Cost-effectiveness and accuracy of cervical cancer screening with a high-risk HPV genotyping assay vs a nongenotyping assay in China: an observational cohort study

Background New screening techniques may affect the optimal approaches for the prevention of cervical cancer. We evaluated the cost-effectiveness and accuracy of alternative screening strategies to provide evidence for cervical cancer screening guidelines in China. Methods In total, 32,306 women were enrolled. The current screening with Cervista® high-risk human papillomavirus (HR-HPV) nongenotyping and cytology cotesting (Cervista® cotesting) was compared with PCR-reverse dot blot HR-HPV genotyping and cytology cotesting (PCR-RDB cotesting). All eligible participants were divided into Arm 1, in which both HR-HPV assays were performed, and Arms 2 and 3, in which the PCR-RDB HPV or Cervista® HR-HPV assay, respectively, was performed. Outcome indicators included the cases, sensitivity, negative predictive value (NPV), colposcopy referral rate and cost of identifying cervical intraepithelial neoplasia of grade 2/3 or worse (CIN2+/CIN3+). Results Among the eligible participants, 18.4% were PCR-RDB HR-HPV-positive, while 16.9% were Cervista® HR-HPV-positive, which reflects good agreement (k = 0.73). PCR-RDB cotesting identified more CIN3+ cases than Cervista® cotesting in the first round of screening in Arm 1 (37 vs 32) and Arms 2/3 (252 vs 165). The sensitivity and NPV of PCR-RDB cotesting for identifying CIN3+ in Arm 1 (sensitivity: 94.9% vs 86.5%; NPV: 99.9% vs 99.7%) and Arms 2/3 (sensitivity: 95.1% vs 80.9%; NPV: 99.9% vs 99.6%) were higher than those of Cervista® cotesting, but the cost was similar. Conclusions The PCR-RDB HR-HPV genotyping and Cervista® HR-HPV assay results were consistent. PCR-RDB cotesting possesses optimal cost-effectiveness for cervical cancer screening in China, which has the highest number of cases globally but low screening coverage.

HPV vaccine was approved in China in 2017, and less than 30% of women over 21 years have been screened for cervical cancer [5]. Although nationwide cervical cancer screening programs began in 2009, due to the large population in China, the proportion of women undergoing programmatic screening is still low [5], and the main procedure for cervical cancer screening is screening at hospitals.
Although cytologic screening has effectively reduced the incidence and mortality, the accuracy of cytological results varies widely, ranging from 55 to 94% [6]. Cervical cancers and their precursors are closely related to persistent infection of high-risk human papillomavirus (HR-HPV) [7,8]. HPV assays are more reliable and less dependent on human expertise [9] and have been applied in cervical cancer screening and management algorithms to maximize the detection rate of high-grade cervical lesions (HSILs) or worse [10]. HPV assays are more sensitive than cytology, and a negative result can better predict a low risk of developing cervical cancer [11]. An HPV test with objective results and high sensitivity that is suitable for use not only in China but also in other developing countries is highly desirable [12]. It has been reported that the negative rate of HR-HPV in cervical cancer patients can be as high as 19.4% [13] to 23.3% [14]. Combined HR-HPV and cytology screening can improve the detection rate of HPV-negative cervical cancer women and reduce missed diagnoses; it is particularly suitable for China, which has low screening coverage. Currently, the screening method of cotesting with HPV and cytology assays is the main strategy for cervical cancer screening in China.
Many studies [18] have shown the great specificity and sensitivity of Cervista ® HR-HPV, similar to those of the HC2 assay. However, the efficacy of the Cervista ® HR-HPV assay combined with a cytology assay (Cervista ® cotesting) for cervical cancer screening is unknown. The PCR-RDB HPV genotyping assay is commonly used for cervical cancer screening in China, but the performance of PCR-RDB HPV genotyping and cytology cotesting (PCR-RDB cotesting) for cervical cancer screening is still unclear. To date, no studies have compared the efficacy of PCR-RDB cotesting and Cervista ® cotesting for the identification of HSIL or worse conditions in a screening population in China.
In this study, we therefore compared the cost-effectiveness of PCR-RDB HR-HPV genotyping and Cervista ® HR-HPV nongenotyping assays for cervical cancer screening among 32,306 Chinese women.

