Abstract
IMPORTANCE
Black and Hispanic or Latina women are more likely than White women to receive a diagnosis of and to die of cervical cancer. Health insurance coverage is associated with diagnosis at an earlier stage of cervical cancer.
OBJECTIVE
To evaluate the extent to which racial and ethnic differences in the diagnosis of advanced-stage cervical cancer are mediated by insurance status.
DESIGN, SETTING, AND PARTICIPANTS
This retrospective, cross-sectional population-based study used data from the Surveillance, Epidemiology, and End Results (SEER) program on an analytic cohort of 23 942 women aged 21 to 64 years who received a diagnosis of cervical cancer between January 1, 2007, and December 31, 2016. Statistical analysis was performed from February 24, 2022, to January 18, 2023.
EXPOSURES
Health inusurance status (private or Medicare insurance vs Medicaid or uninsured).
MAIN OUTCOMES AND MEASURES
The primary outcome was a diagnosis of advanced-stage cervical cancer (regional or distant stage). Mediation analyses were performed to assess the proportion of observed racial and ethnic differences in the stage at diagnosis that were mediated by health insurance status.
RESULTS
A total of 23 942 women (median age at diagnosis, 45 years [IQR, 37–54 years]; 12.9% were Black, 24.5% were Hispanic or Latina, and 52.9% were White) were included in the study. A total of 59.4% of the cohort had private or Medicare insurance. Compared with White women, patients of all other racial and ethnic groups had a lower proportion with a diagnosis of early-stage cervical cancer (localized) (American Indian or Alaska Native, 48.7%; Asian or Pacific Islander, 49.9%; Black, 41.7%; Hispanic or Latina, 51.6%; and White, 53.3%). A larger proportion of women with private or Medicare insurance compared with women with Medicaid or uninsured received a diagnosis of an early-stage cancer (57.8% [8082 of 13 964] vs 41.1% [3916 of 9528]). In models adjusting for age, year of diagnosis, histologic type, area-level socioeconomic status, and insurance status, Black women had higher odds of receiving a diagnosis of advanced-stage cervical cancer compared with White women (odds ratio, 1.18 [95% CI, 1.08–1.29]). Health insurance was associated with mediation of more than half (ranging from 51.3% [95% CI, 51.0%−51.6%] for Black women to 55.1% [95% CI, 53.9%−56.3%] for Hispanic or Latina women) the racial and ethnic inequities in the diagnosis of advanced-stage cervical cancer across all racial and ethnic minority groups compared with White women.
CONCLUSIONS AND RELEVANCE
This cross-sectional study of SEER data suggests that insurance status was a substantial mediator of racial and ethnic inequities in advanced-stage cervical cancer diagnoses. Expanding access to care and improving the quality of services rendered for uninsured patients and those covered by Medicaid may mitigate the known inequities in cervical cancer diagnosis and related outcomes.
Introduction
Cervical cancer incidence and mortality have steadily decreased since the introduction of cervical cytology (Papanicolaou tests) for screening.1 Although the main effect of screening is identification and treatment of cervical precancerous lesions, screening also identifies cancers. Individuals who receive a diagnosis of early-stage (localized) cervical cancer have over a 90% 5-year survival rate, while those who receive a diagnosis of more advanced-stage cancers have much lower 5-year survival rates (59% for regional cancers and 17% for distant or metastatic cancers).1 Despite advances in prevention, screening, and treatment of cervical cancer over the past few decades, Black, Hispanic or Latina, and American Indian or Alaska Native women in the US are more likely to receive a diagnosis of cervical cancer and more likely to receive a diagnosis of a more advanced stage of cervical cancer than White women.2–8 Furthermore, Black women are more likely than White women to receive inadequate treatment for and die of cervical cancer.2,6,9
Previous studies have found that cervical cancer outcomes, including stage at diagnosis and survival, are associated with insurance status.2,9,10 However, these studies did not evaluate insurance status as a mediator for more advanced cervical cancer stage at diagnosis between different racial and ethnic groups. Insurance status, treated as a mediator, can help quantify and explain the association of various racial and ethnic groups with diagnosis of advanced-stage cervical cancer.11 The objective of this study was to evaluate the extent to which health insurance coverage mediated advanced-stage cervical cancer diagnosis among various racial and ethnic groups in the US.
Methods
Selection and Description of Study Participants
This study is a retrospective, population-based, cross-sectional study that used data from the Surveillance, Epidemiology, and End Results (SEER) Program on patients who received a diagnosis of cervical cancer between January 1, 2007, and December 31, 2016 (the most recent year available). For the purpose of this study, we used the SEER Census Tract-level Socioeconomic Status (SES) and Rurality Database.12 The SEER program is funded by the National Institutes of Health National Cancer Institute and includes population-based cancer incidence data. The cancer diagnoses were identified using the International Classification of Diseases for Oncology, Third Edition (eTable 1 in Supplement 1). This specific database includes demographic information that allows for socioeconomic quintiles to be calculated, rurality, and clinical information, including a broader definition of staging (localized, regional, and distant), the American Joint Committee on Cancer (AJCC) stage, and treatment modalities. The University of Illinois at Chicago institutional review board reviewed this study and waived approval and informed consent, determining that it did not involve human participants research. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.
Women aged 21 to 64 years with a diagnosis of localized, regional, or distant cervical cancer were included in our study (n = 24 945). Our analytic cohort excluded women with an unknown exposure variable (unknown or other race and ethnicity [n = 246]), a mediator variable (unknown health insurance status [n = 909]), and/or an outcome variable (unknown stage of cervical cancer [n = 630]). We also excluded those with a cervical cancer diagnosis by autopsy, per death certificate only, and/or nonmicroscopically confirmed cases. We also excluded patients with a diagnosis of multiple cancers and a histologic subtype suggestive of another primary site (eTable 2 in Supplement 1). Our final analytic cohort included 23 492 individuals.
