Abstract
Background Half of the women who develop cervical cancer in the US have never had a cervical cancer screen. Self-sampling is an equivalent technique to the invasive speculum exam. We aim to evaluate the current knowledge, attitudes, and behaviors toward self-sampled primary HPV testing in family medicine.
Methods The annual cross-sectional survey of the Council of Academic Family Medicine’s general membership included knowledge, attitudes, and beliefs about primary human papillomavirus (HPV) screening using self-sampling. The knowledge questions were based on current guidelines, as defined by the 2024 United States Preventive Services Task Force (USPSTF) 1. The attitude and belief questions were based on the Question-Behavior Theory 2. All surveys were emailed to the membership with up to five weekly reminders to complete.
Results We had a 62% survey response rate with 744 respondents. Regardless of demographic descriptors, all respondents significantly changed their intended behavior regarding offering self-sampling for cervical cancer screening (p < 0.001). Of those with a baseline attitude of not offering self-sampling, 88% changed their response to offer self-sampling at the end of the survey. Baseline knowledge of the advantages of primary HPV screening was lowest among underrepresented minority Hispanic respondents (52% correct, compared to 80% among Whites, p < 0.001). Women respondents were 2.96 times more likely to intend to offer self-sampling for cervical cancer screening than men (OR 2.96 (95% CI: 1.18, 7.44, p<0.05).
Conclusions Based on the Question-Behavior Theory, over 90% of family medicine educators intend to offer women self-sampling for cervical cancer screening.
- Cancer Screening
- Cervical Cancer
- Culture
- Family Medicine
- Human Papillomavirus Viruses
- Primary Health Care
- Residency
- Surveys and Questionnaires
Introduction
The incidence per year of cervical cancer in the US is 9.3/100,000,1 twice as high as the World Health Organization (WHO) elimination goal at 4/100,000,2 and the incidence of high-grade cervical lesions is 30.7/100,000.1 However, for many socioeconomic reasons, 20-30% of US-eligible women never screen.3–6
Primary human papillomavirus (HPV) DNA testing has an equivalent sensitivity of CIN3+ detection compared to previous standards of co-testing with cytology and HR HPV testing.7,8 Additional population benefits that make it a preferred test for average risk screening include fewer tests overall, fewer false positive results, and fewer referrals for colposcopy (which has a patient co-pay) for the same incident cervical intraepithelial neoplasias (CIN) detection.9
Over 40 countries worldwide have adopted this technique for organized population screening, including Australia, Denmark, the Netherlands, and the United Kingdom. Many offer self-collection options in both resource-poor and affluent settings.10,11 Several techniques for self-testing include vaginal self-swabs, first void urine testing, tampon/patch insertion, and a lavage-based device.12 In the US, three self-sampling devices have been US Food and Drug Administration (FDA)-approved for cervical cancer screening: two for in-office self-sampling use, the Evalyn Brush and Copan Floq Swab,13 and one for at-home use: the Teal Wand.14
As United States Preventive Services Task Force (USPSTF) guidelines facilitate the uptake of vaginal self-sampling for primary HPV testing as the default cervical cancer screening, we need to assess the readiness of family medicine to adopt this change. We aim to evaluate the current knowledge, attitudes, and behaviors toward self-sampled primary HPV testing in family medicine, with exploratory aims to document behavioral intentions to screen for primary HPV testing and self-sampling using demographic and institutional descriptors.
Methods
Study Design
The general methodology of the cross-sectional annual questionnaire, an omnibus survey, has been previously described.15 This survey was the 2024 Council of Academic Family Medicine (CAFM) Educational Research Alliance (CERA) survey of the general membership, consisting of family medicine physicians (MD/DO), educators, administrators, pharmacists, researchers, and behavioral health clinicians.16 The CERA survey was approved by the American Academy of Family Physicians Institutional Review Board (IRB 19-366 ORG A28) before dissemination in October 2024.
Survey Design
Standardized Demographic Characteristics
All CERA surveys include 13 demographic questions, asking about age, gender, race/ethnicity, self-identified as underrepresented in medicine (URM), highest degree earned, state, years in practice, institution location, and urbanity of the institution (urban, suburban, rural).