Study participants
The study was conducted in accordance with the 2008 Declaration of Helsinki and was approved by the Ethics Committee of Fujian Maternity and Child Health Hospital (No. 2012-031). All participants provided written informed consent. This cohort was a programmatic, large data-based observational cohort of women who underwent cervical cancer screening in China. All participants in this study were from the Fujian Province Cervical Lesions Screening Cohorts (FCLSCs), Fujian, China. FCLSCs are cervical cancer screening cohorts established in Fujian Province, with more than 200,000 cases used to assess the value of introducing HR-HPV testing into screening. Between July 2012 and August 2015, 32, 306 women who participated in a cervical cancer screening program in Fujian Maternity and Child Health Hospital were enrolled for testing with HR-HPV and cytology assays. The participants were required to meet the following criteria: (a) aged 21-65 years; (b) gynecological patient of the hospital; (c) no history of cervical cancer or cervical intraepithelial neoplasia (CIN); and (d) no history of cervical surgery or hysterectomy. Women were excluded based on the following criteria: (a) pregnancy or recent childbirth (within 6 weeks); (b) history of vulvar intraepithelial neoplasia or worse condition or vaginal intraepithelial neoplasia or worse condition; (c) history of other malignancies; (d) history of cervical cancer screening in the past three years; (e) serious autoimmune disease or uremia; and (f ) previous vaccination with an HPV vaccine.
Because the assets needed to compare two HPV assays using the same specimens in a screening population setting were limited, the enrolled participants were divided into three arms. Arm 1 was composed of participants screened for cervical cancer using the Cervista ® HR-HPV assay, PCR-RDB HR-HPV genotyping assay and ThinPrep ® Cytologic Test (TCT) simultaneously to evaluate the consistency and effectiveness of the HR-HPV results. Arm 1 consisted of all eligible participants who met the inclusion criteria and were enrolled between July 2012 and October 2012. Arm 2 was composed of participants screened for cervical cancer using the PCR-RDB HPV genotyping assay and TCT, and Arm 3 was composed of participants screened using the Cervista ® HR-HPV assay and TCT. Arm 2 and Arm 3 were designed to compare the cost-performance of the two HR-HPV assays for cervical cancer screening and consisted of all eligible participants who met the inclusion criteria and were enrolled between November 2012 and August 2015. Arm 2 included patients from Gynecology Districts 1-3 of Fujian Maternity and Child Health Hospital, and Arm 3 included patients from Gynecology Districts 4-5 of Fujian Maternity and Child Health Hospital.

Screening procedures in the Arm 1 cohort
Experienced gynecologists performed all gynecological examinations of the vulva, vagina and cervix for all eligible participants and performed speculum exams, collected specimens of cervical exfoliated cells with a broom brush, and transferred the cells to PreservCyt liquid (Hologic Inc., Boston, USA), which was stored at 4 ℃ until use in the cytology and HR-HPV DNA assays. According to the guidelines [16], Arm 1 included women who simultaneously underwent testing with the Cervista ® HR-HPV assay, PCR-RDB HR-HPV genotyping assay and TCT assay for primary cervical cancer screening. The Cervista ® HR-HPV assay and PCR-RDB HR-HPV genotyping assay were both performed in the same laboratory using the same cervical specimen. The cytology assay was performed in another pathology department; the laboratory staff responsible for each assay were unaware of the results of the other assay(s) when they performed the analyses. In the first round of screening, women with negative HR-HPV and cytology results were instructed to undergo routine screening after 3 years. Individuals whose samples were positive for HR-HPV types other than 16 and 18 (non-16/18) and those who had normal cytology or those with a negative HR-HPV result and cytology demonstrating atypical squamous cells of undetermined significance (ASCUS) underwent repeat HR-HPV genotyping/HR-HPV and cytology assays after 1 year. Women were referred for colposcopy and biopsy if they tested positive for HR-HPV types 16/18 regardless of the cytology results, if they tested positive for any types of HR-HPV and had cytology demonstrating ASCUS, or if the cytology results were indicative of low-grade squamous intraepithelial lesions or worse (≥ LSIL) regardless of the HR-HPV results. Colposcopy was performed by a skilled colposcopy physician. The squamous-columnar junction of the cervix was completely visible by colposcopy, which indicated that the colposcopy was satisfactory. When the colposcopy was normal, 4 sites in the cervix were randomly selected for biopsy. Conversely, when the colposcopy was abnormal, the lesioned tissue was obtained for cervical biopsy. If the colposcopy was unsatisfactory (the squamous-columnar junction was not completely visible), endocervical curettage (ECC) was performed immediately. Women with histologically confirmed cervical intraepithelial neoplasia of grade 2 or worse (CIN2+) were advised to undergo treatment according to the American Society for Colposcopy and Cervical Pathology (ASCCP) guidelines [16,19]. All untreated women were followed up, and a second round of screening was performed 3 years later. A detailed flowchart of the screening procedure is shown in Figs. 1 and 3a, b

