The first biomarker-based cervical cancer screening test, p16/Ki-67 dual-stained cytology (DS), has been clinically validated and approved in the United States for triage of women being screened for cervical cancer who test positive for high-risk human papillomavirus (hrHPV). The primary aim of this work is to evaluate the cost-effectiveness of DS triage after co-testing findings of positive non-16/18 HPV types and atypical squamous cells of undetermined significance or low-grade squamous intraepithelial lesions cytology. A payer-perspective Markov microsimulation model was developed to assess the impact of DS reflex testing. Each comparison simulated 12,250 screening-eligible women through health states defined by hrHPV status and genotype, cervical intraepithelial neoplasia grades 1–3, invasive cervical cancer (ICC) by stage, and cancer-related or non-cancer death. Screening test performance data were from the IMPACT clinical validation trial. Transition probabilities were from population and natural history studies. Costs of baseline medical care, screening visits, tests, procedures, and ICC were included. DS reflex after co-testing was cost-effective with incremental cost-effectiveness ratios per quality-adjusted life-year gained of $15,231 [95% confidence interval (CI), $10,717–$25,400] compared with co-testing with hrHPV pooled primary and genotyped reflex testing, and $23,487 (95% CI, $15,745–$46,175) compared with co-testing with hrHPV genotyping with no reflex test. Screening and medical costs and life-years increased, while ICC costs and risk of ICC death decreased. Incorporating DS reflex into co-testing cervical cancer screening algorithms is projected to be cost-effective.

Prevention Relevance:

The p16/Ki-67 dual-stained cytology (DS) test was recently approved in the United States as a reflex test for cervical cancer screening following positive high-risk human papillomavirus (hrHPV) test results. Adding DS reflex to hrHPV and cervical cytology co-testing strategies in the United States is expected to be cost-effective per life-year or quality-adjusted life-year gained.

The 2018 U.S. Preventive Services Task Force cervical cancer screening guidelines for women with average cervical cancer risk in the United States indicate that primary high-risk human papillomavirus (hrHPV) testing has the most benefits for the least harms (1), and that co-testing with liquid-based cytology and hrHPV testing is an acceptable but expensive option with more harms (2). Most cervical cancer screening is still performed in the United States with the co-testing option where hrHPV testing is either a pooled test (returning a single positive or negative result for any hrHPV), or a partial genotyped test (differentiating between the hrHPV 16/18 types and 12 other hrHPV types) in about equal portions (3, 4).

The CINtec PLUS Cytology test [p16/Ki-67 dual-stained cytology (DS); Roche Diagnostics Solutions, a division of Roche, Basel, Switzerland] was approved by the FDA for reflex triage testing following positive results from the cobas 4800/6800/8800 HPV test (Roche Molecular Systems, Inc., Pleasanton, CA, USA) after primary hrHPV screening or cytology/hrHPV co-testing (5). This is the first next-generation biomarker cytology test approved in the United States for managing initial abnormal cervical cancer screening results.

While there are new risk-based management strategies published in 2019 (6), the currently used management strategy in the United States, a co-testing of hrHPV with cytology, differs at the level of the pathology lab. The first standard of care (SOC 1) uses the pooled hrHPV test option with cytology. For a positive pooled hrHPV test result accompanied by a negative for intraepithelial lesion or malignancy (NILM) cytology result, HPV genotyping in the lab is the only triage step (Fig. 1A). For NILM/hrHPV results that are 16/18 positive, current practice is an immediate referral to colposcopy; for NILM/12 other hrHPV types current practice is a referral to a shortened surveillance. Pooled hrHPV-positive results with any other cytology result leads to colposcopic biopsy.

Figure 1.

Cervical cancer screening algorithms tested in the model: (A) SOC 1: Primary hrHPV pooled testing and cervical cytology co-testing with reflex to hrHPV genotyping; (B) SOC 2: Primary hrHPV genotyping and cervical cytology co-testing with no reflex; (C) Intervention arm: Primary hrHPV genotyping and cervical cytology co-testing with reflex to DS.

Figure 1.

Cervical cancer screening algorithms tested in the model: (A) SOC 1: Primary hrHPV pooled testing and cervical cytology co-testing with reflex to hrHPV genotyping; (B) SOC 2: Primary hrHPV genotyping and cervical cytology co-testing with no reflex; (C) Intervention arm: Primary hrHPV genotyping and cervical cytology co-testing with reflex to DS.

Close modal

The second standard of care co-testing option (SOC 2) is very similar but uses genotyping results earlier in the laboratory process for all women, instead of a portion of women, resulting in the same current practice management (Fig. 1B). While these are commonly used standard clinical and laboratory protocols, these screening algorithms can lead to too many colposcopy referrals, causing harm (1–4).

Using DS as a reflex triage test for women with co-testing results showing positive for 12 other hrHPV types and less than high-grade squamous intraepithelial lesion (HSIL) cytology allows women who test DS negative to be followed at a shortened surveillance interval rather than undergoing colposcopic biopsy (Fig. 1C).

The primary aim is to evaluate the cost-effectiveness of a screening strategy incorporating DS reflex triage testing into co-testing strategies currently being used for cervical cancer screening in the United States.

A decision-analytic Markov microsimulation model with an annual cycle and lifetime time horizon was used to evaluate the cost-effectiveness of primary hrHPV genotyping and cervical cytology co-testing with reflex to DS (intervention algorithm) separately compared with two common standard-of-care co-testing algorithms, SOC 1 and SOC 2 (Fig. 2A). Simulated patients were modeled through health states representing hrHPV-negative with normal histology, hrHPV-positive (16/18+ or 12 other+) with normal histology, hrHPV-positive with cervical intraepithelial neoplasia (CIN) 1, 2, or 3, development of invasive cervical cancer (ICC), and ICC stage progression, from model entry until death (Fig. 2B).

Figure 2.

Markov model design illustrated in a (A) Decision tree diagram and a (B) Health states and annual transitions diagram. aYears between screenings (screening occurs once every 5 years under routine intervals or once every 3 years under shortened intervals), or after screening exit (women diagnosed with ICC or women over 65 years of age, with no abnormal screening results for the last decade and no LEEP in the last 25 years). bWomen who are directed to LEEP are modeled to continue under shortened interval screening for all subsequent screening, while women who have abnormal screening findings not requiring referral to LEEP are returned to routine screening only after 2 sequential screenings with normal findings. cPatients found to have CIN 2 or CIN 3 are directed to LEEP. dPatients with ICC are diagnosed when directed to colposcopy by screening, or when symptomatic. Upon ICC diagnosis, patients exit screening. Subsequent costs, quality of life, and mortality are assigned on the basis of ICC stage at diagnosis.

Figure 2.

Markov model design illustrated in a (A) Decision tree diagram and a (B) Health states and annual transitions diagram. aYears between screenings (screening occurs once every 5 years under routine intervals or once every 3 years under shortened intervals), or after screening exit (women diagnosed with ICC or women over 65 years of age, with no abnormal screening results for the last decade and no LEEP in the last 25 years). bWomen who are directed to LEEP are modeled to continue under shortened interval screening for all subsequent screening, while women who have abnormal screening findings not requiring referral to LEEP are returned to routine screening only after 2 sequential screenings with normal findings. cPatients found to have CIN 2 or CIN 3 are directed to LEEP. dPatients with ICC are diagnosed when directed to colposcopy by screening, or when symptomatic. Upon ICC diagnosis, patients exit screening. Subsequent costs, quality of life, and mortality are assigned on the basis of ICC stage at diagnosis.

Close modal

In both SOC comparisons, screening intervals were defined as one screening every 5 years for routine screening and one screening every 3 years for shortened screening. Patients with CIN 2 or CIN 3 detected by colposcopy and biopsy were directed to loop electrosurgical excision procedure (LEEP) and shortened screening intervals for all subsequent screening. Patients directed to shortened screening or colposcopy by abnormal test results but not to LEEP were required to have 2 sequential screenings at shortened intervals with normal test results before returning to the routine screening interval. Patients exited screening if they were diagnosed with ICC, or if they were over age 65, had no abnormal screening results for the last decade, and had no LEEP in the last 25 years.

