Лечение чернокожих и белых пожилых людей с множественной миеломой: анализ SEER-Medicare.

Introduction

Innovation in Multiple Myeloma

Myeloma is a cancer of the plasma cells that affects primarily older adults. After decades of limited progress, multiple myeloma (MM) treatment entered an era of incredible innovation in the late 1990s that continues to the present day. Over this period, there have been 21 new antimyeloma agents approved by the FDA from 9 novel drug classes [ 1 ]. As a result of this innovation, treatments have become more numerous, more effective, and less toxic [ 2 , 3 ]. Survival has improved dramatically during this time frame, though MM remains incurable [ 4 , 5 ]. Relative 5‐year median survival in the late 1990s was < 30%, and most recent data (2014–2020) estimate relative median 5‐year survival to be 61.1% [ 4 , 5 ].

Innovation and Health Disparities

Fundamental Cause Theory (FCT) proposes that inequity in health outcomes has its root cause in the inequitable distribution of health risks, resources, and opportunities based on social class [ 6 , 7 , 8 ]. FCT further proposes that when innovation in understanding, preventing, or managing illness arises, the potential for disparities to widen increases as individuals with higher social standing (often operationalized as White race, higher education level, higher income) have flexible resources that allow them to better access and benefit from these innovations [ 8 ].

There is evidence that disparities in MM care and outcomes have emerged during this period of rapid innovation as FCT predicts. Waxman et al. observed significant survival improvements among White, but not Black MM patients from 1973 to 2005 [ 9 ]. A more recent analysis by Sun et al. covering the years 1981–2010 similarly reports survival improving for White patients faster than for Black patients [ 7 , 8 , 10 ].

Timely treatment initiation after diagnosis has been identified as an important component of quality cancer care and is a fundamental component of healthcare access [ 11 , 12 , 13 ]. In MM, treatment delay is associated with increased symptom burden, end organ damage, and decreased progression‐free survival [ 14 ]. In other cancers, treatment delays have been associated with tumor progression, upstaging, decreased quality of life, and decreased overall survival [ 15 , 16 , 17 , 18 , 19 ]. There have been few studies of disparities in timely treatment initiation among patients with MM, and none that have explored how these disparities may have changed over time. This study aims to describe MM treatment initiation patterns for Black and White older adults with MM and to test the hypothesis that as treatments have become more effective, Black/White disparities in timely treatment initiation will have widened.

Methods

Data Setting

This study utilized the linked SEER‐Medicare data, which combines Surveillance Epidemiology and End Results program (SEER) cancer registry data with Medicare enrollment claims data. SEER‐Medicare data include a large sample of older adults and provide access to data from SEER on cancer diagnoses, demographics, and survival linked to Medicare claims covering cancer treatment and healthcare utilization [ 20 , 21 ].

Study Population

We assembled a retrospective cohort of Black and White (Hispanic and non‐Hispanic) adults aged 65 years and older diagnosed with MM (ICD‐O3 M‐9732/3) between 2007 and 2017 from the SEER‐Medicare database. Subjects were required to have continuous fee for service Medicare Parts A and B coverage 12 months prior to and following diagnosis (or until death), and Part D coverage 12 months following diagnosis (or until death). Patients diagnosed at death or on autopsy were excluded.

Study Outcome

Our study outcome was time to treatment initiation measured in days from a patient's MM diagnosis (set as the first day of the month of diagnosis) to the date of the first claim for any antimyeloma agent. A list of antimyeloma agents was assembled from National Comprehensive Cancer Network (NCCN) guidelines for MM care from 1998 through 2017. We used the National Cancer Institutes (NCI) Cancer Medications Enquiry Database (CanMED) to identify NDC and HCPCS codes for these agents. See Appendix A in our Supporting Information for a listing of antimyeloma agents and their codes used in this study.

Key Independent Variables

Our key independent variables were race and diagnosis year. Race was categorized as Black or White (Hispanic/Latino or non‐Hispanic/Latino) as identified in the SEER registry. We conceptualize race as a marker of social standing (social advantage/marginalization). This conceptualization of race as a proxy for social standing encompasses the unequal distribution of risk factors (e.g., comorbidities) and resources (e.g., income, access to NCI designated cancer centers), leading to our decision to minimally adjust our results only by age and sex [ 22 , 23 ]. Diagnosis year was categorized into three time periods of similarly sized cohorts (2007–2010, 2011–2014, 2015–2017).

