HPV and Pap testing among white, black, and hispanic women: results from a survey study

While the Pap test, HPV vaccine, and HPV test are important tools to promote cervical health, the American Cancer Society (ACS) recently updated its cervical cancer prevention recommendations to give primacy to HPV testing for cervical cancer screening. We investigate how women’s health providers view these changes, and the extent to which they have reached American women. To do so, we present results from a novel two-survey study of women’s health providers (N = 558) and White, Black, and Hispanic women (N = 1900). We find that a minority of providers report that the HPV test alone is adequate for cervical cancer screening (18%, CI = 14%, 21%) while 96% (CI = 94%, 97%) indicated that women should regularly have both Pap and HPV screenings. Black (88% CI = 86%, 91%) and Hispanic (87% CI = 84%, 89%) women report lower rates of Pap tests compared to White women (94%, CI = 92%, 96%). Only 35% (CI = 33%, 37%) of women report having had an HPV test, and these levels were similar across groups. White (29%, CI = 25%, 33%) and Hispanic women (26% CI = 23%, 29%) report statistically similar levels of provider recommendation for joint Pap-HPV testing, while Black women report statistically lower rates (22%, CI = 18%, 25%). These patterns hold after controlling for several important socio-demographic variables in logistic regression models. The continued lower rates of Pap tests among Black and Hispanic women, combined with low rates of HPV testing is likely to hinder the ability to reduce cervical cancer incidence and mortality, particularly for Hispanic and Black women.


Introduction
Cervical cancer screening guidelines have changed considerably in the last decade. The 2018 U.S. Preventive Services Task Force (USPSTF) and the 2012 version of the American Cancer Society (ACS) guidelines both recommended the use of Pap tests every three years for women ages 21-29, and a variation of acceptable alternative tests for those between 1 3 and White women, as well as the extent to which women have discussed the Pap and the HPV test with a healthcare provider-especially in comparison to the Pap test, given its longer history-and whether there are sociodemographic differences in discussing the Pap test, the HPV test, and the HPV-Pap co-test with a provider. Studying these issues can help us understand the possible impact of transitioning to the most recent American Cancer Society guidelines for cervical cancer screening, and who it may leave behind.

Study sample
The Cervical Cancer Study (CCS) was conducted at NORC at the University of Chicago in 2020 and includes two surveys. The first survey-The Healthcare Providers' survey-is a sample of health care professionals (N = 558) that used Dynata's nonprobability health care panel as the sample. The healthcare providers' survey was offered only in English and on the web. While there may be self-selection bias among healthcare professionals in the Dynata Panel since the sample source is an online nonprobability health care panel, the ages of the sampled health care professionals are evenly distributed (22% are between ages 25-45; 49% are between ages 45-59, and 27% are aged 60 +), indicating a wide range of experience in the women's health field. Gender of respondents was also balanced (49% female, 51% male). Respondents in the provider study were provided an undisclosed monetary incentive for survey completion.
The second survey-The Women's Survey-is a large, nationally representative study of Hispanic, Black, and White women ages 21-65 (N = 1900) that explores women's experiences receiving available cervical cancer related medical examinations, interventions, and recommendations. The completed interviews included White (N = 534), Black (N = 587), and Hispanic (N = 779) women between the ages of 21-65. The CCS Women's survey was offered in English and Spanish and on the web and phone. Data for the CCS Women's survey used NORC's AmeriSpeak® Panel, 3 and respondents for the women's study were provided a $2 incentive after survey completion.
Sampling strata for this study were based on age, race/ethnicity, education, and gender with the size of the selected sample per sampling stratum determined by the population distribution for each stratum. In addition, sample selection considered expected differential survey completion rates by demographic groups so that the set of panel members with a completed interview for a study is a representative sample of the target population. The AmeriSpeak panel sample for the CCS Women's survey included a supplement from Dynata's nonprobability online opt-in panel for Hispanic respondents ages 21 to 29 (N = 139, included in the total Hispanic sample of N = 799). To help reduce potential bias in the nonprobability sample, the nonprobability respondent sample was balanced by age and education via quota sampling and was appropriately weighted. The study was approved by NORC's Institutional Review Board (IRB).