Patient population

Microsimulations generated and analyzed patients between 25 to 65 years of age who were eligible for cervical cancer screening (7). Age distribution was informed by the U.S. Census Bureau 2017 National Population Projections Datasets [https://www.census.gov/data/datasets/2017/demo/popproj/2017-popproj.html (RRID: SCR_011587)]. As women with HIV have elevated risk of cervical cancer and are managed under different guidelines than the routine guidelines for the general population, they were not included in this model of typical screening algorithms and population-based risks (8, 9). Women who were HIV positive, pregnant, had previously undergone a total hysterectomy, or were noncompliant with screening were excluded from the analysis, informed by published age-specific rates (10–13).

Data from the IMPACT clinical trial were used to inform this model (Supplementary Table S1; refs. 14, 15). The prevalence of hrHPV by age, and prevalence of CIN 1–3 or ICC by hrHPV genotype and age were informed by the IMPACT clinical trial (Table 1; Supplementary Tables S1A and S1B; ref. 15). To ensure that rarer patient characteristics in the population (e.g., hrHPV 16/18+, CIN3) were represented among the model-generated patients at model entry, microsimulations were evenly divided across patient groups defined by patient age group (25–29, 30–39, 40–49, 50–59, 60–65 years), hrHPV status (negative, 16/18+, 12 other+), and histologic status (normal, CIN 1, CIN 2, CIN 3, ICC stage 1) at model entry (Supplementary Table S2).

Table 1.

Model entry starting hrHPV and histologic statuses.

hrHPV Prevalence by AgeHistologic Status Prevalence by Age and hrHPV Genotype
Age Group (years of age)GenotypeNon-VaccinatedVaccinatedaNormalCIN 1CIN 2CIN 3ICCTotal
25–29 16/18 6.7% 2.1% 66.7% 11.1% 9.5% 12.7% 0.0% 100% 
 12 Other 18.6% 20.0% 74.6% 11.8% 9.4% 4.1% 0.0% 100% 
 Negative 74.7% 78.0% 100%      
 Total 100% 100%       
30–39 16/18 5.5% 3.7% 68.4% 7.9% 9.1% 13.8% 0.8% 100% 
 12 Other 10.9% 12.6% 76.7% 11.6% 8.3% 3.5% 0.0% 100% 
 Negative 83.5% 83.7% 100%      
 Total 100% 100%       
40–49 16/18 4.1% 3.7% 79.6% 5.3% 7.4% 7.4% 0.4% 100% 
 12 Other 6.9% 12.6% 83.5% 7.2% 7.6% 1.5% 0.2% 100% 
 Negative 89.0% 83.7% 100%      
 Total 100% 100%       
50–59 16/18 3.4% 3.7% 88.3% 3.3% 3.9% 3.9% 0.6% 100% 
 12 Other 6.9% 12.6% 90.0% 4.6% 3.8% 1.6% 0.0% 100% 
 Negative 89.6% 83.7% 100%      
 Total 100% 100%       
60–65 16/18 3.1% 3.7% 89.1% 1.8% 7.3% 1.8% 0.0% 100% 
 12 Other 6.0% 12.6% 93.5% 0.9% 1.9% 3.7% 0.0% 100% 
 Negative 90.9% 83.7% 100%      
 Total 100% 100%       
hrHPV Prevalence by AgeHistologic Status Prevalence by Age and hrHPV Genotype
Age Group (years of age)GenotypeNon-VaccinatedVaccinatedaNormalCIN 1CIN 2CIN 3ICCTotal
25–29 16/18 6.7% 2.1% 66.7% 11.1% 9.5% 12.7% 0.0% 100% 
 12 Other 18.6% 20.0% 74.6% 11.8% 9.4% 4.1% 0.0% 100% 
 Negative 74.7% 78.0% 100%      
 Total 100% 100%       
30–39 16/18 5.5% 3.7% 68.4% 7.9% 9.1% 13.8% 0.8% 100% 
 12 Other 10.9% 12.6% 76.7% 11.6% 8.3% 3.5% 0.0% 100% 
 Negative 83.5% 83.7% 100%      
 Total 100% 100%       
40–49 16/18 4.1% 3.7% 79.6% 5.3% 7.4% 7.4% 0.4% 100% 
 12 Other 6.9% 12.6% 83.5% 7.2% 7.6% 1.5% 0.2% 100% 
 Negative 89.0% 83.7% 100%      
 Total 100% 100%       
50–59 16/18 3.4% 3.7% 88.3% 3.3% 3.9% 3.9% 0.6% 100% 
 12 Other 6.9% 12.6% 90.0% 4.6% 3.8% 1.6% 0.0% 100% 
 Negative 89.6% 83.7% 100%      
 Total 100% 100%       
60–65 16/18 3.1% 3.7% 89.1% 1.8% 7.3% 1.8% 0.0% 100% 
 12 Other 6.0% 12.6% 93.5% 0.9% 1.9% 3.7% 0.0% 100% 
 Negative 90.9% 83.7% 100%      
 Total 100% 100%       

Data informed by the IMPACT trial (Supplementary Tables S1 A and B; ref. 15) Sensitivity analyses varied the all-ages prevalences of hrHPV 16/18, hrHPV 12 other, and lesions by hrHPV genotype by ±25%.

aVaccination against the hrHPV 16/18 strains. Due to few vaccinated patients over 40 years of age, hrHPV prevalence among vaccinated patients over 40 were informed by data from patients 30 to 39 years of age.

Each patient group defined a sub-population of prevalent screening-eligible women within a specified age-range, and with a specific hrHPV status and histologic status at model entry (e.g., 30–39 years of age, hrHPV 16/18+, and normal histology at model entry). Using the published age-based data described above on the U.S. population, screening eligibility, hrHPV prevalence, and CIN 1–3 or ICC prevalence, we calculated the age distribution within each patient group present in the eligible population. From there, we drew a specific single-year age at model start, and a hrHPV vaccination status based on the probability of being vaccinated by age.

Each run of the model simulated 12,250 individual women undergoing cervical cancer screening (i.e., 250 microsimulated patients for each of the 49 patient groups present in the modeled starting population; Supplementary Table S2). Each microsimulated patient was modeled as a pair through both the intervention and the SOC pathway from model entry to death. Patient results were aggregated in each pathway by patient group, and then weighted on the basis of patient group prevalence in the United States screening-addressable population. Comparisons of the intervention with SOC 1 and SOC 2 were assessed separately using independently generated patient populations.

Screening tests and procedures

The model assessed cervical cancer screening with samples for primary screening tests collected in the office setting. Reflex testing for triage based on the primary test results happens at the pathology lab using the specimens already sampled, so it does not require additional office visits. Therefore, the model included office visits for primary testing, but not for reflex testing.

Cervical cancer screening test performance were informed by the IMPACT trial results and published literature (14–18). The sensitivity of hrHPV tests to detect hrHPV was based on the probability of positive hrHPV test results among women with CIN 3+ (CIN 3, adenocarcinoma in situ, ICC) in the IMPACT trial (Table 2; ref. 14). This population was used to inform this sensitivity input because CIN 3 is highly likely to be hrHPV-related, so we assumed that these patients had hrHPV infection. Among these patients, 95.5% received positive hrHPV test results, which was applied in our model as the probability that patients with hrHPV infection would receive positive results from a hrHPV test, regardless of histologic status. In sensitivity analyses, this was varied from 91.1% to 99.9% based on the 95% confidence interval (CI) from a published hrHPV test clinical validation study (17). Cervical cytology or DS test results distributions were evaluated by histologic status (normal, CIN 1, CIN 2, CIN 3, or ICC) and informed by the IMPACT trial results and literature (Table 2; Supplementary Table S1C; refs. 15, 16, 18).

Table 2.

Model input base case values, ranges, and data sources for screening tests performance, screening procedures performance, annual health state transition probabilities, ICC symptoms, quality of life, and costs.

