This cohort study assesses the association between baseline annual household income and financial toxicity, health utility, and survival in patients with head and neck cancer.
Key Points
Question
What is the association between annual household income and financial toxicity, health utility, and survival in head and neck cancer?
Findings
In this prospective cohort study including 555 patients treated for head and neck cancer in Toronto, Ontario, a significant association was observed between household income at time of diagnosis and disease-free survival. Patients with head and neck cancer with an annual household income less than $30 000 appeared to be disproportionately affected.
Meaning
Significant disparities exist for patients with head and neck cancer, even within a system of universal health insurance.
Abstract
Importance
While several studies have documented a link between socioeconomic status and survival in head and neck cancer, nearly all have used ecologic, community-based measures. Studies using more granular patient-level data are lacking.
Objective
To determine the association of baseline annual household income with financial toxicity, health utility, and survival.
Design, Setting, and Participants
This was a prospective cohort of adult patients with head and neck cancer treated at a tertiary cancer center in Toronto, Ontario, between September 17, 2015, and December 19, 2019. Data analysis was performed from April to December 2021.
Exposures
Annual household income at time of diagnosis.
Main Outcome and Measures
The primary outcome of interest was disease-free survival. Secondary outcomes included subjective financial toxicity, measured using the Financial Index of Toxicity (FIT) tool, and health utility, measured using the Health Utilities Index Mark 3. Cox proportional hazards models were used to estimate the association between household income and survival. Income was regressed onto log-transformed FIT scores using linear models. The association between income and health utility was explored using generalized linear models. Generalized estimating equations were used to account for patient-level clustering.
Results
There were 555 patients (mean [SD] age, 62.7 [10.7] years; 109 [20%] women and 446 [80%] men) included in this cohort. Two-year disease-free survival was worse for patients in the bottom income quartile (<$30 000: 67%; 95% CI, 58%-78%) compared with the top quartile (≥$90 000: 88%; 95% CI, 83%-93%). In risk-adjusted models, patients in the bottom income quartile had inferior disease-free survival (adjusted hazard ratio, 2.13; 95% CI, 1.22-3.71) and overall survival (adjusted hazard ratio, 2.01; 95% CI, 0.94-4.29), when compared with patients in the highest quartile. The average FIT score was 22.6 in the lowest income quartile vs 11.7 in the highest quartile. In adjusted analysis, low-income patients had 12-month FIT scores that were, on average, 134% higher (worse) (95% CI, 16%-253%) than high-income patients. Similarly, health utility scores were, on average, 0.104 points lower (95% CI, 0.026-0.182) for low-income patients in adjusted analysis.
Conclusions and Relevance
In this cohort study, patients with head and neck cancer with a household income less than CAD$30 000 experienced worse financial toxicity, health status, and disease-free survival. Significant disparities exist for Ontario’s patients with head and neck cancer.
Introduction
The link between low socioeconomic status (SES) and cancer survival is well established.1 The magnitude of this association, however, varies depending on the geographic region and cancer site.1,2,3 Several nations have reported an inverse association between SES and cancer outcomes among patients with head and neck cancer (HNC).3,4,5,6,7,8,9,10,11,12,13 Potential mechanisms include lack of access to health care, delayed presentation, and higher comorbidity burden of low-SES individuals. In Canada, most health care is provided by a universal health care system. Despite this, cancer-related disparities persist, particularly for those with HNC.2,3 In 1 Ontario study, 5-year overall survival was 14% lower for patients with HNC living in the lowest-income neighborhoods.3
While several studies across the world have documented a link between SES and HNC survival, nearly all have used ecologic (community-based) measures.3,4,5,6,7,8,9,10,11,12,13 Studies using more granular patient-level data are lacking. Patient-level income is likely a better reflection of an individual’s personal circumstances and their potential for financial toxicity.14,15
Financial toxicity is a relatively new concept that references the substantial economic burden imposed by cancer treatment on a household.16 To enhance our understanding of financial vulnerability in HNC, we have systematically collected socioeconomic and financial toxicity data in patients since 2015.17,18 In this study, we examine the association between household income at diagnosis with financial toxicity, health utility (a measure of health status), and survival. We hypothesized that low-income individuals would experience worse survival, financial toxicity, and health utility.
Methods
This was a prospective cohort study of patients treated with curative intent at the Princess Margaret Cancer Centre (Toronto, Ontario) between September 17, 2015, and December 19, 2019. This study was approved by the University Health Network Research Ethics Board. Reporting is in adherence with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.19 Written patient informed consent was obtained from all individuals.