Measures
For the physician study, healthcare providers were asked "should women have both Pap and HPV tests as part of regular screening?" (Yes, No) and "Do you believe HPV screening alone is adequate for screening for cervical cancer?" (Yes, No). These questions are novel and developed and tested by the research team for this study. Supplemental analysis also includes the region of residence of the providers.
For the women's study, the key questions women were asked were whether they have ever had a Pap test 4 and whether they have ever had an HPV test (Yes, No), whether they have discussed each of these medical options with a healthcare provider (Yes, No), whether they have received a recommendation for an HPV-Pap co-test (Yes, No), and the timing 3 Funded and operated by NORC at the University of Chicago, AmeriSpeak® is a probability-based panel designed to be representative of the US household population. Randomly selected US households are sampled using area probability and address-based sampling, with a known, non-zero probability of selection from the NORC National Sample Frame. These sampled households are then contacted by US mail, telephone, and field interviewers (face to face). The panel provides sample coverage of approximately 97% of the U.S. household population. Those excluded from the sample include people with P.O. Box only addresses, some addresses not listed in the USPS.
Delivery Sequence File, and some newly constructed dwellings. While most AmeriSpeak households participate in surveys by web, noninternet households can participate in AmeriSpeak surveys by telephone. Households without conventional internet access but having web access via smartphones are allowed to participate in AmeriSpeak surveys by web. For more information, see [40]. 4 The question "Have you ever had a Pap test?" was drawn from the Behavioral Risk Factor Surveillance System (BRFSS) to maintain statistical validity. Our results for this question (see Table 2) fall within the 95% confidence interval range for BRFSS 2018-the last time this question was asked in BRFSS and available on the BRFSS Web Enabled Analysis Tool. See [41] for questionnaire. of their last Pap test (< 3 years ago, 3 + years ago). Specifically, women were asked "When was your last Pap test?" and answers were coded into two categories: less than 3 years ago (= 1) and 3 years ago or more (= 0). The exact wordings for the remainder of these questions are presented in the results section alongside the findings. Women were also asked to report their race/ethnicity, age, educational level, marital status, employment status, household composition, and income.
All questions were pretested (N = 48 for the Women's Survey and N = 25 for the Healthcare Provider Survey) and reviewed before the data collection started.

Statistical analysis
Statistical analysis was carried out using SAS software 9.4. We first tabulated the proportion of providers who endorsed the ACS's recommendation that HPV testing alone is adequate for cervical cancer screening and the proportion who believed that women should have the HPV-Pap co-test. We then tabulated the proportion of women who had ever had an HPV test, a Pap test, and a provider recommendation for an HPV-Pap co-test.
Differences among Black, Hispanic, and White women in receiving a Pap test, an HPV test, and a provider recommendation for each of the cervical cancer prevention tools were determined by using 95% Confidence Intervals (CI) on the means. We used logistic regression analyses (using SAS PROC LOGISTIC procedure) to determine the likelihood of reporting having had a Pap test, an HPV test, and a provider recommendation for an HPV-Pap co-test among Black, Hispanic and White women from various age groups (see Table 3 for model results). For all the regression models, covariates include race/ethnicity, age, education, household income, and having a provider. All analyses were weighted to represent the general population of Black, Hispanic, and White women.

Provider reports
An overwhelming majority of healthcare providers (96%, CI = 0.94, 0.97) reported that women should have both the Pap and HPV tests as part of their regular healthcare screening, and only 18% (CI = 0.14, 0.21) reported that an HPV test alone was adequate for screening for cervical cancer (Table 1). This reveals that there is a disjuncture between healthcare providers' perceptions about the adequacy of the HPV test alone in screening for cervical cancer and the 2020 recommendations of the American Cancer Society.

Women's reports
Among our sample of women, an overwhelming majority of women reported having talked with a provider about a Pap test (88%) and have ever had a Pap test (92%) ( Table 2). These results reveal that Pap tests have had an impressive reach as a cervical cancer screening method. Given the well-documented importance of Pap tests in reducing cervical cancer incidence and mortality [8], the widespread promotion and use of Pap tests can be understood as a success in public health interventions in women's health.
While most women have talked with a healthcare provider about Pap tests and have had a Pap test, there are nonetheless differences in these outcomes. After controlling for having a provider, age, education and household income, logistic regression models (  [26,27].