Data Collection
Demographic information, including race and ethnicity, was collected at the time of cervical cancer diagnosis from SEER registry data. Our primary exposure of interest was race and ethnicity, which was coded as Hispanic or Latina (all races), non-Hispanic American Indian or Alaska Native (hereafter American Indian or Alaska Native), non-Hispanic Asian or Pacific Islander (hereafter Asian or Pacific Islander), non-Hispanic Black (hereafter Black), and non-Hispanic White (hereafter White). We also collected information on SEER summary diagnosis stage (localized, regional, or distant), AJCC stage, treatment modalities received (surgery, chemotherapy, and radiotherapy), and histologic subtype (squamous cell carcinoma, adenocarcinoma, adenosquamous carcinoma, and other) (eTable 1 in Supplement 1). Area-level variables included time-dependent SES (Yost SES-quintile, census tract-level) and rurality (US Department of Agriculture’s Rural Urban Commuting Area codes, county-level), which were collected at the time of diagnosis and based on the patient’s address.12
The outcome variable of interest for this study was advanced-stage cervical cancer, which we coded dichotomously based on the SEER summary stage (early-stage: localized SEER summary stage; advanced-stage: regional and distant SEER summary stage). Regional and distant summary stages were combined due to the large descrease in survivability.1,6
The mediator variable of interest was insurance status. The insurance recode variable, a SEER defined variable, was used. We specifically collected information and coded for uninsured individuals, those with Medicaid (which includes Indian or Public Health service), and insured individuals (which includes private insurance, Medicare, and military coverage).13 We were unable to ascertain duration of Medicaid coverage in the SEER data; thus, we combined uninsured women and those with Medicaid coverage at the time of diagnosis into 1 group for the purpose of analysis. This is important because there are programs that provide insurance coverage at the time of cancer diagnosis. For example, the Centers for Disease Control and Prevention National Breast and Cervical Cancer Early Detection Program provides Medicaid coverage for cancer treatment to uninsured women with a diagnosis of cervical cancer.14
Statistical Analysis
Statistical analysis was performed from February 24, 2022, to January 18, 2023. We compared the demographic and clinical characteristics of patients by disease stage and race and ethnicity using descriptive statistics. To test for differences between groups, we used the Wilcoxon rank-sum test for median values, Z-test for proportions, and the χ2 test for categorical variables. Similar to previous reports using mediation methods to investigate cancer health inequities,15,16 we applied the product method approach proposed by Baron and Kenny17 and VanderWeele.18,19 In brief, we estimated the presence of mediation and direct and indirect effects using a single mediator model; this was accomplished through a series of regressions by conducting univariable and multivariable analyses of the primary exposure (ie, race and ethnicity) and a priori measured confounders (age, year of diagnosis, and histologic subtype) on the primary outcome variable of advanced-stage cervical cancer (ie, regional or distant stage cervical cancer). We then separately conducted univariable analysis and multivariable analysis, regressing the mediator of interest (ie, health insurance status) on race and ethnicity and a priori measured confounders (age, year of diagnosis, and histologic subtype). With these calculations, the natural direct effects, natural indirect effects, and total effects were estimated.19 Next, we calculated the extent that the exposure-outcome association was associated with the mediator by conducting a proportional mediated calculation using the PARAMED module in Stata,20 release 17 (StataCorp LLC). From the PARAMED module, we used the calculated coefficients from the logistic regression model to define the outcome variables and the calculated coefficients from the logistic regression model to characterize the mediator variables. Interaction from area-level SES with insurance status was accounted for by including this variable in the analysis. We calculated the variance and constructed the 95% CI estimates using the bootstrap method to account for the product method estimator being nonnormal.21,22 Finally, sensitivity analyses were performed to evaluate the proportion of inequity mediated by uninsured status compared with insured status, excluding patients who had Medicaid coverage, and patients with Medicaid insurance coverage compared with privately insured and Medicare-insured patients, excluding uninsured patients. All P values were from 2-sided tests, and results were deemed statistically significant at P < .05.
Results
Our study included 23 492 women with a diagnosis of either localized, regional, or distant cervical cancer (Table 1). The median age at diagnosis for all stages of cancer was 45 years (IQR, 37–53 years). More than half (52.9%) the population was identified as White, 12.8% as Black, 0.8% as American Indian or Alaska Native, 9.0% as Asian or Pacific Islander, and 24.5% as Hispanic or Latina. Approximately two-thirds (67.1%) had a diagnosis of squamous cell carcinoma, 26.2% had a diagnosis of adenocarcinoma, 4.0% had a diagnosis of adenosquamous carcinoma, and 2.8% had a diagnosis of other types of cancer. Most women (59.4%) had insurance, while 8.0% were uninsured and 32.5% were insured by Medicaid. Individuals with private or Medicare insurance compared with those either without insurance or with Medicaid insurance were more likely to be diagnosed with an early-stage (localized) cancer (57.8% [8082 of 13 964] vs 41.1% [3916 of 9528]; P < .001).
Table 1.