Other Variables

Other variables of interest were ascertained at the time of diagnosis, except for comorbidity score which was ascertained in the 12 months preceding diagnosis. These variables included age at diagnosis, sex (male or female), and SEER region (Northeast, South, Midwest, West). Rurality, measured using Rural Urban Commuting Area (RUCA) codes, categorized patients' residence as rural or non‐rural at the census tract level [ 24 ]. Individual and area level poverty were captured by assessing the percentage of the census tract living in poverty from the American Communities Survey and the receipt of Medicare low‐income subsidies (any subsidy/none) [ 25 , 26 ]. We also assessed whether MM patients had received any care within 12 months of diagnosis at an NCI‐designated cancer center (yes/no) as a proxy for access to specialty MM care [ 27 , 28 ]. We estimated comorbidity burden using a claims derived Charlson comorbidity score that excludes cancer related comorbidities [ 29 , 30 ]. Comorbidity scores were categorized as 0, 1–2, and 3+, where higher scores indicate increased comorbidity burden.

Analysis

We described characteristics of our overall and race‐stratified cohort using frequencies and percentages, where chi square tests were used to assess differences in characteristics by race. We evaluated unadjusted associations between patient characteristics and treatment initiation at 3, 6 and 12 months to identify factors associated with treatment initiation.

To estimate treatment initiation, we constructed Cox proportional hazard models stratified by race and diagnosis time period. This statistical approach is suitable for analyzing time‐to‐event data, allowing us to estimate the hazard ratio associated with treatment initiation while accounting for censoring and the timing of events [ 31 ]. Estimates were standardized based on age and sex across race with Black race as the reference category using standardized mortality ratio weights. We used the Fine‐Gray sub distribution hazard model to account for the presence of death as a competing risk in our models [ 32 ]. To assess the proportional hazards assumption we visually inspected the log negative log plots and Schoenfeld residuals [ 31 ]. In addition to reporting hazard ratios, we report cumulative incidence estimates of treatment initiation across several time points: within 3, 6, and 12 months in addition to hazard ratios [ 33 , 34 , 35 ].

All analyses were conducted using SAS v9.4 (Cary, North Carolina, USA). p ‐values for statistical tests below 0.05 were considered statistically significant. This study was deemed exempt by the University of North Carolina Institutional Review Board (IRB #: 21‐2848).

Results

Cohort Characteristics

Our analytic cohort consisted of 10,666 older adults diagnosed, of whom 1649 (15.5%) were Black and 9017 (84.5%) White (Table 1 ). Compared to White MM patients, Black MM patients were younger, more were female (64% vs. 50%), had more comorbidities, were more likely to live in the South (48.2% vs. 29.2%), live in areas with > 20% poverty (49.7% vs. 13.6%), and receive any Medicare low‐income subsidies (58.6% vs. 19.6%). Compared to White MM patients, Black MM patients were less likely to live in rural areas (10.7% vs. 17.1%), live in the West census region (13.7% vs. 30.2%), or receive care in the first 12 months at an NCI designated cancer center (19.8% vs. 28.5%).

Unadjusted Associations With Treatment Access

In unadjusted analyses, we identified several factors associated with treatment initiation across treatment initiation landmarks. MM patients were less likely to initiate treatment across all benchmarks if they were older, Black, had an increased comorbidity score, lived in the South or Northeast, did not receive care at an NCI Cancer Center, lived in census tracts with higher poverty, or received any Medicare low‐income subsidies. Rurality and sex did not appear to be associated with treatment initiation in unadjusted analyses (Table 2 ).

Black, White Treatment Initiation Disparities

We explored Black/White treatment initiation disparities using Cox proportional hazards models. We visually inspected the log negative log plots as well as the Schoenfeld residuals and found that the proportional hazards assumption was met. Hazard ratios (HR) reported in Table 3 , show that White MM patients were significantly more likely to initiate treatment during each time period, with HR and 95% confidence limits (CL) ranging from HR = 1.35 (CL = 1.25, 1.46) for those diagnosed 2007–2010 to HR = 1.36 (CL = 1.27, 1.44) for those diagnosed 2015–2017. We further estimated the cumulative incidence of treatment initiation at several initiation benchmarks (3, 6, and 12 months) (Table 4 ). We found that treatment initiation for both Black and White patients increased over time for each benchmark.

Disparities Over Time

To answer the question of whether Black/White disparities have changed over time we compared hazard ratios (HR) and 95% confidence limits of Black/White treatment initiation differences across our three diagnosis time periods. The HRs for all time periods were strikingly similar and had confidence intervals that were highly overlapping, suggesting that the Black/White treatment initiation disparity has not changed significantly over time (Table 3 ). Results from an analysis using year, and not time period, as our time variable resulted in small cell sizes and imprecise estimates though is included in Supporting Information (Table S1 ).

