Treatment Patterns and Survival Outcomes in Patients With Breast Cancer on Medicaid, Pre- and Post-Expansion.

BACKGROUND: The objective of this study was to evaluate the impact of Medicaid expansion on breast cancer treatment and survival among Medicaid-insured women in Ohio, accounting for the timing of enrollment in Medicaid relative to their cancer diagnosis and post-expansion heterogeneous Medicaid eligibility criteria, thus addressing important limitations in previous studies. METHODS: Using 2011-2017 Ohio Cancer Incidence Surveillance System data linked with Medicaid claims data, we identified women aged 18 to 64 years diagnosed with local-stage or regional-stage breast cancer (n=876 and n=1,957 pre-expansion and post-expansion, respectively). We accounted for women's timing of enrollment in Medicaid relative to their cancer diagnosis, and flagged women post-expansion as Affordable Care Act (ACA) versus non-ACA, based on their income eligibility threshold. Study outcomes included standard treatment based on cancer stage and receipt of lumpectomy, mastectomy, chemotherapy, radiation, hormonal treatment, and/or treatment for HER2-positive tumors; time to treatment initiation (TTI); and overall survival. We conducted multivariable robust Poisson and Cox proportional hazards regression analysis to evaluate the independent associations between Medicaid expansion and our outcomes of interest, adjusting for patient-level and area-level characteristics. RESULTS: Receipt of standard treatment increased from 52.6% pre-expansion to 61.0% post-expansion (63.0% and 59.9% post-expansion in the ACA and non-ACA groups, respectively). Adjusting for potential confounders, including timing of enrollment in Medicaid, being diagnosed in the post-expansion period was associated with a higher probability of receiving standard treatment (adjusted risk ratio, 1.14 [95% CI, 1.06-1.22]) and shorter TTI (adjusted hazard ratio, 1.14 [95% CI, 1.04-1.24]), but not with survival benefits (adjusted hazard ratio, 1.00 [0.80-1.26]). CONCLUSIONS: Medicaid expansion in Ohio was associated with improvements in receipt of standard treatment of breast cancer and shorter TTI but not with improved survival outcomes. Future studies should elucidate the mechanisms at play.

Accounting for Medicaid expansion and regional policy and programs to advance equity in cancer prevention in the United States.

BACKGROUND: Many studies compare state-level outcomes to estimate changes attributable to Medicaid expansion. However, it is imperative to conduct more granular, demographic-level analyses to inform current efforts on cancer prevention among low-income adults. Therefore, the authors compared the volume of patients with cancer and disease stage at diagnosis in Ohio, which expanded its Medicaid coverage in 2014, with those in Georgia, a nonexpansion state, by cancer site and health insurance status. METHODS: The authors used state cancer registries from 2010 to 2017 to identify adults younger than 64 years who had incident female breast cancer, cervical cancer, or colorectal cancer. Multivariable Poisson regression was conducted by cancer type, health insurance, and state to examine the risk of late-stage disease, adjusting for individual-level and area-level covariates. A difference-in-differences framework was then used to estimate the differences in risks of late-stage diagnosis in Ohio versus Georgia. RESULTS: In Ohio, the largest increase in all three cancer types was observed in the Medicaid group after Medicaid expansion. In addition, significantly reduced risks of late-stage disease were observed among patients with breast cancer on Medicaid in Ohio by approximately 7% and among patients with colorectal cancer on Medicaid in Ohio and Georgia after expansion by approximately 6%. Notably, the authors observed significantly reduced risks of late-stage diagnosis among all patients with colorectal cancer in Georgia after expansion. CONCLUSIONS: More early stage cancers in the Medicaid-insured and/or uninsured groups after expansion suggest that the reduced cancer burden in these vulnerable population subgroups may be attributed to Medicaid expansion. Heterogeneous risks of late-stage disease by cancer type highlight the need for comprehensive evaluation frameworks, including local cancer prevention efforts and federal health policy reforms. PLAIN LANGUAGE SUMMARY: This study looked at how Medicaid expansion affected cancer diagnosis and treatment in two states, Ohio and Georgia. The researchers found that, after Ohio expanded their Medicaid program, there were more patients with cancer among low-income adults on Medicaid. The study also found that, among people on Medicaid, there were lower rates of advanced cancer at the time of diagnosis for breast cancer and colon cancer in Ohio and for colon cancer in Georgia. These findings suggest that Medicaid expansion may be effective in reducing the cancer burden among low-income adults.

