Search

PURPOSE Older hospitalized cancer patients face high risks of hospital mortality. Improved risk stratification could help identify high-risk patients who may benefit from future interventions, although we lack validated tools to predict in-hospital mortality for patients with cancer. We evaluated the ability of a high-dimensional machine learning prediction model to predict inpatient mortality and compared the performance of this model to existing prediction indices. METHODS We identified patients with cancer older than 75 years from the National Emergency Department Sample between 2016 and 2018. We constructed a high-dimensional predictive model called Cancer Frailty Assessment Tool (cFAST), which used an extreme gradient boosting algorithm to predict in-hospital mortality. cFAST model inputs included patient demographic, hospital variables, and diagnosis codes. Model performance was assessed with an area under the curve (AUC) from receiver operating characteristic curves, with an AUC of 1.0 indicating perfect prediction. We compared model performance to existing indices including the Modified 5-Item Frailty Index, Charlson comorbidity index, and Hospital Frailty Risk Score. RESULTS We identified 2,723,330 weighted emergency department visits among older patients with cancer, of whom 144,653 (5.3%) died in the hospital. Our cFAST model included 240 features and demonstrated an AUC of 0.92. Comparator models including the Modified 5-Item Frailty Index, Charlson comorbidity index, and Hospital Frailty Risk Score achieved AUCs of 0.58, 0.62, and 0.71, respectively. Predictive features of the cFAST model included acute conditions (respiratory failure and shock), chronic conditions (lipidemia and hypertension), patient demographics (age and sex), and cancer and treatment characteristics (metastasis and palliative care). CONCLUSION High-dimensional machine learning models enabled accurate prediction of in-hospital mortality among older patients with cancer, outperforming existing prediction indices. These models show promise in identifying patients at risk of severe adverse outcomes, although additional validation and research studying clinical implementation of these tools is needed.

Background African American patients with bladder cancer have inferior outcomes compared with non-Hispanic White (White) patients. We hypothesize that access to health care is a primary determinant of this disparity. We compared outcomes by race for patients with bladder cancer receiving care within the predominant hybrid-payer health-care model of the United States captured in the Surveillance, Epidemiology, and End Results (SEER) database with those receiving care within the equal-access model of the Veterans’ Health Administration (VHA). Methods African American and White patients diagnosed with bladder cancer were identified in SEER and VHA. Stage at presentation, bladder cancer–specific mortality (BCM), and overall survival (OS) were compared by race within each health-care system. Results The SEER cohort included 122 449 patients (93.7% White, 6.3% African American). The VHA cohort included 36 322 patients (91.0% White, 9.0% African American). In both cohorts, African American patients were more likely to present with muscle-invasive disease and metastases, but the differences between races were statistically significantly smaller in VHA. In SEER multivariable models, African American patients had worse BCM (hazard ratio [HR] = 1.22, 95% confidence interval [CI] = 1.15 to 1.29) and OS (HR = 1.26, 95% CI = 1.20 to 1.31). In contrast within the VHA, African American patients had similar BCM (HR = 0.97, 95% CI = 0.88 to 1.07) and OS (HR = 0.99, 95% CI = 0.93 to 1.05). Conclusions In this study of contrasting health-care models, receiving medical care in an equal-access system was associated with reduced differences in stage at presentation and eliminated disparities in survival outcomes for African American patients with bladder cancer. Our findings highlight the importance of reducing financial barriers to care to notably improve health equity and oncologic outcomes for African American patients.