Descriptive Characteristics of Patients With Cervical Cancer by Surveillance, Epidemiology, and End Results Program Summary Stage
| Characteristic | Women, No. (%) All (N = 23 492) | P valuea | ||
|---|---|---|---|---|
| Women, No. (%) All (N = 23 492) | Localized cancer (n = 11 998) | Regional and distant cancer (n = 11 494) | ||
| Age, y | ||||
| Median (IQR) | 45 (37–53) | 42 (35–50) | 48 (40–56) | <.001 |
| <30 | 1565 (6.7) | 1066 (8.9) | 499 (4.3) | <.001 |
| 30–49 | 13 601 (57.9) | 7891 (65.8) | 5710 (49.7) | |
| 50–64 | 8326 (35.4) | 3041 (25.3) | 5285 (46.0) | |
| Year of diagnosis | ||||
| 2007–2009 | 7188 (30.6) | 3789 (31.6) | 3399 (29.6) | .003 |
| 2010–2012 | 6940 (29.5) | 3519 (29.3) | 3421 (29.8) | |
| 2013–2016 | 9364 (39.9) | 4690 (39.1) | 4674 (40.7) | |
| Histologic subtype | ||||
| Squamous | 15 760 (67.1) | 7252 (60.4) | 8508 (74.0) | <.001 |
| Adenocarcinoma | 6156 (26.2) | 4062 (33.9) | 2094 (18.2) | |
| Adenosquamous | 929 (4.0) | 446 (3.7) | 483 (4.2) | |
| Other | 647 (2.8) | 238 (2.0) | 409 (3.6) | |
| Race and ethnicity | ||||
| American Indian or Alaska Native | 187 (0.8) | 91 (0.8) | 96 (0.8) | <.001 |
| Asian or Pacific Islander | 2108 (9.0) | 1052 (8.8) | 1056 (9.2) | |
| Black | 3017 (12.8) | 1258 (10.5) | 1759 (15.3) | |
| Hispanic or Latina | 5745 (24.5) | 2966 (24.7) | 2779 (24.2) | |
| White | 12 435 (52.9) | 6631 (55.3) | 5804 (50.5) | |
| Surgery | ||||
| Yes | 14 484 (61.7) | 10 758 (89.7) | 3726 (32.4) | <.001 |
| No | 9008 (38.3) | 1240 (10.3) | 7768 (67.6) | |
| Radiotherapy | ||||
| Yes | 12 597 (53.6) | 2959 (24.7) | 9638 (83.9) | <.001 |
| No or unknown | 10 895 (46.4) | 9039 (75.3) | 1856 (16.1) | |
| Chemotherapy | ||||
| Yes | 11 605 (49.4) | 2184 (18.2) | 9421 (82.0) | <.001 |
| No or unknown | 11 887 (50.6) | 9814 (81.8) | 2073 (18.0) | |
| Insurance status | ||||
| Uninsured | 1882 (8.0) | 751 (6.3) | 1131 (9.8) | <.001 |
| Any Medicaid | 7646 (32.5) | 3165 (26.4) | 4481 (39.0) | |
| Insured | 13 964 (59.4) | 8082 (67.4) | 5882 (51.2) | |
| Marital status | ||||
| Married | 10 508 (44.7) | 5870 (48.9) | 4638 (40.4) | <.001 |
| Not married | 11 941 (50.8) | 5530 (46.1) | 6411 (55.8) | |
| Unknown | 1043 (4.4) | 598 (5.0) | 445 (3.9) | |
| Yost SES quintile | ||||
| First (lowest SES) | 5852 (24.9) | 2652 (22.1) | 3200 (27.8) | <.001 |
| Second | 5007 (21.3) | 2414 (20.1) | 2593 (22.6) | |
| Third | 4328 (18.4) | 2204 (18.4) | 2124 (18.5) | |
| Fourth | 3899 (16.6) | 2192 (18.3) | 1707 (14.9) | |
| Fifth (highest SES) | 3159 (13.4) | 1875 (15.6) | 1284 (11.2) | |
| Unknown | 1247 (5.3) | 661 (5.5) | 586 (5.1) | |
| RUCA category | ||||
| Urban | 20 909 (89.0) | 10 725 (89.4) | 10 184 (88.6) | .03 |
| Rural | 1619 (6.9) | 775 (6.5) | 844 (7.3) | |
| Unknown | 964 (4.1) | 498 (4.2) | 466 (4.1) | |
Abbreviations: RUCA, rural-urban continuum code; SES, socioeconomic status.
To test for differences between groups, we used the Wilcoxon rank sum test for median values and the χ2 test for categorical variables.
A lower proportion of Black (41.7%), American Indian or Alaska Native (48.7%), Asian or Pacific Islander (49.9%), and Hispanic or Latina (51.6%) women received a diagnosis of early-stage cervical cancer (localized) compared with White Women (53.3%) (P < .001) (Table 2). In addition, more White women had private or Medicare insurance at the time of diagnosis (69.4%) compared with all the other racial and ethnic groups in the study (Black, 48.2%; American Indian or Alaska Native, 48.7%; Asian or Pacific Islander, 64.3%; and Hispanic or Latina, 42.5%; P < .001).
Table 2.
Descriptive Characteristics of Patients With Cervical Cancer by Race and Ethnicity
| Characteristic | Women, No. (%) | P valuea | ||||
|---|---|---|---|---|---|---|
| Hispanic or Latina (n = 5745) | American Indian or Alaska Native (n = 187) | Asian or Pacific Islander (n = 2108) | Black (n = 3017) | White (n = 12 435) | ||
| Age, y | ||||||
| Median (IQR) | 43 (36–51) | 44 (37–51) | 47 (39–55) | 46 (38–54) | 45 (37–54) | <.001 |
| <30 | 444 (7.7) | 10 (5.3) | 83 (3.9) | 201 (6.7) | 827 (6.7) | |
| 30–49 | 3624 (63.1) | 123 (65.8) | 1161 (55.1) | 1611 (53.4) | 7082 (57.0) | |
| 50–64 | 1677 (29.2) | 54 (28.9) | 864 (41.0) | 1205 (39.9) | 4526 (36.4) | |
| Year of diagnosis | ||||||
| 2007–2009 | 1743 (30.3) | 46 (24.6) | 602 (28.6) | 946 (31.4) | 3851 (31.0) | .04 |
| 2010–2012 | 1660 (28.9) | 62 (33.2) | 611 (29.0) | 905 (30.0) | 302 (2.4) | |
| 2013–2016 | 2342 (40.8) | 79 (42.2) | 895 (42.5) | 1166 (38.6) | 4882 (39.3) | |
| Histologic subtype | ||||||
| Squamous | 3878 (67.5) | 131 (70.1) | 1335 (63.3) | 2465 (81.7) | 7951 (63.9) | <.001 |
| Adenocarcinoma | 1471 (25.6) | 44 (23.5) | 599 (28.4) | 363 (12.0) | 3679 (29.6) | |
| Adenosquamous | 238 (4.1) | 9 (4.8) | 112 (5.3) | 92 (3.0) | 478 (3.8) | |
