Preferences Among U.S. Women for Cervical Cancer Screening with Self-Collected Specimens for Human Papillomavirus Testing

Abstract

OBJECTIVE:

To examine preferences for human papillomavirus (HPV) specimen self-collection, and collection location, in a nationally representative sample of reproductive-aged women in the United States.

METHODS:

This cross-sectional analysis used household population-based data from the National Survey of Family Growth (January 2022–December 2023) and was limited to women aged 21–49 years without a history of hysterectomy or cervical cancer (sample n=4,465). Survey weights and design variables were applied to generate nationally representative population frequencies and percentages of preference for HPV self-collection compared with clinician collection, and preference for collection location (ie, at home or in office).

RESULTS:

Among eligible U.S. women, 42.9% preferred HPV self-collection, 28.5% preferred clinician collection, and 28.6% expressed no preference. An estimated 41.7 million (71.5%) U.S. women aged 21–49 years were open to HPV self-collection (either preferring it or having no preference), including 9.7 million women who were underscreened or never screened. Among women who were open to HPV self-collection, more than half (52.1%) preferred self-collection at home, 14.7% preferred to do self-collection in a doctor’s office, and 33.2% had no preference for location. More underscreened or never-screened women preferred HPV self-collection (54.0%) and at-home collection (59.3%) compared with those who were up to date with screening (40.3% and 50.2%, respectively, P<.001). Preference for self-collection also varied by race and Hispanic origin, education, income, parity, sexual orientation, and prior experience of nonvoluntary vaginal intercourse.

CONCLUSION:

In this nationally representative study, more than 7 in 10 U.S. women aged 21–49 years were open to HPV self-collection for cervical cancer screening, with more than half favoring at-home collection. Preference was higher among women who were not up to date with screening. These findings provide timely evidence to inform future policy decisions and implementation strategies to improve access to cervical cancer screening.

Approximately 60% of the nearly 13,000 women diagnosed annually with cervical cancer in the United States have not been adequately screened.¹ In addition to access and logistical barriers to screening, women also have reported embarrassment, anxiety, and discomfort with pelvic examinations.² Contemporary screening recommendations include cervical cytology (Pap test) and tests for human papillomavirus (HPV) infection,³ which traditionally are performed by clinicians with a cervical speculum examination during office visits.

The U.S. Food and Drug Administration approved use of self-collected vaginal specimens for HPV testing (hereafter referred to as HPV self-collection) in health care settings in May 2024,⁴ and the first at-home vaginal sample self-collection device in May 2025.⁵ Self-collection has been shown to be feasible, acceptable, and cost-effective in reaching never screened and underscreened populations in both low- and high-resource settings,^2,6 and it is now widely used in several high-income countries (eg, Australia). Substantial evidence indicates that HPV self-collection can improve screening coverage, particularly among underscreened populations.^2,6 However, most of this evidence comes from home settings for self-collection.²

In the United States, women’s perspectives on HPV self-collection are largely unknown. To address this critical gap, we examined preferences of HPV self-collection and collection settings (in office or at home) in a nationally representative sample of reproductive-aged women in the United States. Establishing a national baseline of women’s preferences for self-collected vaginal specimens for HPV testing can inform the implementation of this new screening strategy. Specifically, we examined whether women’s preferences for HPV self-collection and collection settings differed by their screening status and key sociodemographic characteristics.

METHODS

In this observational, cross-sectional study, we analyzed data from the National Survey of Family Growth (NSFG) collected between January 2022 and December 2023. The NSFG gathers detailed information on reproductive health and family formation among the reproductive-aged population in the United States. The 2022–2023 NSFG used a national area, multistage, probability-based sample designed to be representative of the U.S. household population of those aged 15–49 years. Household screeners were used to identify eligible respondents, with oversamples of Black and teen respondents. Surveys were administered online or in person. In-person interviews were conducted by trained female interviewers using tablet computers, which included a computer-assisted self-interview portion for sensitive questions (eg, sexual behavior and history). Additional details of the survey design and methodology are described elsewhere.⁷ Survey protocol and informed consent procedures were approved by the National Center for Health Statistics Ethics Review Board. The 2022–2023 female respondent response rate conditional on screener response was 54.7%. This study included women aged 21–49 years without a history of cervical cancer or hysterectomy.