Understanding Disparities in Receipt of Complex Gastrointestinal Cancer Surgery at a Small Geographic Scale.

OBJECTIVE: To define neighborhood-level disparities in the receipt of complex cancer surgery. BACKGROUND: Little is known about the geographic variation of receipt of surgery among patients with complex gastrointestinal (GI) cancers, especially at a small geographic scale. METHODS: This study included individuals diagnosed with 5 invasive, nonmetastatic, complex GI cancers (esophagus, stomach, pancreas, bile ducts, liver) from the Ohio Cancer Incidence Surveillance System during 2009 and 2018. To preserve patient privacy, we combined US census tracts into the smallest geographic areas that included a minimum number of surgery cases (n=11) using the Max-p-regions method and called these new areas "MaxTracts." Age-adjusted surgery rates were calculated for MaxTracts, and the Hot Spot analysis identified clusters of high and low surgery rates. US Census and CDC PLACES were used to compare neighborhood characteristics between the high- and low-surgery clusters. RESULTS: This study included 33,091 individuals with complex GI cancers located in 1006 MaxTracts throughout Ohio. The proportion in each MaxTract receiving surgery ranged from 20.7% to 92.3% with a median (interquartile range) of 48.9% (42.4-56.3). Low-surgery clusters were mostly in urban cores and the Appalachian region, whereas high-surgery clusters were mostly in suburbs. Low-surgery clusters differed from high-surgery clusters in several ways, including higher rates of poverty (23% vs. 12%), fewer married households (40% vs. 50%), and more tobacco use (25% vs. 19%; all P <0.01). CONCLUSIONS: This improved understanding of neighborhood-level variation in receipt of potentially curative surgery will guide future outreach and community-based interventions to reduce treatment disparities. Similar methods can be used to target other treatment phases and other cancers.

Excess cancer prevalence in men with HIV: A nationwide analysis of Medicaid data.

BACKGROUND: Cancer is one of the most common comorbidities in men living with HIV (MLWH). However, little is known about the MLWH subgroups with the highest cancer burden to which cancer prevention efforts should be targeted. Because Medicaid is the most important source of insurance for MLWH, we evaluated the excess cancer prevalence in MLWH on Medicaid relative to their non-HIV counterparts. METHODS: In this cross-sectional study using 2012 Medicaid Analytic eXtract data nationwide, we flagged the presence of HIV, 13 types of cancer, symptomatic HIV, and viral coinfections using codes from the International Classification of Diseases, Ninth Revision, Clinical Modification. The study population included individuals administratively noted to be of male sex (men), aged 18 to 64 years, with (n = 82,495) or without (n = 7,302,523) HIV. We developed log-binomial models with cancer as the outcome stratified by symptomatic status, age, and race/ethnicity. RESULTS: Cancer prevalence was higher in MLWH than in men without HIV (adjusted prevalence ratio [APR], 1.84; 95% confidence interval [CI], 1.78-1.90) and was higher among those with symptomatic HIV (APR, 2.74; 95% CI, 2.52-2.97) than among those with asymptomatic HIV (APR, 1.73; 95% CI, 1.67-1.79). The highest APRs were observed for anal cancer in younger men, both in the symptomatic and asymptomatic groups: APR, 312.97; 95% CI, 210.27-465.84, and APR, 482.26; 95% CI, 390.67-595.32, respectively. In race/ethnicity strata, the highest APRs were among Hispanic men for anal cancer (APR, 198.53; 95% CI, 144.54-272.68) and for lymphoma (APR, 9.10; 95% CI, 7.80-10.63). CONCLUSIONS: Given the Medicaid program's role in insuring MLWH, the current findings highlight the importance of the program's efforts to promote healthy behaviors and vaccination against human papillomavirus in all children and adolescents and to provide individualized cancer screening for MLWH.

An informatics infrastructure to catalyze cancer control research and practice.

