Social determinants of health have a significant impact on diverse health outcomes, including malignancy. There is emerging awareness of socioeconomic status and socioeconomic deprivation as a primary risk factor for both incidence and adverse outcomes in lung cancer. [16–18]. SES is a broad, heterogeneous yet crucial set of factors used to measure an individual’s social and economic standing, built on parameters such as income, education level, insurance carrier, and geographic location [19, 20]. Robust evidence suggests that individuals with lower SES present with more advanced-stage lung cancer at diagnosis, attenuated response to chemotherapy and obtain less favorable prognosis after diagnosis [16, 18]. Although large trials that investigated lung cancer screening, e.g., the National Lung Screening Trial (NLST) and the Dutch-Belgian lung cancer screening [Nederlands–Leuvens Longkanker Screenings Onderzoek (NELSON)] trial, demonstrated a significant mortality reduction of 20–26% when encompassing all participants [5–7], sociodemographic disparities were present in the trials, and thus there are limitations in generalizing the results to diverse populations with respect to ethnicity, race and income level. Participants in NSLT were predominantly of white race (91%) and had higher educational status (32% with college degrees) and income compared to the general population matched to age and smoking criteria [21]. Therefore, it is not known whether the benefits seen in these trials are applicable to those with lower SES.
While numerous studies have evaluated socioeconomic disparities among individuals enrolled in or eligible for LCS programs, there is a paucity of literature on the clinical impact of SES on lung nodule evaluation and follow up [11]. Lower SES is associated with lower LCS adherence, utilization, and guideline concordant care, yet to date, there is minimal data on the impact of SES on ILNs. In the DELUGE study, which prospectively evaluated both LCS and ILN surveillance, individuals from the lung nodule surveillance program were more likely to be diagnosed with lung cancer [10]. Other studies have highlighted the advantages of combined LCS and ILN programs [9, 10, 12, 22] to capture higher-risk groups, which leads to improved early lung cancer diagnosis and guideline concordant care. However, SES factors in these studies were not consistently reported, and these studies included mostly individuals of White ethnicity, which made it challenging to evaluate the effect of the programs when accounting for SES, race, and ethnic inequalities [23]. To elucidate the influence of SES on lung cancer, we undertook a retrospective analysis of all individuals ≥ 18 years of age with incidental lung nodules on CT chest who were referred to the pulmonary clinic at an urban, tertiary care hospital in Brooklyn, New York over a 3-year period. Such hospitals are an essential safety net serving a socioeconomically diverse population, wherein individuals are typically less likely to be included in screening programs or participate in clinical trials [16]. Specifically, we aimed to evaluate the association between incidental nodule size and SES status using common determinants such as education, insurance type and income, as well as less commonly used indicators such as air quality index, cost of living index and immigration status.
Socioeconomic status and nodule size
The impact of SES on nodule size was assessed with logistic regression analysis using individual and geographical determinants. Categorization of SES by geographic area has gained traction in recent years with the development of indices such as the area-based deprivation index (ADI) [20]. Geography-based determinants reflect the community SES, while individual determinants are specific to the patient. After adjusting for age, sex, and tobacco use, we found that individuals from areas with lower unemployment rates presented with smaller nodules < 6 mm (OR 0.754, p 0.045), while those from areas with higher unemployment rates were associated with larger nodules (OR 1.1, p < 0.357). Patients with noncommercial insurance were more likely to present with larger nodules > 8 mm (OR 2.181, p 0.016) and less likely to have smaller nodules < 6 mm (OR 0.437, p 0.01). These findings highlight that even with incidental nodules, markers of poverty are associated with larger, inherently higher-risk nodules at initial presentation.
Several studies have examined the relationship between poverty and lung cancer outcomes, finding that areas of higher deprivation were associated with higher lung cancer incidence and mortality [18, 24]. These subpar outcomes are likely related to high-risk smoking behaviors and less access to healthcare compared to those with higher SES [25, 26]. Additionally, individuals with lower SES may be more likely to be exposed to secondhand smoke and other environmental toxins that increase the risk of lung cancer [16]. There is growing evidence evaluating the impact of insurance type and lung cancer outcomes. A recent study evaluating ILN surveillance found that underinsured individuals were more likely to be diagnosed with cancer. Individuals with federal insurance are also less likely to complete screening [17, 27] and may not receive full coverage for LDCT.
Disparities in eligibility for LCS due to race have been described [16, 18], but the relationship with SES has been less explored. There was a period of 11 months for CMS to expand LCS eligibility criteria (February 2022) to reflect the USPSTF eligibility update (March 2021) [3, 28]. Consequently, most individuals in our study period would not have benefited from the expanded criteria. CMS limits age to 77 years in determining LCS eligibility [28], although the NLST and NELSON trials demonstrated a mortality benefit in those up to 80 years old [5–7]. Age is a well-established risk factor for lung cancer [2]. Our study showed that increased age was weakly associated with larger nodules > 8 mm, although with a small effect size. Taken together, these findings suggest that older individuals with federal insurance may be ineligible for LCS. These limitations may create additional barriers for LCS in individuals with low SES [16, 17].