Study Cohort
Consecutive adult (age ≥18 years) patients with a newly diagnosed mucosal squamous cell carcinoma of the upper aerodigestive tract were approached. Individuals undergoing noncurative treatment, non-English speakers, and those lacking decisional capacity were excluded. Patients were observed from date of diagnosis and followed up until date of last follow-up, death, or end of study date (February 24, 2021), whichever came first. At study entry, participants completed a validated baseline demographic questionnaire (race and ethnicity options were defined by the investigator). This included questions related to household income.18 Patients also completed the Financial Index of Toxicity (FIT) Instrument at 12 months and 24 months posttreatment. The FIT is a validated measure of subjective financial toxicity and is specific to HNC.17 Scores range from 0 to 100 with higher scores indicating higher financial toxicity. Patients completed the Health Utilities Index-Mark 3 (HUI3). It contains 8 questions/attributes pertaining to vision, hearing, speech, ambulation, dexterity, emotion, cognition, and pain and/or discomfort. Scores are assigned to each attribute and are combined using a formula to generate a health utility score ranging from 0 to 1. The HUI3 measures health utility from item scores using a multiattribute utility function. It has been validated in the HNC patient population and shown to be a valid measure of health status.20,21,22,23 Scores are anchored between 1 (perfect health) and 0 (death). Patients completed the HUI3 at various time points, including prior to treatment, after surgery, during radiation therapy, and at 3, 6, 12, and 24 months posttreatment.
Exposures
Patients were asked to report their total household income in the year prior to diagnosis in Canadian dollars. Responses were binned by increments of $10 000 from $0 to $100 000, then by increments of $25 000 up to $200 000, with a final category for income greater than $200 000. For analysis, income categories were aggregated into quartiles (<$30 000, $30 000-$49 999, $50 000-$89 999, and ≥$90 000).
Outcomes
Our primary outcome of interest was disease-free survival. Secondary outcomes of interest were the patients’ FIT and HUI3 scores.
Covariates
All baseline clinical and demographic variables were measured at date of cancer diagnosis. Age was dichotomized using a cutoff of 65 years or older. This cutoff represents a standard retirement age and is the age at which most Ontario patients become eligible for provincial drug coverage. Sex was reported as either male or female. Cancer stage was reported per the American Joint Committee on Cancer 7th edition and dichotomized as I/II vs III/IV. Human papillomavirus (HPV) status was confirmed through p16 testing and reported as positive, negative, or HPV-unrelated. Eastern Cooperative Oncology Group (ECOG) performance status was treated as a 3-level ordinal variable (0, 1, ≥2).
Statistical Analysis
Baseline characteristics were examined using descriptive statistics for the entire cohort and by availability of income data. Kaplan-Meier survival curves were plotted by household income quartile with log-rank tests. Cox regression models were constructed to evaluate the prognostic significance of household income on survival. Covariates were selected a priori and included age, cancer stage, and HPV status (eFigure 3 in the Supplement). We were limited by a low event rate and therefore selected covariates judiciously (following the 1-in-10 rule).24 In separate models, we added sex and treated age as continuous, though neither substantially affected our parameter estimates. There was no evidence of collinearity between selected covariates. Proportionality assumptions were verified by visual examination of Schoenfeld residual plots and by Grambsch and Therneau’s tests of statistical significance.25
Linear regression models were used to assess the association between FIT scores and income. Given that the FIT scores were skewed and could not be below zero, we opted to log-transform these data. Results are reported as relative increases in FIT scores. The association between baseline income and health utility was explored through a series of generalized linear models. Generalized estimating equations with an autoregressive correlation structure were used to account for patient-level clustering. An interaction term for “<3 months out of treatment” vs “≥3 months out of treatment” and income was also included. This allowed us to report utility changes during the “peritreatment” and “posttreatment” surveillance periods separately. The multivariable model included a time-by-time polynomial term to allow for the modeling of nonlinear trajectories.
We conducted 4 additional analyses. First, we explored the missingness of the data. We compared baseline demographics of those who reported annual income vs those who did not. We also added “prefer not to answer” and “missing” as additional covariates. Second, a series of interaction terms were added to look for potential effect modification between age, stage at diagnosis, and income. Third, we added ECOG performance status as an additional covariate. In our directed acyclic graphs, comorbidity burden was in the causal path between our exposure and outcome of interest (eFigure 3 in the Supplement). While we did not adjust for it in our main analysis, we did explore its effect through a preplanned sensitivity analysis. Finally, we questioned whether the association between income and survival might in part be mediated through treatment interruption. The study was underpowered to do a full mediation analysis.26 We did, however, look at the association between primary radiation treatment completion and income using a logistic regression model adjusted for cancer stage.
All analyses were 2-sided, and statistical significance was set at P < .05. Analyses were conducted using R statistical software package, version 3.6.3 (R Foundation for Statistical Computing).
Results
There were 717 patients who agreed to participate, of whom 555 reported their baseline income. There were no demographic differences between those who reported their income and those who did not (eTable 1 in the Supplement). Among patients with income data, mean (SD) age was 62.7 (10.7) years; 109 (20%) were women and 446 (80%) were men. The majority of participants were White (468 [84%]). Most were married/common law (381 [69%]). Median income was $50 000 to $59 000. Further demographic information can be found in Table 1.