PURPOSE: A disconnect often exists between those with the expertise to manage and analyze complex, multi-source data sets, and the clinical, social services, advocacy, and public health professionals who can pose the most relevant questions and best apply the answers. We describe development and implementation of a cancer informatics infrastructure aimed at broadening the usability of community cancer data to inform cancer control research and practice; and we share lessons learned. METHODS: We built a multi-level database known as The Ohio Cancer Assessment and Surveillance Engine (OH-CASE) to link data from Ohio's cancer registry with community data from the U.S. Census and other sources. Space-and place-based characteristics were assigned to individuals according to residential address. Stakeholder input informed development of an interface for generating queries based on geographic, demographic, and disease inputs and for outputting results aggregated at the state, county, municipality, or zip code levels. RESULTS: OH-CASE contains data on 791,786 cancer cases diagnosed from 1/1/2006 to 12/31/2018 across 88 Ohio counties containing 1215 municipalities and 1197 zip codes. Stakeholder feedback from cancer center community outreach teams, advocacy organizations, public health, and researchers suggests a broad range of uses of such multi-level data resources accessible via a user interface. CONCLUSION: OH-CASE represents a prototype of a transportable model for curating and synthesizing data to understand cancer burden across communities. Beyond supporting collaborative research, this infrastructure can serve the clinical, social services, public health, and advocacy communities by enabling targeting of outreach, funding, and interventions to narrow cancer disparities.

Is Medicaid Expansion Narrowing Gaps in Surgical Disparities for Low-Income Breast Cancer Patients?

BACKGROUND: The objective of this study is to understand the effect of Medicaid expansion under the Affordable Care Act (ACA) on patterns of surgical care among low-income breast cancer patients. Emerging literature suggests cancer patients in Medicaid expansion states are presenting with earlier stages of disease. However, less is known regarding the implications of Medicaid expansion on patterns of surgical care in low-income women. PATIENTS AND METHODS: We compared nonmetastatic 30-64-year-old uninsured or Medicaid-insured Ohio breast cancer patients diagnosed 4 years before and 4 years after the state's 2014 Medicaid expansion (study group); the control group was the privately insured. Time-to-surgery (TTS) was defined as days from diagnosis to surgery. Demographic and treatment variables before and after expansion were examined in multivariate analysis. RESULTS: There was a 10.4% point increase in breast conservation therapy (BCT) in the study group (pre-ACA 26.3%, post-ACA 36.7%; p < 0.01) compared with a 5.8% point increase in the control group (pre-ACA 36.0%, post-ACA 41.8%; p < 0.01). Disparities in reconstruction narrowed between the study (pre-ACA 21.4%, post-ACA 34.5%; p < 0.01) and the control (37.0% pre-ACA, 44.1% post-ACA group p < 0.01) groups. There was no statistically significant change in mean TTS in the study group (pre-ACA 42.1 days, post-ACA 43.1 days p = 0.18) but there was an increase in TTS in the control group (pre-ACA 35.0 days, post ACA 37.0 days; p < 0.01). CONCLUSIONS: Medicaid expansion appears to have narrowed disparities in the utilization of BCT and reconstruction in low-income women. However, there was no improvement in surgical delay.

Post-Affordable Care Act Improvements in Cancer Stage Among Ohio Medicaid Beneficiaries Resulted From an Increase in Stable Coverage.

BACKGROUND: The mechanisms underlying improvements in early-stage cancer at diagnosis following Medicaid expansion remain unknown. We hypothesized that Medicaid expansion allowed for low-income adults to enroll in Medicaid before cancer diagnosis, thus increasing the number of stably-enrolled relative to those who enroll in Medicaid only after diagnosis (emergently-enrolled). METHODS: Using data from the 2011-2017 Ohio Cancer Incidence Surveillance System and Medicaid enrollment files, we identified individuals diagnosed with incident invasive breast (n=4850), cervical (n=1023), and colorectal (n=3363) cancer. We conducted causal mediation analysis to estimate the direct effect of pre- (vs. post-) expansion on being diagnosed with early-stage (-vs. regional-stage and distant-stage) disease, and indirect (mediation) effect through being in the stably- (vs. emergently-) enrolled group, controlling for individual-level and area-level characteristics. RESULTS: The percentage of stably-enrolled patients increased from 63.3% to 73.9% post-expansion, while that of the emergently-enrolled decreased from 36.7% to 26.1%. The percentage of patients with early-stage diagnosis remained 1.3-2.9 times higher among the stably-than the emergently-enrolled group, both pre-expansion and post-expansion. Results from the causal mediation analysis showed that there was an indirect effect of Medicaid expansion through being in the stably- (vs. emergently-) enrolled group [risk ratios with 95% confidence interval: 1.018 (1.010-1.027) for breast cancer, 1.115 (1.064-1.167) for cervical cancer, and 1.090 (1.062-1.118) for colorectal cancer. CONCLUSION: We provide the first evidence that post-expansion improvements in cancer stage were caused by an increased reliance on Medicaid as a source of stable insurance coverage.