Outcomes of this study were measured in two questions: 1) “In the future, you may have the option to use a simple kit to test for HPV infection yourself. To use, you insert a swab about an inch into your vagina and rotate it for 30 seconds. It is easy to use the kit correctly and the results are accurate. If available, would you prefer this option rather than having the test done by a doctor or nurse?” (response options: yes, no, or no preference); those who answered “yes” or “no preference” were then asked; 2) “Would you prefer to do this test at home or at a doctor’s office?” (response options: home, doctor’s office, or no preference). Women who answered “yes” or “no preference” to question 1 were considered open to HPV self-collection.

Female respondents were asked about their past Pap (cytology) and HPV testing. Screening status was categorized as: 1) up to date, defined as having a Pap test within the past 3 years or an HPV test (with or without a Pap test) within the past 5 years; or 2) underscreened or never screened if neither criterion was met.

To generate nationally representative estimates, we used sample weights and design variables to account for NSFG’s complex sample design and differential response rates. Analyses were performed in SAS 9.4. We calculated weighted estimates (frequencies and proportions of women) of HPV self-collection preferences, overall, and stratified by cervical cancer screening status and characteristics known to be associated with cervical cancer screening uptake such as race, ethnicity, and socioeconomic variables. The weighted frequencies are the estimated numbers of eligible U.S. women that the sample represented (ie, U.S. women aged 21–49 years without a history of cervical cancer or hysterectomy). The weighted percentages are the estimated proportions within this population.

Associations between HPV self-collection preferences and population characteristics were assessed using Rao-Scott χ² tests, which adjust the Pearson χ² test for complex survey design, with a significance level of a=0.05. Comparisons between specific categories were considered different when indicated by nonoverlapping 95% CIs. We followed STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) reporting guidelines.

RESULTS

A total of 4,465 female respondents who met the inclusion criteria were included in the analysis, representing 58.3 million women aged 21–49 years in the U.S. household population. Overall, 42.9% preferred HPV self-collection, 28.5% preferred clinician collection, and 28.6% had no preference (Table 1). The estimated number of U.S. women who were open to HPV self-collection was 41.7 million (71.5%), including 9.7 million who reported never being screened or reported being underscreened (Fig. 1). Among those open to HPV self-collection, 52.1% preferred self-collection at home, 14.7% preferred self-collection in a doctor’s office, and 33.2% had no preference for location (Table 2).

Table 1.

Preference for Human Papillomavirus Testing Self-Collection in U.S. Women Aged 21–49 Years, National Survey of Family Growth, 2022–2023^*