Table 1. Demographic Characteristics of the Study Cohort (n = 555).
| Covariate | No. (%) |
|---|---|
| Age, y | |
| Mean (SD) | 62.7 (10.7) |
| Median (range) | 63 (24.7-92.2) |
| Sexa | |
| Female | 109 (20) |
| Male | 446 (80) |
| Race and ethnicitya | |
| Black/African Canadian | 7 (1) |
| East Asian | 28 (5) |
| Filipino | 6 (1) |
| Latino/Hispanic | 2 (<1) |
| Middle Eastern | 6 (1) |
| Mixed | 6 (1) |
| Native Canadian | 1 (<1) |
| South Asian | 25 (5) |
| White | 468 (84) |
| Otherb | 4 (1) |
| Missing | 2 |
| Marital statusa | |
| Married/common law | 381 (69) |
| Never married | 48 (9) |
| Separated/divorced/widowed | 123 (22) |
| Missing | 3 |
| Living arrangementsa,c | |
| Lives with spouse | 365 (68) |
| Lives alone | 106 (20) |
| Lives with friend/other family | 65 (12) |
| Lives in an institution | 4 (1) |
| Missing | 15 |
| Baseline household income ($CAD)a | |
| 0-29 999 | 132 (24) |
| 30 000-49 999 | 102 (18) |
| 50 000-89 999 | 154 (28) |
| ≥90 000 | 167 (30) |
| ECOG performance status | |
| 0 | 309 (50) |
| 1 | 294 (47) |
| 2 | 17 (3) |
| 3 | 3 (<1) |
| Missing | 94 |
| Stage, 7th edition AJCC | |
| 0 | 1 (<1) |
| I | 77 (14) |
| II | 63 (11) |
| III | 90 (16) |
| IVA | 282 (51) |
| IVB | 31 (6) |
| IVC | 11 (2) |
| Disease site | |
| Oropharynx | 244 (44) |
| Lip and oral cavity | 129 (23) |
| Larynx | 88 (16) |
| Nasopharynx | 35 (6) |
| Hypopharynx | 20 (4) |
| Salivary glands | 6 (1) |
| Nasal cavity | 9 (2) |
| Paranasal sinus | 5 (1) |
| Unknown primary | 19 (3) |
| HPV statusc | |
| Positive | 226 (41) |
| Negative | 109 (20) |
| HPV-unrelated | 220 (40) |
| FIT score at 1 y | |
| Mean (SD) | 15.5 (15.4) |
| Median (range) | 11.1 (0-86.1) |
| Missing | 364 |
Abbreviations: AJCC, American Joint Committee on Cancer; ECOG, Eastern Cooperative Oncology Group; FIT, Financial Index of Toxicity; HPV, human papillomavirus.
Self-reported by study participant.
Participants reported Other if their race and ethnicity was not listed as an option on the baseline demographic questionnaire.
Percentages total more than 100% due to rounding.
Median (range) follow-up was 26.2 (0.5-48.5) months. Kaplan-Meier curves are plotted in the Figure. Relative to the patients in the highest income quartile (≥$90 000), patients in the bottom income quartile (<$30 000) had the lowest 2-year disease-free survival (88% vs 67%) and overall survival (94% vs 82%) rates. Univariable and multivariable analyses supported the association between income and disease-free and overall survival (Table 2 and eTable 2 in the Supplement). The results were robust on sensitivity analysis when missing income scores were included as an additional covariate (eFigure 1 in the Supplement). Interaction terms between income and stage as well as income and age were both nonsignificant. The difference in disease-free (2.13; 95% CI, 1.22-3.71) and overall survival (2.01; 95% CI, 0.94-4.29) between the lowest quartile and the highest was strong even after adjustment for stage, HPV status, and age. Low-income individuals were at highest risk of not completing their radiation treatment (25% in the lowest income quartile vs 11% in the highest quartile; odds ratio, 3.19; 95% CI, 1.12-9.10, adjusted for stage).
Figure. Kaplan-Meier Plots of Disease-Free Survival and Overall Survival Stratified by Income.
Table 2. Association Between Income ($CAD) and Disease-Free and Overall Survival.
| Income | 2-y DFS, % (95% CI) | DFS | 2-y OS, % (95% CI) | OS | ||
|---|---|---|---|---|---|---|
| HR (95% CI) | aHR (95% CI)a | HR (95% CI) | aHR (95% CI)a | |||
| ≥$90 000 | 88 (83-93) | 1 [Reference] | 1 [Reference] | 94 (90-98) | 1 [Reference] | 1 [Reference] |
| $50 000-$89 999 | 82 (76-89) | 1.44 (0.82-2.53) | 1.23 (0.70-2.17) | 90 (85-95) | 1.63 (0.76-3.51) | 1.25 (0.57-2.71) |
| $30 000-$49 999 | 84 (77-92) | 1.36 (0.72-2.56) | 1.12 (0.59-2.15) | 92 (86-98) | 1.65 (0.70-3.88) | 1.14 (0.47-2.75) |
| <$30 000 | 67 (58-78) | 2.40 (1.39-4.12) | 2.13 (1.22-3.71) | 82 (74-90) | 2.69 (1.28-5.66) | 2.01 (0.94-4.29) |
Abbreviations: aHR, adjusted hazard ratio (multivariable); DFS, disease-free survival; HR, hazard ratio (univariate); OS, overall survival.