We further explored this question by comparing the Black/White differences in the cumulative incidence of treatment initiation across our three diagnosis time periods for our treatment initiation benchmarks of 3, 6, and 12 months. For all treatment initiation benchmarks, across all time periods, the difference in cumulative incidence between Black and White MM patients remained strikingly consistent, ranging from 0.09 (CL = 0.02–0.15) to 0.11 (CL = 0.05–0.17) (Table 4 ).

Discussion

In this analysis, we present one of the first studies to explore time to treatment initiation among MM patients in the era of multiple highly effective therapies, and the first to examine changes in Black/White treatment initiation disparities over time [ 27 , 33 , 34 , 35 ]. We present clear evidence that while treatment initiation has increased over time for both Black and White MM patients, significant and persistent treatment initiation disparities exist regardless of treatment initiation benchmark and have not changed over time. Contrary to our hypothesis that Black/White disparities would increase over time during an era of treatment innovation, we found no evidence that disparities in treatment initiation are significantly changing over time between 2007 and 2017.

Before proceeding we must discuss a feature of claims‐based studies of MM treatment that affect our estimates of treatment receipt. Active MM for which treatment is recommended and the related condition smoldering multiple myeloma (SMM), for which treatment is not recommended, share the same ICD‐O3 code (9732/3) in registry and claims data [ 36 ]. Ideally our estimates of treatment receipt would be among those appropriate for treatment; though, there are a proportion of patients in our analytic sample with SMM for whom non‐initiation of treatment is appropriate. Our best estimates of the percentage of those with ICD‐O3 9732/3 in claims or registry data who have SMM is between 14% and 16% [ 37 , 38 ].

One way used to address this potential misclassification is to use MM treatment initiation within a time frame (e.g., 120 days, 1 year) as inclusion criteria, thus using treatment itself as evidence that the patient has treatment appropriate active MM [ 33 , 35 ]. This approach to cohort construction precludes the identification of MM patients whose treatment initiation is delayed beyond the prespecified time frame or those who never initiate treatment. For this study of treatment delay, we chose to include all patients with ICD‐O3 9732/3 in our cohort, acknowledging that this approach includes some patients for whom non‐initiation of treatment is the clinically appropriate choice. Evidence suggests similar percentages of Black and White patients have active MM (vs. SMM) in SEER‐Medicare data (84.1% in White patients, 83.9% in Black patients) [ 37 ]. Therefore, we proceeded under the assumption that the proportion of those with SMM among individuals with an IDC‐O3 9732/3 diagnosis code was similar for both Black and White patients, and that treatment initiation estimates for Black and White patients would be affected similarly. This has almost certainly resulted in underestimates of treatment initiation among Medicare‐insured individuals with clinically active MM, though we believe this ascertainment anomaly affected both Black and White treatment estimates equally and has resulted in valid estimates of Black/White disparities.

We found that the percentage initiating treatment increased over time for both Black and White patients. This increase likely represents several converging factors including an improvement in the timeliness of MM diagnostic care, an increased appetite for early treatment among clinicians and patients as treatments have become more effective and less toxic, and the effects of an update to the myeloma diagnostic criteria that took place in 2014 [ 39 ]. This update recategorized some SMM patients as having active MM, increasing the proportion recommended for treatment [ 39 ].

Despite increasing treatment initiation over time, even in our most recent cohort (2015–2017) the proportion initiating treatment by 12 months remains low at 0.58 (CL = 0.53, 0.63) for Black MM patients and 0.69 (CL = 0.67, 0.70) for White MM patients. For reasons stated above, these are likely underestimates of treatment initiation, though the estimated 15% of our sample with SMM wouldn't fully account for the treatment non‐initiation of between 31% and 42% seen here [ 37 , 38 ]. On the other hand, SMM patients likely account for a higher percentage of subjects remaining untreated at later points in time post‐diagnosis, as other subjects move out of the at‐risk pool due to treatment initiation or death. This group of treatment‐ineligible SMM patients may account for the low estimated incidence of treatment initiation at 12 months, although this assumption cannot be verified.