| Other | 158 (2.8) | 3 (1.6) | 62 (2.9) | 97 (3.2) | 327 (2.6) | |
| SEER summary stage | ||||||
| Localized | 2966 (51.6) | 91 (48.7) | 1052 (49.9) | 1258 (41.7) | 6631 (53.3) | <.001 |
| Regional | 2122 (36.9) | 67 (35.8) | 801 (38.0) | 1258 (41.7) | 4169 (33.5) | |
| Distant | 657 (11.4) | 29 (15.5) | 255 (12.1) | 501 (16.6) | 1635 (13.1) | |
| Surgery | ||||||
| Yes | 3546 (61.7) | 111 (59.4) | 1351 (64.1) | 1469 (48.7) | 8007 (64.4) | <.001 |
| No | 2199 (38.3) | 76 (40.6) | 757 (35.9) | 1548 (51.3) | 4428 (35.6) | |
| Radiotherapy | ||||||
| Yes | 3063 (53.3) | 102 (54.5) | 1090 (51.7) | 1890 (62.6) | 6452 (51.9) | <.001 |
| No or unknown | 2682 (46.7) | 85 (45.5) | 1018 (48.3) | 1127 (37.4) | 5983 (48.1) | |
| Chemotherapy | ||||||
| Yes | 2818 (49.1) | 102 (54.5) | 1023 (48.5) | 1707 (56.6) | 5955 (47.9) | <.001 |
| No or unknown | 2927 (50.9) | 85 (45.5) | 1085 (51.5) | 1310 (43.4) | 6480 (52.1) | |
| Insurance status | ||||||
| Uninsured | 700 (12.2) | 3 (1.6) | 116 (5.5) | 341 (11.3) | 722 (5.8) | <.001 |
| Any Medicaid | 2606 (45.4) | 93 (49.7) | 637 (30.2) | 1223 (40.5) | 3087 (24.8) | |
| Insured | 2439 (42.5) | 91 (48.7) | 1355 (64.3) | 1453 (48.2) | 8626 (69.4) | |
| Marital status | ||||||
| Married | 2440 (42.5) | 61 (32.6) | 1249 (59.3) | 701 (23.2) | 6057 (48.7) | <.001 |
| Not married | 3065 (53.4) | 118 (63.1) | 788 (37.4) | 2162 (71.7) | 5808 (46.7) | |
| Unknown | 240 (4.2) | 8 (4.3) | 71 (3.4) | 154 (5.1) | 570 (4.6) | |
| Yost SES quintile | ||||||
| First (lowest SES) | 1917 (33.4) | 40 (21.4) | 269 (12.8) | 1557 (51.6) | 2069 (16.6) | <.001 |
| Second | 1476 (25.7) | 52 (27.8) | 336 (15.9) | 622 (20.6) | 2521 (20.3) | |
| Third | 984 (17.1) | 33 (17.6) | 388 (18.4) | 379 (12.6) | 2544 (20.5) | |
| Fourth | 686 (11.9) | 17 (9.1) | 531 (25.2) | 243 (8.1) | 2422 (19.5) | |
| Fifth (highest SES) | 413 (7.2) | 7 (3.7) | 491 (23.3) | 104 (3.4) | 2144 (17.2) | |
| Unknown | 269 (4.7) | 38 (20.3) | 93 (4.4) | 112 (3.7) | 735 (5.9) | |
| RUCA category | ||||||
| Urban | 5400 (94.0) | 120 (64.2) | 1998 (94.8) | 2792 (92.5) | 10 599 (85.2) | <.001 |
| Rural | 134 (2.3) | 29 (15.5) | 56 (2.7) | 151 (5.0) | 1249 (10.0) | |
| Unknown | 211 (3.7) | 38 (20.3) | 54 (2.6) | 74 (2.5) | 587 (4.7) | |
Abbreviations: RUCA, rural-urban continuum code; SEER, Surveillance, Epidemiology, and End Results Program; SES, socioeconomic status.
To test for differences between groups, we used the Wilcoxon rank sum test for median values and the χ2 test for categorical variables.
A larger proportion of White women than Black women underwent surgery for early-stage cervical cancer (localized) (91.1% vs 82.8%; P < .001) (eTable 3 in Supplement 1). This trend continued for more advanced-stage (regional or distant) cancers, with 33.9% of White women and 24.3% of Black women undergoing surgery (P < .001).
In the multivariable regression analysis, we found that Black women and Hispanic or Latina women had a higher odds of diagnosis with advanced-stage cervical cancer compared with White women, after adjustment for age, year of diagnosis, and histologic subtype (Black: adjusted odds ratio [aOR], 1.39 [95% CI, 1.27–1.51]; Hispanic or Latina: OR, 1.13 [95% CI, 1.06–1.21]) (Table 3). These associations were attenuated when the models included area-level SES and insurance status for Black women (aOR, 1.18 [95% CI, 1.08–1.29]) and were reversed for Hispanic or Latina women (aOR, 0.93 [95% CI, 0.87–0.99]). In our second analysis, after adjusting for age, year of diagnosis, and area-level SES, all racial and ethnic groups had statistically significant increased odds of being uninsured or having Medicaid insurance at the time of cervical cancer diagnosis compared with White women (Black: aOR, 1.64 [95% CI, 1.50–1.79]; American Indian or Alaska Native: aOR, 1.98 [95% CI, 1.48–2.67]; Asian or Pacific Islander: aOR, 1.44 [95% CI, 1.30–1.59]; and Hispanic or Latina: aOR, 2.54 [95% CI, 2.37–2.71]) (Table 4).
Table 3.
Multivariable Logistic Regression Analyses for Associations Between Race and Ethnicity and Diagnosis With Regional or Distant Cervical Cancer
| Race and ethnicity | Unadjusted model | Confounder-adjusted modela | Confounder- and mediator-adjusted modelb | |||
|---|---|---|---|---|---|---|
| OR (95% CI) | P value | OR (95% CI) | P value | OR (95% CI) | P value | |
| American Indian or Alaska Native | 1.21 (0.90–1.61) | .21 | 1.22 (0.91–1.65) | .19 | 1.07 (0.79–1.45) | .66 |
| Asian or Pacific Islander | 1.15 (1.05–1.26) | .01 | 1.07 (0.97–1.17) | .20 | 1.06 (0.96–1.17) | .25 |
| Black | 1.60 (1.47–1.73) | <.001 | 1.39 (1.27–1.51) | <.001 | 1.18 (1.08–1.29) | <.001 |
| Hispanic or Latina | 1.07 (1.01–1.14) | .03 | 1.13 (1.06–1.21) | <.001 | 0.93 (0.87–0.99) | .04 |
| White | 1 [Reference] | NA | 1 [Reference] | NA | 1 [Reference] | NA |
Abbreviations: NA, not applicable; OR, odds ratio.
