Article Text
PDF
Abstract
Objectives Limited evidence exists on the association of Pap test utilisation with comorbidity and functional impairment among middle-aged non-Hispanic black (NHB) women in the USA. We aimed to assess whether middle-aged NHB women with a higher burden of comorbidity and functional impairment have a lower rate of Pap test utilisation.
Design Nationwide cross-sectional survey in the USA.
Setting 2018 Behavioral Risk Factor Surveillance System.
Participants 6359 middle-aged NHB women.
Exposures and outcome The primary exposures were comorbidity and functional impairment. The outcome of interest was whether a woman reported having a Pap test in the last 3 years.
Data analysis We fit unadjusted and multivariable logistic regression models to calculate ORs and 95% CI for comorbidity and functional impairment. Sensitivity analysis was restricted to women without a history of hysterectomy or cancer. We added interaction terms between exposures and age, as well as lifestyle indicators.
Results Of the 6359 women, 4141 (65.1%) had comorbidity and 2429 (38.2%) had functional impairment. Middle-aged NHB women with comorbidity (≥2 vs 0, aOR=0.72, 95% CI=0.61 to 0.85, p trend<0.01) or functional impairment (≥2 vs 0, aOR=0.69, 95% CI=0.57 to 0.83, p trend<0.01) had a lower rate of Pap test utilisation compared with healthier counterparts, regardless of histories of hysterectomy and prior cancer. The analyses for age and lifestyle indicators subgroup difference indicated no statistically significant effect (p interaction>0.05). However, the magnitude of these associations was stronger among women with adverse lifestyle factors (eg, comorbidity ≥2 v.s. 0, aOR=0.53, 95% CI=0.40, to 0.71; functional impairment ≥2 v.s. 0, aOR=0.35, 95% CI=0.16, to 0.72 among binge drinkers).
Conclusion Comorbidity or functional impairment could be a potential barrier to Pap test utilisation among middle-aged NHB women in the USA. Our study highlights the importance of implementing targeted intervention programmes and prioritised health resource allocation to promote Pap test utilisation. Cohort studies with clear temporality and indicators reflecting disease severity will be essential for further understanding this association.
- health equity
- risk factors
- epidemiology
- behavior
- chronic disease
- health services accessibility
Data availability statement
Data are available in a public, open access repository.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
STRENGTHS AND LIMITATIONS OF THIS STUDY
The composite exposure variables were constructed from nine types of comorbidities and four types of functional impairment, providing a comprehensive assessment.
In contrast to previous studies on similar topic, our multivariable models incorporated lifestyle factors, which further minimised residual confounding and enhanced the validity of the estimates.
Comorbidity and functional impairment were self-report, which may be subject to bias.
Our study does not have indicators reflecting disease control or severity.
Introduction
Racial and ethnic disparities in cervical cancer incidence and mortality have been well documented, especially for non-Hispanic black (NHB) women in the USA.1–5 Despite an overall decline in cervical cancer incidence, NHB women aged over 40 and over 50 years had a higher rate of cervical cancer in 2011–2015 compared with non-Hispanic white and Hispanic women in the same age group.6
The US Preventive Services Task Force (USPSTF) recommends cervical cytology (Pap test) every 3 years alone or in combination with a 5 yearly test for the presence of high-risk human papilloma virus (HPV) to detect precancerous cervical lesions and cervical cancer in women aged 30–65 years.7 However, screening rates are low among NHB women in this age group. For example, 3-year cervical cancer screening rate in national health plan was consistently low among women aged 30–64 years living in predominantly black neighbourhoods (>75% blacks) compared with women living in other neighbourhoods between 2001 and 2010.8 Similarly, among women aged 30–64 years eligible to receive Pap test and HPV cotesting services through the Connecticut Breast and Cervical Cancer Early Detection Program between 2013 and 2015, only 49% of NHB women were tested compared with 58%–68% of women who identified with other racial/ethnic groups.9 This race-specific disparity in cervical cancer screening highlights the need to investigate factors that may be unique among middle-aged and older women.
