Abstract
Objective
In 2018, the United States Preventive Services Task Force promoted shared decision making between healthcare provider and patient for men aged 55 to 69. This study aimed to analyze rates of prostate-specific antigen (PSA) testing across racial and ethnic groups following this new recommendation.
Methods
A secondary analysis was conducted of the 2020–2021 Behavioral Risk Factor Surveillance System database to assess men aged 55 or older without a history of prostate cancer. We defined four race-ethnicity groups: non-Hispanic Whites (NHWs), non-Hispanic Blacks (NHBs), Hispanics, and Other. The primary outcome was the most recent PSA test (MRT), defined as the respondent's most recent PSA test occurring pre-2018 or post-2018 guidelines. Logistic regression adjusted for covariates including age, socioeconomic status factors, marital status, smoking history, and healthcare access factors.
Results
In the age 55 to 69 study sample, NHW men had the greatest proportion of MRT post-2018 guidelines (n = 15,864, 72.5%). NHB men had the lowest percentage of MRT post-2018 guidelines (n = 965, 66.6%). With NHW as referent, the crude odds of the MRT post-2018 guidelines was 0.68 (95% confidence interval (CI) = 0.53–0.90) for NHB. The maximally adjusted odds ratio was 0.78 (0.59–1.02).
Conclusions
We found that NHB aged 55 to 69 reported decreased rates of PSA testing after 2018 when compared to NHW. This was demonstrated on crude analysis but not after adjustment. Such findings suggest the influence of social determinants of health on preventative screening for at-risk populations.
Introduction
Prostate cancer remains a prevalent malignancy among men in the United States, accounting for 174,650 new cases and 34,500 deaths in 2022. 1 A notable concern is the significant racial and ethnic disparities observed in its incidence rates; non-Hispanic black (NHB) men encounter a 1.7-fold higher incidence than non-Hispanic white (NHW) men, whereas non-Hispanic Asian and Pacific Islander men exhibit approximately half the incidence rate of NHW men.2,3 These disparities arise from a complex interplay of biological, social, environmental, and healthcare-related factors. 4 This includes issues like access to care and screening. Current practice guidelines underscore the necessity of tailoring screening decisions to an individual's values, preferences, overall health, and life expectancy.
In a significant change during 2018, the US Preventive Services Task Force (USPSTF) revised its recommendation for prostate cancer screening using the prostate-specific antigen (PSA) test. 5 The recommendation previously advised against PSA screening and was grade D, indicating a moderate to high certainty of no net benefit or harm outweighing benefits due to overtreatment. It was upgraded to grade C, recommending a shared decision making (SDM) process to carefully assess the potential benefits and drawbacks of screening for men aged 55 to 69. For men aged 70 and above, the recommendation remained at grade D and advised against screening. This revision emphasizes the necessity of in-depth discussions regarding potential harms, including false positives, overdiagnosis, overtreatment, surgical complications, and adverse outcomes.
Despite these changes, ongoing research consistently indicates the persistence of racial and ethnic disparities in the utilization of PSA screening. African American and Hispanic individuals have been found to be less likely to undergo PSA screening compared to their Caucasian counterparts.6,7 This inequity accentuates the public health concern for screening strategies that encompass race, ethnicity, and other risk factors. Therefore, scrutinizing the role of race and ethnicity in prostate cancer screening decisions for adults aged 55 and older—the age brackets specifically addressed by the 2018 USPSTF guidelines—is paramount. To capture screening participation post-2018 guidelines in a population previously adherent to screening, we focused our inquiry on men who self-reported previous PSA testing. Comprehending the intersections between race, ethnicity, and screening decisions can guide the creation of tailored screening strategies that resonate with the needs of diverse populations and promote involvement in their care. This study aims to assess disparities in preventive care for prostate cancer following the 2018 USPSTF guideline revision, focusing on racial-ethnic variations in PSA screening.
