Cervical Cancer Screening and Incidence by Age: Unmet Needs Near and After the Stopping Age for Screening

Abstract

Introduction

Leading professional organizations recommend cervical cancer screening for average-risk women aged 21–65 years. For average-risk women aged >65 years, routine screening may be discontinued if “adequate” screening with negative results is documented. Screening is recommended after age 65 years for women who do not meet adequate prior screening criteria or are at special risk.

Methods

Authors examined the most recent cervical cancer incidence data from two federal cancer surveillance programs for all women by age and race, corrected for hysterectomy status. The 2013 and 2015 National Health Interview Surveys were analyzed in 2016 to examine the proportion of women aged 41–70 years without a hysterectomy who reported that they never had a Pap test or that their most recent Pap test was >5 years ago (not recently screened).

Results

The incidence rate for cervical cancer among older women, corrected for hysterectomy status, did not decline until age ≥85 years. The proportion not recently screened increased with age, from 12.1% for women aged 41–45 years to 18.4% for women aged 61–65 years.

Conclusions

Even among women within the recommended age range for routine screening, many are not up to date, and a substantial number of women approach the “stopping” age for cervical cancer screening without an adequate prior screening history. Efforts are needed to reach women who have not been adequately screened, including women aged >65 years, to prevent invasive cervical cancer cases and deaths among older women.

INTRODUCTION

One fifth of cervical cancer cases and one third of cervical cancer deaths occurred among women aged ≥65 years in the U.S. in 2013.¹ Cervical cancer screening is effective at preventing invasive cases and deaths from cervical cancer among older women.^2,3 Professional organizations recommend that routine cervical cancer screening be discontinued for average-risk women aged >65 years after three consecutive negative cytology results or two consecutive negative co-test results within the previous 10 years, with the most recent test performed within the past 5 years.^4–6 Substantially lower risk of cervical cancer after age 65 years was associated with adequate screening at age 50–64 years, suggesting benefits of increasing screening uptake prior to and, if needed for catch-up, after age 65 years.⁷

Data from the 2013 and 2015 National Health Interview Survey (NHIS) were used to examine cervical cytology test use as women approach the cervical cancer screening “stopping age.” Cervical cancer incidence rates were examined for all women by age, corrected by hysterectomy status, based on data from the 2013 NHIS and the latest cancer incidence data from 2013.

METHODS

Data Sample

Survey data came from the 2013 and 2015 NHIS, a cross-sectional household survey, conducted in person and representative of the civilian, non-institutionalized U.S. population (www.cdc.gov/nchs/nhis/index.htm). Information was collected about the household, each family in the household, and a randomly selected sample adult and child in each family. Response rates for the 2013 and 2015 NHIS sample adult section were 61.2% and 55.2%, respectively, for the final samples taking household nonresponse into account.^8,9 Data were combined for these 2 years to improve estimate stability.

Survey respondents were asked which group best represents their race and if they consider themselves to be Hispanic/Latino. The sample adult data file included specific race and ethnicity recodes. Participants were categorized as non-Hispanic white, non-Hispanic black/African American, American Indian/Alaska Native, Asian/Pacific Islander (Asian), or Hispanic. In 2013 and 2015, the survey asked female respondents if they ever had a hysterectomy, ever had a Pap smear or Pap test, and when their most recent Pap test was performed. These responses were used to determine the proportion of the female population aged 41–70 years who were not recently screened, defined as never having a Pap test or the last Pap test was >5 years ago.^10,11 Only women with no hysterectomy were included in the analyses for the combined years (n=12,518).

Data on the number of invasive cervical cancer cases were obtained from combined registry data from the Centers for Disease Control and Prevention’s National Program of Cancer Registries and the National Cancer Institute’s Surveillance, Epidemiology, and End Results Program for 2013.¹ All contributing cancer registries meet high data quality criteria (www.cdc.gov/cancer/npcr/uscs/technical_notes/criteria.htm).

Statistical Analysis

Weighted proportions and 95% CIs were calculated for populations that were not recently screened by age and race/ethnicity and prevalence of hysterectomy by age, using SAS, version 9.3, with SUDAAN, version 11, to adjust for the NHIS complex sampling design and the combined survey years. American Indian/Alaska Native women were included in the combined sample of all women, but estimates for American Indian/Alaska Native women were unreliable (relative SE >30%) and not reported. All analyses were conducted in 2016.

For 2013, combined registry data from the National Program of Cancer Registries and the Surveillance, Epidemiology, and End Results Program cover approximately 99% of the U.S. population. Data for specific racial or ethnic populations other than white and black were not examined because of misclassification of race or ethnicity or suppressed cell counts.¹ Following a method similar to Rositch et al.,¹² the authors used age- and race-specific estimates of hysterectomy prevalence based on data from the 2013 NHIS to adjust the age- and race-specific number-at-risk denominators to exclude women without an intact cervix. These adjusted denominators were used to calculate cervical cancer incidence rates corrected for hysterectomy with 95% CIs.