Medicaid Expansion Associated With Reduction in Geospatial Breast Cancer Stage at Diagnosis Disparities.

CONTEXT: Prior studies demonstrate that Medicaid expansion has been associated with earlier-stage breast cancer diagnosis among women with low income, likely through increased access to cancer screening services. However, how this policy change has impacted geospatial disparities in breast cancer stage at diagnosis is unclear. OBJECTIVE: To examine whether there were reductions in geospatial disparities in advanced stage breast cancer at diagnosis in Ohio after Medicaid expansion. DESIGN: The study included 33 537 women aged 40 to 64 years diagnosed with invasive breast cancer from the Ohio Cancer Incidence Surveillance System between 2010 and 2017. The space-time scan statistic was used to detect clusters of advanced stage at diagnosis before and after Medicaid expansion. Block group variables from the Census were used to describe the contextual characteristics of detected clusters. RESULTS: The percentage of local stage diagnosis among women with breast cancer increased from 60.2% in the pre-expansion period (2010-2013) to 62.6% in the post-expansion period (2014-2017), while the uninsured rate among those women decreased from 13.7% to 7.5% during the same period. Two statistically significant ( P < .05) and 6 nonsignificant spatial clusters ( P > .05) of advanced stage breast cancer cases were found in the pre-expansion period, while none were found in the post-expansion period. These clusters were in the 4 largest metropolitan areas in Ohio, and individuals inside the clusters were more likely to be disadvantaged along numerous socioeconomic factors. CONCLUSIONS: Medicaid expansion has played an important role in reducing geospatial disparities in breast cancer stage at diagnosis, likely through the reduction of advanced stage disease among women living in socioeconomically disadvantaged communities.

The effect of Medicaid expansion among adults from low-income communities on stage at diagnosis in those with screening-amenable cancers.

BACKGROUND: Several states have opted to expand Medicaid under the Patient Protection and Affordable Care Act (ACA), which offers insurance coverage to low-income individuals up to 138% of the federal poverty level. This expansion of Medicaid to a medically vulnerable population potentially can reduce cancer outcome disparities, especially among patients with screening-amenable cancers. The objective of the current study was to estimate the effect of Medicaid expansion on the percentage of adults from low-income communities with screening-amenable cancers who present with metastatic disease. METHODS: Using state cancer registry data linked with block group-level income data, a total of 12,760 individuals aged 30 to 64 years who were diagnosed with incident invasive breast (female), cervical, colorectal, or lung cancer from 2011 through 2016 and who were uninsured or had Medicaid insurance at the time of diagnosis were identified. This sample was probability weighted based on income to reflect potential Medicaid eligibility under the ACA's Medicaid expansion. A multivariable logistic model then was fitted to examine the independent association between the exposure (pre-expansion [years 2011-2013] vs postexpansion [years 2014-2016]) and the outcome (metastatic vs nonmetastatic disease at the time of diagnosis). RESULTS: After adjusting for potential confounders, individuals who were diagnosed postexpansion were found to have 15% lower odds of having metastatic disease compared with those who were diagnosed pre-expansion (adjusted odds ratio, 0.85; 95% confidence interval, 0.77-0.93). As a control, a separate analysis that focused on individuals with private insurance who resided in high-income communities found nonsignificant postexpansion (vs pre-expansion) changes in the outcome (adjusted odds ratio, 1.02; 95% confidence interval, 0.96-1.09). CONCLUSIONS: Medicaid expansion is associated with a narrowing of a critical cancer outcome disparity in adults from low-income communities.