InputBase caseLowHighSource
hrHPV Test Sensitivitya 95.5% 91.1% 99.9% (15,17
Cervical Cytology Test, % Results     
Normal histology 
 NILM 71.1%   (15)b 
 ASCUS+ 28.9% 8.1% 36.1% c 
 ASCUS 17.3%   (15)b 
 LSIL 9.8%   (15)b 
 HSIL 1.8%   (15)b 
CIN 1 
 NILM 40.0%   (15)b 
 ASCUS+ 60.0% 45.0% 75.1% d 
 ASCUS 26.0%   (15)b 
 LSIL 31.5%   (15)b 
 HSIL 2.5%   (15)b 
CIN 2 
 NILM 40.1%   (15)b 
 ASCUS+ 59.9% 44.9% 74.9% d 
 ASCUS 20.3%   (15)b 
 LSIL 28.0%   (15)b 
 HSIL 11.6%   (15)b 
CIN 3 
 NILM 27.2%   (15)b 
 ASCUS+ 72.8% 54.6% 91.0% d 
 ASCUS 14.0%   (15)b 
 LSIL 13.6%   (15)b 
 HSIL 45.1%   (15)b 
ICC 
 NILM 14.3%   (15)b 
 ASCUS+ 85.7% 64.3% 100.0% d 
 ASCUS 4.2%   (15,18)b 
 LSIL 6.3%   (15,18)b 
 HSIL 75.3%   (15,18)b 
DS Results, % Positive 
 No Lesions 30.9% 23.1% 38.6% (15)b,d 
 CIN 1 60.7% 45.5% 75.8% (15)b,d 
 CIN 2 82.2% 61.7% 100.0% (15)b,d 
 CIN 3 89.8% 67.3% 100.0% (15)b,d 
 ICC 100.0% 75.0% 100.0% (15)b,d 
Colposcopy and Biopsy Sensitivity 
 CIN 2 95.8% 91.4% 99.0% (19
 CIN 3 96.1% 91.4% 100.0% (19
 ICC, stage 1e 100.0% 91.4% 100.0% (19
 LEEP, % Successf 100.0% 95.5% 100.0% (20
Annual Health State Transition Probabilities 
 Efficacy of hrHPV vaccine to prevent hrHPV 16/18 infection 95.4% 89.4% 98.5% (23
 From hrHPV-negative & normal histology to    
 hrHPV 16/18 1.0% 0.5% 1.8% (22
 hrHPV 12 other 2.6% 2.0% 3.2% (22
From hrHPV 16/18 to 
 hrHPV-negative & normal histology 29.7% 22.2% 39.8% (21
 CIN 1 6.2% 3.6% 9.2% (24
 CIN 2 2.2% 0.9% 4.1% (24
 CIN 3 2.1% 0.7% 3.6% (25
From hrHPV 12 other to 
 hrHPV-negative & normal histology 48.4% 37.3% 58.2% (21
 CIN 1 3.9% 1.6% 6.9% (24
 CIN 2 1.4% 0.1% 3.4% (24
 CIN 3 0.2% 0.1% 0.3% (25
From CIN 1 to 
 hrHPV-negative & normal histology 27.6% 25.1% 31.0% (26,27
 hrHPV-positive 3.1% 2.8% 3.4% (26,27
 CIN 2 8.2% 3.3% 14.8% (27
 CIN 3 1.8% 0.7% 3.3% (27
From CIN 2 to 
 hrHPV-negative & normal histology 14.8% 5.1% 28.4% (27,28
 CIN 1 12.4% 4.3% 23.7% (27,28
 CIN 3 19.4% 2.5% 49.0% (27
From CIN 3 to 
 hrHPV-negative & normal histology 10.2% 7.6% 12.7% (29
 CIN 1 4.1% 3.1% 5.1% (29
 ICC Stage 1 2.7% 1.7% 4.5% (29
 From ICC Stage 1 to 2 27.8% 20.8% 34.7% (30
 From ICC Stage 2 to 3 29.4% 22.1% 36.8% (30
 From ICC Stage 3 to 4 40.0% 30.0% 50.0% (30
 Cancer-Specific Mortality, diagnosed ICC 
 Stage 1 1.4% 1.2% 1.5% (32
 Stage 2 5.8% 5.2% 6.4% (32
 Stage 3 11.1% 10.4% 11.8% (32
 Stage 4 25.9% 24.4% 27.4% (32
 Added risk with undiagnosed ICC 3.0% 2.3% 3.8% (33
Probability of Symptoms with ICC 
 Stage 1 7.5% 5.6% 9.4% (30
 Stage 2 11.3% 8.4% 14.1% (30
 Stage 3 30.0% 22.5% 37.5% (30
 Stage 4 45.0% 33.8% 56.3% (30
Quality of Life 
 Utility without ICC diagnosis 0.90 0.84 0.95 (34
Disutility with ICC diagnosis 
 Initial phase, by stage of diagnosisg     
 Stage 1 0.07 0.02 0.12 (34
 Stage 2 0.19 0.03 0.36 (34
 Stage 3 0.19 0.03 0.36 (34
 Stage 4 0.19 0.03 0.36 (34
 Continuing phaseg 0.08 0.05 0.11 (34
 Terminal phase, cancer deathg 0.54 0.50 0.57 (34
Annual Costs 
 Baseline medical costs     
 25-44 years of age $7,137 $5,353 $8,921 (37)d 
 45-64 years of age $12,408 $9,306 $15,510 (37)d 
 65-84 years of age $20,144 $15,108 $25,181 (37)d 
 85+ years of age $40,663 $30,498 $50,829 (37)d 
Additional annual cost of diagnosed ICC 
 Initial phase, by stage of diagnosisg     
 Stage 1 $42,696 $32,022 $53,370 (41)d 
 Stage 2 $68,689 $51,517 $85,861 (41)d 
 Stage 3 $68,689 $51,517 $85,861 (41)d 
 Stage 4 $80,902 $60,677 $101,128 (41)d 
 Continuing phase, by stage of diagnosisg    
 Stage 1 $2,401 $1,801 $3,001 (41)d 
 Stage 2 $3,549 $2,662 $4,437 (41)d 
 Stage 3 $3,549 $2,662 $4,437 (41)d 
 Stage 4 $15,241 $11,431 $19,051 (41)d 
Terminal phaseg 
 Cancer death, by stage of diagnosis    
 Stage 1 $84,347 $63,260 $105,434 (41)d 
 Stage 2 $96,561 $72,421 $120,701 (41)d 
 Stage 3 $96,561 $72,421 $120,701 (41)d 
 Stage 4 $115,456 $86,592 $144,319 (41)d 
 Non-cancer death $28,707 $21,530 $35,884 (41)d 
Procedure Costsh 
 Primary screening office visits     
 25-39 years of age $120.03 $90.02 $150.04 (39)d 
 40-64 years of age $127.71 $95.78 $159.64 (39)d 
 65+ years of age $137.13 $102.85 $171.41 (39)d 
 Reflex screening office visit $92.47 $69.35 $115.59 (39)d 
 hrHPV test, pooled $35.09 $26.32 $43.86 (40)d 
 hrHPV test, with genotyping $35.09 $26.32 $43.86 (40)d 
 Cervical cytology test $26.61 $19.96 $33.26 (40)d 
 DS test $178.30 $133.73 $222.88 (39)d 
 Colposcopy and biopsiesi $387.31 $201.07 $502.02 (19,39)d 
 LEEPj $2,115.07 $1,586.30 $2,643.84 (38,39)d 
Annual Time Preference Discount Rate 
 Costs 3.0% 0.0% 5.0% (35
 QALYs 0.0% 0.0% 5.0% (35
InputBase caseLowHighSource
hrHPV Test Sensitivitya 95.5% 91.1% 99.9% (15,17
Cervical Cytology Test, % Results     
Normal histology 
 NILM 71.1%   (15)b 
 ASCUS+ 28.9% 8.1% 36.1% c 
 ASCUS 17.3%   (15)b 
 LSIL 9.8%   (15)b 
 HSIL 1.8%   (15)b 
CIN 1 
 NILM 40.0%   (15)b 
 ASCUS+ 60.0% 45.0% 75.1% d 
 ASCUS 26.0%   (15)b 
 LSIL 31.5%   (15)b 
 HSIL 2.5%   (15)b 
CIN 2 
 NILM 40.1%   (15)b 
 ASCUS+ 59.9% 44.9% 74.9% d 
 ASCUS 20.3%   (15)b 
 LSIL 28.0%   (15)b 
 HSIL 11.6%   (15)b 
CIN 3 
 NILM 27.2%   (15)b 
 ASCUS+ 72.8% 54.6% 91.0% d 
 ASCUS 14.0%   (15)b 
 LSIL 13.6%   (15)b 
 HSIL 45.1%   (15)b 
ICC 
 NILM 14.3%   (15)b 
 ASCUS+ 85.7% 64.3% 100.0% d 
 ASCUS 4.2%   (15,18)b 
 LSIL 6.3%   (15,18)b 
 HSIL 75.3%   (15,18)b 
DS Results, % Positive 
 No Lesions 30.9% 23.1% 38.6% (15)b,d 
 CIN 1 60.7% 45.5% 75.8% (15)b,d 
 CIN 2 82.2% 61.7% 100.0% (15)b,d 
 CIN 3 89.8% 67.3% 100.0% (15)b,d 
 ICC 100.0% 75.0% 100.0% (15)b,d 
Colposcopy and Biopsy Sensitivity 
 CIN 2 95.8% 91.4% 99.0% (19
 CIN 3 96.1% 91.4% 100.0% (19
 ICC, stage 1e 100.0% 91.4% 100.0% (19
 LEEP, % Successf 100.0% 95.5% 100.0% (20
Annual Health State Transition Probabilities 
 Efficacy of hrHPV vaccine to prevent hrHPV 16/18 infection 95.4% 89.4% 98.5% (23
 From hrHPV-negative & normal histology to    
 hrHPV 16/18 1.0% 0.5% 1.8% (22
 hrHPV 12 other 2.6% 2.0% 3.2% (22
From hrHPV 16/18 to 
 hrHPV-negative & normal histology 29.7% 22.2% 39.8% (21
 CIN 1 6.2% 3.6% 9.2% (24
 CIN 2 2.2% 0.9% 4.1% (24
 CIN 3 2.1% 0.7% 3.6% (25
From hrHPV 12 other to 
 hrHPV-negative & normal histology 48.4% 37.3% 58.2% (21
 CIN 1 3.9% 1.6% 6.9% (24
 CIN 2 1.4% 0.1% 3.4% (24
 CIN 3 0.2% 0.1% 0.3% (25
From CIN 1 to 
 hrHPV-negative & normal histology 27.6% 25.1% 31.0% (26,27
 hrHPV-positive 3.1% 2.8% 3.4% (26,27
 CIN 2 8.2% 3.3% 14.8% (27
 CIN 3 1.8% 0.7% 3.3% (27
From CIN 2 to 
 hrHPV-negative & normal histology 14.8% 5.1% 28.4% (27,28
 CIN 1 12.4% 4.3% 23.7% (27,28
 CIN 3 19.4% 2.5% 49.0% (27
From CIN 3 to 
 hrHPV-negative & normal histology 10.2% 7.6% 12.7% (29
 CIN 1 4.1% 3.1% 5.1% (29
 ICC Stage 1 2.7% 1.7% 4.5% (29
 From ICC Stage 1 to 2 27.8% 20.8% 34.7% (30
 From ICC Stage 2 to 3 29.4% 22.1% 36.8% (30
 From ICC Stage 3 to 4 40.0% 30.0% 50.0% (30
 Cancer-Specific Mortality, diagnosed ICC 
 Stage 1 1.4% 1.2% 1.5% (32
 Stage 2 5.8% 5.2% 6.4% (32
 Stage 3 11.1% 10.4% 11.8% (32
 Stage 4 25.9% 24.4% 27.4% (32
 Added risk with undiagnosed ICC 3.0% 2.3% 3.8% (33
Probability of Symptoms with ICC 
 Stage 1 7.5% 5.6% 9.4% (30
 Stage 2 11.3% 8.4% 14.1% (30
 Stage 3 30.0% 22.5% 37.5% (30
 Stage 4 45.0% 33.8% 56.3% (30
Quality of Life 
 Utility without ICC diagnosis 0.90 0.84 0.95 (34
Disutility with ICC diagnosis 
 Initial phase, by stage of diagnosisg     
 Stage 1 0.07 0.02 0.12 (34
 Stage 2 0.19 0.03 0.36 (34
 Stage 3 0.19 0.03 0.36 (34
 Stage 4 0.19 0.03 0.36 (34
 Continuing phaseg 0.08 0.05 0.11 (34
 Terminal phase, cancer deathg 0.54 0.50 0.57 (34
Annual Costs 
 Baseline medical costs     
 25-44 years of age $7,137 $5,353 $8,921 (37)d 
 45-64 years of age $12,408 $9,306 $15,510 (37)d 
 65-84 years of age $20,144 $15,108 $25,181 (37)d 
 85+ years of age $40,663 $30,498 $50,829 (37)d 
Additional annual cost of diagnosed ICC 
 Initial phase, by stage of diagnosisg     
 Stage 1 $42,696 $32,022 $53,370 (41)d 
 Stage 2 $68,689 $51,517 $85,861 (41)d 
 Stage 3 $68,689 $51,517 $85,861 (41)d 
 Stage 4 $80,902 $60,677 $101,128 (41)d 
 Continuing phase, by stage of diagnosisg    
 Stage 1 $2,401 $1,801 $3,001 (41)d 
 Stage 2 $3,549 $2,662 $4,437 (41)d 
 Stage 3 $3,549 $2,662 $4,437 (41)d 
 Stage 4 $15,241 $11,431 $19,051 (41)d 
Terminal phaseg 
 Cancer death, by stage of diagnosis    
 Stage 1 $84,347 $63,260 $105,434 (41)d 
 Stage 2 $96,561 $72,421 $120,701 (41)d 
 Stage 3 $96,561 $72,421 $120,701 (41)d 
 Stage 4 $115,456 $86,592 $144,319 (41)d 
 Non-cancer death $28,707 $21,530 $35,884 (41)d 
Procedure Costsh 
 Primary screening office visits     
 25-39 years of age $120.03 $90.02 $150.04 (39)d 
 40-64 years of age $127.71 $95.78 $159.64 (39)d 
 65+ years of age $137.13 $102.85 $171.41 (39)d 
 Reflex screening office visit $92.47 $69.35 $115.59 (39)d 
 hrHPV test, pooled $35.09 $26.32 $43.86 (40)d 
 hrHPV test, with genotyping $35.09 $26.32 $43.86 (40)d 
 Cervical cytology test $26.61 $19.96 $33.26 (40)d 
 DS test $178.30 $133.73 $222.88 (39)d 
 Colposcopy and biopsiesi $387.31 $201.07 $502.02 (19,39)d 
 LEEPj $2,115.07 $1,586.30 $2,643.84 (38,39)d 
Annual Time Preference Discount Rate 
 Costs 3.0% 0.0% 5.0% (35
 QALYs 0.0% 0.0% 5.0% (35