Adjusted for stage, human papillomavirus status, and age.
Median FIT scores were 11.1 out of 100 (range, 0-86.1) at the 12-month assessment and 11.1 (range, 0-70.3) at the 24-month assessment. Financial toxicity was significantly associated with baseline income. The median FIT score for the lowest income quartile was 17.6 (IQR, 11.1-29.8) compared with 8.3 (IQR, 2.8-13.9) for the highest income quartile (eTable 3 and eFigure 2 in the Supplement). In risk-adjusted models, low-income patients had 12-month FIT scores that were, on average, 134% higher (95% CI, 16%-253%) than that of high-income patients (Table 3). During the peritreatment treatment window, patients in the lowest income quartile had health utility scores that were, on average, 0.104 points lower (95% CI, 0.026-0.182) in adjusted analysis. In posttreatment surveillance, health utility scores remained lower for patients in the lowest income quartile (−0.127; 95% CI, −0.190 to −0.064). For all patients in posttreatment surveillance, a dose–response association was observed between income and health utility (Table 4).
Table 3. Financial Index of Toxicity Scoresa.
| Income | FIT score, mean (SD) | Univariable change in FIT, % (95% CI)b | Multivariable change in FIT, % (95% CI)c |
|---|---|---|---|
| ≥$90 000 | 11.7 (12.7) | 1 [Reference] | 1 [Reference] |
| $50 000-$89 999 | 15.2 (14.8) | 25.3 (−19.2 to 69.9) | 35.1 (−13.6 to 83.7) |
| $30 000-$49 999 | 17.8 (15.2) | 63.7 (−8.3 to 135.7) | 75.2 (−7.2 to 157.6) |
| <$30 000 | 22.6 (20.0) | 116.2 (21.1 to 211.3) | 134.6 (16.3 to 252.9) |
Abbreviation: FIT, Financial Index of Toxicity.
Association between income (exposure) and FIT score at 12 months (outcome). Results were analyzed using linear models.
Regression fit on outcome = log(FIT score + 1), where FIT score is out of 100. This allows us to interpret the exponentiated parameter estimate as the percent change in FIT scores compared with the reference category, income $90 000 or greater. Positive values indicate an increase in FIT scores (eg, 25% change FIT in this table represents a 25% increase in FIT scores compared with the highest income quartile) while negative values represent a decrease in FIT scores.
Adjusted for stage, human papillomavirus status, age, sex, race and ethnicity, marital status, living situation (alone vs with others), and disease site.
Table 4. Health Utilitya.
| Income | Utility changeb (95% CI) | |
|---|---|---|
| Univariablec | Multivariabled | |
| Peritreatment window | ||
| ≥$90 000 | 1 [Reference] | 1 [Reference] |
| $50 000-$89 999 | −0.006 (−0.060 to −0.048) | −0.006 (−0.061 to 0.048) |
| $30 000-$49 999 | −0.040 (−0.105 to 0.026) | −0.025 (−0.092 to 0.041) |
| <$30 000 | −0.113 (−0.191 to −0.034) | −0.104 (−0.182 to −0.026) |
| Posttreatment surveillance | ||
| ≥$90 000 | 1 [Reference] | 1 [Reference] |
| $50 000-$89 999 | −0.048 (−0.093 to −0.003) | −0.049 (−0.094 to −0.005) |
| $30 000-$49 999 | −0.080 (−0.137 to −0.023) | −0.064 (−0.122 to −0.007) |
| <$30 000 | −0.141 (−0.205 to −0.077) | −0.127 (−0.190 to −0.064) |
Abbreviation: HUI3, Health Utilities Index–Mark 3.
Association between income (exposure) and health utility (outcome). Results were analyzed using linear models under a generalized estimating equations approach.
Parameter estimate from generalized linear models, reports the average change in utility for patients relative to the reference of $90 000 or greater.
Model structure: Response (HUI3) ~ Beta0 + Beta1(Income) + Beta2(≥3 months after treatment) + Beta3(Income × ≥3 months after treatment).
Model structure: Response (HUI3) ~ Beta0 + Beta1(Income) + Beta2(≥3 months after treatment) + Beta3(Income × ≥3 months after treatment) + Beta4(Treatment modality) + Beta5(time since treatment started) + Beta6(squared time since treatment started) + Beta7(Age).
Discussion
In this cohort study, we observed a significant association between household income at time of cancer diagnosis and survival. This association was monotonic but not necessarily linear. Patients with HNC with an annual household income less than $30 000 appeared to be disproportionately affected. Community-based measures approximate income but make some assumptions about the homogeneity of income levels in an area and over time. This more granular data set confirms what has been reported in prior ecologic studies: that even within settings of universal health insurance, low-income individuals have worse oncologic outcomes. This vulnerable population experiences increased financial toxicity, reduced health status, and inferior disease-free survival.