Background Despite higher risks associated with prostate cancer, young African American men are poorly represented in prostate-specific antigen (PSA) trials, which limits proper evidence-based guidance. We evaluated the impact of PSA screening, alongside primary care provider utilization, on prostate cancer outcomes for these patients. Methods We identified African American men aged 40-55 years, diagnosed with prostate cancer between 2004 and 2017 within the Veterans Health Administration. Inverse probability of treatment-weighted propensity scores were used in multivariable models to assess PSA screening on PSA levels higher than 20, Gleason score of 8 or higher, and metastatic disease at diagnosis. Lead-time adjusted Fine-Gray regression evaluated PSA screening on prostate cancer–specific mortality (PCSM), with noncancer death as competing events. All statistical tests were 2-sided. Results The cohort included 4726 patients. Mean age was 51.8 years, with 84-month median follow-up. There were 1057 (22.4%) with no PSA screening prior to diagnosis. Compared with no screening, PSA screening was associated with statistically significantly reduced odds of PSA levels higher than 20 (odds ratio [OR] = 0.56, 95% confidence interval [CI] = 0.49 to 0.63; P < .001), Gleason score of 8 or higher (OR = 0.78, 95% CI = 0.69 to 0.88; P < .001), and metastatic disease at diagnosis (OR = 0.50, 95% CI = 0.39 to 0.64; P < .001), and decreased PCSM (subdistribution hazard ratio = 0.52, 95% CI = 0.36 to 0.76; P < .001). Primary care provider visits displayed similar effects. Conclusions Among young African American men diagnosed with prostate cancer, PSA screening was associated with statistically significantly lower risk of PSA levels higher than 20, Gleason score of 8 or higher, and metastatic disease at diagnosis and statistically significantly reduced risk of PCSM. However, the retrospective design limits precise estimation of screening effects. Prospective studies are needed to validate these findings.

OBJECTIVE Recurrence is known to predict laryngeal squamous cell cancer (LSCC) survival. Recurrence patterns in T4a LSCC are poorly characterized and represent a possible explanation for observed survival discrepancies by treatment rendered. STUDY DESIGN Retrospective database review. SETTING Veterans Affairs national database. METHODS Patients with T4a LSCC between 2000 and 2017 were identified and stratified by treatment (chemoradiotherapy [CRT] vs total laryngectomy + neck dissection + adjuvant therapy [surgical]). Primary outcomes were locoregional and distant recurrence. Secondary outcomes of overall mortality, larynx cancer mortality, and noncancer mortality were evaluated in Cox and Fine-Gray models. RESULTS A total of 1043 patients had comparable baseline demographics: 438 in the CRT group and 605 in the surgical group. Patients undergoing CRT had higher proportions of node positivity (64.6% vs 53.1%, P < .001). Locoregional and distant recurrence were less common in the surgical group (23.0% vs 37.2%, P < .001; 6.8% vs 13.3%, P < .001, respectively); however, distant metastatic rates did not differ within the N0 subgroup (P = .722). On multivariable regression, surgery demonstrated favorable locoregional recurrence (hazard ratio [HR], 0.49; 95% CI, 0.39-0.62; P < .001), distant recurrence (HR, 0.47; 95% CI, 0.31-0.71; P < .001), overall mortality (HR, 0.75; 95% CI, 0.64-0.87; P < .001), and larynx cancer mortality (HR, 0.69; 95% CI, 0.56-0.85; P < .001). CONCLUSION T4a LSCC survival discrepancies between surgical and nonsurgical treatment are influenced by varying recurrence behaviors. Surgery was associated with superior disease control and improved survival. Beyond the known benefit in locoregional control with surgery, there may be a protective effect on distant recurrence that depends on regional disease burden.

BACKGROUND Radiation therapy plays an important role for symptom palliation for intrathoracic malignancies ineligible for curative-intent therapy. Limited data exists regarding the role of stereotactic body radiation therapy (SBRT) versus conformal radiation in intrathoracic tumors for palliation. We report the efficacy of hypofractionated RT (or palliative SBRT) in the symptom management and durable control of lung and non-lung intrathoracic tumors. METHODS We performed a retrospective review of ninety-two thoracic lesions across 76 patients who completed palliative SBRT with doses ranging 25-50 Gy in 5-10 fractions between 2009 and 2019. Symptoms (cough, chest pain, hemoptysis, shortness of breath) were assessed at consult and 1-6 months follow-up. Local control was evaluated using follow-up CT imaging via RECIST criteria. Descriptive statistics were used to evaluate symptom palliation and Kaplan-Meier method to analyze local control. RESULTS Of primary lung (Cohort P) lesions, 40% showed stable symptoms, 30% never developed symptoms, and 19% showed symptom relief. CT imaging 1-6 months post-SBRT showed 91% with partial response (PR) or stable disease (SD) in Cohort P and 87% with PR or SD in metastatic (Cohort M) lesions. In patients with initial PR/SD, local control until death was achieved in 71% of Cohort P and 84% of Cohort M. Of our symptomatic patients (Cohort S), 98% showed no symptom progression post-radiotherapy. All patients with hemoptysis at presentation achieved hemostasis post-radiotherapy. CONCLUSIONS Palliative SBRT has the advantage of higher biologic dose without protracted course for patients with limited prognosis. Patients showed significant symptom palliation and long-term local control. Palliative SBRT represents a reasonable treatment modality for incurable thoracic malignancies.