In the USA, NHB individuals disproportionately suffer from comorbid conditions.10–12 The risk of these conditions begins to increase in middle age (45–64 years). Previous studies suggest that comorbid conditions may serve as a barrier to screening utilisation.13 14 This implies that middle-aged NHB women with a higher rate of comorbid conditions are at an increased risk for late cervical cancer diagnosis and delayed treatment due to underutilisation of screening, despite experiencing higher rates of cervical cancer relative to other age or racial and ethnic groups.6
Moreover, disabilities often result from chronic conditions,15–17 which have also been shown to influence Pap test rates.18–20 In 2020, women with basic action (movement, sensory, emotional or cognitive) difficulties or complex action (self-care, social or work) limitations had Pap test rates 3%–20% lower than women without disability.20 This difference increased with the severity of impairment. Given that chronic conditions and attendant disabilities peak at mid-age,15 and available evidence is limited, investigating their association with cervical cancer screening among middle-aged NHB women is warranted. Thus, we aimed to assess whether middle-aged NHB women with a higher burden of comorbid conditions and functional impairment have a lower rate of Pap test utilisation.
Methods
Data source
We used data from the 2018 Behavioral Risk Factor Surveillance System (BRFSS), an annual nationwide telephone survey of residents aged 18 years and older in the USA.21 The BRFSS collects information on health-related risk behaviours, chronic health conditions and the use of preventive services. The survey includes all 50 states, the District of Columbia, Puerto Rico, the U.S. Virgin Islands, Guam, American Samoa and Palau. You could visit the website www.edc.gov/brfss/about/index.htm for comprehensive details about BRFSS. For the current study, we included a total of 6359 middle-aged NHB women aged 45–64 years at review without missing values for exposure, outcome and other covariates.
Measures
Our outcome of interest was whether a woman reported having a Pap test in the last 3 years, measured as a dichotomous variable (no vs yes). Women who underwent a Pap test within the last 3 years were considered to have had a Pap test based on the 2012 USPSTF screening guideline22 and criteria used in published articles.23 24 Pap test utilisation was assessed using the following two questions: ‘Have you ever had a Pap test?’ and ‘How long has it been since you had your last Pap test?’ Non-informative answers, including do not know, not sure and refused were treated as missing data, and statistical imputation was not performed.
Comorbidity and functional impairment were our primary exposures of interest. In the 2018 BRFSS survey, participants reported nine major comorbidities including myocardial infarction, coronary heart disease, stroke, asthma, chronic obstructive pulmonary disease, arthritis, depressive disorder, chronic kidney disease and diabetes mellitus. These comorbid conditions approximate those in the Charlson Comorbidity Index,25 which has been widely used as a comprehensive measure of comorbidities. Functional impairment was measured in three dimensions (communication, psychological and somatic/physical function impairment) using the following four questions: (1) Are you deaf or do you have serious difficulty hearing? (2) Are you blind or do you have serious difficulty seeing, even when wearing glasses? (3) Because of a physical, mental, or emotional condition, do you have serious difficulty concentrating, remembering, or making decisions? and (4) Do you have serious difficulty walking or climbing stairs? We integrated these questions into a composite variable.26 We then categorised the number of comorbidities and functional impairments as ordinal variables: (1) comorbidity: 0, 1 and ≥2; (2) functional impairment: 0, 1 and ≥2.
We examined sociodemographic factors as covariates, including age, education level and marital status. Age was divided into two levels by 10-year intervals: 45–54 and 55–64 years. Education was categorised as an ordinal variable: attended high school or less, attended college/technical school or graduated from college/technical school. Marital status was reclassified as married (married couple or member of an unmarried couple) and not married (never married, divorced, widowed or separated), as those classified as married might have stronger feelings of obligation to their spouses impacting screening utilisation.27 Adverse lifestyle indicators, including current smoker, binge drinking, obesity and physically inactive, were included for analysis because they could be associated with screening utilisation and burden of comorbidity or functional impairment.28–30 Smoking status was categorised as never smoked, current smoker or former smoker. Binge drinking was defined as consuming four or more drinks on one occasion. Obesity was defined as body mass index≥30 kg/m2.31 Physical inactivity was defined as having no physical activity or exercise (eg, running, callisthenics, golf, gardening or walking) during the past month. A composite variable ‘no adverse lifestyle indicators’ was created to describe participants who reported none of these unhealthy behaviours. Participants self-reported their histories of routine clinical check-ups and healthcare coverage. Participants were considered as having healthcare coverage if they had health insurance, prepaid plans (eg, Health Maintenance Organization plan), government plans (eg, Medicare) or Indian Health Service.