The association between SES and cancer survival is complex and likely multifactorial. In part, observed disparities may be associated with delayed presentation and subsequent advanced cancer staging among low-income individuals.1,27 However, the delayed presentation mechanism is inconsistently observed.1,28 In our own analysis, disparities persisted even after adjusting for cancer stage. The association between SES and survival might also be explained by comorbidity burden. Comorbid conditions have been more commonly identified in low-income individuals and may lead to worsened outcomes.1 We conceptualized comorbidity burden as being a part of the causal path between income and survival. Nonetheless, the income–survival association persisted even after adjustment for performance status. Third, the association might be explained by the experience and skill of clinical teams, which could differ by a patient’s socioeconomic status.1 For instance, high-SES individuals might be more likely to seek out care in high-volume centers, which tend to have better oncologic outcomes.29,30 We controlled for this by restricting the study population to a single high-volume center. Finally, it may be that the inherent financial toxicity faced by low-SES patients leads to suboptimal treatment decisions and worsened oncologic outcomes.31
Financial toxicity is the objective financial burden and subjective financial distress experienced by patients because of their cancer treatment.15,32 This adverse effect of cancer treatment is amplified in low-income persons. Low-income individuals have the lowest level of financial reserves. They are more likely to work in an unsupportive environment and to experience a cancer-related job loss.33 Financial toxicity in this setting may lead to maladaptive coping strategies, including nonadherence to cancer treatment.31 In our own study, low-income patients were least likely to complete radiotherapy treatment, which has been associated with inferior tumor control.34
Compared with patients with other cancers, patients with HNC are likely to be at higher risk of financial toxicity because they are disproportionately poorer, less educated, and experience heightened levels of uncontrolled symptom burden and distress.35,36,37,38,39 Additionally, patients with HNC are faced with chronic treatment-related toxic effects that can further limit their ability to return to work. In 1 study, more than half of all HNC survivors either reduced their work hours or left the workforce entirely.40 Financial toxicity literature specific to HNC is small but growing.17,18,41,42,43 Identified risk factors include larynx and hypopharynx site, age younger than 65 years, and median county income less than $50 000 USD.42,43 In patients with HNC, financial toxicity has been associated with skipped clinic visits and medication nonadherence.43 Defining a high-risk cohort is a critical step toward targeted intervention.
The bulk of the literature examining associations between socioeconomic status and cancer survival comes from the US.3,44 While it is tempting to attribute disparities to a private payer model, comparative studies have shown that the association between income and cancer survival is only slightly reduced in countries with comprehensive health insurance programs.10 In Ontario, residents make no payments for medical services or hospitalization. There is no parallel private sector. Despite this, patients with HNC still incur many costs, including prescription drugs, travel costs, and allied health services. Some of these costs may be offset by private insurance; however, for many, substantial financial burden exists. For patients with HNC in Ontario, direct out-of-pocket costs varied from $360 to $1600 depending on treatment modality.17 These estimates likely underestimate the true financial burden because they do not account for indirect costs, such as lost income. Severe financial distress requiring bankruptcy protection after cancer diagnosis has been shown to be an independent risk factor for mortality.31
The present study identifies individuals with HNC with an income less than $30 000 as being at high risk for financial toxicity and inferior survival. Strategies to mitigate the detrimental effects of low income and financial toxicity were beyond the scope of this study. Nonetheless, other groups have described multipronged strategies to mitigate health disparities at the patient, physician, hospital, and insurer levels.15,45,46,47 This work serves as a call to action for cancer centers and provincial governments. Despite living in a system of universal health insurance, low-income individuals with HNC appear to have worse oncologic and functional outcomes. Targeting individuals with incomes less than $30 000 with added supports and financial navigation resources may help close current care gaps.16
Limitations
These results must be interpreted in the context of a few limitations. Baseline household income is a more accurate reflection of SES than many ecologic measures. However, it is not perfect. Income is age dependent, more unstable than other SES metrics, and carries a high nonresponse rate. It also does not reflect all key assets, such as family resources and disability insurance. Second, while there are many similarities between US, Canadian, and European health care systems, financial stressors likely differ somewhat by jurisdiction. Third, there is the potential for residual confounding. Due to the relatively low event rate in our survival models, we had a modest number of degrees of freedom. Covariates needed to be introduced judiciously. We could not adjust for all potential confounders, such as tobacco and alcohol use, psychosocial factors, and race and ethnicity.35,48,49,50 The fact that our study population was predominantly White also limits generalizability. The ECOG performance status is not an all-encompassing tool for assessing comorbidity burden. The study was also not sufficiently powered to assess the association between score on the FIT and survival. Finally, it is important to recognize that the data are not missing completely at random. Low-SES patients were more likely to be lost to follow-up. This differential censoring likely biases us toward the null, leading to an underestimation of the true effect.