Abbreviations: ASCUS+, ASCUS or LSIL or HSIL; LSIL, low-grade squamous intraepithelial lesion.

aFor the presence of hrHPV.

bIMPACT trial. For more detail, see Supplementary Table S1 C.

cRange based on Cuzick et al. for the low value (16), and +25% for the high value.

dRange established as +/- 25% of the base case value.

eAssumed to be 100% at base case. Colposcopy and biopsy sensitivity for ICC stage 2+ is assumed to be 100%.

fRemoving CIN 2 or CIN 3 lesions.

gTerminal phase is the year prior to death, initial phase the year after diagnosis (not including the terminal phase), and continuing phase is time in between (if any).

hProcedure codes, descriptions, and costs are detailed in Supplementary Table S3.

iIncludes three biopsies and pathology at base case, range 1-4 biopsies, per the Biopsy Study (19).

jIncludes pathology for one specimen.

The mean number of biopsies per colposcopy and the sensitivity of colposcopy and biopsy for CIN 2 and CIN 3 were extracted from the Biopsy Study (19), while this sensitivity was assumed to be 100% for ICC (varied to 95% for stage 1 ICC in sensitivity analyses; ref. 20). Patients found to have CIN 2 or CIN 3 by colposcopy and biopsy were directed to LEEP, which was assumed at base case to completely remove the CIN, returning the patient to a health state defined by their underlying hrHPV status. Patients found to have ICC were assumed to be diagnosed at the current stage, treated and exited population screening.

Annual health state transitions

The annual probabilities of contracting or clearing hrHPV (16/18 or 12 other) have been previously published (Table 2; refs. 21, 22). Vaccinated patients had a lower probability of entering the model with hrHPV 16/18 infection. Furthermore, a relative reduction of 95.4% was applied to the annual risk of contracting hrHPV 16/18 for vaccinated patients based on efficacy from the PATRICIA randomized trial (23).

We used the following data sources for our state transition probabilities. The probabilities of patients with hrHPV progressing to CIN 1–3 were assessed in observational and retrospective studies (24, 25). Natural history of patients with CIN 1 or CIN 2 were evaluated in registries, meta-analyses, and analyses for the U.S. Preventive Services Task Force (26–28). Natural history data for untreated CIN 3 were from a retrospective review of data collected in a prospective study (29). Stage progression among patients with undiagnosed ICC, and the probabilities of ICC symptoms among patients with ICC by stage were informed by a published U.S. mathematical model (30). Patients with pathognomonic symptoms were assumed to be diagnosed with ICC.

