Adherence to National Guidelines on Cervical Screening: A Population-Based Evaluation from a Statewide Registry

Factors Associated With Guideline-concordant and Excessive Cervical Cancer Screening: A Mixed Methods Study

INTRODUCTION

National guidelines recommend cervical cancer screening with Papanicolaou (Pap) testing at 3-year intervals or with human papillomavirus (HPV) testing alone or HPV/Pap cotesting at 5-year intervals for average-risk individuals aged 30-65 years.

METHODS

We explored factors associated with clinician-reported guideline-concordant screening, as well as facilitators and barriers to appropriate cervical cancer screening.

RESULTS

A national sample of clinicians (N = 1,251) completed surveys; a subset (n = 55) completed interviews. Most (94%) reported that they screened average-risk patients aged 30-65 years with cotesting. Nearly all clinicians who were categorized as nonadherent to national guidelines were overscreening (98%). Guideline concordant screening was reported by 47% and 82% of those using cotesting and HPV testing, respectively (5-year intervals), and by 62% of those using Pap testing only (3-year intervals). Concordant screening was reported more often by clinicians who were aged <40 years, non-Hispanic, and practicing in the West or Midwest, and less often by obstetrician-gynecologists and private practice physicians. Concordant screening was facilitated by beliefs that updated guidelines were evidence-based and reduced harms, health care system dissemination of guidelines, and electronic medical record prompts. Barriers to concordant screening included using outdated guidelines, relying on personal judgment, concern about missing cancers, inappropriate patient risk assessment, and lack of support for guideline adoption through health care systems or electronic medical records.

CONCLUSIONS

Most clinicians screened with Pap/HPV cotesting and approximately one-half endorsed a 5-year screening interval. Clinician knowledge gaps include understanding the evidence underlying 5-year intervals and appropriate risk assessment to determine which patients should be screened more frequently. Education and tracking systems can promote guideline-concordant screening.

Factors associated with shorter-interval cervical cancer screening for young women in three United States healthcare systems

Frequently changing cervical cancer screening guidelines over the past two decades have been inconsistently adopted in the United States. Current guidelines set the recommended screening interval to three years for average-risk women aged 21-29 years. Few studies have evaluated how patient and provider factors are associated with implementation of cervical cancer screening intervals among younger women. This study evaluated multilevel factors associated with screening interval length among 69,939 women aged 21-29 years with an initial negative Pap screen between 2010 and 2015 across three large health systems in the U.S. Shorter-interval screening was defined as a second screening Pap within 2.5 years of an initial negative Pap. Mixed-effects logistic regression was performed for each site to identify provider and patient characteristics associated with shorter-interval screening. The odds of shorter-interval screening decreased over the study period across all sites, though the proportion of patients screened within 2.5 years remained between 7.5% and 20.7% across sites in 2014-2015. Patient factors including insurance, race/ethnicity, and pregnancy were associated with shorter-interval screening, though the patterns differed across sites. At one site, the variation in shorter-interval screening explained by the provider was 10.6%, whereas at the other two sites, the provider accounted for < 2% of the variation in shorter-interval screening. Our results highlight the heterogeneity in factors driving cervical cancer screening interval across health systems and point to the need for tailored approaches targeted to both providers and patients to improve guideline-concordant screening.

Impact of HPV testing in opportunistic cervical screening: Support for primary HPV screening in the United States

Human papillomavirus (HPV) testing for cervical screening increases diagnosis of precancer and reduces the incidence of cervical cancer more than cytology alone. However, real-world evidence from diverse practice settings is lacking for the United States (U.S.) to support clinician adoption of primary HPV screening. Using a population-based registry, which captures all cervical cytology (with or without HPV testing) and all cervical biopsies, we conducted a real-world evidence study of screening in women aged 30 to 64 years across the entire state of New Mexico. Negative cytology was used to distinguish cotests from reflex HPV tests. A total of 264 198 cervical screening tests (with exclusions based on clinical history) were recorded as the first screening test between 2014 and 2017. Diagnoses of cervical intraepithelial neoplasia grades 2 or 3 or greater (CIN2+, CIN3+) from 2014 to 2019 were the main outcomes. Of cytology-negative screens, 165 595 (67.1%) were cotests and 4.8% of these led to biopsy within 2 years vs 3.2% in the cytology-only group. Among cytology-negative, HPV tested women, 347 of 398 (87.2%) CIN2+ cases were diagnosed in HPV-positive women, as were 147 of 164 (89.6%) CIN3+ cases. Only 29/921 (3.2%) CIN3+ and 67/1964 (3.4%) CIN2+ cases were diagnosed in HPV-negative, cytology-positive women with biopsies. Under U.S. opportunistic screening, across a diversity of health care delivery practices, and in a population suffering multiple disparities, we show adding HPV testing to cytology substantially increased the yield of CIN2+ and CIN3+. CIN3+ was rarely diagnosed in HPV-negative women with abnormal cytology, supporting U.S. primary HPV-only screening.

Updated estimate of the annual direct medical cost of screening and treatment for human papillomavirus associated disease in the United States

The annual direct medical cost attributable to human papillomavirus (HPV) in the United States over the period 2004-2007 was estimated at $9.36 billion in 2012 (updated to 2020 dollars). The purpose of this report was to update that estimate to account for the impact of HPV vaccination on HPV-attributable disease, reductions in the frequency of cervical cancer screening, and new data on the cost per case of treating HPV-attributable cancers. Based primarily on data from the literature, we estimated the annual direct medical cost burden as the sum of the costs of cervical cancer screening and follow-up and the cost of treating HPV-attributable cancers, anogenital warts, and recurrent respiratory papillomatosis (RRP). We estimated the total direct medical cost of HPV to be $9.01 billion annually over the period 2014-2018 (2020 U.S. dollars). Of this total cost, 55.0% was for routine cervical cancer screening and follow-up, 43.8% was for treatment of HPV-attributable cancer, and less than 2% was for treating anogenital warts and RRP. Although our updated estimate of the direct medical cost of HPV is slightly lower than the previous estimate, it would have been substantially lower had we not incorporated more recent, higher cancer treatment costs.

Trends in the Incidence of Human Papillomavirus-Associated Cancers by County-Level Income and Smoking Prevalence in the United States, 2000-2018

Human papillomavirus (HPV)-associated cancer burden is rising in the United States. Trends in the incidence by county-level income and smoking prevalence remain undescribed. We used the Surveillance, Epidemiology, and End Results 21 database to ascertain HPV-associated cancers during 2000-2018. Trends were estimated by county-level income and smoking prevalence quartiles. Anal and vulvar cancer incidence among women and anal cancer incidence among men increased markedly in the lowest-income counties, whereas the increases were slower in the highest-income counties (eg, for vulvar cancer, incidence increased 1.9% per year, 95% confidence interval [CI] = 0.9% to 2.9%, in the lowest-income counties vs 0.8% per year, 95% CI = 0.6% to 1.1%, in the highest-income counties). In recent years, cervical cancer incidence plateaued (0.0% per year [95% CI = -0.5% to 0.5%]) in the highest-income counties; in the lowest-income counties, the annual percentage change was 1.6% per year (95% CI = -0.7% to 4.0%). Counties with high smoking prevalence had marked increases in incidence compared with their counterparts (eg, anal cancer among men increased 4.4% per year [95% CI = 2.7% to 6.0%] for those living in counties with the highest smoking prevalence vs 1.2% per year [95% CI = 0.7% to 1.7%] for those living in counties with the lowest smoking prevalence). Improved and targeted prevention is needed to combat the widening disparities.