Included multivariable adjustment for age, year of diagnosis, and histologic subtype.
Included multivariable adjustment for age, year of diagnosis, histologic subtype, insurance status, and analysis with area-level socioeconomic status as an interaction variable.
Table 4.
Multivariable Logistic Regression Analyses for Associations Between Race and Ethnicity and Being Uninsured or Having Medicaid Coverage
| Race and ethnicity | Unadjusted model | Confounder-adjusted modela | Confounder- and mediator-adjusted modelb | |||
|---|---|---|---|---|---|---|
| OR (95% CI) | P value | OR (95% CI) | P value | OR (95% CI) | P value | |
| American Indian or Alaska Native | 2.39 (1.79–3.19) | <.001 | 2.41 (1.80–3.21) | <.001 | 1.98 (1.48–2.67) | <.001 |
| Asian or Pacific Islander | 1.26 (1.14–1.39) | <.001 | 1.25 (1.14–1.38) | <.001 | 1.44 (1.30–1.59) | <.001 |
| Black | 2.44 (2.25–2.64) | <.001 | 2.43 (2.24–2.64) | <.001 | 1.64 (1.50–1.79) | <.001 |
| Hispanic or Latina | 3.07 (2.88–3.27) | <.001 | 3.10 (2.90–3.31) | <.001 | 2.54 (2.37–2.71) | <.001 |
| White | 1 [Reference] | NA | 1 [Reference] | NA | 1 [Reference] | NA |
Abbreviations: NA, not applicable; OR, odds ratio.
Included multivariable adjustment for age and year of diagnosis.
Included multivariable adjustment for age, year of diagnosis, and analysis with area-level socioeconomic status as an interaction variable.
In the mediation analysis, we found that being uninsured or insured by Medicaid accounted for more than half (ranging from 51.3% [95% CI, 51.0%−51.6%] for Black women to 55.1% [95% CI, 53.9%−56.3%] for Hispanic or Latina women) the estimated inequity in advanced-stage (regional or distant) cervical cancer diagnoses across racial and ethnic groups compared with White women (Table 5). Sensitivity analyses evaluating uninsured vs insured status and Medicaid insurance vs private or Medicare insurance status found similar proportions mediated across all racial and ethnic groups compared with White women (eTable 4 in Supplement 1).
Table 5.
Multivariable Product Method Estimates for Proportion of Stage-Related Inequities Mediated by Insurance Status Across Racial and Ethnic Groupsa
| Race and ethnicity | % Mediated (95% CI) |
|---|---|
| American Indian or Alaska Native | 52.5 (51.9–53.2) |
| Asian or Pacific Islander | 53.8 (52.9–54.7) |
| Black | 51.3 (51.0–51.6) |
| Hispanic or Latina | 55.1 (53.9–56.3) |
| White | [Reference] |
Estimates including area-level socioeconomic status as an interaction term were similar to our primary approach; results presented here incorporate area-level socioeconomic status as an interaction variable of the observed racial and ethnic inequities.
Discussion
Our study confirmed previous findings that, compared with White women, Black women, as well as women from all other racial and ethnic groups included in the study, were more likely to receive a diagnosis of advanced-stage cervical cancer.23–28 We also found that insurance status mediated more than half of the inequities in the diagnosis of advanced-stage cervical cancer among all other racial and ethnic groups compared with White women. This finding could explain known differences in the stage of diagnosis for different racial and ethnic groups. Although previous studies have found inequities associated with health insurance and stage of cervical cancer,10,28,29 we quantify the association of insurance status with the diagnosis of more advanced-stage cancer by using population-based data and formally treating insurance status as a mediator. Health insurance status is a modifiable risk factor that could reduce these persistent cervical cancer mortality inequities among Black individuals.30 Extrapolating our results, even among populations that do not possess differences in cervical cancer mortality, far fewer individuals of all races and ethnicities would receive a diagnosis of advanced-stage cancer and, thus, would have a higher 5-year survival rate.
Despite the Patient Protection and Affordable Care Act and programs such as the Centers for Disease Control and Prevention’s National Breast and Cervical Cancer Early Detection Program, there is an increasing proportion of the population that has not received up-to-date cervical cancer screening services.31,32 Uninsured individuals and those with Medicaid insurance are the most likely to be underscreened.31–33 Even if these populations receive screening, both uninsured individuals and individuals with Medicaid are more likely to experience delays in accessing the following diagnostic procedures necessary to diagnose actual cancer.28,34–36 This lack of access to initial screening and barriers to follow-up care could explain why, in racial and ethnic minority groups, women who are uninsured or covered by Medicaid present with more cases of advanced-stage cervical cancer.
These findings are important in the context of previous studies investigating the survivability of patients from racial and ethnic groups after a diagnosis of cervical cancer. One study found that insurance status mediated less than 19% of the association with excess mortality among Black women compared with White women.2 That study also found that a more important mediator of survivability was the stage of diagnosis. In context with our findings, if private or Medicare insurance coverage was expanded or if Medicaid eased the administrative burden physicians face,37 far fewer women from racial and ethnic minority groups would receive a diagnosis of advanced-stage cancer and, thus, would have a higher survivability.
Although this study found that insurance status is a significant mediator of the racial and ethnic inequities in advanced-stage cervical cancer diagnosis, there are numerous additional modifiable factors that play a major role in the prevention of cancer. Societal, community, and individual factors are associated with differential uptake of cervical cancer screening services and affect the ability to follow up after an abnormal test result.36 Stigma related to human papillomavirus infection or inability to receive screening can prevent further follow-up care,38 which can result in more advanced-stage cancer diagnoses. Also, human papillomavirus vaccination is an effective primary prevention intervention, and efforts should focus on ensuring that all eligible populations are vaccinated.39
Nevertheless, this study adds to the growing literature that highlights the poor outcomes for individuals who are underinsured.2,9,15,40 Future policy changes need to focus on expanding access to care and improving the quality of services rendered for uninsured patients and those covered by Medicaid to specifically address this cancer inequity for historically marginalized racial and ethnic groups. Furthermore, to reduce inequities in the incidence of cervical cancer and in mortality and achieve the World Health Organization’s mandate to eliminate the disease, a multimodal approach is necessary to address all modifiable factors. Adopting a multifactorial approach that integrates clinical and financial care navigation to improve access to health care services,41 eliminates racism in medicine or health care,42–44 and adapts a socioecological model to address social determinants of health that prevent access to health care44,45 will be necessary to fully address these cervical cancer inequities.