All-cause mortality data by age was from the Centers of Disease Control and Prevention (31). Annual cancer-specific mortality among patients diagnosed with ICC by stage at diagnosis was from the Surveillance, Epidemiology, and End Results (SEER) database (Table 2; ref. 32). Undiagnosed ICC had an additional 3% annual mortality risk based on an analysis of treated and untreated patients with cervical cancer in the Alberta Cancer Board Alberta Registry Database (33). Patients who died due to cancer without having their ICC otherwise found were considered to have been diagnosed at stage IV in their last year of life.

Quality of life

Health utilities measure quality of life on a scale of zero to one, with zero representing death and one representing perfect health. All cytology results were collapsed into one baseline utility of 0.90 for the screening population without a cancer diagnosis (Table 2; ref. 34). The impact of diagnosed ICC was accounted for with reductions, or disutilities, by stage at diagnosis and initial, continuing, or terminal phase. The terminal phase was the last year of life. The initial phase was the first year following diagnosis, excluding the terminal phase. The continuing phase encompassed the time in between. The terminal phase disutility was applied only if the patient died of cancer-related mortality. Otherwise, the continuing phase disutility was applied in the last year of life. Undiagnosed ICC did not incur disutilities. At base case, no time preference discounting was applied to quality-adjusted life-years (QALY; ref. 35).

Costs

This analysis considered direct medical costs from the perspective of a third-party U.S. payer, including baseline medical costs, screening test and procedures costs, and ICC costs from model entry until death (Table 2). Costs were reported in 2021 United States Dollars (36). Annual baseline healthcare costs for women by age group were reported by the Centers for Medicare and Medicaid Services’ (CMS) National Health Expenditure Data (37). Costs of screening visits, tests, procedures, and pathology were informed by the 2021 CMS Physician Fee and Clinical Laboratory Fee Schedules, and the 2020 Physician/Supplier Procedure Summary File (Supplementary Table S3; refs. 38–40). Annual additional costs associated with diagnosed ICC by stage at diagnosis and phase were from a published analysis of the SEER-Medicare linked database (41). At base case, an annual 3%-time preference discount rate was applied to costs (35).

Willingness-to-pay

Willingness-to-pay thresholds of $50,000 to $150,000 per QALY or life-year are often discussed for cost-effectiveness in the United States, and are used by the Institute for Clinical and Economic Review, a U.S.-based health technology assessment group (42). Their 2020–2023 Value Assessment Framework judged that the most recent research supported a $100,000 operational cost-effectiveness threshold based on both opportunity cost and willingness-to-pay paradigms, and they continued to evaluate $50,000 as a low end and $150,000 as a flexible upper end. Incremental cost-effectiveness ratios (ICER) below these thresholds may be considered cost-effective.

Additional analysis

Bootstrap 95% CIs were calculated to assess the impact of patient variability using 1,000 bootstrap samples of 12,250 microsimulated patients. In parameter sensitivity analyses, related input parameters were grouped and varied together. One-way sensitivity analysis was used to evaluate the most influential model parameters on costs, QALYs gained, and life-years gained by varying inputs independently across their individual ranges. A probabilistic sensitivity analysis assessed the impact of both patient and parameter variability on the ICER using 1,000 Monte Carlo simulations with 2,450 new patient microsimulations each (50 per patient group). For each Monte Carlo simulation, a new set of model parameter values were randomly drawn from their individual distributions. Beta distributions were applied to percentages and risks, while gamma distributions were applied to costs, similar to previously published cost-effectiveness analyses, and in accordance with Good Research Practices guidelines from the International Society for Pharmacoeconomics and Outcomes Research and Society of Medical Decision Making Task Force (33, 43). Probabilities of cost-effectiveness were evaluated at different willingness-to-pay thresholds.

Data availability

Supporting information for the inputs and results of this study are available in the Supplementary Materials of this article. Further data that support the findings of this study are available from the corresponding author upon reasonable request.

Base case

Base case analyses showed that primary hrHPV genotyping and cervical cytology co-testing with reflex to DS was expected to improve QALY and life-year outcomes and increase lifetime cost compared with hrHPV and cervical cytology co-testing alone (Table 3). In the comparison with SOC 1 (hrHPV pooled testing and cervical cytology co-testing with reflex to hrHPV genotyping), switching to the intervention algorithm (primary hrHPV genotyping and cytology co-testing with reflex DS) was expected to increase costs by $49 (95% CI, $31–$64), QALYs by 0.0025 years (95% CI, 0.0007–0.0044), and life-years by 0.0022 years (95% CI, 0.0003–0.0043) per patient on average. The ICERs were $19,892 per QALY (95% CI, $10,426–$61,987), and $21,891 per life-year (95% CI, $10,737–$85,502; Fig. 3).

Table 3.

Change in mean costs, change in survival, and ICERs for adding DS triage to primary hrHPV and cervical cytology co-testing for cervical cancer screening.

Intervention: hrHPV genotyping and cervical cytology co-testing with reflex to DS
vs. SOC 1vs. SOC 2
hrHPV pooled and cervical cytology co-testing with reflex to hrHPv genotypinghrHPV genotyping and cervical cytology co-testing with no reflex
Change per average patient with intervention instead of SOCMean95% CIMean95% CI
Costs $49 ($31–$64) $62 ($50–$74) 
Baseline $15 ($3–$28) $16 ($5–$29) 
Screening visits $1 ($0–$1) $1 ($0–$1) 
Screening tests $36 ($33–$39) $45 ($42–$48) 
Colposcopiesa $7 ($3–$12) $6 ($2–$10) 
LEEP treatmentsa $7 ($2–$12) $6 ($4–$9) 
ICC −$17 (−$31 to −$8) −$13 (−$19 to −$7) 
Survival 
Risk of ICC death −0.007% (−0.012% to −0.002%) −0.007% (−0.012% to −0.003%) 
Life-years 0.0022 (0.0003–0.0043) 0.0024 (0.0007–0.0045) 
QALYs 0.0025 (0.0007–0.0044) 0.0024 (0.0009–0.0042) 
Cost per 
Life-year $21,891 ($10,737–$85,502) $25,354 ($15,563–$70,190) 
QALY $19,892 ($10,426–$61,987) $25,348 ($16,095–$58,310) 
Intervention: hrHPV genotyping and cervical cytology co-testing with reflex to DS
vs. SOC 1vs. SOC 2
hrHPV pooled and cervical cytology co-testing with reflex to hrHPv genotypinghrHPV genotyping and cervical cytology co-testing with no reflex
Change per average patient with intervention instead of SOCMean95% CIMean95% CI
Costs $49 ($31–$64) $62 ($50–$74) 
Baseline $15 ($3–$28) $16 ($5–$29) 
Screening visits $1 ($0–$1) $1 ($0–$1) 
Screening tests $36 ($33–$39) $45 ($42–$48) 
Colposcopiesa $7 ($3–$12) $6 ($2–$10) 
LEEP treatmentsa $7 ($2–$12) $6 ($4–$9) 
ICC −$17 (−$31 to −$8) −$13 (−$19 to −$7) 
Survival 
Risk of ICC death −0.007% (−0.012% to −0.002%) −0.007% (−0.012% to −0.003%) 
Life-years 0.0022 (0.0003–0.0043) 0.0024 (0.0007–0.0045) 
QALYs 0.0025 (0.0007–0.0044) 0.0024 (0.0009–0.0042) 
Cost per 
Life-year $21,891 ($10,737–$85,502) $25,354 ($15,563–$70,190) 
QALY $19,892 ($10,426–$61,987) $25,348 ($16,095–$58,310) 

aIncludes biopsy and pathology costs.

Figure 3.

Cost-effectiveness of adding DS triage to primary hrHPV and cervical cytology co-testing for cervical cancer screening. ICERs with the addition of DS triage were below $30,000 per QALY or life-year gained at base case, and 95% upper CIs were below $90,000 per QALY or life-year gained.

Figure 3.

Cost-effectiveness of adding DS triage to primary hrHPV and cervical cytology co-testing for cervical cancer screening. ICERs with the addition of DS triage were below $30,000 per QALY or life-year gained at base case, and 95% upper CIs were below $90,000 per QALY or life-year gained.