Screening procedures in the Arm 2 cohort (PCR-RDB cotesting)
For PCR-RDB HR-HPV genotyping and cytology cotesting for primary cervical screening in Arm 2 women, the two cervical exfoliated cell samples were simultaneously collected from all Arm 2 women for PCR-RDB HR-HPV genotyping and cytology assays. Cervical exfoliated cells were collected and preserved in the same manner as that in the Arm 1 women. Women with negative HR-HPV genotyping and cytology results were instructed to undergo a second routine screening after 3 years. Individuals whose samples were positive for HR-HPV non-16/18 and who had normal cytology or those with a negative HR-HPV result and cytology demonstrating ASCUS underwent repeat HR-HPV genotyping and cytology assays after 1 year. Women were referred for colposcopy and/or biopsy within 10 weeks of the collection of exfoliated cells if they tested positive for HR-HPV types 16/18 regardless of the cytology results, if they tested positive for any types of HR-HPV and had cytology demonstrating ASCUS, or if the cytology results were indicative of ≥ LSIL regardless of the HR-HPV results. Women with histologically confirmed CIN2+ were treated according to the ASCCP guidelines [16,19]. All untreated women were followed up, and a second round of screening was performed 3 years later. A detailed flowchart of the screening procedure is shown in Figs. 2 and 3c.

Screening procedures in the Arm 3 cohort (Cervista ® cotesting)
Cervical exfoliated cells were collected and preserved in the same manner as in Arm 1 women. Arm 3 included women who were simultaneously tested by the Cervista ® HR-HPV nongenotyping assay and cervical cytology assay for primary cervical cancer screening. In the first round of screening, women with negative HR-HPV and cytology results were routinely screened 3 years later. Women with a positive HR-HPV result and normal cytology or a negative HR-HPV result and cytology demonstrating ASCUS underwent repeat testing with HR-HPV and cytology assays after 1 year. If both the HR-HPV and cytology results were negative at the 1-year follow-up, a second round of screening was performed 3 years later; however, if the HR-HPV result was positive or cytological abnormalities were noted, the patient was immediately referred for colposcopy and/or biopsy. Women with a positive HR-HPV result and cytology demonstrating ASCUS and those with cytology results classified as ≥ LSIL regardless of the HR-HPV result immediately underwent colposcopy and/or biopsy. Subsequent management of women with histologically confirmed CIN2+ was similar to that in Arm 1. A detailed flowchart of the screening procedure is shown in Figs. 2 and 3d.

Outcomes
The study compared key program indicators in terms of consistency, effectiveness and cost. When assessing consistency (defined as the degree of agreement between two   similar assays applied to the same specimen), our goal was to evaluate whether the Cervista ® HR-HPV assay and PCR-RDB HR-HPV genotyping assay can identify HR-HPV infection in the same cervical exfoliated cells. We defined consistency according to the percentage and kappa value of similar results for HR-HPV detection using the same specimen, which was detected by the PCR-RDB HPV genotyping assay and the Cervista ® HR-HPV assay. When evaluating effectiveness, the purpose of this study was to compare the effectiveness of the PCR-RDB cotesting (PCR-RDB HPV genotyping assay) and the Cervista ® cotesting (Cervista ® HR-HPV assay) strategies in identifying CIN2+/CIN3+ lesions. The outcome was reflected in the proportion of women with pathologically identified CIN2+/CIN3+ among all screened women in the first round, 1-year follow-up and second round of screening in Arm 1 and Arm 2/Arm 3, as well as the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), positive likelihood ratio (PLR) and negative likelihood ratio (NLR) of the screening strategy. Sensitivity was defined as the percentage of women identified as positive by the PCR-RDB cotesting (or Cervista ® cotesting) among women with a cervical histopathological diagnosis of CIN2+/ CIN3+. Specificity was defined as the percentage of women identified as negative by PCR-RDB cotesting (or Cervista ® cotesting) among women with a cervical histopathological diagnosis of normal/CIN1. The PPV was defined as the percentage of women who had a CIN2+/ CIN3+ outcome (identified by cervical histopathology) out of all women who tested "positive" on PCR-RDB cotesting (or Cervista ® cotesting). The NPV was defined as the percentage of women who had a normal/CIN1 outcome out of all women who tested "negative" on PCR-RDB cotesting (or Cervista ® cotesting). Cost was defined as the cost of performing cervical cancer screening (per 1000 screened women) [20].    [21,22] have demonstrated that the Cervista ® HR-HPV assay is an effective HR-HPV detection method.