Methods
Data source
We conducted a secondary analysis of the Behavioral Risk Factor Surveillance System (BRFSS) survey data from 2020–2021. 8 The BRFSS is a cross-sectional telephone survey of adults in the United States regarding health-related factors and preventative services. The BRFSS administers varying survey question modules to various states each year. The study was deemed exempt by the Institutional Review Board at Florida International University.
Study population
The study population included men who were age 55 or older and had completed survey questions in the 2020 and/or 2021 BRFSS surveys, as these individuals are specifically addressed in the USPSTF guidelines. These individuals were those who had self-identified as “Male.” We then excluded males with a history of prostate cancer. Our cohort was narrowed to men who had answered Yes to the question “Have you ever had a P.S.A. test?” In the flow of the BRFSS survey, these respondents were subsequently asked the question of interest for the primary outcome variable, which refers to when the patient most recently received a PSA test. Specifically: “How long has it been since you had your last P.S.A. test?” in 2020 and “About how long has it been since your most recent P.S.A. test?” in 2021. Additional BRFSS questions included the following: age, race, and ethnicity. Age-eligible males with missing values for race/ethnicity or the question of interest were excluded.
We stratified our analysis, based on age, to address the guideline's specific language: Men aged 55 to 69 years old (y.o.), within the recommended age for PSA testing by the 2018 USPSTF guidelines (“55–69 y.o.”), and men aged 70 or older, who were older than the recommended age for PSA testing (“≥70 y.o.”). The exposure variable was defined as race-ethnicity group, with the following four categories: NHW, NBH, Hispanic, and Other races (including American Indian or Alaskan Native and non-Hispanic; Asian and non-Hispanic, Native Hawaiian or other Pacific Islander and Non-Hispanic; Other race and non-Hispanic; Multiracial and non-Hispanic). NHW was the referent category.
Measures
The primary outcome was the “most recent PSA test” (MRT). This was defined as ever having a PSA test, measured by the approximate timing of that MRT. We treated this as binary to identify whether the answer to the question of interest suggested the respondent's MRT was pre-2018 guidelines or post-2018 guidelines. A MRT post-2018 guidelines was defined as the respondent answering “Within the past year (anytime <12 months ago)” for the 2020 BRFSS survey. It was also defined as “Within the past years (anytime < 12 months ago)” or “Within the past 2 years (1 year but < 2 years ago)” for the 2021 BRFSS survey. A MRT pre-2018 guidelines was defined as an answer representing the respondent's MRT occurred earlier than the aforementioned answers.
Statistical analysis
All analyses were employed in consideration of the complex sampling design of the BRFSS, applying sampling weights appropriately. 9 Missing data (all <1% except for income level) were excluded from analysis. We used counts and percentages to present frequencies of categorical variables. Differences in proportions were tested using Pearson's chi-square test. We used simple and multivariable logistic regression to explore the association between race-ethnicity and the reported MRT. This allowed us to obtain crude and adjusted odds ratios (ORs) and corresponding 95% confidence intervals (CIs). We sequentially ran minimally adjusted models, adjusting for age, and annual household income, followed by fully adjusted models additionally adjusting for marital status, smoking status, body mass index (BMI), education level, health insurance status, primary care provider (PCP) status, cost barriers to primary care, most recent PCP visit, and residing in metropolitan areas. A p-value of <0.05 was considered significant. All analyses were performed using Stata (Version 16, Statacorp., College Station, TX, USA) statistical software.
Results
There were 288,560 males in the eligible age range during the study years. Of these, 3,472 (1.2%) were excluded for having a history of prostate cancer, leaving 288,560 males. Then, 88,514 of these males were asked if they had ever received a PSA test. The 5,436 (6.1%) respondents who were unsure or refused to answer the question, and the 31,948 (36.1%) who reported never having a PSA test, were further excluded. Of the 51,130 (57.8%) males who reported ever having a PSA test, none were excluded for missing race/ethnicity, but 9,046 (17.7%) did not answer the question about the timing of their test. This resulted in a final analytical sample of 42,084 males. This study population was comprised of 22,077 (52.5%) 55–69 y.o. males and 20,007 (47.5%) ≥ 70 y.o. males (Table 1). Table 1 shows baseline demographics by racial-ethnic group for 55–69 y.o. and ≥70 y.o. cohorts.