We also evaluated less commonly studied markers of SES, such as the air quality index and cost of living index (COLI). COLI is a relative marker of living expenses compared to United States estimates, with values above 100 conferring higher than average costs of living. COLI has not been extensively studied in relation to LCS or lung nodule outcomes; however, prior studies in individuals with hepatocellular cancer found that those with lower COLI presented with more advanced cancers, while higher COLI was associated with improved survival [29]. We found a weak association between individuals from lower COLI areas and smaller nodules < 6 mm (OR 0.933, p 0.056), while higher COLI areas were weakly associated with nodules between 6–8 mm (PR 1.084, p 0.048), the significance of which is undetermined. Individuals with lower SES may be more likely to reside in lower COLI areas; however, NYC on average has the highest COLI in the United States, and all individuals in this study resided in areas with a COLI > 100. Studies suggest that increased living costs may compete with other financial burdens, which may disproportionately affect individuals with lower SES, leading to less guideline concordant care [5]. Larger studies utilizing the cost of living index may help further define these relationships in individuals with lung nodules.
The air quality index is a measure of air pollution, standardized based on the Clean Air Act [30]. Metropolitan areas such as New York City are required to report air quality daily. In our study, there were no associations with nodule size and air quality index, and all individuals resided in areas with satisfactory air quality. The association of the air quality index and ILN size has not been explored, but previous studies have demonstrated significant associations with a higher air quality index and increased lung cancer incidence [31]. While no relationship was present in our study, it will be interesting to assess how SES and air quality impact ILN size in areas with more pollution.
Education level has been assessed as a risk factor for worse lung cancer outcomes. Health literacy may be lower in individuals with lower educational attainment [32], and these individuals may face more barriers to screening despite being at higher risk [33]. Our study showed that individuals from areas with educational attainment less than high school were more likely to have smaller nodules (OR 1.243, p < 0.008). There was no association between education and nodules > 8 mm. Prior population studies have found that educational attainment lower than high school level was associated with decreased LCS eligibility, in contrast to those with college education or higher (which compromised a significant percentage of NLST participants) [16]. Our assessment is limited by including only one measure of educational attainment, having no comparison group with those of higher attainment, and a small sample size. The association between education and ILNs should be further explored, as individuals with ILNs represent a distinct population from those in LCS programs and are less likely to be eligible for LCS.
Race/ethnicity and nodule size
The distribution of race and ethnicity in our sample was diverse, with mostly nonwhite Hispanic individuals (56%) followed by African American (26.5%) and White (15.3%) individuals. This significantly differs from the NLST, which comprised 91% White individuals [5, 6]. Our study found that nonwhite individuals were 0.67 times less likely to have smaller nodules < 6 mm and 1.63 times more likely to have nodules > 8 mm, although these relationships were not statistically significant. It is crucial to examine the intersection of socioeconomic status (SES) and racial disparities in lung cancer outcomes, given its strong correlation. Studies have shown that individuals belonging to racial and ethnic minorities have the lowest SES, resulting in lower eligibility and LCS utilization [16, 34]. However, there is a lack of research on how the combination of low SES and minority racial status impacts ILN surveillance, LCS eligibility, utilization, or outcomes. Other studies found that African American participants had a lower screening rate than White participants, and unscreened individuals had a lower annual household income [34]. This suggests that African American individuals with low annual household income may have an even lower screening rate. Additionally, our study showed that when adjusting for race/ethnicity, SES determinants demonstrated stronger associations with ILN size. Other studies have shown that lower SES and ethnic-minority groups have significantly lower overall lung cancer patient survival rates. [2, 17]. These findings suggest that SES represents an important driver of ILN size and ultimately lung cancer risk.
Limitations:
Our study has several limitations. As with all retrospective studies, our data reveal associations but does not provide evidence for causation. Several SES determinants, such as income and poverty level, did not have significant relationships with lung nodule size. This may be due to the overall high rate of deprivation in the population served by the safety-net hospital, making it more difficult to find significant relationships, since most areas may be similarly deprived.
As a community hospital located in an area facing a shortage of healthcare professionals during the COVID-19 pandemic, our center did not have an established formal surveillance program for LCS or ILNs during the study period. We included individuals with ILNs who were referred to the pulmonary service for any indication. However, primary care providers in the community may manage ILNs without mandating a referral to the pulmonary service. Since our study only included patients referred to the pulmonary services, patients with ILNs managed by PCPs were excluded, which reduced the power of the study. It is conceivable that variability in the medical and socioeconomic history of these excluded patient with comparison to those referred to the pulmonary service and could have influenced the findings.
Conversely, our study included many patients with small, lower risk ILNs. This distinguishes our study from others that include patients from formal LCS or ILN programs, which generally include nodules at least 6mm in size. Nonetheless, our sample represents a socially diverse at-risk group that is typically excluded from research studies.
Additionally, geographical markers of socioeconomic status were derived from census data according to zip code. These markers do not reflect the individual socioeconomic status of patients but rather reflect the deprivation faced by their neighborhoods. The individual patient may have markers of higher or lower SES compared to their neighborhood.