Our results suggest remarkably consistent disparities in treatment initiation between Black and White MM patients, regardless of when a patient was diagnosed or at what time benchmark treatment is assessed. Our data suggest that Black MM patients remain at higher risk of delayed treatment compared to White MM patients and are less likely to ever initiate treatment. Patient navigation interventions have been shown to mitigate the effects of social marginalization on cancer care access, though they have not been studied extensively in hematologic malignancies [ 40 , 41 ]. Patient navigation models usually include systematically identifying patients at risk, assessing barriers, developing individualized plans to overcome barriers, and following up through treatment completion [ 42 , 43 ]. Some successful navigation interventions have also included proactive surveillance of patients' journeys and quality and equity performance feedback [ 44 , 45 ].

We hypothesized that as MM treatment has improved, disparities in care would have widened. Contrary to our hypothesis, Black‐White racial disparities in timely treatment initiation did not increase over time; though we note they are substantial and have not decreased. One possible explanation for why disparities have not widened is that important innovations in MM care occurred before 2007. For example, the agents bortezomib and lenalidomide were FDA approved in 2003 and 2006, respectively, and remain the backbone of therapy for those with newly diagnosed MM to this day [ 1 ]. Our reliance on Medicare Part D (which began in 2006) to identify treatment initiation limited us to assessing those diagnosed during or after 2007. That disparities have not improved is less surprising given evidence of limited progress on health care equity over the past two decades and rising income inequality across the United States [ 46 , 47 , 48 , 49 ].

This current study was designed to plainly describe differences in time to treatment initiation for Black and White MM patients, and not to examine what is driving or causing these differences. With that caveat, associations with time to treatment offer some hints as to what may be involved in increased time to treatment initiation for Black patients. Our unadjusted associations with treatment initiation suggest factors associated with lower levels of treatment initiation are over‐represented among Black patients. These include increased rates of 3+ comorbidities (37.1% in Black older adults, 22.6% in White), less NCI Cancer Center care (19.8% in Black adults, 28.5% in White), living in the south (48.2% of Black older adults, 29.2% of White), living in high poverty census tracts (49.7% in Black older adults, 13.6% in White), and receiving Medicare low‐income subsidies (an individual marker of poverty, present in 58.6% in Black older adults and 19.6% in White) (Tables 3 and 4 ). These results are striking though they should be interpreted cautiously. Future work is needed to explore the respective contributions of factors on treatment initiation, and further to develop and test interventions to improve care access and equity for all patients.

Limitations

This work has important limitations that should be considered. Like other studies utilizing SEER Medicare data, our findings may not be generalizable to other patient populations. Results presented here represent the experiences of older adults in the U.S. with fee‐for‐service Medicare insurance coverage which included Medicare Part D. Our study does not represent the experiences of younger patients, those without Medicare Part D, those with other insurance coverage, those without continuous coverage or those who lacked insurance coverage, potentially biasing our results. Finally, registry linked claims data are excellent resources to describe what is happening in real world cancer care but have limited ability to help us understand how and why the phenomenon we describe occurs. The substantial, significant, and persistent Black/White disparities in treatment initiation described here deserve to be further explored and this would be best accomplished through community‐engaged research aimed at uncovering the multilevel determinants of treatment delay.

Conclusion

In summary, we report substantial, significant, and persistent Black/White disparities in timely MM treatment initiation. These disparities are not significantly improving over time, though they do not appear to be widening as we had hypothesized. Our results suggest that Black patients are not only more likely to have delayed treatment compared to White patients, but may also be less likely to ever initiate treatment for their MM. We identified several demographic, socioeconomic, and geographic risk factors associated with delayed treatment initiation that are overrepresented among Black MM patients. Further work is needed to investigate how and why these risk factors contribute to disparate treatment initiation to inform the development of effective interventions.

Author Contributions

Matthew R. LeBlanc: conceptualization (equal), formal analysis (equal), investigation (equal), methodology (equal), project administration (equal), writing – original draft (lead), writing – review and editing (lead). Xi Zhou: conceptualization (equal), formal analysis (equal), methodology (equal), writing – review and editing (equal). Christopher D. Baggett: conceptualization (equal), funding acquisition (equal), methodology (equal), supervision (equal), writing – review and editing (equal). Jennifer L. Lund: conceptualization (equal), data curation (equal), formal analysis (equal), funding acquisition (equal), investigation (equal), methodology (equal), supervision (equal), writing – review and editing (equal). Christopher E. Jensen: conceptualization (equal), investigation (equal), methodology (equal), writing – review and editing (equal). Tzy‐Mey Kuo: formal analysis (equal), investigation (equal), methodology (equal), writing – review and editing (equal). Bradford E. Jackson: conceptualization (equal), methodology (equal), writing – review and editing (equal). Mya L. Roberson: conceptualization (equal), writing – review and editing (equal). Sascha A. Tuchman: conceptualization (equal), investigation (equal), writing – review and editing (equal). Samuel M. Rubinstein: conceptualization (equal), investigation (equal), writing – review and editing (equal). Eben I. Lichtman: conceptualization (equal), writing – review and editing (equal). Laura E. Green: data curation (equal), project administration (equal), resources (equal), writing – review and editing (equal). Katherine E. Reeder‐Hayes: conceptualization (equal), formal analysis (equal), funding acquisition (equal), investigation (equal), methodology (equal), supervision (equal), writing – review and editing (equal).