Close modal

In the SOC 2 comparison (hrHPV genotyping and cervical cytology co-testing with no reflex), adding reflex DS was expected to increase costs by $62 (95% CI, $50–$74), QALYs by 0.0024 (95% CI, 0.0009–0.0042), and life-years by 0.0024 years (95% CI, 0.0007–0.0045) per patient on average. The ICERs were $25,348 per QALY (95% CI, $16,095–$58,310), and $25,354 per life-year (95% CI, $15,563–$70,190).

One-way sensitivity

Change in cost

In one-way sensitivity analyses of DS reflex testing following primary co-testing in comparison with SOC 1 or SOC 2, two parameters were among the top five most influential on the mean change in costs in both comparisons: the time preference discount rate for costs, and the probability of clearing the hrHPV 12 other+ infection given normal histology (Supplementary Table S4, Supplementary Fig. S1A). In the SOC 1 comparison, costs of the hrHPV pooled and genotyping tests and the risks of progression from hrHPV 12 other infection to CIN 1–3 within 1 year were also influential. In the SOC 2 comparison, the probability of atypical squamous cells of undetermined significance (ASCUS)+ among women with any grade CIN or ICC, cost of the DS test, and probability of ASCUS+ among women with normal histology were also influential.

Change in effectiveness

The probability of ASCUS+ among women with any grade CIN or ICC, and probability of positive DS test result among women with any grade CIN or ICC were among the top five most influential variables on the mean change in QALYs or life-years per patient in both comparisons with SOC 1 or SOC 2 (Supplementary Table S4, Supplementary Fig. S1B and C). In the comparison with SOC 1, the risk of progression from CIN 1 to CIN 2 or CIN 3 was the third-most influential variable on change in either outcome. The time preference discount rate and risk of progression from hrHPV to any grade CIN were also influential on the change in QALYs, while the risk of progression from CIN 2 to CIN 3, or CIN 3 to ICC stage 1 were also influential on the change in life-years. In the comparison with SOC 2, the risk of progression from CIN 2 to CIN 3 was the second most influential input on change in either outcome, risk of progression from CIN 3 to ICC stage 1 was fourth, and probability of CIN 2 regression to CIN 1 or hrHPV-negative with normal histology was fifth.

Probabilistic sensitivity

In probabilistic sensitivity analyses, 60.8% to 80.8% showed increased QALYs with ICERs below $50,000 to $150,000 per QALY for DS compared with SOC 1 (Supplementary Fig. S2). On a per-life-year basis, this was 67.3% to 80.3%, respectively. In the SOC 2 comparison, 56.4% to 79.4% and 62.9% to 78.1% of simulations’ ICERs were below $50,000 to $150,000 per QALY and per life-year, respectively.

Adding DS reflex screening to hrHPV genotyping and cervical cytology co-testing for cervical cancer screening was expected to be cost-effective in the United States compared with commonly used SOC co-testing algorithms, with mean ICERs less than even the lowest commonly discussed willingness-to-pay threshold (<$50,000 per QALY or life-year), and ICER 95% CIs below the operational willingness-to-pay threshold in the United States (<$100,000 per QALY or life-year). Compared with primary pooled hrHPV and cervical cytology co-testing with reflex hrHPV genotyping (SOC 1), switching hrHPV genotyping to primary testing and adding DS reflex testing had an ICER of $19,892 per QALY gained or $21,891 per life-year gained (Table 3). Adding DS reflex testing to primary hrHPV genotyping and cervical cytology co-testing (SOC 2) resulted in ICERs of $25,348 per QALY gained and $25,354 per life-year gained.

Both comparisons projected small increases in QALYs, life-years, and costs with hrHPV genotyping and cervical cytology co-testing with DS reflex. The increases in costs were largely due to increased costs of all screening tests and of longer survival, while ICC-related costs and ICC-related death decreased. CIs across these findings did not cross zero, supporting the reliability of the mean base case results despite variability across patient microsimulations.

Willingness to pay thresholds often discussed for cost-effectiveness in the United States range from $50,000 to $150,000 per QALY, and the Institute for Clinical and Economic Review in the United States uses an operational threshold of $100,000 per QALY (42). ICERs below these thresholds may be considered cost-effective. At base case, the ICERs per QALY and per life-year were below the lowest $50,000 threshold, and their CIs did not cross the $100,000 operational threshold, supporting the cost-effectiveness of adding DS reflex to existing co-testing algorithms with hrHPV genotyping and cervical cytology.

The probabilistic sensitivity analysis simulations demonstrated that DS was likely to be cost-effective at typical thresholds even when considering both patient microsimulation variability and parameter variability (Supplementary Fig. S2). In both comparisons, the majority of simulations showed cost-effective results at the $50,000 threshold per QALY or life-year, increasing to approximately three-fourths of simulations at the $100,000 threshold. Despite patient-level and parameter-level uncertainty, a majority of simulations were cost-effective at typical willingness-to-pay thresholds in the United States, further supporting the cost-effectiveness of adding DS reflex testing to common hrHPV genotyping and cervical cytology co-testing algorithms.

This model analysis relied on currently available data. The proportion of women of screening-age who were vaccinated, hrHPV prevalence, and the risk of contracting hrHPV were based on the current population. As vaccination rates against hrHPV 16/18 or 12 other strains increase, the prevalence and risk of contracting hrHPV would be expected to decrease, but this model focused on the current population based on currently available data.

Quality of life decrements were modeled only for patients diagnosed with ICC, and were not applied on the basis of test results, use of colposcopy, or LEEP. Utilities data were not available for cervical cancer screening patients by positive or negative DS result, and other studies of quality of life with other screening methods have been unclear as to whether there were meaningful differences based on screening results without ICC diagnosis. One study asking women 21 to 65 years of age to evaluate hypothetical health states defined by hrHPV and/or cervical cytology test results, colposcopy findings, and need for LEEP excision via time trade-off methods suggested differences in health state utilities (44). However, a different study using validated surveys with women who received actual cervical cancer screening results and procedures found that women directed to colposcopy or triage without colposcopy (similar to management options based on DS findings) had similar utilities (34). Additional research to assess the utility of health states by DS test result in combination with other screening characteristics would be helpful to determine whether there are quality of life implications of cervical cancer screening test results and procedures for women without ICC diagnosis. Because DS testing happens at the pathology lab level on specimens already taken, the DS test procedure itself is unlikely to affect the woman's quality of life.

Cervical cytology results among patients with normal histology were from the IMPACT trial, in which 71% of such patients had NILM results. Other studies have suggested this may be as high as 92%, showing variability in the specificity of cervical cytology to differentiate patients with normal histology (16). Sensitivity analyses tested a wide range for the proportion of NILM versus ASCUS+ results among patients with normal histology (range: 8.1%–36.1% ASCUS+), but this was not a highly influential variable on results in any SOC comparison except for the change in costs in the comparison with SOC 2 where it was the fifth most influential input on costs but none of the changes in costs, QALYs, or life-years crossed zero.

The health states in our analysis differentiated patients with hrHPV and normal histology by genotype (16/18+ vs. 12 other+), but health states for any grade CIN or ICC were not differentiated by hrHPV genotype, and risks were applied based on CIN 1–3 or ICC stage 1–4 status alone. Where available, we used transition probabilities data specific to women positive for hrHPV. A network meta-analysis of the natural history of CIN 1 and CIN 2 found that studies specific to women who were positive for hrHPV were few and small, and reported substantial intra-study heterogeneity (27). Furthermore, the natural history of CIN 3 is estimable only by inference to a historical study. Even so, while the one-way sensitivity analysis identified the risks of progression from CIN 1, CIN 2, and CIN 3 as influential parameters on the change in outcomes in both SOC comparisons, and further research on the natural history and risks of CIN would be beneficial, the large majority of probabilistic sensitivity analysis simulations remained cost-effective at the operational willingness-to-pay threshold in the United States, even as transition probabilities were all varied simultaneously across reported 95% CIs and ranges.