Demographics and interaction by racial-ethnic groups for 55–69 y.o. and ≥70 y.o. males.
NHW: Non-Hispanic White; NHB: Non-Hispanic Black; BMI: body mass index; GED: general educational diploma; PCP: primary care provider. Boldface indicates statistical significance p < 0.05.
Among 55–69 y.o. males
The MRT post-2018 guidelines was 71.9% (n = 15,864) and pre-2018 guidelines was 28.1% (n = 6,213) for all 55–69 y.o. males (Table 2). NHW men and Other men had the greatest proportion of MRT post-2018 guidelines (n = 15,864, 72.5% and n = 736, 70.1%, respectively), while NHB men had the lowest percentage of MRT post-2018 guidelines (n = 965, 66.6%) (p = 0.0178) (Table 2). The 55–69 y.o. males who more frequently received their MRT post-2018 guidelines had at least one PCP, did not have any cost barriers associated with seeing their PCP, and had seen their PCP in the last year. Statistical analysis of differences in MRT by demographic information is presented in Table 2.
Reported most recent PSA screening test by racial-ethnic group for 55–69 y.o. and ≥70 y.o. males.
NHW: Non-Hispanic White; NHB: Non-Hispanic Black; BMI: body mass index; GED: general educational diploma; PCP: primary care provider; PSA: prostate-specific antigen. Boldface indicates statistical significance p < 0.05.
The crude OR for MRT post-2018 guidelines compared to NHW males was 0.68 (95% CI = 0.53–0.90) for NHB 55–69 y.o. males (Table 3). There were no significantly differing odds of MRT post-2018 guidelines for NHB 55–69 y.o. males compared to NHW males upon minimal or maximal adjustment. In all models, there were no significantly different odds of MRT post-2018 guidelines for Hispanic or Other 55–69 y.o. men compared to NHW men.
Logistic regression odds ratios for self-reported most recent PSA test by racial-ethnic group.
NHW: Non-Hispanic White; NHB: Non-Hispanic Black; H: Hispanic; O: Other; PSA; prostate-specific antigen; PCP: primary care provider.
(a) Crude analysis.
(b) Model 2 adjusted for age and income level.
(c) Model 3 additionally adjusted for marital status, smoking status, body mass index, education level, health insurance status, PCP (primary care provider) status, cost barriers to primary care, most recent PCP visit, and residing in metropolitan areas.
Among ≥70 y.o. males
The MRT post-2018 guidelines was 67.5% (n = 13,512) and pre-2018 guidelines was 32.5% (n = 6,495) for all ≥70 y.o. males (Table 2). Men of Hispanic ethnicities had the highest percentage of receiving a PSA post-2018 guidelines (n = 403, 71.0%), while NHB men had the lowest rate (n = 550, 61.2%), but the difference was not statistically significant across racial-ethnic groups (Table 2). The odds of having the most recent PSA test post-2018 guidelines were not different between NHW ≥70 y.o. men and NHB, Hispanic, or Other racial-ethnic groups (Table 3).
Discussion
In the present study, we sought to determine whether racial and ethnic differences played a significant role in PSA testing in the context of the 2018 USPSTF guideline changes. 5 We found a significant decrease of 32% in the relative likelihood of the MRT being post-2018 guidelines for NHB 55–69 y.o. males in comparison to NHW. Simply, NHB 55–69 y.o. males were less likely to undergo screening in the 2 years after the 2018 USPSTF guideline changes, yet this change was attenuated to non-significant levels upon minimal and full adjustment. Moreover, there was no decreased likelihood of having a PSA test after 2018 for the Hispanic or Other groups in the 55–69 y.o. cohort when compared to NHW men. There was no difference in likelihood of PSA screening after the 2018 guidelines for any racial-ethnic group in the ≥70 y.o. cohort when compared to NHW men. Our analyses provide a unique perspective on factors influencing NHB males in their receipt of PSA testing immediately following the USPSTF guidelines update in 2018.