3
However, among women who have ever had a Pap test in their lifetime, a larger proportion of Black and Hispanic women had more recent Pap tests. 81% of Black women (CI = 77%, 86%) and 80% of Hispanic women (CI = 76%, 85%) had a Pap test less than three years ago, while only 74% of White women (CI = 69%, 79%) had a Pap test within the past three years (Table 4).
In contrast to the near universal adoption of Pap tests, we found that only 37% of women have talked with their provider about an HPV test, and only 35% have ever had an HPV test. These results show that many women are neither receiving an HPV test nor having discussions with their healthcare providers about this important screening test.
While we found that Black and Hispanic women were less likely to have ever had a Pap test than White women, our logistic regression analysis estimating having had an HPV test (Table 3) revealed that there were no statistical differences between Black, Hispanic, and White women in HPV test uptake. We found that women with a regular healthcare provider were twice as likely to have had an HPV test compared with those who do not have a regular healthcare provider (OR = 2.33, CI = [1.18, 1.66]). We also found that younger women were more likely to have had an HPV test than older women ( Although the percentage of women who have ever talked with a provider about an HPV test is small, the proportion of women who have had an HPV-Pap co-test recommendation from a provider was even smaller, with only 27% (CI = 25%, 29%) of women reporting that they have received a provider recommendation for having the Pap and HPV test at the same time. Yet in our physician survey, we found that 96% of providers believed that women should have the HPV-Pap cotest as part of their regular screening. This could suggest that even though providers believe in the co-test, they may not be discussing this option with their patients, or that some women were not aware that one sample is used for both tests.
Descriptive statistics show that White (29%, CI = 25%, 33%) and Hispanic women (26%, CI = 23%, 29%) report statistically similar levels of receiving provider recommendation for the HPV-Pap co-test, while Black women report statistically lower rates (22%, CI = 18%, 25%). Our logistic regression analysis (Table 3) shows that, even when controlling for covariates, Black women were still less likely than White and Hispanic women to receive a provider recommendation for an HPV-Pap co-test. In addition, we found that women who have a regular healthcare provider were more likely to receive