Statistical analysis
We descriptively summarised the numbers and percentages of study characteristics in the overall sample and subpopulations defined by the numbers of comorbidities and functional impairments. We used Pearson’s χ2 tests to investigate if the distributions of these covariates differed by the numbers of comorbidities and functional impairments. We reported the number of observations and column percentages, as well as calculating the number and rate of Pap test utilisation for different levels of comorbidity (0, 1 and ≥2) and functional impairment (0, 1 and ≥2).
We then fit unadjusted and multivariable logistic regression models to calculate crude OR (cOR) and adjusted OR (aOR) and 95% CI for each exposure. We conducted two sets of multivariable logistic regressions: one adjusted for age only, and the other further adjusted for marital status, obesity, education, smoking, binge drinking, physical activity, clinical check-up and healthcare coverage. The multivariable logistic model for functional impairment included functional impairment and comorbidity simultaneously, adjusting for the aforementioned covariates. Because functional impairment might be downstream events of comorbidity (eg, vision impairment and forefoot ulcers induced by diabetes and mobility decline induced by arthritis) and act as the mediator between comorbidity and Pap test utilisation, we estimated effect measures for comorbidity using multivariable logistic regression without functional impairment. Tests for trend were conducted by treating exposures as continuous variables in the model. We did not correct for the BRFSS sampling weight in our analysis because the sample only accounted for a small proportion of the original data (approximately 1.5%) and middle-aged NHB women could be sociodemographically and biologically different from the overall general adults in the USA.
In the sensitivity analysis, we performed the same analyses as described above in women without a history of hysterectomy (n=4081) to account for those who could not have a Pap test due to the removal of the uterus. Another sensitivity analysis was restricted to women without a prior cancer history (n=6315) to explore if effect measures differed between women with and without a history of cancer. Subgroup analysis was stratified by two age categories (≤54 vs ≥55 years) as screening utilisation associations with comorbidity and functional impairment may vary by age. Additionally, we conducted several subgroup analyses to examine if adverse lifestyle indicators acted as effect modifiers. Five subgroups were generated based on the following indicators: (1) no adverse lifestyle indicators, (2) former/current smoker, (3) binge drinking, (4) obesity and (5) physically inactive. The multivariable model with comprehensive covariates was used to calculate aORs and 95% CI in each subgroup. Interaction terms between exposures and these covariates (with no adverse lifestyle indicators as the referent group) were added into the multivariable model, and Wald tests were used to determine if the interaction terms were statistically significant.
For current analysis, two-sided p values<0.05 were considered statistically significant. All statistical analyses were conducted using Stata, V.15.0 (College Station, Texas, USA: StataCorp).
Results
Table 1 presents the distribution of variables by exposures of interest. Of the 6359 women in our sample, 4141 (65.1%) had comorbidities (1785 had 1 and 2356 had at least 2) and 2429 (38.2%) had functional limitations (1415 had 1 and 1014 had at least 2). Overall, the mean age was 55.4 years (SD=5.7 years) and over half of the participants (56.6%) were aged≥55 years at the interview. More than one-third of the study participants (39.1%) had a high school education or less and the majority (65.9%) reported being unmarried. In terms of lifestyle indicators, one-third (33.6%) had a history of smoking (current 16.4%; former 17.2%) and 6.8% were binge drinkers. The prevalence of obesity and physical inactivity was 54.2% and 34.2%, respectively. Most women had a clinical check-up during the past year (90.8%) and had healthcare coverage (90.6%) (table 1). Overall, older age (55–64 years), lower education, unmarried, current or former smoker, obesity, physical inactivity, clinical check-up last year, healthcare coverage, hysterectomy and cancer history were significantly and positively associated with comorbidity and functional impairment (p<0.05). Binge drinking was positively associated with the number of comorbidities, but not functional impairments.