Limitations
This study has several limitations. First, the population sampled for our analysis may not be generalizable to the whole population of the US.46 Second, insurance information may be flawed, especially as patients who initially received a diagnosis of cervical cancer may have enrolled in Medicaid by the time they were enrolled in the registry data,47,48 potentially underestimating the effectiveness of Medicaid insurance coverage at preventing advanced-stage cervical cancer. In addition, SEER only reports insurance status data up to 2016; thus, more recent analysis investigating insurance status and disease stage at diagnosis of cervical cancer is not possible at this time.13 Third, while SEER has AJCC staging information, there is a higher frequency of missing AJCC stage data. In comparison, SEER has much more robust data for the 3 summary stages: localized, regional, and distant. This may limit the specificity of our findings due to the large grouping of the AJCC and International Federation of Gynecology and Obstetrics stages within these 3 categories. Fourth, SEER lacks data regarding cervical precancers; this limits our ability to conduct further analysis comparing precancer and early-stage cancer. Future studies could consider evaluating insurance status and odds of precancerous diagnosis compared with cancer diagnosis. Fifth, the analysis was able to account for only area-level SES; future studies using individual-level SES could provide more insight into the odds of developing advanced-stage cancer. Sixth, SEER data on race and ethnicity are coded using their underlying cancer registry, which has a history of misclassifications of race and ethnicity compared with the criterion standard of self-report.49,50 Future efforts to improve the accuracy of the information we collect, especially regarding race and ethnicity, is critical for successful health equity research.51
Conclusions
Our study found that insurance status mediated more than half of the advanced-stage cervical cancers diagnosed among women from racial and ethnic minority groups. Although our findings suggest that a large proportion of the cancer inequities was associated with insurance status, we also acknowledge that equity in insurance coverage will not eliminate cervical cancer unilaterally.
Supplementary Material
eTable 1. Inclusion and Classification of Cancer Subtype ICD-O-3
eTable 2. Exclusion ICD-O-3 Codes
eTable 3. Rates of Treatment With Surgery, Radiation, and Chemotherapy Among Women With Cervical Cancer by Race/Ethnicity and Summary Stage
eTable 4. Multivariable Product Method Estimates for Proportion of Stage-Related Inequities Mediated by Insurance Status Across Racial/Ethnic Groups
Data Sharing Statement
Key Points.
Question
How much of the racial and ethnic differences in diagnosis of advanced-stage (regional or distant stage) cervical cancer is mediated by insurance status?
Findings
This cross-sectional study of 23 942 individuals found that more than half of the diagnoses of advanced-stage cervical cancer among individuals from minority racial and ethnic groups were mediated by insurance coverage or type.
Meaning
This study suggests that insurance is a modifiable risk factor that plays an important role in the racial and ethnic inequities observed in the diagnosis of advanced-stage cervical cancer.
Role of the Funder/Sponsor:
The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Funding/Support:
Dr Holt was supported by the University of Illinois at Chicago Building Interdisciplinary Research Careers in Women’s Health grant K12HD101373 from the National Institutes of Health Office of Research on Women’s Health.
Footnotes
Conflict of Interest Disclosures: Dr Guadamuz reported being employed by Flatiron Health, an independent subsidiary of Roche, outside the submitted work. Dr Calip reported being employed by and holding stock in Flatiron Health outside the submitted work. No other disclosures were reported.
Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Data Sharing Statement: See Supplement 2.
Supplemental content
Author affiliations and article information are listed at the end of this article.
Contributor Information
Hunter K. Holt, Department of Family and Community Medicine, University of Illinois at Chicago, Chicago.
Caryn E. Peterson, Department of Epidemiology and Biostatistics, University of Illinois at Chicago, Chicago.
Shannon MacLaughlan David, Department of Obstetrics and Gynecology, University of Illinois at Chicago, Chicago.
Abdullah Abdelaziz, Department of Pharmacy Systems, Outcomes and Policy, University of Illinois at Chicago, Chicago.
George F. Sawaya, Department of Obstetrics, Gynecology, and Reproductive Sciences, University of California, San Francisco.
Jenny S. Guadamuz, Flatiron Health, New York, New York, Program on Medicines and Public Health, University of Southern California, Los Angeles.
Gregory S. Calip, Department of Pharmacy Systems, Outcomes and Policy, University of Illinois at Chicago, Chicago, Flatiron Health, New York, New York.