RESULTS

Hysterectomy prevalence increased with age, from about 12% for women aged 40–44 years to >40% after age 70 years (data not shown). Cervical cancer incidence rates corrected for hysterectomy increased with age until about age 70 years and then declined after age 85 years (Figure 1). The absolute magnitude of the difference between uncorrected and corrected cervical cancer incidence rates increased with age and was larger for black women than white women (Figure 1).

Figure 1.

Age-specific incidence and 95% CI of cervical cancer in the U.S., 2013, corrected and uncorrected for hysterectomy, by race.

Among women with no hysterectomy, the proportion not recently screened increased with age, from about 12% for women in their 40s to nearly 24% for women aged 66–70 years (Table 1). For most age groups, this proportion was slightly higher for Hispanic and Asian women than for other women, but the CIs were fairly wide. Among women aged 66–70 years, 5.1% (95% CI=3.7%, 6.8%) reported never being screened (data not shown).

Table 1.

Women Aged 41–70 Years Not Recently Screened for Cervical Cancer, NHIS 2013 and 2015

Age group, years	Not recently screened,a % (95% CI), n
	Total	Non-Hispanic white	Non-Hispanic black	Hispanic	Asian
41–45	12.1 (10.6, 13.8), 2,338	11.3 (9.2, 13.7), 1,279	7.1 (4.6, 10.7), 348	16.4 (12.9, 20.6), 534	14.5 (10.1, 20.5), 157
46–50	11.7 (10.2, 13.4), 2,255	10.5 (8.7, 12.7), 1,312	13.1 (9.1, 18.3), 332	14.9 (11.3, 19.4), 436	12.1 (8.1, 17.7), 154
51–55	11.3 (9.8, 12.9), 2,227	9.6 (7.9, 11.6), 1,355	13.3 (10.0, 17.6), 384	17.5 (12.7, 23.5), 344	14.0 (8.2, 22.8), 124
56–60	17.1 (14.0, 19.4), 2,175	16.5 (14.0, 19.3), 1,434	16.9 (12.4, 22.6), 326	18.9 (13.6, 25.6), 283	20.0 (12.2, 30.9), 111
61–65	18.4 (16.2, 20.8), 1,937	17.9 (15.4, 20.9), 1,326	22.4 (16.4, 29.9), 270	16.2 (11.5, 22.4), 231	22.3 (13.5, 34.7), 95
66–70	23.6 (20.8, 26.6), 1,586	22.4 (19.1, 26.1), 1,070	24.3 (17.5, 32.8), 218	28.1 (19.8, 38.1), 205	32.0 (20.9, 45.6), 79

^aWomen were identified as not recently screened if they reported never having a Pap test or if the last Pap test was >5 years ago. NHIS, National Health Interview Survey.

DISCUSSION

Before reaching age 65 years, many women are not up to date with routine cervical cancer screening, and a substantial number approach the “stopping” age without an adequate screening history. Based on the 2015 NHIS data, about 845,000 American women aged 61–65 years were not recently screened. For these women, catch-up screening may be needed after age 65 years to achieve an adequate negative screening history. Many women aged 66–70 years also may still need to be screened, especially those never screened.

A recommended upper age limit for routine screening may lead women and providers to assume that cervical cancer is a younger women’s disease.¹³ Some of the highest cervical cancer incidence rates occur among women aged >65 years, with notably higher rates among older black women. These results, representing 99% of the U.S. population, confirm those of Rositch and colleagues¹² using data from 4–13 years earlier and 13 states. Although corrected incidence rates were not calculated for Asian women, their observed lower cervical cancer screening is of concern and incidence rates may be higher for Asians than uncorrected rates suggest.¹

Limitations

These data are subject to several limitations. Self-reported data from NHIS on screening history may be inaccurate,¹⁴ error may vary by race/ethnicity, and hysterectomy status could not be confirmed. The sample may under-represent disadvantaged population groups most likely not to be recently screened. Sample sizes in specific age and race or ethnic groups were small, leading to unstable estimates. Additional women might have had a Pap test within the past 5 years, but not three negative tests within the past 10 years. Screening use among older women at special risk was not examined.

CONCLUSIONS

Premature discontinuation of routine screening among women in the years before age 65 years could contribute to preventable cases of invasive cervical cancer and deaths. Catch-up screening may be needed for underscreened women after age 65 years, taking into account the physical and psychological issues associated with cervical cancer screening decisions for older women.^13,15 In the short term, efforts could be undertaken to clarify misperceptions about the risk of cervical cancer among older women and providers. Messages about a “stopping age” should emphasize the recommendation for an adequate screening history of previous negative tests before screening is discontinued, not just chronologic age. Age itself may need to be reconsidered, given high cervical cancer incidence rates after age 65 years, increases in life expectancy, and different human papillomavirus exposures by birth cohort.¹⁶