Conclusions
In this cohort study, we observed a significant association between income, financial toxicity, health status, and survival. This work enriches our understanding of social inequalities for patients with HNC. The population with an annual income less than $30 000 is at high risk for worsened oncologic outcomes, poor health status, and financial toxicity. Future research efforts should be directed toward defining strategies and targeted interventions to support these individuals.
eTable 1. Demographic characteristics for those with and without reported incomes
eTable 2. Factors associated with overall and disease-free survival in univariable analysis
eTable 3. Mean and median reported health utility for a given income band
eFigure 1. Kaplan Meier plots of disease-free survival and overall survival stratified by income
eFigure 2. Mean health utility score by income band
eFigure 3. Causal directed acyclic graph of the association between household income (exposure) and survival (outcome)
References
- 1.Woods LM, Rachet B, Coleman MP. Origins of socio-economic inequalities in cancer survival: a review. Ann Oncol. 2006;17(1):5-19. doi: 10.1093/annonc/mdj007 [DOI] [PubMed] [Google Scholar]
- 2.Booth CM, Li G, Zhang-Salomons J, Mackillop WJ. The impact of socioeconomic status on stage of cancer at diagnosis and survival: a population-based study in Ontario, Canada. Cancer. 2010;116(17):4160-4167. doi: 10.1002/cncr.25427 [DOI] [PubMed] [Google Scholar]
- 3.Mackillop WJ, Zhang-Salomons J, Groome PA, Paszat L, Holowaty E. Socioeconomic status and cancer survival in Ontario. J Clin Oncol. 1997;15(4):1680-1689. doi: 10.1200/JCO.1997.15.4.1680 [DOI] [PubMed] [Google Scholar]
- 4.Bedir A, Abera SF, Efremov L, Hassan L, Vordermark D, Medenwald D. Socioeconomic disparities in head and neck cancer survival in Germany: a causal mediation analysis using population-based cancer registry data. J Cancer Res Clin Oncol. 2021;147(5):1325-1334. doi: 10.1007/s00432-021-03537-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.McDonald JT, Johnson-Obaseki S, Hwang E, Connell C, Corsten M. The relationship between survival and socio-economic status for head and neck cancer in Canada. J Otolaryngol Head Neck Surg. 2014;43(1):2. doi: 10.1186/1916-0216-43-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Chu KP, Shema S, Wu S, Gomez SL, Chang ET, Le QT. Head and neck cancer-specific survival based on socioeconomic status in Asians and Pacific Islanders. Cancer. 2011;117(9):1935-1945. doi: 10.1002/cncr.25723 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Nutting CM, Robinson M, Birchall M. Survival from laryngeal cancer in England and Wales up to 2001. Br J Cancer. 2008;99(1)(suppl 1):S38-S39. doi: 10.1038/sj.bjc.6604582 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Andersen ZJ, Lassen CF, Clemmensen IH. Social inequality and incidence of and survival from cancers of the mouth, pharynx and larynx in a population-based study in Denmark, 1994-2003. Eur J Cancer. 2008;44(14):1950-1961. doi: 10.1016/j.ejca.2008.06.019 [DOI] [PubMed] [Google Scholar]
- 9.Reitzel LR, Nguyen N, Zafereo ME, Li G, Wei Q, Sturgis EM. Neighborhood deprivation and clinical outcomes among head and neck cancer patients. Health Place. 2012;18(4):861-868. doi: 10.1016/j.healthplace.2012.03.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Boyd C, Zhang-Salomons JY, Groome PA, Mackillop WJ. Associations between community income and cancer survival in Ontario, Canada, and the United States. J Clin Oncol. 1999;17(7):2244-2255. doi: 10.1200/JCO.1999.17.7.2244 [DOI] [PubMed] [Google Scholar]
- 11.Groome PA, Schulze KM, Keller S, et al. Explaining socioeconomic status effects in laryngeal cancer. Clin Oncol (R Coll Radiol). 2006;18(4):283-292. doi: 10.1016/j.clon.2005.12.010 [DOI] [PubMed] [Google Scholar]
- 12.Auluck A, Walker BB, Hislop G, Lear SA, Schuurman N, Rosin M. Population-based incidence trends of oropharyngeal and oral cavity cancers by sex among the poorest and underprivileged populations. BMC Cancer. 2014;14(1):316. doi: 10.1186/1471-2407-14-316 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Megwalu UC. Impact of county-level socioeconomic status on oropharyngeal cancer survival in the United States. Otolaryngol Head Neck Surg. 2017;156(4):665-670. doi: 10.1177/0194599817691462 [DOI] [PubMed] [Google Scholar]
- 14.Shavers VL. Measurement of socioeconomic status in health disparities research. J Natl Med Assoc. 2007;99(9):1013-1023. [PMC free article] [PubMed] [Google Scholar]
