Background
The guidelines for managing abnormal cervical cancer screening tests changed from a results-based approach in 2012 to a risk-based approach in 2019.
Objective
We estimated the cost-effectiveness of the 2019 management guidelines and the changes in resource utilization moving from 2012 to 2019 guidelines.
Study Design
We utilized a previously published model of cervical cancer natural history and screening to estimate and compare the lifetime costs and the number of screens, colposcopies, precancer treatments, cancer cases, and cancer deaths associated with the 2012 vs 2019 management guidelines. We assessed these guidelines under the scenarios of observed screening practice and perfect screening adherence to 3-year cytology starting at age 21, with a switch to either 3-year or 5-year cytology plus human papillomavirus cotesting at age 30. In addition, we estimated the lifetime costs and life years to determine the cost-effectiveness of shifting to the 2019 management guidelines.
Results
Under the assumptions of both observed screening practice and perfect screening adherence with a strategy of 3-year cytology at ages 21 to 29 and switching to 3-year cotesting at age 30, the management of the screening tests according to the 2019 guidelines was less costly and more effective than the 2012 guidelines. For 3-year cytology screening at ages 21 to 29 and switching to 5-year cotesting at age 30, the 2019 guidelines were more cost-effective at a willingness-to-pay threshold of $100,000 per life year gained. Across all scenarios, the 2019 management guidelines were associated with fewer colposcopies and cancer deaths.
Conclusion
Our model-based analysis suggests that the 2019 guidelines are more effective overall and also more cost-effective than the 2012 guidelines, supporting the principle of “equal management of equal risks.”
Introduction
Over the past 20 years, cervical cancer screening guidelines have evolved from annual cytology-based (pap smear) screening in 2003 to biennial cytology-based screening in 2009 and to triennial cytology-based screening in 2012. In addition, human papillomavirus (HPV) testing was included in the 2012 US Preventive Services Task Force (USPSTF) guidelines as a part of “cotesting,” which involves both cytology and HPV testing every 5 years in women aged 30 and over (preceded by triennial cytology screening starting at age 21 years). In the most recent USPSTF guidelines of 2018, the screening options were expanded to include switching to 5-year primary HPV testing alone at age 30 years as well. , As the number of testing options and consequently the number of possible test result combinations has increased, the guidelines for the management of individuals with abnormal cervical cancer screening results have become more complex, as the decision-makers must determine the best action for a particular combination of results. One simplifying principle is that of “equal management of equal risks,” in which the strategy for the follow-up of a particular individual depends on the risk of precancer estimated using the past screening history and the current test results. ,
Why was this study conducted?
New guidelines for the management of abnormal cervical cancer screens utilize a risk-based approach, but the implications of this approach on the costs and clinical outcomes are not known.
Key findings
Risk-based guidelines are more cost-effective than previous algorithm-based guidelines, potentially offering better health outcomes with cost-savings.
What does this add to what is known?
This study illustrates how a risk-based approach to patient management can improve the efficiency of cervical cancer screening by accounting for a patient’s screening history when determining follow-up protocols.
The American Society for Colposcopy and Cervical Pathology (ASCCP) is a professional organization consisting of physicians, researchers, and other stakeholders representing numerous societies, health organizations, and federal agencies. ASCCP has convened 4 consensus conferences of the medical organizations involved in cervical cancer screening to create and revise national guidelines for the management of abnormal cervical cancer screening tests and cancer precursors in the years 2001, 2006, 2012, and most recently in 2019. As more follow-up data on individuals who undergo screening have become available, studies detailing the downstream risks of cervical disease and cancer for combinations of test results have been published. The goal of the 2019 update is to use this evidence base to increase the accuracy of the risk estimates and actions for different test results and to create a management framework based on precancer risk that follows the principle of “equal management of equal risks.” .
Simulation modeling is a useful tool to assess the future impact of guideline changes. This study aims to adapt and utilize an existing model of cervical cancer to project changes in the health and cost outcomes and to estimate the cost-effectiveness of the 2019 ASCCP Risk-Based Management Consensus Guidelines for clinical care among individuals with abnormal screening results. Whereas the actions recommended by the 2012 management guidelines were primarily based on current test results, the new guidelines estimate the risk, based on both current and prior results and more fully utilize the “equal management of equal risks” principle. The results of this analysis aim to shed light on the efficiency of the new guidelines and on the potential impact it may have on healthcare resources and outcomes.
Methods
Model
We used a previously developed model of cervical cancer that has been used extensively to evaluate the impact and cost-effectiveness of cervical cancer prevention strategies in the United States, including to inform recent USPSTF guidelines. This is an individual-based microsimulation model of HPV infections and cervical cancer natural history. The model simulates individuals from age 9 years until death, and incorporates HPV, precancer (cervical intraepithelial neoplasia grades 2 and 3 [CIN2, CIN3]) and invasive cancer health states. Preclinical cancers may progress in stages or may be detected by symptoms. Transitions are based on multiple factors such as age, the duration of infection or precancer, the HPV genotype, and history of HPV infection.
Previous literature provides detailed descriptions of the model’s parameter estimation and calibration and the data sources used in model development. These include the medical costs associated with the screening, diagnosis, and treatment and nonmedical costs such as patient time. The states of the model are depicted in Figure 1 . Screening strategies and management protocols for individuals with abnormal screening results can be modified within the model, including screening start and stop ages, interval since the last screen, screening modality, and the probability of adhering to the recommended protocols. The model then produces the relevant outputs of cost, precancer incidence, cancer incidence, life years, and test counts.