Self-reported PSA screening frequency has responded to alterations in the guidelines, as inferred by previous population-based studies using the BRFSS. For example, Kensler and colleagues studied similar racial-ethnic groups from 2012 to 2018 over 2-year intervals, among men aged 40–74. 6 They found an absolute decrease in PSA testing of 11% annually across the study period. While this decline in PSA testing was steeper for NHB men compared to NHW men, this change was mainly driven by men aged 40–54, with no statistically significant differences in the 55–69 age group. Our pre-post 2018 guideline findings are consistent with and are extending the pre-post 2012 guideline study, in the 55–69 age bracket. The 2012 guidelines resulted in an increase in the incidence of regional and distant prostate cancer. 10 It is yet to be seen how the 2018 guidelines will affect the incidence of this disease.
We found a significant difference in the unadjusted odds of PSA testing between NHB and NHW men; thus it is worth exploring why after adjustment this association was no longer statistically significant. Social determinants of health, including income level and access to care, may be potentially driving the racial-ethnic group differences in PSA screening for 55–69-year-old men. In our bivariate analyses, NHB were more frequently impoverished, uninsured, having more than one PCP, and having cost barriers to seeing their PCP when compared to NHW, suggesting, time and again, the social nature of the race-ethnicity constructs. 11 Interestingly, while NHB men had more PCPs than NHW, there was no difference in the time since their last PCP visit, yet upon crude analysis they had lower rates of PSA testing after 2018. This finding highlights the complex, multifactorial nature of the screening behavior, and the interplay between access to healthcare, quality of PCP visit, quality of the shared decision-making process, trust, health beliefs, and other social-cultural factors influencing the effectiveness of preventative care in the “real world.” 12
While in the crude analysis, there was a significant decrease in screening rates for NHB compared to NHW after the 2018 guideline release, this association was no longer statistically significant after adjusting for age and income. This suggests that the association was driven by these two variables. Notably, individuals in the lowest income bracket had the highest frequency of having their MRT before the 2018 guidelines. This highlights previous research that identifies socioeconomic status as an influencing factor of the disparities in preventative care for prostate cancer.13,14 Moreover, further adjusting for metrics of healthcare access did not significantly change the estimated OR, suggesting the critical role of socioeconomic status as a major social determinant of health.
Yet our findings may also draw attention to the complex interplay between healthcare access and distribution of equitable care. In our study sample, NHB had more PCPs than NHW and had no difference in the time since their last PCP visit. This draws attention specifically to the visit during which providers engage in discussion about preventative screening with their patients. These visits are opportunities for meaningful engagement to understand a patient's goals and share information. During such visits, assessment of health literacy and establishing trust and rapport are paramount for individuals of disadvantaged socioeconomic status or for racial-ethnic minority groups who are at high risk of prostate cancer. 15 Prior study reveals the historical inequity in screening for NHB due to health literacy and social determinants of health, which have extended as far as the incidence of fatal prostate cancer. 16 NHB men have been found to present with more aggressive disease and advanced staging at diagnosis. 16 Yet for NHB, given historical inequities in their care, providers must ensure that they appropriately address the benefits and limitations of screening. 17 NHB men have been reported to less frequently perceive their risk of prostate cancer as higher than average. 18 Also, NHB men have expressed reservations about the accuracy and costs of such preventative tests. 19 Our study highlights the opportunity each provider has for ensuring equitable screening across racial-ethnic groups by accurately discussing the pros and cons of screening during their visits with patients.