Disclosure

This study used the linked SEER‐Medicare database. The interpretation and reporting of these data are the sole responsibility of the authors. The authors acknowledge the efforts of the National Cancer Institute; Information Management Services (IMS) Inc.; and the Surveillance, Epidemiology, and End Results (SEER) Program tumor registries in the creation of the SEER‐Medicare database. The collection of cancer incidence data used in this study was supported by the California Department of Public Health pursuant to California Health and Safety Code Section 103885; Centers for Disease Control and Prevention's (CDC) National Program of Cancer Registries, under cooperative agreement 1NU58DP007156; the National Cancer Institute's Surveillance, Epidemiology and End Results Program under contract HHSN261201800032I awarded to the University of California, San Francisco, contract HHSN261201800015I awarded to the University of Southern California, and contract HHSN261201800009I awarded to the Public Health Institute. The ideas and opinions expressed herein are those of the author(s) and do not necessarily reflect the opinions of the State of California, Department of Public Health, the National Cancer Institute, and the Centers for Disease Control and Prevention or their Contractors and Subcontractors.

Conflicts of Interest

Jennifer L. Lund: Dr. Lund receives unrelated research support from Janssen and F‐Hoffman La Roche to the University of North Carolina at Chapel Hill. Chris E. Jensen: Consulting work with Abbvie. Xi Zhou, Chris D. Baggett, Tzy‐Mey Kuo, Bradford E. Jackson, Eben I. Lichtman, Laura E. Green: None to report. Mya L. Roberson: Consulting work with National Committee for Quality Assurance Outside the Scope of the Submitted Work. Sascha Tuchman: Honoraria/advisory board: Sanofi, GSK, Prothena, Kedrion; Research funding: Karyopharm, Sanofi, Caelum, Abbvie, BMS, Janssen. Sam Rubinstein: Consulting work with LAVA therapeutics, Janssen, Sanofi, EUSA Pharma, and Roche Diagnostics. Katherine E. Reeder‐Hayes: Grant support to institution from Pfizer Global Medical Foundation, unrelated to the submitted work.

Supporting information

Data S1: cam471563‐sup‐0001‐Supinfo.docx.

Table S1: cam471563‐sup‐0002‐TableS1.docx.

Acknowledgments

This work was supported by the UNC Cancer Information and Population Health Resource New Investigator program with funding provided by the University Cancer Research Fund via the state of North Carolina.

Data Availability Statement

The SEER‐Medicare data linkage used in this study is overseen by the National Cancer Institute (NCI). Release of SEER‐Medicare data is project specific. You must have an approved data use agreement with the NCI to access these data.

References

1 U.S. Food and Drug Administration , Hematology/Oncology (Cancer) Approvals & Safety Notifications ( U.S. Food and Drug Administration ), accessed March 15, 2023, https://www.fda.gov/drugs/resources‐information‐approved‐drugs/oncology‐cancer‐hematologic‐malignancies‐approval‐notifications .

2 K. C. Anderson , “ Progress and Paradigms in Multiple Myeloma ,” Clinical Cancer Research 22 , no. 22 ( 2016 ): 5419 – 5427 , 10.1158/1078-0432.ccr-16-0625 . 28151709 PMC5300651

3 S. K. Kumar , A. Dispenzieri , M. Q. Lacy , et al., “ Continued Improvement in Survival in Multiple Myeloma: Changes in Early Mortality and Outcomes in Older Patients ,” Leukemia 28 , no. 5 ( 2014 ): 1122 – 1128 , 10.1038/leu.2013.313 . 24157580 PMC4000285

4 National Cancer Institute , SEER Cancer Stat Facts: Myeloma ( National Cancer Institute ), accessed May 17, 2024, https://seer.cancer.gov/statfacts/html/mulmy.html .