PCR-RDB HR-HPV genotyping assay
The PCR-RDB HPV genotyping assay (Yaneng Bioscience Co., Ltd., China) can detect 14 HR-HPV genotypes (HPV-16, -18, -31, -33, -35, -39, -45, -51, -52, -56, -58, -59, -66, -68), and the type distribution was the same as that of the Cervista ® HR-HPV assay. The PCR-RDB HR-HPV genotyping assay is based on the PCR and reverse dot hybridization method for detecting the 14 HR-HPV genotypes. The PCR-RDB HPV genotyping assay was performed according to the manufacturer's instructions. Briefly, an aliquot of a 5-μL DNA sample was used. HPV was amplified in a thermal cycler under the following conditions: 50 °C for 15 min and 95 °C for 10 min, followed by a total of 40 cycles of 94 °C for 10 s, 45 °C for 90 s, and 72 °C for 30 s. The PCR products were immobilized onto a nitrocellulose membrane and hybridized with typing probes. After amplification, HPV genotyping was performed by RDB hybridization on nitrocellulose membrane strips fixed with different HPV type-specific probes. Our study and previous studies [17,23] have demonstrated that the PCR-RDB HR-HPV genotyping assay is an effective HR-HPV genotyping detection method.

Liquid-based cytology
All liquid-based cytology specimens were blinded and independently evaluated by 2 experienced cytopathologists. If the diagnoses differed, the sample was reviewed again, and a consensus diagnosis was obtained. The results were analyzed using the Bethesda system. Samples were classified as negative for intraepithelial lesion or malignancy (NILM), ASCUS, LSIL, atypical squamous cells, and it was not possible to exclude high-grade squamous intraepithelial lesions (ASC-H), HSIL, squamous cervical cancer (SCC) and atypical glandular cells (AGC).

Histology
Women with a punch biopsy diagnosis of HSIL or higher were treated with cold knife conization or a loop electrosurgical excision procedure (LEEP). Formalin (10%) was used to fix specimens, which were routinely processed for paraffin embedding. Subsequently, nearly 4-μm-thick histological sections were cut and stained with hematoxylin and eosin according to the standard instructions. Then, cervical biopsy specimens were histologically examined and classified according to the CIN system.

Sequencing
Products of the HPV L1 gene amplified from samples by nested PCR using type-specific primers were purified using the QIAquick PCR Purification Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions and sequenced by Sangon Biotech Co., Ltd. The resulting DNA sequences were compared with the sequences of known HPV types using the basic local alignment search tool from the NCBI website (https :// www.ncbi.nlm.nih.gov/BLAST ).

Statistical analysis
The consistency of the PCR-RDB HPV genotyping and Cervista ® HR-HPV assays was assessed using Cohen's kappa value. Kappa (k) values of 0-0.

Comparison of the consistency of the PCR-RDB HPV genotyping and Cervista ® HR-HPV assays for detecting HR-HPV infection
In total, 32,306 women who participated in the cervical cancer screening program were enrolled for testing with HR-HPV and cytology assays in this study. Overall, 5565 women were excluded because they met the exclusion criteria, and 1124 women were excluded because of inadequate cervical exfoliated cell specimens, invalid HR-HPV results, or invalid cytology results. (744/13,505) in the Cervista ® non-A9-positive group (Cervista ® HR-HPV A9 group negative, but A5/A6 group and/or A7 group positive). The HR-HPV positivity rate of the PCR-RDB genotyping assay showed two age peaks, 21-24 (20.1%) and 50-65 (24.0%) years, but PCR-RDB HPV-16/18 positivity rates did not show an age association. Similar results were obtained for the Cervista ® HR-HPV assay, as shown in Additional file 2: Figure  S1A. Detailed results for the two HPV assays in different age subgroups, cytology subgroups and pathology subgroups are shown in Additional file 2: Figure S1. Of the 25,617 participants, 2.9% (748/25,617) of the women were diagnosed with CIN2+ and 2.0% were diagnosed (501/25,617) with CIN3+ during the first-round screening and 1-year follow-up. Consistency analysis indicated that the results of the two HPV assays showed good agreement (k = 0.73, 95% confidence interval [CI] 0.69-0.76). This consistency remained for women in different age groups. However, the consistency of the two HPV assays for LSIL (k = 0.83, 95% CI 0.64-1.00)/HSIL (k = 0.85, 95% CI 0.56-1.00) cytology samples was higher than that in the cytologynegative samples (k = 0.65, 95% CI 0.61-0.69). Detailed results for the consistency analysis are shown in Fig. 4. In total, 94 samples were PCR-RDB HR-HPV-positive but Cervista ® HR-HPV-negative, and 78 samples were PCR-RDB HR-HPV-negative but Cervista ® HR-HPV-positive. Sixty samples were PCR-RDB HPV-16/31/33/35/52/58 (A9 group)-positive but Cervista ® HR-HPV A9 groupnegative, and 42 samples were PCR-RDB HPVA9 groupnegative but Cervista ® HR-HPV A9 group-positive (Additional file 3: Table S2).