The intricate balance between prostate cancer detection and overdiagnosis is left to the decision of the provider and patient, underscoring the importance of determining a patient's health goals. Consideration of the optimal interval for screening is an on-going discussion across the scientific, medical community. For example, a microsimulation model by Nyame et al. explored the impact of changing the frequency of PSA screening for NHB men aged 55 to 69 and found that increasing the frequency to every year could reduce mortality by up to 19% but at the risk of increasing overdiagnosis. 20 Thus, a balanced approach is needed. Their study suggested that intensified screening would most benefit younger men and men with a family history of prostate cancer. 20 Indeed, the landscape of preventative health is perhaps changing. The most current 2023 American Urological Association guidelines state that providers should offer regular prostate cancer screening every 2–4 years for people aged 50 to 69. This is updated from their 2018 guideline suggesting that providers may offer screening at intervals of 2 or more years. 21 Moreover, men aged 40 to 45 with an increased risk of developing prostate cancer, like those of NHB ancestry or those with a family history of prostate cancer, are recommended to begin testing.
The present work suggests that the intended benefit of the 2018 USPSTF guideline change, to decrease overdiagnosis and false positives, had a similar effect across racial-ethnic groups when adjusting for social determinants of health such as age, socioeconomic status, and factors affecting healthcare access. The 2018 USPSTF guidelines on PSA screening are now in the process of being updated 5 ; it is left to be seen how these changes will be served across the United States. Future work should continue to explore the influence of social determinants of health on the implementation of practice guidelines and thus the equitable administration of preventative healthcare.
Limitations
Our study results need to be interpreted in the light of several inherent limitations. The study population is confined to the population of age-eligible men that report having previous PSA screening as our intention was to capture the timing of their most recent screening. Thus, we cannot make inferences about the application of the guidelines to all individuals including those who did not undergo PSA testing. The reliance on self-reported data introduces reporting biases that can undermine the study's internal validity due to potential errors in recall bias. However, it is worth noting that a previous study has reported a high level of validity in older BRFSS datasets (2004–2011). 22 Additionally, the BRFSS questionnaire modules are administered differently each year; PSA testing questions were asked in all 50 states in 2020 but only in New Jersey during 2021; thus our results may not be generalizable to the entire population across the United States. The cross-sectional nature of these data restricts capacity to establish causal relationships or capture temporal changes, as it merely offers a snapshot of information at a specific point in time. While missing data were excluded at the time of analysis, missing data for each co-variable accounted for less than 1% of the sample size, except for income level (∼12%), which is understandable given the sensitive nature of the question. It is unlikely that this low magnitude of missing data could have significantly biased our results.
As the BRFSS survey is administered over the course of the calendar year, it may not exactly reflect screening before or after the USPSTF guidelines were updated and implemented in May of 2018. Moreover, the question used for “time since the last PSA test” did not differentiate between indications like prostate cancer screening or monitoring of prostatic hyperplasia. This may limit the direct association of the study to identify appropriateness of the screening. Lastly, response rate remained low for both 2020 (47.9%) and 2021 (43.4%), which could possibly result in inappropriately attributed outcomes for one ethnic group over another.23,24
Conclusions
The present study found that NHB men aged 55 to 69, a group reflecting the 2018 USPSTF guideline's recommended age for SDM about prostate cancer screening, reported decreased rates of PSA testing after 2018 when compared to NHW men. This association was observed in the crude analysis but not after adjustment for variables that represent socioeconomic status and healthcare access. Such findings suggest the influence of social determinants of health on preventative screening for at-risk populations.
Footnotes
Acknowledgment
Chrisnel Lamy at Herbert Wertheim College of Medicine, Florida International University, Miami, Florida, for help with statistical analysis.
Authors’ Contributions
The content of this article has not been previously presented or published elsewhere.
Data availability statement
The datasets generated during and/or analyzed during the current study are not publicly available due to not having a hosting site, but are available from the corresponding author on reasonable request.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