5 L. J. Costa , I. K. Brill , J. Omel , K. Godby , S. K. Kumar , and E. E. Brown , “ Recent Trends in Multiple Myeloma Incidence and Survival by Age, Race, and Ethnicity in the United States ,” Blood Advances 1 , no. 4 ( 2017 ): 282 – 287 , 10.1182/bloodadvances.2016002493 . 29296944 PMC5727774

6 B. G. Link and J. Phelan , “ Social Conditions as Fundamental Causes of Disease ,” Journal of Health and Social Behavior ( 1995 ): 80 – 94 , 10.2307/2626958 . 7560851

7 J. C. Phelan , B. G. Link , A. Diez‐Roux , I. Kawachi , and B. Levin , “ ‘Fundamental Causes’ of Social Inequalities in Mortality: A Test of the Theory ,” Journal of Health and Social Behavior 45 , no. 3 ( 2004 ): 265 – 285 , 10.1177/002214650404500303 . 15595507

8 J. C. Phelan and B. G. Link , “ Controlling Disease and Creating Disparities: A Fundamental Cause Perspective ,” Journals of Gerontology. Series B, Psychological Sciences and Social Sciences 60 ( 2005 ): 27 – 33 , 10.1093/geronb/60.special_issue_2.s27 . 16251587

9 A. J. Waxman , P. J. Mink , S. S. Devesa , et al., “ Racial Disparities in Incidence and Outcome in Multiple Myeloma: A Population‐Based Study ,” Blood 116 , no. 25 ( 2010 ): 5501 – 5506 , 10.1182/blood-2010-07-298760 . 20823456 PMC3031400

10 T. Sun , S. Wang , H. Sun , J. Wen , G. An , and J. Li , “ Improved Survival in Multiple Myeloma, With a Diminishing Racial Gap and a Widening Socioeconomic Status Gap Over Three Decades ,” Leukemia & Lymphoma 59 , no. 1 ( 2018 ): 49 – 58 , 10.1080/10428194.2017.1335398 . 28595471

11 E. B. Cone , M. Marchese , M. Paciotti , et al., “ Assessment of Time‐to‐Treatment Initiation and Survival in a Cohort of Patients With Common Cancers ,” JAMA Network Open 3 , no. 12 ( 2020 ): e2030072 , 10.1001/jamanetworkopen.2020.30072 . 33315115 PMC7737088

12 J. Abraham and B. Bolwell , Time to Treatment Is a Priority. The ASCO Post ( 2019 ), https://ascopost.com/issues/december‐25‐2019/time‐to‐treatment‐is‐a‐priority/ .

13 S. E. Scott , F. M. Walter , A. Webster , S. Sutton , and J. Emery , “ The Model of Pathways to Treatment: Conceptualization and Integration With Existing Theory ,” British Journal of Health Psychology 18 , no. 1 ( 2013 ): 45 – 65 , 10.1111/j.2044-8287.2012.02077.x . 22536840

14 C. C. Kariyawasan , D. A. Hughes , M. M. Jayatillake , and A. B. Mehta , “ Multiple Myeloma: Causes and Consequences of Delay in Diagnosis ,” QJM 100 , no. 10 ( 2007 ): 635 – 640 , 10.1093/qjmed/hcm077 . 17846059

15 T. P. Hanna , W. D. King , S. Thibodeau , et al., “ Mortality due to Cancer Treatment Delay: Systematic Review and Meta‐Analysis ,” BMJ 371 ( 2020 ): m4087 , 10.1136/bmj.m4087 . 33148535 PMC7610021

16 M. Alexander , R. Beattie‐Manning , R. Blum , et al., “ Guidelines for Timely Initiation of Chemotherapy: A Proposed Framework for Access to Medical Oncology and Haematology Cancer Clinics and Chemotherapy Services ,” Internal Medicine Journal 46 , no. 8 ( 2016 ): 964 – 969 , 10.1111/imj.13157 . 27553996

17 M. Alexander , R. Blum , K. Burbury , et al., “ Timely Initiation of Chemotherapy: A Systematic Literature Review of Six Priority Cancers—Results and Recommendations for Clinical Practice ,” Internal Medicine Journal 47 , no. 1 ( 2017 ): 16 – 34 , 10.1111/imj.13190 . 27396268

18 A. Coca‐Pelaz , R. P. Takes , K. Hutcheson , et al., “ Head and Neck Cancer: A Review of the Impact of Treatment Delay on Outcome ,” Advanced Therapeutics 35 , no. 2 ( 2018 ): 153 – 160 , 10.1007/s12325-018-0663-7 . 29396681