Characteristic	Sample Size†	Weighted U.S. Population (×1,000)	Preference for HPV Self-Collection			P‡
			Yes	No	No Preference
Overall	4,465	58 ,311	42.9 (40.3–45.4)	28.5 (26.2–30.9)	28.6 (26.7–30.5)	NA
Cervical cancer screening status§
Up to date	3,622	45 ,991	40.3 (37.7–42.9)	31.0 (28.4–33.7)	28.7 (26.5–30.9)	<.001
Underscreened or never screened	824	11 ,936	54.0 (49.8–58.2)	18.9 (15.6–22.1)	27.1 (23.9–30.4)
Age (y)
21–29	1,257	18 ,956	43.8 (39.4–48.3)	29.7 (25.8–33.6)	26.5 (23.6–29.4)	.07
30–39	1,894	21 ,429	44.8 (41.5–48.2)	26.5 (23.7–29.3)	28.7 (26.3–31.0)
40–49	1,314	17 ,927	39.5 (36.2–42.7)	29.9 (26.4–33.4)	30.7 (27.2–34.1)
Race and Hispanic origin
Black, non-Hispanic	672	7,685	39.7 (34.2–45.3)	37.2 (32.3–42.1)	23.0 (18.7–27.4)	.004
Hispanic	921	12 ,339	44.8 (38.8–50.8)	29.5 (24.5–34.5)	25.7 (21.0–30.4)
None of the above, non-Hispanic	586	7,521	45.4 (39.2–51.7)	24.9 (20.3–29.5)	29.6 (23.9–35.4)
White, non-Hispanic	2,286	30 ,767	42.2 (39.4–45.0)	26.9 (24.4–29.4)	30.9 (28.6–33.1)
Health insuranceǁ
Private	2,802	36 ,243	42.7 (40.0–45.5)	26.3 (23.8–28.8)	31.0 (28.9–33.1)	<.001
Medicaid, CHIP, or state-sponsored	903	12,054	41.1 (36.0–46.3)	36.7 (31.8–41.7)	22.1 (18.2–26.0)
Medicare, military health, or other government	252	3,110	48.9 (41.2–56.6)	26.5 (19.5–33.4)	24.6 (17.1–32.1)
None	508	6,905	43.7 (36.4–51.0)	27.2 (22.0–32.4)	29.1 (23.7–34.5)
Education
High school graduate or less	994	15 ,573	41.7 (37.0–46.5)	35.4 (31.0–39.9)	22.8 (19.2–26.4)	<.001
Some college	1,308	16 ,609	43.8 (40.2–47.3)	29.8 (26.1–33.5)	26.5 (23.6–29.3)
Bachelor’s degree or more	2,163	26 ,129	43.0 (39.6–46.3)	23.7 (21.1–26.3)	33.4 (30.8–35.9)
Family income (% FPL)¶
Less than 150	1,199	15 ,902	43.6 (39.3–47.8)	33.1 (28.8–37.3)	23.4 (19.7–27.1)	<.001
150–299	1,071	13 ,141	40.7 (36.7–44.8)	32.1 (28.1–36.0)	27.2 (24.0–30.4)
300 or more	2,195	29 ,269	43.4 (40.5–46.3)	24.5 (22.0–27.0)	32.1 (29.4–34.7)
Parity
0	1,936	24 ,680	46.2 (42.5–49.9)	24.7 (21.4–28.0)	29.1 (26.5–31.7)	.001
1 or more	2,529	33 ,631	40.4 (37.7–43.1)	31.4 (28.6–34.1)	28.2 (25.6–30.8)
Nonvoluntary vaginal intercourse ever
No	3,463	45 ,515	41.8 (39.0–44.5)	28.9 (26.4–31.3)	29.4 (27.5–31.2)	.02
Yes	888	11 ,219	49.3 (43.5–55.0)	26.7 (21.6–31.7)	24.0 (19.7–28.4)
Sexual orientation
Heterosexual or straight	3,483	44 ,949	40.4 (37.7–43.0)	29.7 (27.0–32.4)	29.9 (28.0–31.9)	<.001
Lesbian, bisexual, or something else	845	11 ,441	54.7 (49.8–59.5)	23.8 (19.8–27.8)	21.5 (18.0–25.1)

HPV, human papillomavirus; NA, not applicable; CHIP, Children’s Health Insurance Program; FPL, federal poverty level.

Data are weighted % (95% CI) unless otherwise specified.

^*Respondents were asked, “In the future, you may have the option to use a simple kit to test for HPV infection yourself. To use, you insert a swab about an inch into your vagina and rotate it for 30 seconds. It is easy to use the kit correctly and the results are accurate. If available, would you prefer this option rather than having the test done by a doctor or nurse?”

^†Sample frequencies for the following variables do not total 4,465 owing to answers of “not ascertained,” “refused,” or “don’t know”: cervical cancer screening status (n=19; 0.4%), nonvoluntary vaginal intercourse ever (n=114; 2.6%), sexual orientation (n=137; 3.1%).

^‡Rao-Scott χ² test of association between HPV self-collection preference and each population characteristic at a=0.05.