Specific Content
The ten knowledge, attitudes, and beliefs questions were designed to educate and assess current and future cervical cancer screening behaviors. Knowledge statements and questions were derived from the current literature about primary HPV screening and clinical guidelines using the 2024 USPSTF gold standard definition.17–21
The Question-Behavior Theory,22 a subset of the Theory of Planned Behavior,23 is the foundation of our attitude and belief of intended behavior change questions, where attitudes and intentions of future behavior influence the actualized behavior.
A knowledge statement was presented before the questions and was available for review at any time during the survey. Before widespread distribution, all closed-ended questions on attitude and behavioral intentions were trialed among primary care providers in the community for clarity.
Survey Dissemination
The CERA survey dissemination occurred from October 15 to November 22, 2024. The survey was delivered to a sample of 4,844 general members of the CAFM parent organizations.16 The survey was conducted via Survey Monkey; nonrespondents received five weekly reminders to complete it. The entire CERA survey set, including this survey, took approximately 15 minutes to complete.
Survey Variables and Definitions
The survey is presented in Appendix 1. The first content question provided the current baseline screening services offered (Q0). Two knowledge questions were about cervical cancer testing techniques and testing frequency (Q1 and Q2). Five belief questions concerned patients and clinical staff accepting self-sampling (Q3, Q4, Q5, Q8, and Q9). The two behavioral intentions questions were Q6 and Q7.
Statistical Analysis
Descriptive statistics of means (SD) and frequencies (%) were calculated for each population descriptor. Content knowledge questions were analyzed using proportions compared with the Chi-Square or Fisher’s Exact Test, with a two-sided 0.05 level of significance. Logistic regression was used to assess the likelihood of the intention of offering primary HPV testing with self-sampling based on population descriptors.
Results
Of the 1,194 respondents who completed their demographic questions, 744 (62.3% response rate) also completed the survey content questions. Most respondents were clinicians (98.6%). Most responders were 36-55 (51.1%), with a mean age of 47.5 (SD 11.8). The majority of respondents were women (63.5%), White (70.1%), not an underrepresented minority (81.7%), residing in the North (70.1%), having more than ten years in practice (59.4%), being part of a many residency institution (69.7%), located in a city (47.5%) and in an underserved area (62.1%).
Increase in Intent to Use Self-Sampling for Cervical Cancer Screening
Table 1 shows a significant increase in intent to use primary HPV screening with self-sampling post-survey compared to pre-survey for all population descriptors (p<0.001). The greatest change in intentions to use self-sampling for cervical cancer screening was observed among the oldest individuals (30%), those self-identifying as Black (34%), clinicians (21%), and those in practice for more than 10 years (23%). The greatest change in intentions to use self-sampling came from respondents whose institutions were located in the South (27%), had unopposed residency (19%), were in rural areas (27%), and were in underserved areas (23%). Following the survey, several groups achieved 100% intent to use self-sampling, and all were above 90% intent to change.
Of those who initially indicated that they were unlikely to offer self-sampling for primary HPV cervical cancer screening, 88.2% (149/169) changed their behavioral intent toward offering self-sampling (p < 0.001) (Table 2). Less than 1% (4/549) stated that they initially intended to provide primary HPV testing with self-sampling but changed their behavioral intent post-survey to not offer self-sampling.
Baseline Knowledge to Implement Primary HPV Testing
The first knowledge question pertained to the advantages of using primary HPV testing for screening, and the second one pertained to the frequency of testing in a general population. Neither knowledge question involved self-sampling; instead, only the concept of primary HPV testing was queried.
Correct knowledge about the advantages of primary HPV testing differed among people of color who also identified themselves as URM (Table 3). Among the self-identified Black and Hispanic URM respondents, the correct response rates were 62% and 52%, respectively, compared to White respondents (80%, p < 0.001).
However, for the knowledge of the 5-year screening frequency, only respondents in unopposed family medicine residencies had a significantly higher correct response rate than those at institutions with multiple residencies (78% vs. 70%, p < 0.05).