19 N. R. Mayne , S. S. Bajaj , J. Powell , et al., “ Extended Delay to Treatment for Stage III‐IV Non‐Small‐Cell Lung Cancer and Survival: Balancing Risks During the COVID‐19 Pandemic ,” Clinical Lung Cancer 23 , no. 6 ( 2022 ): e362 – e376 , 10.1016/j.cllc.2022.05.001 . 35660355 PMC9068243

20 T. M. Kuo and L. R. Mobley , “ How Generalizable Are the SEER Registries to the Cancer Populations of the USA? ,” Cancer Causes & Control 27 , no. 9 ( 2016 ): 1117 – 1126 , 10.1007/s10552-016-0790-x . 27443170 PMC5180598

21 L. Enewold , H. Parsons , L. Zhao , et al., “ Updated Overview of the SEER‐Medicare Data: Enhanced Content and Applications ,” Journal of the National Cancer Institute Monographs 55 ( 2020 ): 3 – 13 , 10.1093/jncimonographs/lgz029 . PMC7225666 32412076

22 C. P. Jones , “ Invited Commentary: ‘Race,’ Racism, and the Practice of Epidemiology ,” American Journal of Epidemiology 154 , no. 4 ( 2001 ): 299 – 304 , 10.1093/aje/154.4.299 . 11495851

23 J. W. Jackson , “ Meaningful Causal Decompositions in Health Equity Research: Definition, Identification, and Estimation Through a Weighting Framework ,” Epidemiology 32 , no. 2 ( 2021 ): 282 – 290 , 10.1097/ede.0000000000001319 . 33394809 PMC8478117

24 United States Department of Agriculture. 2020 Rural‐urban commuting area (RUCA) codes, census tracts . ( 2020 ), https://www.ers.usda.gov/data‐products/rural‐urban‐commuting‐area‐codes .

25 United States Census Bureau , How the Census Bureau Measures Poverty ( 2023 ), https://www.census.gov/topics/income‐poverty/poverty/guidance/poverty‐measures.html .

26 M. E. Stahlman , The Medicare Drug Benefit: Update on the Low‐Income Subsidy , vol. 833 ( Issue Brief (George Washington University, National Health Policy Forum) , 2009 ), 1 – 22 . 19685590

27 S. Ailawadhi , P. Advani , D. Yang , et al., “ Impact of Access to NCI‐ and NCCN‐Designated Cancer Centers on Outcomes for Multiple Myeloma Patients: A SEER Registry Analysis ,” Cancer 122 , no. 4 ( 2016 ): 618 – 625 , 10.1002/cncr.29771 . 26565660

28 A. T. Freeman , M. Kuo , L. Zhou , et al., “ Influence of Treating Facility, Provider Volume, and Patient‐Sharing on Survival of Patients With Multiple Myeloma ,” Journal of the National Comprehensive Cancer Network 17 , no. 9 ( 2019 ): 1100 – 1108 , 10.6004/jnccn.2019.7298 . 31487686

29 C. N. Klabunde , A. L. Potosky , J. M. Legler , and J. L. Warren , “ Development of a Comorbidity Index Using Physician Claims Data ,” Journal of Clinical Epidemiology 53 , no. 12 ( 2000 ): 1258 – 1267 . 11146273 10.1016/s0895-4356(00)00256-0

30 M. E. Charlson , P. Pompei , K. L. Ales , and C. R. MacKenzie , “ A New Method of Classifying Prognostic Comorbidity in Longitudinal Studies: Development and Validation ,” Journal of Chronic Diseases 40 , no. 5 ( 1987 ): 373 – 383 . 3558716 10.1016/0021-9681(87)90171-8

31 D. Machin , Y. B. Cheung , and M. Parmar , Survival Analysis: A Practical Approach , 2nd ed. ( Wiley , 2006 ), 260 .

32 J. P. Fine and R. J. Gray , “ A Proportional Hazards Model for the Subdistribution of a Competing Risk ,” Journal of the American Statistical Association 94 , no. 446 ( 1999 ): 496 – 509 , 10.2307/2670170 .

33 V. Kumar , M. Alhaj‐Moustafa , L. Bojanini , et al., “ Timeliness of Initial Therapy in Multiple Myeloma: Trends and Factors Affecting Patient Care ,” JCO Oncology Practice 16 , no. 4 ( 2020 ): e341 – e349 , 10.1200/jop.19.00309 . 32048939

34 T. Jayakrishnan , V. Bakalov , N. S. Callander , S. Sadashiv , R. Wagner , and S. Ailawadhi , “ Impact of the Affordable Care Act on Timeliness to Treatment for Patients With Multiple Myeloma ,” Anticancer Research 40 , no. 10 ( 2020 ): 5727 – 5734 , 10.21873/anticanres.14587 . 32988898

35 T. Jayakrishnan , V. Bakalov , R. E. Wegner , and S. Sadashiv , “ Impact of Socioeconomic Factors on Timeliness of First Line Chemotherapy and Survival in Patients With Active Multiple Myeloma ,” Blood 134 ( 2019 ): 4729 , 10.1182/blood-2019-129147 .