Development and validation of a model to predict outcomes of colon cancer surveillance.

PURPOSE: Clinical trials suggest that intensive surveillance of colon cancer (CC) survivors to detect recurrence increases curative-intent treatment, although any survival benefit of surveillance as currently practiced appears modest. Realizing the potential of surveillance will require tools for identifying patients likely to benefit and for optimizing testing regimens. We describe and validate a model for predicting outcomes for any schedule of surveillance in CC survivors with specified age and cancer stage. METHODS: A Markov process parameterized based on individual-level clinical trial data generates natural history events for simulated patients. A utilization submodel simulates surveillance and diagnostic testing. We validate the model against outcomes from the follow-up after colorectal surgery (FACS) trial. RESULTS: Prevalidation sensitivity analysis showed no parameter influencing curative-intent treatment by > 5.0% or overall five-year survival (OS5) by > 1.5%. In validation, the proportion of recurring subjects predicted to receive curative-intent treatment fell within FACS 95% CI for carcinoembryonic antigen (CEA)-intensive, computed tomography (CT)-intensive, and combined CEA+CT regimens, but not for a minimum surveillance regimen, where the model overestimated recurrence and curative treatment. The observed OS5 fell within 95% prediction intervals for all regimens. CONCLUSION: The model performed well in predicting curative surgery for three of four FACS arms. It performed well in predicting OS5 for all arms.

Factors associated with prescription pain reliever misuse: An analysis of statewide data.

OBJECTIVES: Characterize the individual-level risk factors associated with recent prescription pain reliever misuse (PPRM). DESIGN: A multivariable logistic regression was developed with 21 covariates. SETTING AND PARTICIPANTS: The 2015 Ohio Medicaid Assessment Survey (OMAS) was analyzed. The 2015 OMAS was a complex-designed representative survey of all adult Ohioans, regardless of Medicaid status (n = 42,876). OUTCOME MEASURES: The main outcome was PPRM within the past year ("yes" vs. "no"). RESULTS: An estimated 2.75% of adults in Ohio reported PPRM in the past year (approximately 242,232 individuals). The regression model identified 11 covariates with significantly higher odds of PPRM within the past year. The 5 covariates with the largest, significant effect sizes were binge drinking habits (adjusted odds ratio [AOR] for binge drinking in the past month: 2.35, 95% CI 1.97-2.81; AOR for drinking in the past month but no bingeing: 1.43, 95% CI 1.22-1.67), did not get medical examination, medical supplies, mental health care, or eyeglasses in the past year because of cost (AOR 1.74, 95% CI 1.44-2.11), no insurance (AOR 1.67, 95% CI 1.15-2.45), emergency room visits in the past year (AOR 1.66, 95% CI 1.42-1.93), and had problems paying medical bills in the past year (AOR 1.52, 95% CI 1.29-1.80). CONCLUSION: In this cross-sectional study of 42,847 adult Ohioans, the covariates associated with PPRM in the past year were identified and quantified. The significant covariates can be broadly grouped into substance use history, poor health care access, demographic characteristics, and chronic health conditions. The single covariate with the largest effect size was binge drinking in the past month; those individuals were more than twice as likely to have endorsed PPRM in the past year compared with those who did not drink at all in the past month.

US Cancer Detection Decreased Nearly 9 Percent During The First Year Of The COVID-19 Pandemic.

We investigated the impact of the COVID-19 pandemic on cancer detection, using data from the Surveillance, Epidemiology, and End Results Program, which recently released data through the first year of the pandemic (2020). Across all cancer sites, cancer incidence fell by 8.7 percent. The most common cancers that experienced the largest disruptions were lung and bronchus, melanoma of the skin, and thyroid cancer.

Geographic Variation and Risk Factor Association of Early Versus Late Onset Colorectal Cancer.

