Characterizing Cervical Cancer Screening in the US: Preparing for the Era of Self-Collection

Discussion

Our results are the first to show, through validated medical record documentation of a large representative sample of US health care systems, how patients are currently receiving CCS. We show that 70.5% of eligible individuals had at least one CCS between 2017 and 2022 and that most of these screenings were provided by nongynecologists (61.9%). The portion of women screened agrees with reporting from national surveillance programs,⁷ commercial insurers, Medicaid HMOs,⁶ and Healthy People 2030.¹⁶ Furthermore, prior analyses from 3 large US health care settings (2010 to 2019) showed overall that FM provided 32.3% of CCS, OG 32.4%, and GIM 23.3% of CCS in 3 diverse settings, where the actual medical specialty varied widely.¹⁷ Our more detailed data include advanced practice providers and their subtaxonomies from a nationally representative population.

One surprising finding in this analysis was the mean number of cervical samples per patient was 2.6 (SD 2.7). In general, while about 90% of the screening-eligible population are considered average-risk, asymptomatic women recommended to receive 1 screen every 5 years, our results show a remarkably high average of CCS per patient over the study time frame. We speculate that more routine screens may occur more frequently than the USPSTF recommends.¹⁸ We also hypothesize that a larger proportion of the population than anticipated continues to test positive for HPV, requiring yearly HPV screening tests until clearance or disease development. Our results are hypothesis-generating: perhaps only half of the CCS received are preventive screens, with the other half being surveillance after an abnormal. We cannot verify this until we have a database that provides detailed laboratory results of every screen.

Our report also offers the first analysis of the medical specialties offering CCS. We anticipate this will change as primary HPV testing and self-sampling become the dominant screening technique, which FM and GIM physicians and APPs will most often do.

In addition, this report offers the first US estimate of colposcopy rates ranging from 1.8% to 7.9%, depending on the group's age and race/ethnicity. This study's overall 3.9% rate is significantly higher than in British Columbia, Canada.¹⁹ When British Columbia shifted to primary HPV screening with reflex cytology above the Atypical Squamous Cells of Undetermined Significance (ASCUS) threshold for colposcopy referral, the colposcopy rate increased from 0.8% to 3.2% initially as the prevalent disease was uncovered, dropping down to an expected background rate of 1.1% after 2 rounds of primary HPV screening, now detecting only incident disease.¹⁹ Our results reflect the US reality of cotesting with high false positive rates, resulting in high colposcopy rates,²⁰ which should be reduced with the future shift to primary HPV testing and self-sampling.

Most importantly, our results reveal continued low rates of screening among women older than 50, the median age of cervical cancer detection.²¹ Screening among the 55 to 64 yo detects 19% of the cervical cancers, and screening among women older than 65 identifies 20% of the cervical cancers.²¹ Over 50% of all cervical cancers are detected in women over 50 years old. The most common reason for cervical cancer detection in this age-group is inadequate documentation of 2 prior consecutive CCS at 5-year intervals that were both consecutively negative/normal before age 65.²²

As this study points out, the proportion of CCS performed by OB drops significantly in patients above age 50, while FM and IM screening rates increase. As FM and GIM already facilitate screening in this age-group for other cancers (such as colon, breast, and lung), cervical screening is a natural addition to ensure that exit screenings are appropriately completed. Thus, these specialties are also in an excellent position to modify current practices, including recommended workflow, and implement new testing strategies, including self-screening and informatics, within their electronic health records. This outcome will require FM and GIM leadership in this space, which OG has previously overseen.

With self-sampling newly an option for individuals 25 to 65, the historic requirement of disrobing to undergo an invasive procedure can be eliminated, giving both the patient and the busy clinician increased opportunities for screening.

Another possible contributor to the lack of screening in this age-group is the patient decision to await age qualification for Medicare, usually at 65 years, to avoid medical costs. Only 41% of women 55 years and older are employed with private insurance,²³ and many, even if insured, avoid health care until they are under Medicare coverage.²⁴ When she turns 65, the automatic electronic alert will be suppressed, often not requiring evidence of 2 prior normal CCS to be documented, leaving her at risk for cancer development.

From the health care system perspective, while all medical schools, allopathic and osteopathic, teach students how to do a human speculum examination using simulation and standardized patients, very few graduates use the technical skills in their professional lives. Similarly, training for the midwife, NP, and PA to use the speculum is individualized in their training and according to their practice interests. Our results show the long list of professionals who offered patients a speculum-based CCS in 2017 to 2022, but the vast majority were OG, FM, or GIM. The portion of practicing FM physicians not doing CCS has remained constant between 17 to 20%²⁵ over the past 2 decades, while 44% of practicing internists do not do CCS.²⁵ Self-sampling does not require speculum skills and offers an opportunity for OG, FM, and GIM to increase screening, both among those never before screened and for those who prematurely stop screening at 50. It also allows FM and GIM to join OG as experts in managing abnormal CCS and assume most CCS care.

An abnormal CCS requires technical training in colposcopy if the risk of CIN3 is greater than 4%.¹⁵ FM residencies offer colposcopy training to 42% of their graduates, but only 12% use these skills in practice.²⁶ IM, universally, does not provide colposcopy training in residency but does offer Women's Health tracks that teach CCS and colposcopy.^27–29 Both FM and GIM are in a position to increase their technical skills and empower their patients to screen by self-sampling, improving screening and follow-up among a greater number of women.

The lack of an OG presence in more than half of US counties, the dramatic drop in CCS by OG among women 50 and older, and the residency emphasis on technical operating skills in 7 subspecialties³⁰ all indicate a need for increased ownership in cervical cancer prevention by FM and GIM. The USPSTF already incorporates FM and GIM professionals proportionately in considering new evidence to change CCS guidelines, such as the shift to self-screening. The committees dedicated to managing abnormal results must be inclusive in their membership.