^§Up to date defined as having a Pap test within the past 3 years or an HPV test within the past 5 years; underscreened or never screened if neither criterion was met.

^ǁRespondents with private health insurance or Medigap at the time of the survey were categorized as having private health insurance. Women with no insurance and those with only a single-service plan or only the Indian Health Service coverage were considered uninsured.

^¶Family income as a percentage of the FPL, which is calculated from family income and family size using the United States Census Bureau’s poverty thresholds for the previous calendar year (https://www.census.gov/data/tables/time-series/demo/income-poverty/historical-poverty-thresholds.html).

Fig. 1.

Estimated percentage and number of U.S. women aged 21–49 years open* to human papillomavirus (HPV) self-collection overall and by screening status, National Survey of Family Growth, 2022–2023. *Includes women who answered “yes” or “no preference” to the survey question, “In the future, you may have the option to use a simple kit to test for HPV infection yourself. To use, you insert a swab about an inch into your vagina and rotate it for 30 seconds. It is easy to use the kit correctly and the results are accurate. If available, would you prefer this option rather than having the test done by a doctor or nurse?”

Qin. Preferences for HPV Self-collection in U.S. Women. Obstet Gynecol 2026.

Table 2.

Preference for Location of Self-Collected Human Papillomavirus Tests in U.S. Women Aged 21–49 Years, National Survey of Family Growth, 2022–2023^*

Characteristic	Sample Size†	Weighted U.S. Population (×1,000)	Preference for HPV Self-Collection Location			P‡
			Home	Doctor’s Office	No Preference
Overall	3,209	41 ,554	52.1 (49.5–54.7)	14.7 (12.8–16.6)	33.2 (30.9–35.5)	NA
Cervical cancer screening status§
Up to date	2,535	31 ,717	50.2 (47.2–53.2)	15.9 (13.8–18.0)	33.9 (31.2–36.6)	.001
Underscreened or never screened	662	9,636	59.3 (54.9–63.8)	10.6 (7.3–13.8)	30.1 (25.6–34.6)
Age (y)
21–29	879	13 ,295	54.4 (49.9–58.9)	15.3 (12.3–18.4)	30.3 (26.6–34.0)	.08
30–39	1,391	15 ,745	53.3 (50.0–56.6)	14.3 (11.7–16.9)	32.4 (28.9–35.8)
40–49	939	12 ,514	48.2 (43.9–52.5)	14.5 (11.4–17.5)	37.4 (33.3–41.5)
Race and Hispanic origin
Black, non-Hispanic	395	4,817	49.6 (41.9–57.2)	23.3 (18.1–28.6)	27.1 (21.7–32.5)	<.001
Hispanic	661	8,691	47.9 (41.1–54.6)	20.0 (15.9–24.0)	32.1 (25.9–38.3)
None of the above, non-Hispanic	436	5,603	53.6 (47.5–59.6)	14.3 (10.7–17.9)	32.2 (27.0–37.3)
White, non-Hispanic	1,717	22 ,443	53.9 (50.6–57.2)	10.9 (8.7–13.1)	35.2 (32.1–38.3)
Health insuranceǁ
Private	2,078	26 ,635	51.6 (48.9–54.4)	13.8 (11.6–16.0)	34.6 (32.4–36.8)	.49
Medicaid, CHIP, or state-sponsored	581	7,623	54.4 (47.6–61.2)	15.8 (12.6–19.0)	29.8 (23.5–36.0)
Medicare, military health, or other government	182	2,286	50.4 (38.8–61.9)	14.4 (7.4–21.3)	35.2 (24.8–45.7)
None	368	5,010	51.8 (44.7–58.9)	18.0 (12.2–23.8)	30.2 (24.3–36.0)
Education
High school graduate or less	624	10 ,023	53.5 (48.0–58.9)	16.6 (12.8–20.4)	30.0 (25.0–34.9)	.25
Some college	926	11 ,655	53.5 (49.3–57.7)	14.4 (11.2–17.6)	32.1 (27.9–36.3)
Bachelor’s degree or more	1,659	19 ,876	50.6 (47.4–53.8)	13.9 (11.4–16.4)	35.5 (32.9–38.1)
Family income (% FPL)¶
Less than 150	797	10 ,624	52.2 (46.7–57.7)	17.5 (13.9–21.2)	30.3 (25.3–35.2)	.07
150–299	747	8,926	53.9 (49.1–58.7)	15.5 (11.9–19.2)	30.6 (25.7–35.4)
300 or more	1,665	22 ,004	51.3 (48.0–54.7)	13.0 (10.7–15.2)	35.7 (32.8–38.6)
Parity
0	1,487	18 ,536	55.2 (51.4–59.0)	12.8 (10.0–15.5)	32.1 (28.8–35.3)	.03
1 or more	1,722	23 ,018	49.6 (46.4–52.9)	16.3 (14.0–18.5)	34.1 (30.9–37.3)
Nonvoluntary vaginal intercourse ever
No	2,463	32 ,347	51.2 (48.5–54.0)	15.3 (13.1–17.6)	33.4 (31.1–35.8)	.04
Yes	671	8,193	58.4 (52.9–63.8)	11.8 (8.6–15.1)	29.8 (24.3–35.3)
Sexual orientation
Heterosexual or straight	2,451	31 ,578	49.6 (47.0–52.2)	16.3 (14.0–18.7)	34.1 (31.7–36.5)	<.001
Lesbian, bisexual, or something else	665	8,681	63.8 (58.6–69.0)	8.6 (6.0–11.2)	27.6 (22.4–32.7)