- 15.Carrera PM, Kantarjian HM, Blinder VS. The financial burden and distress of patients with cancer: understanding and stepping-up action on the financial toxicity of cancer treatment. CA Cancer J Clin. 2018;68(2):153-165. doi: 10.3322/caac.21443 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Offodile AC II, Gallagher K, Angove R, Tucker-Seeley RD, Balch A, Shankaran V. Financial navigation in cancer care delivery: state of the evidence, opportunities for research, and future directions. J Clin Oncol. 2022;40(21):2291-2294. doi: 10.1200/JCO.21.02184 [DOI] [PubMed] [Google Scholar]
- 17.Hueniken K, Douglas CM, Jethwa AR, et al. Measuring financial toxicity incurred after treatment of head and neck cancer: development and validation of the Financial Index of Toxicity questionnaire. Cancer. 2020;126(17):4042-4050. doi: 10.1002/cncr.33032 [DOI] [PubMed] [Google Scholar]
- 18.Khan MN, Hueniken K, Manojlovic-Kolarski M, et al. Out-of-pocket costs associated with head and neck cancer treatment. Cancer Rep (Hoboken). 2022;5(7):e1528. doi: 10.1002/cnr2.1528 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP; STROBE Initiative . The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Ann Intern Med. 2007;147(8):573-577. doi: 10.7326/0003-4819-147-8-200710160-00010 [DOI] [PubMed] [Google Scholar]
- 20.Stephens RF, Noel CW, Su JS, et al. Mapping the University of Washington quality of life questionnaire onto EQ-5D and HUI-3 indices in patients with head and neck cancer. Head Neck. 2020;42(3):513-521. doi: 10.1002/hed.26031 [DOI] [PubMed] [Google Scholar]
- 21.Noel CW, Stephens RF, Su JS, et al. Mapping the EORTC QLQ-C30 and QLQ-H&N35, onto EQ-5D-5L and HUI-3 indices in patients with head and neck cancer. Head Neck. 2020;42(9):2277-2286. doi: 10.1002/hed.26181 [DOI] [PubMed] [Google Scholar]
- 22.Noel CW, Keshavarzi S, Forner D, et al. Construct validity of the EuroQoL-5 Dimension and the Health Utilities Index in head and neck cancer. Otolaryngol Head Neck Surg. 2022;166(5):877-885. doi: 10.1177/01945998211030173 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Noel CW, Lee DJ, Kong Q, et al. Comparison of health state utility measures in patients with head and neck cancer. JAMA Otolaryngol Head Neck Surg. 2015;141(8):696-703. doi: 10.1001/jamaoto.2015.1314 [DOI] [PubMed] [Google Scholar]
- 24.Harrell FE Jr, Lee KL, Califf RM, Pryor DB, Rosati RA. Regression modelling strategies for improved prognostic prediction. Stat Med. 1984;3(2):143-152. doi: 10.1002/sim.4780030207 [DOI] [PubMed] [Google Scholar]
- 25.Grambsch PM, Therneau TM. Proportional hazards tests and diagnostics based on weighted residuals. Biometrika. 1994;81(3):515-526. doi: 10.1093/biomet/81.3.515 [DOI] [Google Scholar]
- 26.Lee H, Cashin AG, Lamb SE, et al. ; AGReMA group . A guideline for reporting mediation analyses of randomized trials and observational studies: the AGReMA statement. JAMA. 2021;326(11):1045-1056. doi: 10.1001/jama.2021.14075 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Olsen MH, Bøje CR, Kjær TK, et al. Socioeconomic position and stage at diagnosis of head and neck cancer—a nationwide study from DAHANCA. Acta Oncol. 2015;54(5):759-766. doi: 10.3109/0284186X.2014.998279 [DOI] [PubMed] [Google Scholar]
- 28.Johnson S, Corsten MJ, McDonald JT, Chun J. Socio-economic factors and stage at presentation of head and neck cancer patients in Ottawa, Canada: a logistic regression analysis. Oral Oncol. 2010;46(5):366-368. doi: 10.1016/j.oraloncology.2010.02.010 [DOI] [PubMed] [Google Scholar]
- 29.Eskander A, Irish J, Groome PA, et al. Volume-outcome relationships for head and neck cancer surgery in a universal health care system. Laryngoscope. 2014;124(9):2081-2088. doi: 10.1002/lary.24704 [DOI] [PubMed] [Google Scholar]
- 30.Eskander A, Merdad M, Irish JC, et al. Volume-outcome associations in head and neck cancer treatment: a systematic review and meta-analysis. Head Neck. 2014;36(12):1820-1834. doi: 10.1002/hed.23498 [DOI] [PubMed] [Google Scholar]
- 31.Ramsey SD, Bansal A, Fedorenko CR, et al. Financial insolvency as a risk factor for early mortality among patients with cancer. J Clin Oncol. 2016;34(9):980-986. doi: 10.1200/JCO.2015.64.6620 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Olver I, ed. The MASCC Textbook of Cancer Supportive Care and Survivorship. Springer; 2018. doi: 10.1007/978-3-319-90990-5 [DOI] [Google Scholar]
- 33.Blinder V, Eberle C, Patil S, Gany FM, Bradley CJ. Women with breast cancer who work for accommodating employers more likely to retain jobs after treatment. Health Aff (Millwood). 2017;36(2):274-281. doi: 10.1377/hlthaff.2016.1196 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Ohri N, Rapkin BD, Guha C, Kalnicki S, Garg M. Radiation therapy noncompliance and clinical outcomes in an urban academic cancer center. Int J Radiat Oncol Biol Phys. 2016;95(2):563-570. doi: 10.1016/j.ijrobp.2016.01.043 [DOI] [PubMed] [Google Scholar]