36 National Comprehensive Cancer Network , NCCN Guidelines Version 1.2024 Multiple Myeloma ( National Comprehensive Cancer Network Inc ., 2023 ).

37 M. A. Fiala , J. Dukeman , S. A. Tuchman , M. Keller , R. Vij , and T. M. Wildes , “ Development of an Algorithm to Distinguish Smoldering Versus Symptomatic Multiple Myeloma in Claims‐Based Data Sets ,” JCO Clinical Cancer Informatics 1 ( 2017 ), 10.1200/cci.17.00089 . PMC6289182 30547156

38 A. Ravindran , A. C. Bartley , S. J. Holton , et al., “ Prevalence, Incidence and Survival of Smoldering Multiple Myeloma in the United States ,” Blood Cancer Journal 6 , no. 10 ( 2016 ): e486 , 10.1038/bcj.2016.100 . 27768092 PMC5098258

39 S. V. Rajkumar , “ Updated Diagnostic Criteria and Staging System for Multiple Myeloma ,” American Society of Clinical Oncology Educational Book/ASCO American Society of Clinical Oncology Meeting 35 ( 2016 ): e418 – e423 , 10.14694/edbk_15900910.1200/edbk_159009 . 27249749

40 B. M. Bernardo , X. Zhang , C. M. Beverly Hery , R. J. Meadows , and E. D. Paskett , “ The Efficacy and Cost‐Effectiveness of Patient Navigation Programs Across the Cancer Continuum: A Systematic Review ,” Cancer 125 , no. 16 ( 2019 ): 2747 – 2761 , 10.1002/cncr.32147 . 31034604

41 H. P. Freeman , “ The History, Principles, and Future of Patient Navigation: Commentary ,” Seminars in Oncology Nursing 29 , no. 2 ( 2013 ): 72 – 75 , 10.1016/j.soncn.2013.02.002 . 23651676

42 K. M. Freund , “ Implementation of Evidence‐Based Patient Navigation Programs ,” Acta Oncologica 56 , no. 2 ( 2017 ): 123 – 127 , 10.1080/0284186x.2016.1266078 . 28033027

43 K. M. Freund , T. A. Battaglia , E. Calhoun , et al., “ National Cancer Institute Patient Navigation Research Program: Methods, Protocol, and Measures ,” Cancer 113 , no. 12 ( 2008 ): 3391 – 3399 , 10.1002/cncr.23960 . 18951521 PMC2698219

44 S. Cykert , E. Eng , M. A. Manning , et al., “ A Multi‐Faceted Intervention Aimed at Black‐White Disparities in the Treatment of Early Stage Cancers: The ACCURE Pragmatic Quality Improvement Trial ,” Journal of the National Medical Association 112 , no. 5 ( 2020 ): 468 – 477 , 10.1016/j.jnma.2019.03.001 . 30928088

45 S. Cykert , E. Eng , P. Walker , et al., “ A System‐Based Intervention to Reduce Black‐White Disparities in the Treatment of Early Stage Lung Cancer: A Pragmatic Trial at Five Cancer Centers ,” Cancer Medicine 8 , no. 3 ( 2019 ): 1095 – 1102 , 10.1002/cam4.2005 . 30714689 PMC6434214

46 K. E. Pickett and R. G. Wilkinson , “ Income Inequality and Health: A Causal Review ,” Social Science & Medicine 128 ( 2015 ): 316 – 326 , 10.1016/j.socscimed.2014.12.031 . 25577953

47 B. L. Hardy , E. Krause , and J. P. Ziliak , “ Income Inequality in the United States, 1975–2022 ,” Fiscal Studies 45 , no. 2 ( 2024 ): 155 – 171 , 10.1111/1475-5890.12368 .

48 National Academies of Sciences, Engineering, and Medicine , Unequal Treatment Revisited: The Current State of Racial and Ethnic Disparities in Health Care: Proceedings of a Workshop ( National Academies Press , 2024 ), 80 . 38315785

49 National Academies of Sciences, Engineering, and Medicine , Ending Unequal Treatment: Strategies to Achieve Equitable Health Care and Optimal Health for All ( National Academies Press , 2024 ), 374 . 38924457