HPV, human papillomavirus; NA, not applicable; CHIP, Children’s Health Insurance Program; FPL, federal poverty level. Data are weighted % (95% CI) unless otherwise specified.

^*The question “Would you prefer to do this test at home or in a doctor’s office?” was asked only among those who answered “yes” or “no preference” to the survey question: “In the future, you may have the option to use a simple kit to test for HPV infection yourself. To use, you insert a swab about an inch into your vagina and rotate it for 30 seconds. It is easy to use the kit correctly and the results are accurate. If available, would you prefer this option rather than having the test done by a doctor or nurse?”

^†Sample frequencies for the following variables do not total 3,209 owing to answers of “not ascertained,” “refused,” or “don’t know”: cervical cancer screening status (n=12; 0.4%), nonvoluntary vaginal intercourse ever (n=75; 2.3%), sexual orientation (n=93; 2.9%).

^‡Rao-Scott χ² test of association between HPV self-collection location preference and each population characteristic at a=0.05.

^§Up to date, defined as having a Pap test within the past 3 years or an HPV test within the past 5 years; underscrened or never screened, if neither criterion was met.

^ǁRespondents with private health insurance or Medigap at the time of the survey were categorized as having private health insurance. Women with no insurance and those with only a single-service plan or only the Indian Health Service coverage were considered uninsured.

^¶Family income as a percentage of the FPL, which is calculated from family income and family size using the United States Census Bureau’s poverty thresholds for the previous calendar year (https://www.census.gov/data/tables/time-series/demo/income-poverty/historical-poverty-thresholds.html).

A higher estimated percentage of women who were underscreened or were never screened preferred HPV self-collection (54.0%), compared with those who were up to date on screening (40.3%, P<.001). In addition, a higher estimated percentage of women who were underscreened or were never screened preferred at-home collection (59.3%), compared with those who were up to date on screening (50.2%, P=.001). More than one-third of non-Hispanic Black women; women with Medicaid, CHIP (Children’s Health Insurance Program), or state-sponsored health insurance; women with a high school education or less; and women with low incomes answered “no” to preference for HPV self-collection question (Table 1). Women who were nulliparous; women who had ever experienced nonvoluntary vaginal intercourse; and women who identified as lesbian, bisexual, or something else more often preferred HPV self-collection and self-collection at home (Tables 1 and 2). Additional univariate analysis showed that having a usual source of care, place of birth (U.S.-born or foreign-born), and rural or urban residence were not associated with preference for self-collection (Appendix 1, available online at http://links.lww.com/AOG/E452). Women in rural areas were less likely than women in urban areas to prefer HPV self-collection at a doctor’s office (9.6% vs 15.5%), and a higher proportion of U.S.-born women reported “no preference” for collection location compared with foreign-born women (34.5% vs 27.5%, Appendix 2, available online at http://links.lww.com/AOG/E452).