REFERENCES
- 1.National Cancer Institute, Surveillance, Epidemiology, and End Results Program. Cancer stat facts: cervical cancer Accessed February 17, 2022. https://seer.cancer.gov/statfacts/html/cervix.html [Google Scholar]
- 2.Markt SC, Tang T, Cronin AM, et al. Insurance status and cancer treatment mediate the association between race/ethnicity and cervical cancer survival. PLoS One 2018;13(2):e0193047. doi: 10.1371/journal.pone.0193047 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Singh GK. Rural-urban trends and patterns in cervical cancer mortality, incidence, stage, and survival in the United States, 1950–2008. J Community Health 2012;37(1):217–223. doi: 10.1007/s10900-011-9439-6 [DOI] [PubMed] [Google Scholar]
- 4.Beavis AL, Gravitt PE, Rositch AF. Hysterectomy-corrected cervical cancer mortality rates reveal a larger racial disparity in the United States. Cancer 2017;123(6):1044–1050. doi: 10.1002/cncr.30507 [DOI] [PubMed] [Google Scholar]
- 5.Yoo W, Kim S, Huh WK, et al. Recent trends in racial and regional disparities in cervical cancer incidence and mortality in United States. PLoS One 2017;12(2):e0172548. doi: 10.1371/journal.pone.0172548 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Brookfield KF, Cheung MC, Lucci J, Fleming LE, Koniaris LG. Disparities in survival among women with invasive cervical cancer: a problem of access to care. Cancer 2009;115(1):166–178. doi: 10.1002/cncr.24007 [DOI] [PubMed] [Google Scholar]
- 7.Islami F, Fedewa SA, Jemal A. Trends in cervical cancer incidence rates by age, race/ethnicity, histological subtype, and stage at diagnosis in the United States. Prev Med 2019;123:316–323. doi: 10.1016/j.ypmed.2019.04.010 [DOI] [PubMed] [Google Scholar]
- 8.National Cancer Institute, Surveillance, Epidemiology, and End Results Program. SEER*Explorer: an interactive website for SEER cancer statistics Accessed October 11, 2022. https://seer.cancer.gov/statistics-network/explorer/ [Google Scholar]
- 9.Churilla T, Egleston B, Dong Y, et al. Disparities in the management and outcome of cervical cancer in the United States according to health insurance status. Gynecol Oncol 2016;141(3):516–523. doi: 10.1016/j.ygyno.2016.03.025 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Davis M, Strickland K, Easter SR, et al. The impact of health insurance status on the stage of cervical cancer diagnosis at a tertiary care center in Massachusetts. Gynecol Oncol 2018;150(1):67–72. doi: 10.1016/j.ygyno.2018.05.002 [DOI] [PubMed] [Google Scholar]
- 11.Farland LV, Correia KFB, Dodge LE, et al. The importance of mediation in reproductive health studies. Hum Reprod 2020;35(6):1262–1266. doi: 10.1093/humrep/deaa064 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.National Cancer Institute, Surveillance, Epidemiology, and End Results Program. Census Tract-level SES and Rurality Database (2006–2018) Accessed July 14, 2022. https://seer.cancer.gov/seerstat/databases/census-tract/index.html [Google Scholar]
- 13.National Cancer Institute, Surveillance, Epidemiology, and End Results Program. Insurance recode (2007+) Accessed January 11, 2023. https://seer.cancer.gov/seerstat/variables/seer/insurance-recode/ [Google Scholar]
- 14.Centers for Disease Control and Prevention. National Breast and Cervical Cancer Early Detection Program (NBCCEDP) Updated February 15, 2022. Accessed June 23, 2022. https://www.cdc.gov/cancer/nbccedp/index.htm [Google Scholar]
- 15.Ko NY, Hong S, Winn RA, Calip GS. Association of insurance status and racial disparities with the detection of early-stage breast cancer. JAMA Oncol 2020;6(3):385–392. doi: 10.1001/jamaoncol.2019.5672 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Guadamuz JS, Ozenberger K, Qato DM, et al. Mediation analyses of socioeconomic factors determining racial differences in the treatment of diffuse large B-cell lymphoma in a cohort of older adults. Medicine (Baltimore) 2019;98(46):e17960. doi: 10.1097/MD.0000000000017960 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Baron RM, Kenny DA. The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. J Pers Soc Psychol 1986;51(6):1173–1182. doi: 10.1037/0022-3514.51.6.1173 [DOI] [PubMed] [Google Scholar]
- 18.VanderWeele TJ. Mediation and mechanism. Eur J Epidemiol 2009;24(5):217–224. doi: 10.1007/s10654-009-9331-1 [DOI] [PubMed] [Google Scholar]
- 19.VanderWeele TJ. Mediation analysis: a practitioner’s guide. Annu Rev Public Health 2016;37(1):17–32. doi: 10.1146/annurev-publhealth-032315-021402 [DOI] [PubMed] [Google Scholar]
- 20.Emsley R, Liu H. PARAMED: Stata module to perform causal mediation analysis using parametric regression models 2013. Accessed February 24, 2022. https://econpapers.repec.org/software/bocbocode/s457581.htm [Google Scholar]
- 21.Nevo D, Liao X, Spiegelman D. Estimation and inference for the mediation proportion. Int J Biostat 2017;13 (2):1–31. doi: 10.1515/ijb-2017-0006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.VanderWeele TJ. Explanation in Causal Inference: Methods for Mediation and Interaction Oxford University Press; 2015. [Google Scholar]
- 23.Newmann SJ, Garner EO. Social inequities along the cervical cancer continuum: a structured review. Cancer Causes Control 2005;16(1):63–70. doi: 10.1007/s10552-004-1290-y [DOI] [PubMed] [Google Scholar]
- 24.Coker AL, Desimone CP, Eggleston KS, White AL, Williams M. Ethnic disparities in cervical cancer survival among Texas women. J Womens Health (Larchmt) 2009;18(10):1577–1583. doi: 10.1089/jwh.2008.1342 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Sheppard CS, El-Zein M, Ramanakumar AV, Ferenczy A, Franco EL. Assessment of mediators of racial disparities in cervical cancer survival in the United States. Int J Cancer 2016;138(11):2622–2630. doi: 10.1002/ijc.29996 [DOI] [PubMed] [Google Scholar]