The proportion of patients diagnosed with colorectal cancer (CRC) at age < 50 (early-onset CRC, or EOCRC) has steadily increased over the past three decades relative to the proportion of patients diagnosed at age ≥ 50 (late-onset CRC, or LOCRC), despite the reduction in CRC incidence overall. An important gap in the literature is whether EOCRC shares the same community-level risk factors as LOCRC. Thus, we sought to (1) identify disparities in the incidence rates of EOCRC and LOCRC using geospatial analysis and (2) compare the importance of community-level risk factors (racial/ethnic, health status, behavioral, clinical care, physical environmental, and socioeconomic status risk factors) in the prediction of EOCRC and LOCRC incidence rates using a random forest machine learning approach. The incidence data came from the Surveillance, Epidemiology, and End Results program (years 2000-2019). The geospatial analysis revealed large geographic variations in EOCRC and LOCRC incidence rates. For example, some regions had relatively low LOCRC and high EOCRC rates (e.g., Georgia and eastern Texas) while others had relatively high LOCRC and low EOCRC rates (e.g., Iowa and New Jersey). The random forest analysis revealed that the importance of community-level risk factors most predictive of EOCRC versus LOCRC incidence rates differed meaningfully. For example, diabetes prevalence was the most important risk factor in predicting EOCRC incidence rate, but it was a less important risk factor of LOCRC incidence rate; physical inactivity was the most important risk factor in predicting LOCRC incidence rate, but it was the fourth most important predictor for EOCRC incidence rate. Thus, our community-level analysis demonstrates the geographic variation in EOCRC burden and the distinctive set of risk factors most predictive of EOCRC.

Risk factors and geographic disparities in premature cardiovascular mortality in US counties: a machine learning approach.

Disparities in premature cardiovascular mortality (PCVM) have been associated with socioeconomic, behavioral, and environmental risk factors. Understanding the "phenotypes", or combinations of characteristics associated with the highest risk of PCVM, and the geographic distributions of these phenotypes is critical to targeting PCVM interventions. This study applied the classification and regression tree (CART) to identify county phenotypes of PCVM and geographic information systems to examine the distributions of identified phenotypes. Random forest analysis was applied to evaluate the relative importance of risk factors associated with PCVM. The CART analysis identified seven county phenotypes of PCVM, where high-risk phenotypes were characterized by having greater percentages of people with lower income, higher physical inactivity, and higher food insecurity. These high-risk phenotypes were mostly concentrated in the Black Belt of the American South and the Appalachian region. The random forest analysis identified additional important risk factors associated with PCVM, including broadband access, smoking, receipt of Supplemental Nutrition Assistance Program benefits, and educational attainment. Our study demonstrates the use of machine learning approaches in characterizing community-level phenotypes of PCVM. Interventions to reduce PCVM should be tailored according to these phenotypes in corresponding geographic areas.

An Analysis of Cancer Center-Provided Rideshare Utilization for Radiation Therapy.

OBJECTIVES: The potential of rideshare services to facilitate timely radiation therapy (RT), especially for resource-limited patients, is understudied. METHODS: Patients (n = 63) who received 73 courses of RT (1,513 fractions) and utilized free hospital-provided rideshare service (537 rides) were included in this retrospective study. A multidimensional analysis was conducted including a comparison of demographic, disease characteristics, and treatment completion data; a revenue analysis to evaluate the financial impact of rideshare services; and a geospatial analysis to evaluate community-level characteristics of patients. RESULTS: Median age was 59; most were female (56%) and self-identified as Black or African American (56%), not working (91%), not partnered (83%), high school educated or less (78%), and insured with Medicaid (51%). Geospatial analysis revealed that patients lived in communities with significantly higher rates of resource deprivation. Median rideshare distance was 6.4 miles (interquartile range 3.4-11.2) with a median cost of $13.04 per rideshare (interquartile range 9-19). Of the rideshare-facilitated treatments, 100% were completed, with an overall course completion rate of 97.3% compared with 85.4% for those who did not use rideshare (P = .001); two patients discontinued RT for reasons unrelated to transportation. High rideshare utilization (n = 32), defined as utilization ≥ 45% of the treatment course, was associated with significantly shorter treatment courses and lower radiation doses compared with low rideshare utilization (P = .04). Total rideshare cost for high utilizers and whole cohort was $11,589 and $16,895, facilitating an estimated revenue of $401,952 and $1,175,119, respectively. CONCLUSIONS: Free hospital-provided rideshare service is economically feasible and associated with high RT completion rates. It may help enhance quality radiation care for those who come from resource-limited communities.