ImportancePrimary care physicians (PCPs) are significant contributors of early cancer detection, yet few studies have investigated whether consistent primary care translates to improved downstream outcomes.ObjectiveTo evaluate the association of prediagnostic primary care use with metastatic disease at diagnosis and cancer-specific mortality (CSM).Design, Setting, and ParticipantsThis cohort study used databases with primary care and referral linkage from multiple Veterans’ Affairs centers from 2004 to 2017 and had a 68-month median follow-up. Analysis was completed between July 2021 and September 2022. Participants included veterans older than 39 years who had been diagnosed with 1 of 12 cancers. Inclusion criteria included known clinical staging, survival follow-up, cause of death, and receiving care at the Veterans Affairs health system (VA).ExposuresPrediagnostic PCP use, measured in the 5 years prior to diagnosis. PCP visits were binned into none (0 visits), some (1-4 visits), and annual (5 visits).Main Outcomes and MeasuresMetastatic disease at diagnosis, cancer-specific mortality (CSM) for entire cohort and stratified by tumor subtype.ResultsAmong 245 425 patients representing 12 tumor subtypes, mean age was 65.8 (9.3) years, and the cohort skewed male (97.6%), and White (76.1%), with higher levels of comorbidity (58.6% with Charlson Comorbidity Index scores ≥2). Compared with no prior visit, some PCP use was associated with 26% decreased odds of metastatic disease at diagnosis (odds ratio [OR], 0.74; 95% CI, 0.71-0.76; P P P P P P Conclusions and RelevanceIn this cohort study, increased primary care use before cancer diagnosis was associated with significant decreases in metastatic disease at diagnosis and cancer-related death, with potentially the greatest difference from annual use. PCPs play a vital role in cancer prevention, and additional resources should be allocated to assist these physicians.

Purpose Cancer patients frequently utilize the emergency department (ED) for a variety of diagnoses both related to and unrelated to their cancer, yet ED outcomes for cancer patients are not well documented. This study sought to define risks and identify predictors for inpatient admission and hospital mortality among cancer patients presenting to the ED. Patients and Methods We utilized the National Emergency Department Sample to identify patients with and without a diagnosis of cancer presenting to the ED between January 2016 and December 2018. We used multivariable mixed‐effects logistic regression models to assess the influence of cancer on outcomes of hospital admission after the ED visit and hospital mortality for the whole patient cohort and individual presenting diagnoses. Results There were 340 million weighted ED visits, of which 8.3 million (2.3%) were associated with a cancer diagnosis. Compared to non‐cancer patients, patients with cancer had an increased risk of inpatient admission (64.7% vs. 14.8%; p, On average, 2.8 million (2.3%) patients visit the emergency department. For each of the top 15 presenting diagnoses cancer patients had increased risks of hospitalization and death compared to non‐cancer patients. Among cancer patients, cancer site was the most robust individual predictor associated with risk of hospitalization or death.