- 26.Benard VB, Watson M, Saraiya M, et al. Cervical cancer survival in the United States by race and stage (2001–2009): findings from the CONCORD-2 study. Cancer 2017;123(suppl 24):5119–5137. doi: 10.1002/cncr.30906 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Yu L, Sabatino SA, White MC. Rural–urban and racial/ethnic disparities in invasive cervical cancer incidence in the United States, 2010–2014. Prev Chronic Dis 2019;16:E70. doi: 10.5888/pcd16.180447 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Benard VB, Jackson JE, Greek A, et al. A population study of screening history and diagnostic outcomes of women with invasive cervical cancer. Cancer Med 2021;10(12):4127–4137. doi: 10.1002/cam4.3951 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.O’Malley CD, Shema SJ, Clarke LS, Clarke CA, Perkins CI. Medicaid status and stage at diagnosis of cervical cancer. Am J Public Health 2006;96(12):2179–2185. doi: 10.2105/AJPH.2005.072553 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Robbins AS, Han X, Ward EM, Simard EP, Zheng Z, Jemal A. Association between the Affordable Care Act Dependent Coverage Expansion and cervical cancer stage and treatment in young women. JAMA 2015;314(20): 2189–2191. doi: 10.1001/jama.2015.10546 [DOI] [PubMed] [Google Scholar]
- 31.Suk R, Hong Y-R, Rajan SS, Xie Z, Zhu Y, Spencer JC. Assessment of US Preventive Services Task Force guideline–concordant cervical cancer screening rates and reasons for underscreening by age, race and ethnicity, sexual orientation, rurality, and insurance, 2005 to 2019. JAMA Netw Open 2022;5(1):e2143582. doi: 10.1001/jamanetworkopen.2021.43582 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Watson M, Benard V, King J, Crawford A, Saraiya M. National assessment of HPV and Pap tests: changes in cervical cancer screening, National Health Interview Survey. Prev Med 2017;100:243–247. doi: 10.1016/j.ypmed.2017.05.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Sabik LM, Dahman B, Vichare A, Bradley CJ. Breast and cervical cancer screening among Medicaid beneficiaries: the role of physician payment and managed care. Med Care Res Rev 2020;77(1):34–45. doi: 10.1177/1077558718771123 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Chase DM, Osann K, Sepina N, Wenzel L, Tewari KS. The challenge of follow-up in a low-income colposcopy clinic: characteristics associated with noncompliance in high-risk populations. J Low Genit Tract Dis 2012;16(4): 345–351. doi: 10.1097/LGT.0b013e318249640f [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Fish LJ, Moorman PG, Wordlaw-Stintson L, Vidal A, Smith JS, Hoyo C. Factors associated with adherence to follow-up colposcopy. Am J Health Educ 2013;44(6):293–298. doi: 10.1080/19325037.2013.838881 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Smith JS. Ethnic disparities in cervical cancer illness burden and subsequent care: a prospective view in managed care. Am J Manag Care 2008;14(6)(suppl 1):S193–S199. [PubMed] [Google Scholar]
- 37.Dunn A, Gottlieb JD, Shapiro A, Sonnenstuhl DJ, Tebaldi P. A denial a day keeps the doctor away. National Bureau of Economic Research working paper 29010 July 2021. Updated January 2023 doi: 10.3386/w29010 [DOI] [Google Scholar]
- 38.Peterson CE, Silva A, Goben AH, et al. Stigma and cervical cancer prevention: a scoping review of the U.S. literature. Prev Med 2021;153:106849. doi: 10.1016/j.ypmed.2021.106849 [DOI] [PubMed] [Google Scholar]
- 39.Lei J, Ploner A, Elfström KM, et al. HPV vaccination and the risk of invasive cervical cancer. N Engl J Med 2020;383(14):1340–1348. doi: 10.1056/NEJMoa1917338 [DOI] [PubMed] [Google Scholar]
- 40.Halpern MT, Ward EM, Pavluck AL, Schrag NM, Bian J, Chen AY. Association of insurance status and ethnicity with cancer stage at diagnosis for 12 cancer sites: a retrospective analysis. Lancet Oncol 2008;9(3):222–231. doi: 10.1016/S1470-2045(08)70032-9 [DOI] [PubMed] [Google Scholar]
- 41.Freeman HP, Wingrove BK. Excess Cervical Cancer Mortality: A Marker for Low Access to Health Care in Poor Communities National Cancer Institute, Center to Reduce Cancer Health Disparities. 2005. NIH publication; 05–5282. [Google Scholar]
- 42.Ford CL, Griffith DM, Bruce MA, Gilbert KL, eds. Racism: Science & Tools for the Public Health Professional American Public Health Association; 2019. doi: 10.2105/9780875533049 [DOI] [Google Scholar]
- 43.Ansell DA, James B, De Maio FG. A call for antiracist action. N Engl J Med 2022;387(1):e1. doi: 10.1056/NEJMp2201950 [DOI] [PubMed] [Google Scholar]
- 44.Thomas SB, Quinn SC, Butler J, Fryer CS, Garza MA. Toward a fourth generation of disparities research to achieve health equity. Annu Rev Public Health 2011;32(1):399–416. doi: 10.1146/annurev-publhealth-031210-101136 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Braveman P, Egerter S, Williams DR. The social determinants of health: coming of age. Annu Rev Public Health 2011;32(1):381–398. doi: 10.1146/annurev-publhealth-031210-101218 [DOI] [PubMed] [Google Scholar]
- 46.Kuo T-M, Mobley LR. How generalizable are the SEER registries to the cancer populations of the USA? Cancer Causes Control 2016;27(9):1117–1126. doi: 10.1007/s10552-016-0790-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Bradley CJ, Gardiner J, Given CW, Roberts C. Cancer, Medicaid enrollment, and survival disparities. Cancer 2005;103(8):1712–1718. doi: 10.1002/cncr.20954 [DOI] [PubMed] [Google Scholar]
- 48.Sabik LM, Bradley CJ. Understanding the limitations of cancer registry insurance data—implications for policy. JAMA Oncol 2018;4(10):1432–1433. doi: 10.1001/jamaoncol.2018.2436 [DOI] [PubMed] [Google Scholar]
- 49.Clarke LC, Rull RP, Ayanian JZ, et al. Validity of race, ethnicity, and national origin in population-based cancer registries and rapid case ascertainment enhanced with a Spanish surname list. Med Care 2016;54(1):e1–e8. doi: 10.1097/MLR.0b013e3182a30350 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Atekruse SF, Cosgrove C, Cronin K, Yu M. Comparing cancer registry abstracted and self-reported data on race and ethnicity. J Registry Manag 2017;44(1):30–33. [PubMed] [Google Scholar]
- 51.Chin MH. Using patient race, ethnicity, and language data to achieve health equity. J Gen Intern Med 2015;30 (6):703–705. doi: 10.1007/s11606-015-3245-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eTable 1. Inclusion and Classification of Cancer Subtype ICD-O-3
eTable 2. Exclusion ICD-O-3 Codes
eTable 3. Rates of Treatment With Surgery, Radiation, and Chemotherapy Among Women With Cervical Cancer by Race/Ethnicity and Summary Stage
eTable 4. Multivariable Product Method Estimates for Proportion of Stage-Related Inequities Mediated by Insurance Status Across Racial/Ethnic Groups
Data Sharing Statement