Cancer burden in women with HIV on Medicaid: A nationwide analysis.

BACKGROUND: Cancer is the leading cause of death in people living with HIV. In the United States, nearly 1 in 4 people living with HIV are women, more than half of whom rely on Medicaid for healthcare coverage. OBJECTIVE: The objective of this study is to evaluate the cancer burden of women living with HIV on Medicaid. DESIGN: We conducted a cross-sectional study of women 18-64 years of age enrolled in Medicaid during 2012, using data from Medicaid Analytic eXtract files. METHODS: Using International Classification of Diseases, Ninth Revision, Clinical Modification diagnosis codes, we identified women living with HIV (n = 72,508) and women without HIV (n = 17,353,963), flagging the presence of 15 types of cancer and differentiating between AIDS-defining cancers and non-AIDS-defining cancers. We obtained adjusted prevalence ratios and 95% confidence intervals for each cancer and for all cancers combined, using multivariable log-binomial models, and additionally stratifying by age and race/ethnicity. RESULTS: The highest adjusted prevalence ratios were observed for Kaposi's sarcoma (81.79 (95% confidence interval: 57.11-117.22)) and non-Hodgkin's lymphoma (27.69 (21.67-35.39)). The adjusted prevalence ratios for anal and cervical cancer, both of which were human papillomavirus-associated cancers, were 19.31 (17.33-21.51) and 4.20 (3.90-4.52), respectively. Among women living with HIV, the adjusted prevalence ratio for all cancer types combined was about two-fold higher (1.99 (1.86-2.14)) in women 45-64 years of age than in women 18-44 years of age. For non-AIDS-defining cancers but not for AIDS-defining cancers, the adjusted prevalence ratios were higher in older than in younger women. There was no significant difference in the adjusted prevalence ratios for all cancer types combined in the race/ethnicity-stratified analyses of the women living with HIV cohort. However, in cancer type-specific sub-analyses, differences in adjusted prevalence ratios between Hispanic versus non-Hispanic women were observed. For example, the adjusted prevalence ratio for Hispanic women for non-Hodgkin's lymphoma was 2.00 (1.30-3.07) and 0.73 (0.58-0.92), respectively, for breast cancer. CONCLUSION: Compared to their counterparts without HIV, women living with HIV on Medicaid have excess prevalence of cervical and anal cancers, both of which are human papillomavirus related, as well as Kaposi's sarcoma and lymphoma. Older age is also associated with increased burden of non-AIDS-defining cancers in women living with HIV. Our findings emphasize the need for not only cancer screening among women living with HIV but also for efforts to increase human papillomavirus vaccination among all eligible individuals.

Phenotype Discovery and Geographic Disparities of Late-Stage Breast Cancer Diagnosis across U.S. Counties: A Machine Learning Approach.

BACKGROUND: Disparities in the stage at diagnosis for breast cancer have been independently associated with various contextual characteristics. Understanding which combinations of these characteristics indicate highest risk, and where they are located, is critical to targeting interventions and improving outcomes for patients with breast cancer. METHODS: The study included women diagnosed with invasive breast cancer between 2009 and 2018 from 680 U.S. counties participating in the Surveillance, Epidemiology, and End Results program. We used a machine learning approach called Classification and Regression Tree (CART) to identify county "phenotypes," combinations of characteristics that predict the percentage of patients with breast cancer presenting with late-stage disease. We then mapped the phenotypes and compared their geographic distributions. These findings were further validated using an alternate machine learning approach called random forest. RESULTS: We discovered seven phenotypes of late-stage breast cancer. Common to most phenotypes associated with high risk of late-stage diagnosis were high uninsured rate, low mammography use, high area deprivation, rurality, and high poverty. Geographically, these phenotypes were most prevalent in southern and western states, while phenotypes associated with lower percentages of late-stage diagnosis were most prevalent in the northeastern states and select metropolitan areas. CONCLUSIONS: The use of machine learning methods of CART and random forest together with geographic methods offers a promising avenue for future disparities research. IMPACT: Local interventions to reduce late-stage breast cancer diagnosis, such as community education and outreach programs, can use machine learning and geographic modeling approaches to tailor strategies for early detection and resource allocation.