CC) screening in Mexico. Methods: Pap smears, self-collected vaginal specimens (SS) for HPV testing, and clinician-collected cervical specimens (CS) for HPV testing were obtained from 7868 women, aged 15–85 years old, attending CC screening at the Mexican Institute of Social Security (IMSS) between May and October, 1999. SS and CS specimens were screened for oncogenic HPV DNA by Hybrid Capture 2. Women who received cytological interpretations of atypical squamous cells of undetermined significance (ASCUS), and/or a positive HPV test were referred for colposcopy and histologic studies. The relative estimates for sensitivity, specificity and predictive values of each test were calculated using histological diagnoses of cervical intraepithelial neoplasia (CIN) grades 2 or 3, or CC histological diagnosis. Results: Oncogenic HPV detection rate was 11.6% for SS, and 9.3% for CS. Pap smear abnormalities were observed in 2.4% of the women. Of 1147 women who had at least one abnormal test result, 88.5% underwent colposcopy, and 101 biopsy-confirmed CIN2/3 or cancer cases were identified. The relative sensitivity estimates for the Pap test, SS and CS were 59.4% (95% CI: 49.2–68.9), 71.3% (95% CI: 61.3–79.6), and 93.1% (95% CI: 85.8–96.9), respectively, while the specificities were 98.3% (95% CI: 98.0–98.6), 89.2% (95% CI: 88.5–89.9), and 91.8% (95% CI: 91.2–92.4), respectively. The positive predictive values of Pap, SS and CS were 36.1, 9.1 and 14.9, the colposcopy referrals needed to detect a case of CIN2/3 or cancer were 2.8, 11.0 and 6.7, respectively. Discussion: Both HPV assays detected more cases of CIN2/3 or CC than Pap cytology alone. However, the HPV assays increased the number of colposcopy referrals. Our study suggests that HPV testing could be an effective way to improve the performance of CC screening.