Characteristics of 6359 middle-aged non-Hispanic black women from 2018 BRFSS
There were inverse associations of Pap test utilisation with comorbidity and functional impairment, regardless of covariate adjustment, with the statistical test indicating a significant trend (p trend<0.05, table 2). Specifically, having comorbidities≥2 was associated with a 29% relative decrease in the odds of Pap test utilisation compared with comorbidities=0 (aOR=0.71, 95% CI=0.61 to 0.82, p trend<0.01) when adjusting for age. This negative association remained consistent after further adjustment (aOR=0.72, 95% CI=0.61 to 0.85, p trend<0.01). Middle-aged NHB women with functional impairments appeared to have a lower rate of Pap test utilisation (cOR=0.63, 95% CI=0.53 to 0.74; aOR=0.69, 95% CI=0.57 to 0.83) in both the crude model and the model adjusted for age. Among women without a history of hysterectomy or cancer, aORs were largely unchanged compared with effect estimates of the overall sample (table 3).
Associations of Pap test utilisation in the last 3 years with comorbidity and functional impairment among middle-aged non-Hispanic black women in the USA
Sensitivity analyses of associations between comorbidity and functional impairment and Pap test utilisation
The test for age subgroup difference indicated that there is no statistically significant subgroup effect, suggesting that age does not modify the association of Pap test utilisation with comorbidity and functional impairment (p interaction=0.53, p interaction=0.49, respectively, table 4). The associations between comorbidity and functional impairment and Pap test utilisation did not reach statistical significance in women without adverse lifestyle indicators (≥2 vs 0: aOR=0.87, 95% CI=0.57 to 1.32, aOR=0.97, 95% CI=0.48 to 1.96, respectively). However, the magnitude of these associations was stronger among women with adverse lifestyle factors, with point estimates of aORs showing a more pronounced decrease, although all p interaction>0.05. For example, the aOR was 0.53 (95% CI=0.40 to 0.71) for comorbidity≥2 compared with 0, and the aOR was 0.35 (95% CI=0.16 to 0.72) for functional impairment≥2 compared with 0 among binge drinkers.
Subgroup analyses by age and lifestyle indicators
Discussion
To our knowledge, this is the first epidemiologic study examining associations of Pap test utilisation with comorbidity and functional impairment among middle-aged NHB women in the USA. Our research suggested that middle-aged NHB women with comorbidity or functional impairment had a lower rate of Pap test utilisation than their healthier counterparts regardless of histories of hysterectomy and prior cancer. The analyses for age and lifestyle indicators subgroup difference indicated no statistically significant effect. However, the magnitude of these associations was stronger among women with adverse lifestyle factors. Active lifestyle behaviours may imply favourable health awareness and could attenuate the influence of functional impairment on Pap test utilisation.32
Our results are consistent with previous studies. For example, an epidemiologic study using data of the National Health Interview Surveys, which included 3073 US women aged≥35 years, reported an inverse association between the Charlson comorbidity index score and chronic disability and Pap test utilisation within the past 3 years.33 Another cross-sectional study conducted in France, involving 4226 women aged 25–65 years, reported lower adherence to Pap test recommendations among those with chronic respiratory diseases, diabetes or obesity.34 Similar outcomes have been documented in various epidemiologic studies.13 14 35–38 Regarding functional impairment, Iezzoni et al reported negative associations between Pap test utilisation and severe movement difficulty among 9073 women aged 21–65 years who had no history of cervical cancer or hysterectomy in the 2010 National Health Interview Survey data.20 While these studies analysed populations of broader age ranges and included different racial groups compared with our research, the consistency in findings supports the inverse association between Pap test utilisation with coexisting unfavourable health conditions, indirectly validating our outcomes to some extent.