In 2019, the American Society for Colposcopy and Cervical Pathology (ASCCP) released risk-based guidelines for management following cervical cancer screening tests based on the immediate and 5-year risks of CIN 3, adenocarcinoma in situ, or any cancer (collectively CIN 3+) inferred by results from historical and current screening test results (6). This risk-based approach accounts for both test sensitivity and test specificity by the level of risk from 0% to 100% certainty of CIN 3+ disease associated with the test result. As cervical cancer screening tools and testing in clinical practice continues to advance, these enduring guidelines can and are intended to be continually updated with new testing strategies evaluated against CIN 3+ risk thresholds. For instance, the probability of CIN 3+ based on persistent hrHPV result with or without mildly abnormal cytology used as primary co-testing results can be further informed by the DS test result and thus moved above or below the threshold for management by colposcopy/treatment or shortened surveillance. While the current analysis does not apply ASCCP risk-based management, future analyses to consider the cost-effectiveness of the new DS test in the context of these risk-based management guidelines would be welcomed.

Since its advent, cervical cancer screening has substantially decreased cervical cancer rates among women in the United States, and improvements in screening methods have continued to be developed through management guidelines and advances in testing tools (6, 15, 45). The co-testing algorithms evaluated as SOC 1 and SOC 2 in this analysis are commonly used in the United States, and this analysis aimed to inform current clinical practice by comparing the addition of reflex DS testing to these contemporary cervical cancer screening approaches.

Conclusion

The addition of the DS test to common current co-testing algorithms including hrHPV genotyping and cervical cytology for cervical cancer screening is expected to be cost-effective. Lifetime cost of screening tests were projected to increase, but risk of ICC death and ICC-related costs were projected to decrease, with increases in QALYs and life-years per patient on average.

R.J. Anderson reports personal fees from F. Hoffmann-La Roche AG during the conduct of the study. E. Baker reports Roche Employee. T.M. Yu reports grants from Roche Molecular Systems during the conduct of the study; and Employment by Guidehouse Inc. D.M. Harper's institution received a grant from Roche Diagnostics Solutions to provide content expertise to the project. No disclosures were reported by the other author.

D.M. Harper: Conceptualization, data accuracy, supervision, methodology, writing–review and editing. R.J. Anderson: Conceptualization, formal analysis, visualization, methodology, writing–original draft, writing–review and editing. E. Baker: Conceptualization, methodology, writing–review and editing. T.M. Yu: Conceptualization, data curation, formal analysis, supervision, visualization, methodology, writing–original draft, project administration, writing–review and editing.

The authors would like to thank Dr. Ruediger Ridder, Teressa Egeler, Lauren Perlaza, Mark Torowus, Dr. Carolyn Kay, Dr. Charles Cachoeira, and Dr. James Karichu for their advice and feedback that they provided to this analysis. Financial support for this research was provided by Roche Diagnostics Solutions to the University of Michigan so that Dr. D.M. Harper could provide content expertise to Guidehouse Inc., which employed R. Anderson and T.M. Yu at the time of this research.

The publication costs of this article were defrayed in part by the payment of publication fees. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.

Note: Supplementary data for this article are available at Cancer Prevention Research Online (http://cancerprevres.aacrjournals.org/).