- 35.Noel CW, Eskander A, Sutradhar R, et al. ; Enhanced Supportive Psycho-oncology Canadian Care (ESPOC) Group . Incidence of and factors associated with nonfatal self-injury after a cancer diagnosis in Ontario, Canada. JAMA Netw Open. 2021;4(9):e2126822. doi: 10.1001/jamanetworkopen.2021.26822 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Massa ST, Osazuwa-Peters N, Adjei Boakye E, Walker RJ, Ward GM. Comparison of the financial burden of survivors of head and neck cancer with other cancer survivors. JAMA Otolaryngol Head Neck Surg. 2019;145(3):239-249. doi: 10.1001/jamaoto.2018.3982 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Noel CW, Sutradhar R, Zhao H, et al. Patient-reported symptom burden as a predictor of emergency department use and unplanned hospitalization in head and neck cancer: a longitudinal population-based study. J Clin Oncol. 2021;39(6):675-684. doi: 10.1200/JCO.20.01845 [DOI] [PubMed] [Google Scholar]
- 38.Bubis LD, Davis L, Mahar A, et al. Symptom burden in the first year after cancer diagnosis: an analysis of patient-reported outcomes. J Clin Oncol. 2018;36(11):1103-1111. doi: 10.1200/JCO.2017.76.0876 [DOI] [PubMed] [Google Scholar]
- 39.Halpern MT, de Moor JS, Yabroff KR. Impact of pain on employment and financial outcomes among cancer survivors. J Clin Oncol. 2022;40(1):24-31. doi: 10.1200/JCO.20.03746 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Giuliani M, Papadakos J, Broadhurst M, et al. The prevalence and determinants of return to work in head and neck cancer survivors. Support Care Cancer. 2019;27(2):539-546. doi: 10.1007/s00520-018-4343-6 [DOI] [PubMed] [Google Scholar]
- 41.Baddour K, Fadel M, Zhao M, et al. The cost of cure: examining objective and subjective financial toxicity in head and neck cancer survivors. Head Neck. 2021;43(10):3062-3075. doi: 10.1002/hed.26801 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Mady LJ, Lyu L, Owoc MS, et al. Understanding financial toxicity in head and neck cancer survivors. Oral Oncol. 2019;95:187-193. doi: 10.1016/j.oraloncology.2019.06.023 [DOI] [PubMed] [Google Scholar]
- 43.Beeler WH, Bellile EL, Casper KA, et al. Patient-reported financial toxicity and adverse medical consequences in head and neck cancer. Oral Oncol. 2020;101:104521. doi: 10.1016/j.oraloncology.2019.104521 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Berg JW, Ross R, Latourette HB. Economic status and survival of cancer patients. Cancer. 1977;39(2):467-477. doi: [DOI] [PubMed] [Google Scholar]
- 45.Tefferi A, Kantarjian H, Rajkumar SV, et al. In support of a patient-driven initiative and petition to lower the high price of cancer drugs. Mayo Clin Proc. 2015;90(8):996-1000. doi: 10.1016/j.mayocp.2015.06.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.American Society of Clinical Oncology . American Society of Clinical Oncology position statement on addressing the affordability of cancer drugs. Accessed October 20, 2022. https://www.asco.org/sites/new-www.asco.org/files/content-files/advocacy-and-policy/documents/2017-ASCO-Position-Statement-Affordability-Cancer-Drugs-footer.pdf [DOI] [PubMed]
- 47.Boby JM, Rajappa S, Mathew A. Financial toxicity in cancer care in India: a systematic review. Lancet Oncol. 2021;22(12):e541-e549. doi: 10.1016/S1470-2045(21)00468-X [DOI] [PubMed] [Google Scholar]
- 48.Noel CW, Sutradhar R, Li Q, et al. Association of immigration status and Chinese and South Asian ethnicity with incidence of head and neck cancer. JAMA Otolaryngol Head Neck Surg. 2020;146(12):1125-1135. doi: 10.1001/jamaoto.2020.4197 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Noel CW, Sutradhar R, Li Q, et al. Chinese and South Asian ethnicity, immigration status and head and neck cancer outcomes: a population based study. Oral Oncol. 2021;113:105118. doi: 10.1016/j.oraloncology.2020.105118 [DOI] [PubMed] [Google Scholar]
- 50.Gudi S, O’Sullivan B, Hosni A, et al. Outcome and treatment toxicity in East-Indian versus White-Canadian patients with oral cavity cancer following postoperative (chemo-)radiotherapy delivered under similar multidisciplinary care: a propensity-matched cohort study. Oral Oncol. 2021;120:105419. doi: 10.1016/j.oraloncology.2021.105419 [DOI] [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. Demographic characteristics for those with and without reported incomes
eTable 2. Factors associated with overall and disease-free survival in univariable analysis
eTable 3. Mean and median reported health utility for a given income band
eFigure 1. Kaplan Meier plots of disease-free survival and overall survival stratified by income
eFigure 2. Mean health utility score by income band
eFigure 3. Causal directed acyclic graph of the association between household income (exposure) and survival (outcome)