Predictors of Intent to Offer Self-Sampling for Cervical Cancer Screening
Table 4 indicates an equal likelihood of offering self-sampling for cervical cancer screening across various demographic characteristics, including age, race, self-identified URM status, years in practice, location, number of family medicine residencies at an institution, and urbanity. The only significant predictor was a respondent self-identifying as woman, where the chances were nearly three times more likely that they would offer self-sampling for primary HPV testing for cervical cancer screening than men ((OR=2.96 (1.18, 7.44), p<0.05) (Table 4).
Discussion
Our work shows that the general membership of CAFM shows a readiness to change to primary HPV testing with self-sampling as an option for women for cervical cancer screening. While only about 75% of all respondents across all categories recognized the three advantages of primary HPV screening over co-testing —fewer tests overall, fewer false-positive tests, and fewer unnecessary referrals to colposcopy for the same incident disease detected24 —97% indicated their intention to change their screening behavior.
Others have shown that innovating the screening test to allow the detection of persistent HPV infection, the cause of most cervical cancers25,26 by self-sampling, improved cervical cancer control.27 In particular, detecting HPV16 and HPV18, the most aggressive types,28 is critical. However, among vaccinated women, HPV types 51, 39, 59, 56, and 68 have replaced the vaccine-specific HPV type infections, causing the same high-grade CIN.29–31 Therefore, self-screening must use a test that detects all high-risk HPV types to be effective.32
Previous US studies identified several reasons for hesitation of the adoption of vaginal self-swab by clinicians including uncertainty of effectiveness of the test and concern for lack of insurance coverage of self-collection. Directed education would help make aware that self-sampling is non-inferior to co-testing33; and that the primary HPV testing is a Healthcare Effectiveness Data and Information Set (HEDIS) measure and National Committee for Quality Assurance (NCQA) approved, thus covered by the vast majority of insurances including Medicaid/Medicare, and the top 25 private payors in the United States which covers over 95% insured patients.34
Recent qualitative research has explored women’s perspectives on cervical cancer screening and documented descriptors of the in-office speculum exam as painful, embarrassing, cold, invasive, humiliating, murder, tortured, and traumatizing.6 Others stated multiple issues that self-sampling would solve. These include finding time to go to the office, transgender biology, and using the time the screening activity replaces to discuss the test’s meaning and how to interpret its results.3,4 Women across five continents have already demonstrated patient preferences for self-collection, where patients are twice as likely to agree to self-collection as a clinician performs the swab with a speculum.35
While self-sampling provides equivalent accuracy and significant ease for screening, it also offers a positive impact on climate control. The ecological benefits of self-sampling for cervical cancer screening demonstrate a significantly lower carbon footprint compared to the speculum, whether plastic or metal, providing additional environmental support for self-sampling.36,37
Strengths and Limitations
The strength of our study lies in its large number of respondents and the result of the unequivocal acceptance of the intention to change cervical cancer screening technique.
The limitation of our study mirrors that of other quantitative surveys, where we do not know the reasons behind their rankings or choices. Additionally, we recognize that the intention to perform a behavior and the actual execution of that behavior are not equivalent. All clinical changes must be reinforced, reiterated, recognized, reaffirmed, and revisited to ensure process implementation. The survey was quantitative and not a mixed methods survey and thus could not determine if knowledge disparities across racial/ethnic groups reflect differences in exposure, training or systemic barriers. The population of respondents may be biased towards those who are more interested in the most efficient evidence-based primary care delivery.
Conclusions
Family medicine faculty intend to change to self-sampling for primary HPV testing for cervical cancer screening now.
Conflicts of Interest
None.
Ethics Approval
AAFP IRB for CERA surveys: 19-366 ORG A28.
Corresponding Author
Alisa P. Young, MD, Department of Family Medicine, University of Michigan, alisay{at}med.umich.edu
This article was externally peer reviewed.
Appendix
- Received for publication August 19, 2025.
- Accepted for publication December 8, 2025.