1.
Kim
JJ
,
Burger
EA
,
Regan
C
,
Sy
S
.
Screening for cervical cancer in primary care: a decision analysis for the US preventive services task force
.
JAMA
2018
;
320
:
706
14
.
2.
U.S. Preventive Services Task Force
,
Curry
SJ
,
Krist
AH
,
Owens
DK
,
Barry
MJ
,
Caughey
AB
, et al
.
Screening for cervical cancer: US Preventive Services Task Force recommendation statement
.
JAMA
2018
;
320
:
674
86
.
3.
Cuzick
J
,
Du
R
,
Adcock
R
,
Kinney
W
,
Joste
N
,
McDonald
RM
, et al
.
Uptake of co-testing with HPV and cytology for cervical screening: a population-based evaluation in the United States
.
Gynecol Oncol
2021
;
162
:
555
9
.
4.
Silver
MI
,
Rositch
AF
,
Phelan-Emrick
DF
,
Gravitt
PE
.
Uptake of HPV testing and extended cervical cancer screening intervals following cytology alone and Pap/HPV co-testing in women aged 30–65 years
.
Cancer Causes Control
2018
;
29
:
43
50
.
5.
US Food & Drug Administration
.
2020 Dec 29
.
Premarket Approval (PMA): CINtec PLUS Cytology
.
Available from:
<https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfpma/pma.cfm?id=P190024>.
6.
Egemen
D
,
Cheung
LC
,
Chen
X
,
Demarco
M
,
Perkins
RB
,
Kinney
W
, et al
.
Risk estimates supporting the 2019 ASCCP risk-based management consensus guidelines
.
J Low Genit Tract Dis
2020
;
24
:
132
43
.
7.
Fontham
ETH
,
Wolf
AMD
,
Church
TR
,
Etzioni
R
,
Flowers
CR
,
Herzig
A
, et al
.
Cervical cancer screening for individuals at average risk: 2020 guideline update from the American Cancer Society
.
CA Cancer J Clin
2020
;
70
:
321
46
.
8.
Perkins
RB
,
Guido
RS
,
Castle
PE
,
Chelmow
D
,
Einstein
MH
,
Garcia
F
, et al
. 2019
ASCCP risk-based management consensus guidelines for abnormal cervical cancer screening tests and cancer precursors
.
J Low Genit Tract Dis
2020
;
24
:
102
31
.
9.
World Health Organization
.
2021 Jun 08
.
New WHO recommendations on screening and treatment to prevent cervical cancer among women living with HIV: Policy Brief
. Available from: <https://www.who.int/publications/i/item/9789240030961>.
10.
Agency for Healthcare Research and Quality
.
2016 Oct 29
.
Trends in Hysterectomies and Oophorectomies in Hospital Inpatient and Ambulatory Settings, 2005–2013
.
Available from:
<https://www.hcup-us.ahrq.gov/reports/statbriefs/sb214-Hysterectomy-Oophorectomy-Trends.pdf>.
11.
Centers for Disease Control and Prevention
.
2017 Jan 24
.
Cancer screening test use - United States, 2015. In Morbidity and Mortality Weekly Report
.
Vol
.
66
/
No. 8
. Available from: <https://www.cdc.gov/mmwr/volumes/66/wr/pdfs/mm6608a1.pdf>.
12.
Centers for Disease Control and Prevention
.
2019 Oct 29
.
National Vital Statistics Report - Births: Final Data for 2018
.
Available from:
<https://www.cdc.gov/nchs/data/nvsr/nvsr68/nvsr68_13-508.pdf>.
Accessed 2021 Oct 29
.
13.
Centers for Disease Control and Prevention
.
2019 Oct 29
.
Diagnoses of HIV Infection in the United States and Dependent Areas, 2018
. Available from: <https://www.cdc.gov/hiv/pdf/library/reports/surveillance/cdc-hiv-surveillance-report-2018-preliminary-vol-30.pdf>.
14.
Safaeian
M
,
Wright
TC
Jr.
,
Stoler
MH
,
Ranger-Moore
J
,
Rehm
S
,
Aslam
S
, et al
.
The IMproving Primary Screening And Colposcopy Triage trial: human papillomavirus, cervical cytology, and histopathologic results from the baseline and 1-year follow-up phase
.
Am J Obstet Gynecol
2021
;
225
:
278
.
15.
Wright
TC
Jr.
,
Stoler
MH
,
Ranger-Moore
J
,
Fang
Q
,
Volkir
P
,
Safaeian
M
, et al
.
Clinical validation of p16/Ki-67 dual-stained cytology triage of HPV-positive women: results from the IMPACT trial
.
Int J Cancer
2022
;
150
:
461
71
.
16.
Cuzick
J
,
Myers
O
,
Hunt
WC
,
Robertson
M
,
Joste
NE
,
Castle
PE
, et al
.
A population-based evaluation of cervical screening in the United States: 2008–2011
.
Cancer Epidemiol Biomarkers Prev
2014
;
23
:
765
73
.
17.
Saville
M
,
Sultana
F
,
Malloy
MJ
,
Velentzis
LS
,
Caruana
M
,
Ip
ELO
, et al
.
Clinical validation of the cobas HPV test on the cobas 6800 system for the purpose of cervical screening
.
J Clin Microbiol
2019
;
57
:
e01239
18
.
18.
Subramaniam
A
,
Fauci
JM
,
Schneider
KE
,
Whitworth
JM
,
Erickson
BK
,
Kim
K
, et al
.
Invasive cervical cancer and screening: what are the rates of unscreened and underscreened women in the modern era?
J Low Genit Tract Dis
2011
;
15
:
110
3
.
19.
Wentzensen
N
,
Walker
JL
,
Gold
MA
,
Smith
KM
,
Zuna
RE
,
Mathews
C
, et al
.
Multiple biopsies and detection of cervical cancer precursors at colposcopy
.
J Clin Oncol
2015
;
33
:
83
9
.
20.
Del Mistro
A
,
Matteucci
M
,
Insacco
EA
,
Onnis
G
,
Da Re
F
,
Baboci
L
, et al
.
Long-term clinical outcome after treatment for high-grade cervical lesions: a retrospective monoinstitutional cohort study
.
Biomed Res Int
2015
;
2015
:
984528
.
21.
Bulkmans
NW
,
Berkhof
J
,
Bulk
S
,
Bleeker
MC
,
van Kemenade
FJ
,
Rozendaal
L
, et al
.
High-risk HPV type-specific clearance rates in cervical screening
.
Br J Cancer
2007
;
96
:
1419
24
.
22.
Ferris
DG
,
Brown
DR
,
Giuliano
AR
,
Myers
E
,
Joura
EA
,
Garland
SM
, et al
.
Prevalence, incidence, and natural history of HPV infection in adult women ages 24 to 45 participating in a vaccine trial
.
Papillomavirus Res
2020
;
10
:
100202
.
23.
Apter
D
,
Wheeler
CM
,
Paavonen
J
,
Castellsague
X
,
Garland
SM
,
Skinner
SR
, et al
.
Efficacy of human papillomavirus 16 and 18 (HPV-16/18) AS04-adjuvanted vaccine against cervical infection and precancer in young women: final event-driven analysis of the randomized, double-blind PATRICIA trial
.
Clin Vaccine Immunol
2015
;
22
:
361
73
.
24.
Insinga
RP
,
Perez
G
,
Wheeler
CM
,
Koutsky
LA
,
Garland
SM
,
Leodolter
S
, et al
.
Incident cervical HPV infections in young women: transition probabilities for CIN and infection clearance
.
Cancer Epidemiol Biomarkers Prev
2011
;
20
:
287
96
.
25.
Khan
MJ
,
Castle
PE
,
Lorincz
AT
,
Wacholder
S
,
Sherman
M
,
Scott
DR
, et al
.
The elevated 10-year risk of cervical precancer and cancer in women with human papillomavirus (HPV) type 16 or 18 and the possible utility of type-specific HPV testing in clinical practice
.
J Natl Cancer Inst
2005
;
97
:
1072
9
.
26.
Kulasingam
SL
,
Havrilesky
L
,
Ghebre
R
,
Myers
ER
.
August 12. Screening for Cervical Cancer: A Decision Analysis for the U.S. Preventive Services Task Force. Rockville (MD): Agency for Healthcare Research and Quality (US); 2011 May. (Evidence Syntheses, No. 86s. Appendix B Table 1)
.<https://www.ncbi.nlm.nih.gov/books/NBK92546/>.
27.
Loopik
DL
,
Bentley
HA
,
Eijgenraam
MN
,
IntHout
J
,
Bekkers
RLM
,
Bentley
JR
.
The natural history of cervical intraepithelial neoplasia grades 1, 2, and 3: a systematic review and meta-analysis
.
J Low Genit Tract Dis
2021
;
25
:
221
31
.
28.
Skorstengaard
M
,
Lynge
E
,
Suhr
J
,
Napolitano
G
.
Conservative management of women with cervical intraepithelial neoplasia grade 2 in Denmark: a cohort study
.
BJOG
2020
;
127
:
729
36
.
29.
McCredie
MR
,
Sharples
KJ
,
Paul
C
,
Baranyai
J
,
Medley
G
,
Jones
RW
, et al
.
Natural history of cervical neoplasia and risk of invasive cancer in women with cervical intraepithelial neoplasia 3: a retrospective cohort study
.
Lancet Oncol
2008
;
9
:
425
34
.
30.
Goldie
SJ
,
Kohli
M
,
Grima
D
,
Weinstein
MC
,
Wright
TC
,
Bosch
FX
, et al
.
Projected clinical benefits and cost-effectiveness of a human papillomavirus 16/18 vaccine
.
J Natl Cancer Inst
2004
;
96
:
604
15
.
31.
Arias
E
,
Xu
J
.
Centers for Disease Control and Prevention
.
2019 Oct 29
.
National Vital Statistics Reports - United States Life Tables, 2017
. Available from: <https://www.cdc.gov/nchs/data/nvsr/nvsr68/nvsr68_07-508.pdf>.
32.
Surveillance, Epidemiology, and End Results (SEER) Program (www.seer.cancer.gov) SEER*Stat Database: Incidence - SEER 18 Regs Research Data + Hurricane Katrina Impacted Louisiana Cases, Nov 2018 Sub (2000–2016) <Katrina/Rita Population Adjustment>- Linked To County Attributes - Total U.S., 1969–2017 Counties, National Cancer Institute, DCCPS, Surveillance Research Program
,
released April 2019
,
based on the November 2018 submission
.
33.
Chuck
A
.
Cost-effectiveness of 21 alternative cervical cancer screening strategies
.
Value Health
2010
;
13
:
169
79
.
34.
de Kok
I
,
Korfage
IJ
,
van den Hout
WB
,
Helmerhorst
TJM
,
Habbema
JDF
,
Essink-Bot
ML
, et al
.
Quality of life assumptions determine which cervical cancer screening strategies are cost-effective
.
Int J Cancer
2018
;
142
:
2383
93
.
35.
Ramsey
SD
,
Willke
RJ
,
Glick
H
,
Reed
SD
,
Augustovski
F
,
Jonsson
B
, et al
.
Cost-effectiveness analysis alongside clinical trials II-An ISPOR Good Research Practices Task Force report
.
Value Health
2015
;
18
:
161
72
.
36.
US Bureau of Labor Statistics
.
2021 July
.
Consumer Price Index: Medical Services
.
Available from:
<https://www.bls.gov/cpi/factsheets/medical-care.htm>.
37.
Centers for Medicare & Medicaid Services
.
2019 Jan 13
.
Health Expenditures by Age and Gender
. Available from: <https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/NationalHealthExpendData/Age-and-Gender>.
38.
Centers for Medicare & Medicaid Services
.
2020 March 31
.
Physician/Supplier Procedure Summary
. Available from: <https://data.cms.gov/summary-statistics-on-use-and-payments/physiciansupplier-procedure-summary/data/2020>.
39.
Centers for Medicare & Medicaid Services
.
2021 Jan 13
.
Medicare Physician Fee Schedule
. Available from: <https://www.cms.gov/medicare/physician-fee-schedule/search/overview>.
40.
Centers for Medicare & Medicaid Services
.
2021 Jan 28
.
Clinical Laboratory Fee Schedule
. Available from: <https://www.cms.gov/medicaremedicare-fee-service-paymentclinicallabfeeschedclinical-laboratory-fee-schedule-files/21clabq1>.
41.
Mariotto
AB
,
Enewold
L
,
Zhao
J
,
Zeruto
CA
,
Yabroff
KR
.
Medical care costs associated with cancer survivorship in the United States
.
Cancer Epidemiol Biomarkers Prev
2020
;
29
:
1304
12
.
42.
Institute for Clinical and Economic Review
.
2020 Aug 31
.
2020–2023 Value Assessment Framework
. Available from: <https://icer.org/wp-content/uploads/2020/10/ICER_2020_2023_VAF_102220.pdf>.
43.
Briggs
AH
,
Weinstein
MC
,
Fenwick
EA
,
Karnon
J
,
Sculpher
MJ
,
Paltiel
AD
, et al
.
Model parameter estimation and uncertainty analysis: a report of the ISPOR-SMDM modeling good research practices task force working group-6
.
Med Decis Making
2012
;
32
:
722
32
.
44.
Sawaya
GF
,
Sanstead
E
,
Alarid-Escudero
F
,
Smith-McCune
K
,
Gregorich
SE
,
Silverberg
MJ
, et al
.
Estimated quality of life and economic outcomes associated with 12 cervical cancer screening strategies: a cost-effectiveness analysis
.
JAMA Intern. Med.
2019
;
179
:
867
78
.
45.
American Cancer Society
.
Cancer Facts & Figs. 2021
.
Atlanta
:
American Cancer Society
;
2021
;
27
8
.
This open access article is distributed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) license.