DISCUSSION

In this nationally representative sample, an estimated 71.5% of U.S. women aged 21–49 years were open to HPV self-collection. In its draft recommendations published in December 2024, the U.S. Preventive Services Task Force included self-collected HPV tests every 5 years starting at age 30 years.² Recently, the Enduring Guidelines Committee published additional recommendations for self-collected vaginal specimens for HPV testing, including a 3-year screening interval when self-collected samples are used.⁸ In our study, more than half of women who were open to HPV self-collection preferred to do it at home, an approach currently under evaluation by the National Institutes of Health.⁴

Our findings were consistent with those reported in a nonrandomized clinical trial, where nearly one-half (46%) of study participants preferred at-home HPV self-collection, followed by in-clinic speculum collection (34%), with in-clinic self-collection least preferred (20%).⁹ In addition, 93% of participants in the clinical trial indicated they would prefer an at-home HPV self-collection option if they knew the results were equivalent to clinician-collected specimens.

A greater proportion of women who were not up to date on screening preferred self-collection over clinician collection, compared with those who were up to date. Among women who were open to self-collection, those who were not up to date were also more likely to prefer at-home collection rather than collection in a doctor’s office. In 2021, 26% of eligible U.S. women were not up to date with screening,¹⁰ including 12 million women aged 21–49 years estimated in this study. Our findings suggest that self-collection may be an effective strategy to reach women who are overdue for screening and likely at increased risk for cervical cancer.

Women with lower education attainment and household income have lower cervical cancer screening coverage¹¹; however, we found they were more likely to not prefer HPV self-collection compared with women with higher education and income levels. Concerns about accuracy and lack of confidence in performing self-collection¹² may explain these responses.

Women who have experienced nonvoluntary vaginal intercourse or identify as nonheterosexual also have lower adherence to recommended cervical cancer screening.^13,14 Distress related to pelvic examinations and fear of stigma are commonly cited barriers.^13,14 Our findings suggest that HPV self-collection, particularly when conducted at home, may improve screening uptake in these populations.

With multiple options available for cervical cancer screening, our findings underscore the importance of shared, informed decision-making between patients and clinicians, as well as the value of offering flexible screening approaches that reflect individual preferences. Some women prefer in-person care, valuing the opportunity to discuss concerns with a trusted health care professional and address multiple health needs during the same visit.¹⁵ Women also report feeling more comfortable with self-collection when it is recommended by their clinician.^15,16 As such, patient education and clinician support will be critical to successfully implement HPV self-collection if it is included as a new option for cervical cancer screening.^12,15,16 In addition to the introduction of HPV self-collection, it is important to inform patients about the need for in-person follow-up if the screening result is positive.

This study has several limitations. First, the NSFG excludes women aged 50 years and older, although screening is recommended through at least age 65 years. Second, all data were self-reported, which may introduce recall or reporting bias. Third, the reasons behind women’s preferences were not assessed, limiting the interpretation of underlying motivations. Future studies that explore these motivations, including among older women, could further inform cervical cancer screening efforts. Fourth, the survey item described the HPV self-sampling kit as “easy to use,” which may cause bias because perceptions of usability are subjective, and not all women may share this view. However, previous studies have found that most women consider using HPV self-sampling kit to be easy and convenient.^12,17

HPV self-collection for cervical cancer screening is preferred by millions of women of reproductive age in the United States, with most favoring an at-home test. These findings provide a national baseline of women’s preferences for HPV self-collection and identify subpopulations that may benefit most from this option. Screening programs could use this evidence to inform future implementation strategies and improve access to cervical cancer screening.