Several potential reasons may explain the observed associations between comorbidity or functional impairment and Pap test utilisation. Healthcare providers might prioritise managing chronic diseases or functional impairment over recommending preventive health services because coexisting illnesses can directly cause adverse health events in a shorter time period and typically receive more attentions.39–41 Additionally, the Pap test procedure involves a series of movements, including getting onto standard fixed-height examination tables with knees bent and heels rest in stirrups, which can be physically strenuous for women with movement disabilities.42 43 Consequently, these individuals may encounter practical challenges when accessing the Pap test. Moreover, comorbidity and functional impairment (eg, depressive disorder and cognitive impairment) can impede effective communication and the gathering of medical histories.44 Providers may hold negative attitudes towards this population, potentially resulting in a reluctance to recommend Pap test.45
Our study had major strengths. First, we considered nine types of comorbidities and four types of functional impairment to construct the composite exposure variables, providing a comprehensive assessment. Second, unlike many previous studies of similar topics, we included lifestyle factors in the multivariable models, which further reduced residual confounding and enhanced the validity of the estimates. Additionally, we conducted sensitivity analyses restricted to women without a history of hysterectomy or cancer to test the robustness of the results in primary models. Subgroup analyses also allowed us to identify potential heterogeneity related to age or adverse lifestyle factors.
However, several limitations in this research should be noted. First, comorbidity and functional impairment were obtained via self-report, which is reportedly less accurate compared with medical record review or registry data abstraction.46 Second, we were unable to examine the effect of individual comorbidities or functional impairments due to the small sample sizes. Third, our study does not have data on indicators reflecting disease control or severity (eg, haemoglobin A1c can show glycaemic control in diabetes patients), suggesting that people with the same number of illnesses may have differing health status and healthcare-seeking behaviours, including Pap test utilisation. At last, given the cross-sectional nature of the study, we were unable to ascertain the temporality of the observed relationship between exposure and outcome, which limits our ability to infer causality.
Our findings highlight important health implications. While older women with comorbidity or functional impairment might have a limited life expectancy to benefit from screening and may experience a greater risk of overdiagnosis and overtreatment,7 middle-aged women with comorbid conditions still have a considerable life expectancy and may drive substantial benefits from screening. Moreover, women in this age group have a high risk of cervical cancer, with 42.2% of women aged 45–64 years being diagnosed with cervical cancer according to Surveillance, Epidemiology, and End Results data in 2013–2017.47 Therefore, it is imperative for health practitioners to establish targeted and cost-effective prevention programmes to enhance Pap test utilisation among middle-aged NHB women to reduce the burden of cervical cancer.48 Given that nearly all participants reported undergoing clinical check-up (90.8%) in the current study, healthcare providers and primary care systems provide an ideal platform to engage individuals with comorbidity or functional impairment, encouraging and actively supporting them to undergo Pap tests.49
Based on our findings, comorbidity or functional impairment could be a potential barrier to Pap test utilisation among middle-aged NHB women in the USA. Policymakers should consider implementing targeted interventions to ensure equitable access to Pap test. Furthermore, prioritised allocation of health resource to this unique population is essential. In the future, cohort studies are required to establish temporal relationships and indicators reflecting disease severity will be needed to further explore the impact of coexisting illnesses on Pap test utilisation in NHB women.
Data availability statement
Data are available in a public, open access repository.
Ethics statements
Patient consent for publication
Ethics approval
Ethics approval was not required as we used publicly available deidentified data.
Acknowledgments
Dr Dongyu Zhang, the Associate Director of Epidemiology and Real-world Data Sciences at Johnson & Johnson, provided relevant suggestions of statistical analysis. We thank Yingxi Chen from Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health for language editing and proofreading. The authors appreciate research staff participating in BRFSS data collection, management and dissemination.
References
Footnotes
X @bourneliyang
ZZ and YW contributed equally.
Contributors Conceptualisation: ZZ and YL. Methodology: ZZ and YW. Formal analysis: ZZ. Writing—original draft preparation: ZZ and YW. Writing—review and editing: ZZ, JH and YL. Funding acquisition: JH. Guarantor: ZZ. All authors have read and agreed to the published version of the manuscript.
Funding This research was funded by National Natural Sciences Foundation of China (grant number 82003275 to JH), Beijing Municipal Administration of Hospitals' Youth Program (grant number QML20210504 to JH) and Management Special Research Fund Project of Beijing Tiantan Hospital (grant number TYGL202308 to ZZ).
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.