Discussion
The Pap test has been used for cervical cancer screening since the first half of the twentieth century and can detect cell changes on the cervix at risk of becoming cervical cancer if left untreated [4,42]. It appears that public health campaigns around the Pap test were efficacious, as demonstrated by the fact that not only have most women had a Pap test in their lifetime, but most importantly, that cervical cancer incidence and mortality rates have significantly decreased in recent years. These benefits, however, have not reached all women [20][21][22]. Our findings, consistent with previous research [23], reveal that in spite of the increased access to the Pap test, Black and Hispanic women are still less likely to have ever had a Pap test or to have spoken with a provider about a Pap test. At the same time, we also find that among women who have ever had a Pap test, a larger proportion of Black and Hispanic women have had a Pap test within the past three years than White women; yet Black and Hispanic women are still more likely to be diagnosed later with, and to die of, cervical cancer [17][18][19]. While our data cannot directly address why this is so, despite the more recent Pap tests, Black and Hispanic women have worse cervical cancer outcomes than White women. The literature provides several possible explanations: First, having ever received a Pap test may not address the critical access issue of having a regular source of care. Black and Hispanic women are less likely to have regular healthcare providers compared to White women [39], so if they had an abnormal cervical screening test, they may not receive adequate follow-up care, such as a referral for or follow-up with a colposcopy; indeed, research finds that Black and Hispanic individuals are less likely to have follow-up appointments for test results than White individuals [43]. Second, Black and Hispanic women have lower HPV vaccination rates compared to White Women [44]. The lower HPV vaccination rates among Black and Hispanic women may also contribute to higher rates of cervical cancer among these women.
Previous studies on cervical cancer prevention have mostly focused on Pap tests, HPV vaccinations, and discussions with providers about Pap tests and HPV vaccinations, yet little is known about women's experiences with HPV testing, HPV-Pap co-testing, and discussions with providers about HPV tests. Our study attempts to address this gap in the literature by describing the experiences of women with HPV testing and HPV-Pap co-testing, as well as providers' perceptions of these tests. We found that women's participation in HPV testing is quite limited; while most women have had a Pap test, less than half of the women surveyed have had an HPV test or discussed this test with their healthcare providers. That HPV is a sexually transmitted infection (STI) may help explain this finding; women may not feel comfortable discussing an STI, even one that is as common as HPV, with their healthcare providers. Further, research has shown that even patients seen at STI clinics are sometimes unaware of which tests are performed [45]. While rates of cervical cancer incidence and mortality are higher for Black and Hispanic women compared to White women [20-22, 24, 25, 46, 47], we find that Black and Hispanic women are no more likely than White women to have had an HPV test in their lifetimes.
While the new ACS guidelines promote the primary HPV test for cervical cancer screening, findings from the Provider Survey show that only 18% of physicians believe that the HPV test alone is adequate for cervical cancer screening. This may explain why HPV test uptake is universally low for all women. Additionally, while the HPV DNA test was first approved in 2011 to be specifically used alongside the Pap test, the primary HPV test was not approved by the FDA until 2020-which is quite a recent development [48]. Unlike the first iteration of the HPV test and those used in co-testing, the newer primary HPV test is able to detect more HPV genotypes-including high risk types 16 and 18 [49].We hypothesize that some physicians may not want to depend solely on HPV testing for cancer screening, as these tests are mainly used to determine follow-ups with a colposcopy or Pap test, depending on the HPV type [48]. Additionally, as HPV is quite common and has no treatment-and for many women with lower-risk HPV types goes away on its own without significant health issues [50]-HPV testing may pose risk for over-treatment. However, long-term infections with high-risk forms of HPV can cause cervical cancer and genital warts [51].
Supplementary Table S1 shows that provider opinion on the adequacy of HPV testing is higher in the Northeast and West, compared to the Midwest and South regions of the U.S. While regional variation may explain some of the differences in patient screening, more research is necessary to understand these differences. Further research should study HPV testing and co-testing availability by different geographic levels-especially for rural clinics-as well as how such availability may be spatially correlated with underserved populations. However, we also found that 96% of providers believe that an HPV-Pap co-test, where typically one sample is taken and tested for both the HPV DNA and precancerous cells, is important. It may be the case that providers are indeed performing an HPV test at the same time as a Pap test but are failing to communicate this to their patients. Further research is needed to determine whether the low reports of HPV testing among women stems from a lack of access to this test, or whether women have had them but understand them as Pap tests instead of as an HPV-Pap combined test.

Limitations
One limitation of this study is that the survey is based on women's self-reports of their healthcare experiences, which may be subject to retrospective bias. Having access to medical records or insurance reimbursement could strengthen these self-reports. However, even though these are self-reports, they reflect women's perceptions of their healthcare encounters and experiences with each of the different cervical cancer prevention tools, and these perceptions may inform women's healthcare seeking behavior.
While we surveyed physicians' beliefs about the adequacy of the HPV test alone as a cervical cancer screening method, we did not measure how often physicians performed these screening tests with their patients or whether they reported discussing this option with their patients. However, physicians' beliefs could inform actual clinical practices and if providers do not believe that HPV testing alone is adequate for cervical cancer screening, they may not recommend this testing option to their patients. Because our survey also includes women's reports of the use of the HPV test and their reports of having received a provider recommendation for the use of these tests, we are able to confirm that both only a minority of physicians believe in the benefit of HPV testing alone and only a minority of women have had an HPV test or discussed this option with a healthcare provider, providing support that physicians' beliefs could be a proxy for actual clinical behavior.
Additionally, while this study has shown that disparities persist for Black and Hispanic women around the use of the Pap test and receiving an HPV-Pap co-testing recommendation from a provider, it did not include women of other backgrounds such as Asian American, Middle Eastern, and Native American. To fully understand how access to and discussions with providers about cervical cancer prevention strategies affect all women, future studies should expand their sample to include women from other backgrounds.

Conclusion
With changing cervical cancer screening guidelines, it is imperative that further research evaluate how these new guidelines impact Black, Hispanic, and White women's experiences with cervical cancer screening and whether they further reduce cervical cancer incidence and mortality rates. While the Pap test, the HPV vaccine, and the HPV test are important tools to prevent cervical cancer, Black and Hispanic women still bear the burden of higher cervical cancer rates and mortality. Therefore, we encourage public health experts to continue to promote cervical cancer screenings through outreach-including through web, social media, TV, print media, and radio messages-to all women, and especially to Black and Hispanic women.