Validation/calibration
Data from Kaiser Permanente Northern California (KPNC) that informed the 2019 management guidelines were used to provide the estimates of cumulative CIN3+ risk (the probability that an individual develops CIN grade 3 or cervical cancer over time) for distinct sets of index screening results, including multiple rounds of screening. , In the model, we tracked the 5-year CIN3+ risk in women who received different combinations of cotesting results from primary screening to ensure that the risks inferred in the model corresponded to the KPNC observed data. Person-level model outputs were processed to calculate 5-year CIN 3+ estimates consistent with the KPNC empirical analysis. To account for discrepancies in the model fit, we adjusted the input parameters that affected the distribution of test results until a visual fit near or within the 95% confidence intervals of the data was achieved.
Model Scenario
Table 1 shows the 5-year CIN3+ risks stratified by preceding the cotest results in the KPNC cohort and the recommended management based on the 2012 and 2019 guidelines. It is to be noted that for individuals receiving their first screening test, the management recommendations, as outlined in Table 1 , were largely the same in both 2012 and 2019. The key changes in the 2019 management guidelines were primarily around more aggressive management for very high-risk results and less aggressive management for lower-risk results. The 2019 management guidelines are more aggressive in referring to immediate treatment for individuals with screen results that indicate a high risk (≥60%) of CIN3+, compared with the 2012 guidelines that recommended confirmation of CIN3+ via colposcopy/biopsy in most cases. To model this, we allowed individuals with high-grade squamous intraepithelial lesion (HSIL)/HPV-positive results to proceed directly to treatment if their HPV was identified as HPV type 16 or if they were aged 45 years or above. The rationale for the treatment of HSIL that is positive for type 16 is that the immediate CIN3+ risk exceeds the 60% treatment threshold. The decision to send the HSIL/HPV-positive individuals aged 45 years and older to treatment instead of colposcopy was made because future fertility concerns would be unlikely to impact shared decision-making.
| Cotest result (cytology/HPV) | N | Proportion of cotest results (%) | CIN3+ 5 y risk (%) | 95% CI | Management |
|---|---|---|---|---|---|
| NILM/Neg | 1,388,153 | 89.76 | 0.12 | (0.12, 0.13) | 5- y |
| ASCUS/Neg | 25,331 | 1.64 | 0.40 | (0.31, 0.48) | 3-y |
| LSIL/Neg | 3300 | 0.21 | 1.96 | (1.40, 2.52) | 1-y |
| HSIL/Neg | 183 | 0.01 | 27.37 | (20.27, 34.46) | Colposcopy |
| NILM/Pos | 63,541 | 4.11 | 4.80 | (4.58, 5.03) | 1-y |
| ASCUS/Pos | 30,506 | 1.97 | 7.27 | (6.87, 7.67) | Colposcopy |
| LSIL/Pos | 23,659 | 1.53 | 6.94 | (6.49, 7.40) | Colposcopy |
| HSIL/Pos | 3980 | 0.26 | 53.16 | (51.18, 55.13) | Colposcopy or treatment a |
a Individuals with HPV16+ or age 45 years or older received treatment in the 2019 guidelines model; all individuals received colposcopy in the 2012 guidelines model
In addition, the 2019 management guidelines contained several changes compared with the 2012 guidelines that allowed less aggressive management when considering two-screen result combinations. We utilized the KPNC data on a second round of screening and the test results conditional on preceding colposcopy result and/or treatment to determine the actions for two-screen combinations ( Table 2 ). For example, an HPV-positive Atypical Squamous Cells of Undetermined Significance (ASCUS) or Low-grade Squamous Intraepithelial Lesion (LSIL) cytology preceded by an HPV-negative and negative for intraepithelial lesion or malignancy (NILM) cytology or an HPV-negative ASCUS cytology in the first round of screening returned in 1 year using the 2019 guidelines rather than receiving immediate referral to colposcopy as indicated by the 2012 guidelines. In the model, we assumed that once a third screening result was observed, the first result was no longer considered and the action was based on the 2 most recent screens. An exception to this rule was a delayed return to 5-year screening. Individuals who had a high-grade result (HSIL cytology and CIN2 or CIN3 biopsy) stayed in a 3-year screening interval until 25 years passed. For other result combinations, the individuals could return to 5-year screening after 4 consecutive NILM/HPV-negative results.
| Test | Second test result | |||||||
|---|---|---|---|---|---|---|---|---|
| First test result | NILM/Neg | ASCUS/Neg | LSIL/Neg | HSIL/Neg | NILM/Pos | ASCUS/Pos | LSIL/Pos | HSIL/Pos |
| NILM/Neg | 5-y | 3-y | 3-y | Colposcopy | 1-y | 1-y | 1-y | Colposcopy |
| ASCUS/Neg | 5-y | 1-y | 1-y | Colposcopy | 1-y | 1-y | 1-y | Colposcopy |
| LSIL/Neg | 3-y | 1-y | 1-y | Colposcopy | 1-y | Colposcopy | Colposcopy | Colposcopy |
| Treated CIN2 or CIN3 (independent of preceding test results) | 1-y×3 | 1-y×3 | 1-y×3 | Colposcopy | Colposcopy | Colposcopy | Colposcopy | Treatment |
| NILM/Pos | 1-y | 1-y | 1-y | Colposcopy | Colposcopy | Colposcopy | Colposcopy | Colposcopy/Treatment a |
| ASCUS/Pos or LSIL/pos followed by colposcopy finding <CIN2 | 3-y | 1-y | 1-y | Colposcopy | 1-year | 1-year | 1-year | Colposcopy |
| HSIL/Pos or Neg followed by colposcopy finding <CIN2 | 1 year | 1-year | 1-year | Colposcopy | Colposcopy | Colposcopy | Colposcopy | Treatment |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