Discussion
As the largest developing country, cervical cancer poses a significant health burden in China. The PCR-RDB HPV genotyping assay is a PCR-based method used to detect and identify viral DNA from 14 HR-HPV genotypes. Some studies have confirmed that HPV-16/18 genotyping has excellent triage performance in HPV-positive women [24]. Therefore, in 2012, the ASCCP recommended that HPV-16/18-positive women be directly referred for colposcopy, regardless of their cytology results [16]. The PCR-RDB HPV genotyping assay is widely used in China because it can detect 14 HR-HPV genotypes separately. Our previous research [17] revealed that the PCR-RDB HPV genotyping assay provided a sensitive and reliable method for clinical applications of cervical cancer screening. The Chinese Society of Colposcopy & Cervical Pathology (CSCCP) guidelines recommend combined screening using HR-HPV and cytology as the primary method for cervical cancer screening in China [17,25]. However, no studies have evaluated the efficacy of PCR-RDB cotesting in the identification of HSIL or worse conditions in China.
In this study, we used Cervista ® cotesting as a reference screening strategy and compared the cost-effectiveness of PCR-RDB cotesting for cervical cancer screening in China. In Arm 1 and Arms 2/3, we found that PCR-RDB cotesting had a higher sensitivity for identifying CIN2+ in the first round of screening. PCR-RDB cotesting can result in the detection of more CIN2+ women than Cervista ® cotesting in the first round of screening; thus, implementing the PCR-RDB cotesting screening strategy will result in earlier detection of precancerous lesions and cervical cancer.  , 95% confidence interval for the kappa value prompt intervention or treatment of patients, which can help prevent progression to severe cervical cancer and the subsequent loss of the chance for a cure [26]. Our research confirms that the implementation of the PCR-RDB cotesting screening strategy in regions with a high burden of cervical cancer will help that region recognize the WHO cervical cancer elimination targets as soon as possible [26].
The positivity rate of 14 genotypes of HR-HPV was 18.4% with the PCR-RDB HR-HPV genotyping assay, which is slightly higher than that with the Cervista ® HR-HPV (16.9%) and higher than that reported by Xiumin Zhao et al. [27] (9.9%). This result may have been obtained because the participants in the study were screened in hospitals rather than in the community. After age stratification, the positivity rate of the PCR-RDB HR-HPV genotyping and Cervista ® HR-HPV assays had two age peaks, 21-24 years and 50-65 years, which was similar to the findings from other studies from China [17,[27][28][29], Japan [30] and Chile [31], but HR-HPV prevalence Table 1 The performances of PCR-RDB genotyping cotesting and Cervista ® cotesting for identifying CIN2+. (N = 24,819) PCR-RDB cotesting, primarily screens women with both cytology and PCR-RDB HR-HPV genotyping assays, and then refers those with cytology ASCUS and PCR-RDB HR-HPV positive/PCR-RDB HPV16 or HPV18 positive/cytology LSIL or worse to colposcopy; Cervista ® cotesting, primarily screens women with both cytology and Cervista ® HR-HPV assays, and then refers those with cytology ASC-US and HR-HPV positive/cytology LSIL or worse to colposcopy; Arm 1, composed of participants screened for cervical cancer using the Cervista ® HR-HPV assay, PCR-RDB HPV genotyping assay and ThinPrep ® Cytologic Test (TCT) simultaneously; Arm 2, composed of participants screened for cervical cancer using the PCR-RDB HR-HPV genotyping assay and the TCT; Arm 3, composed of participants screened using the Cervista ® HR-HPV assay and the TCT CIN2+, cervical intraepithelial neoplasia grade 2 or worse; PPV, positive predictive value; NPV, negative predictive value; PLR, positive likelihood ratio; NLR, negative likelihood ratio a Calculate only the cost of screening women for cervical cancer screening in hospitals b Compare Arm 2′s screening strategy with Arm 3′s screening strategy  [32,33]. The main cause of the first peak of HPV infection is that women aged 21-24 are sexually active. The second peak of HR-HPV infection may be due to factors such as increased extramarital sexual behavior of the women or their husbands or decreased immunity, which could lead to infection with HPV with a low replication status [34]. In our study, good consistency was found between the PCR-RDB HR-HPV genotyping and Cervista ® HR-HPV assays (k = 0.73). However, the consistency among HPV test results was lowest in the cytology NILM group (k = 0.65). This finding may be explained by the very low viral load that is near the cutoff value of the HPV test, leading to poorly judged results in this subgroup [35]. Some studies [36,37] have also confirmed this problem. Previous studies have shown that Cervista ® and HC2 have similar properties in terms of HSIL identification, although their principles are different [18]. Therefore, the results of this study indicated that the efficiency of PCR-RDB cotesting was significantly higher than that of Cervista ® HR-HPV cotesting for predicting CIN2+/ CIN3+, but the total cost of the two screening methods was similar for every 1000 women screened ($57,932 vs $57,922 in Arm 1; $56,993 vs $56,902 in Arms 2/3). However, the cost of PCR-RDB cotesting was lower than that of Cervista ® cotesting for each CIN2+/CIN3+ patient identified ($2527 vs $2632 in Arm 1; $1654 vs $1867 in Arms 2/3). Cost is an important factor to consider when formulating screening strategies, especially in low-and middle-income countries where resources are scarce. This study revealed that PCR-RDB cotesting may be a more cost-effective alternative to cervical cancer screening strategies in low-and middle-income countries, which have low screening coverage and large populations.
The PCR-RDB HR-HPV genotyping cotesting strategy revealed a subgroup of HPV-16/18-positive but cytologically normal women. In this study, 25.3% (407/1609) of HR-HPV-positive but cytologically normal women were HPV-16/18-positive in Arm 1 and Arm 2/3, and all of them were directly referred for colposcopy. This procedure resulted in the detection of CIN2+ lesions in an additional 19.1% (81/425) of patients during the first round of screening, confirming the importance of a PCR-RDB genotyping cotesting program for early detection of  [38]. Therefore, HR-HPV genotyping cotesting can be used to directly refer HPV-16/18-positive women for colposcopy to achieve early diagnosis and reduce misdiagnosis of CIN2+ lesions, especially in lowand middle-income countries, which have low screening coverage.
The follow-up of cervical cancer screening is difficult in low-and middle-income countries. In this study, all women with serious abnormal screening results in the first round of screening were directly referred for colposcopy, while those with slight abnormalities will undergo repeat HR-HPV and TCT testing one year later. To increase the follow-up rate of women with abnormal screening results, we have established telephone follow-up and online follow-up for women with abnormal screening to detect the progression of cervical disease over time and reduce the rate of missed follow-up. All colposcopy and pathological examinations in this study were performed by two expert colposcopy physicians or pathologists. We also regularly provide technical training for colposcopy physicians and pathologists to improve the accuracy of colposcopy diagnosis and pathological diagnosis.
There are some limitations of this study that should be noted. First, some women with HR-HPV (non-HPV-16/18)-positive or cytology-positive results alone were not referred for colposcopy; thus, the evaluation of the specificity and PPV of the HPV assay may have been affected. Second, some women were not followed up after receiving abnormal screening results, and the loss to follow-up rate in our study was 3.1% (798/25,617), which is similar to what has been reported in other studies [39]. Although our loss to follow-up was within the allowable range, these women were at high risk of CIN2+/CIN3+, which may have led to reduced sensitivity of the study results. In future studies, telephone follow-up is needed to reduce the rate of loss of follow-up and detect disease development of patients to show more realistic results.

Conclusions
In summary, the results of this study indicate that the PCR-RDB HPV genotyping assay and the Cervista ® HR-HPV assay are effective for detecting CIN2+ and CIN3+. Additionally, the strategy of PCR-RDB HPV cotesting for the 16/18 genotypes presents a higher sensitivity and NPV than the standard strategy of Cervista ® HR-HPV cotesting without increasing the cost. The PCR-RDB HPV genotyping assay as a cotest for the 16/18 genotypes possesses optimal cost-effectiveness for screening and should be recommended for cervical cancer screening programs in China and other developing countries.