Purpose: Neutrophil-lymphocyte ratio has been explored as a prognosticator in several cancer types, but its association with larynx cancer outcomes is not well known. We aimed to identify an optimal NLR cutoff point and examine the prognostic utility of this biomarker in patients with locoregionally advanced larynx cancer treated with curative intent. Methods: In the Veterans Affairs’ (VA) national database, we identified patients with locoregionally advanced (T3-4N0-3M0) laryngeal squamous cell carcinoma diagnosed between 2000 and 2017 and treated with curative intent. NLR cutoff points were calculated using Contal/O’Quigley’s method. Outcomes of larynx cancer-specific survival (CSS), overall survival (OS), and non-larynx cancer survival (NCS) were evaluated in multivariable Cox and Fine-Gray models. Results: In 1047 patients, the optimal pretreatment NLR cutoff was identified as 4.17 - 722 patients with NLR ⩽ 4.17, 325 patients with NLR > 4.17. The elevated NLR cohort had a higher proportion of T4 disease (39.4% vs 28.4%), node positive disease (52.3% vs 43.1%), and surgical treatment (43.7% vs 35.2%). In multivariable analysis, NLR > 4.17 was independently associated with worse OS (HR 1.31, 95% CI 1.12-1.54, P = .001) and worse CSS (HR 1.46, 95% CI 1.17-1.83, P Conclusion: In locoregionally advanced larynx cancer treated with curative intent, we identified elevated NLR to be associated with inferior OS and CSS. Further prospective studies are needed to investigate pretreatment NLR and our identified 4.17 cutoff as a potential larynx cancer-specific marker for this high risk population.

To evaluate the predictive utility of the Hospital Frailty Risk Score (HFRS), a stratification tool based on theRetrospective database review.Nationwide Readmissions Database (2017).Patients with head and neck cancer who underwent major surgical procedures were identified from the 2017 Nationwide Readmissions Database, representing 116 medical centers nationwide. Bivariate and multivariable logistic regression methods were used to identify factors associated with unplanned 30-day readmission, 30-day readmission mortality, and increased length of hospital stay.A total of 14,420 patients underwent major head and neck cancer surgery. Unplanned readmission occurred in 11% of patients. The most common reasons for unplanned readmission were procedural complications (26.5%), sepsis (7.3%), and respiratory failure (3.9%). Elevated frailty index (HFRS ≥5) was identified in 22% of patients. Frailty was associated with higher 30-day readmission rates (18.0% vs 9.5%,In this head and neck cancer surgical population, HFRS significantly predicted unplanned readmission. HFRS is a potential risk stratification tool and should be compared with other methods and explored in other cancer populations. Beyond the challenge of identifying at-risk patients, future work should explore potential interventions aimed at mitigating readmission.

The coronavirus disease 2019 (COVID-19) pandemic has raised concerns about delaying treatment for localized cancer and its impact on long-term outcomes.We aimed to investigate the impact of time to chemoradiation (CRT) on recurrence and survival outcomes for patients with muscle-invasive bladder cancer (MIBC).In the national Veterans Affairs' database, we identified patients with urothelial histology, MIBC (T2-4a/N0-3/M0) diagnosed between 2000 to 2018 and treated with definitive CRT. Time to treatment was defined as the number of days between date of diagnosis and start date of CRT. The cohort was stratified into90 (early) or ≥ 90 days (delayed) groups. Endpoints of locoregional failure (LRF), distant failure (DF), overall survival (OS), and bladder cancer-specific survival (BCS) were evaluated in multivariable Cox and Fine-Gray models.305 patients with MIBC underwent CRT - 190 (62.3%) received early CRT, 115 (37.7%) received delayed CRT. Multivariable analysis (including success of transurethral resection of bladder tumor and type of chemotherapy) revealed no difference in recurrence between groups - LRF HR 1.12 (95%CI 0.76-1.67, P = 0.56) and DF HR 1.03 (95%CI 0.70-1.53, P = 0.88). Similarly, there were no differences in survival outcomes. The lack of association was maintained at both earlier and later time cutoffs (60-120 days).Our findings suggest that a short-term delay in definitive therapy may not affect long-term outcomes for patients with MIBC undergoing CRT. This study does not endorse delays in therapy, but rather provides information to aid patients and clinicians navigate the unique challenges of MIBC care in both pandemic and non-pandemic times.

Purpose: Chemoradiation (CRT) is a definitive treatment option for muscle-invasive bladder cancer (MIBC). Despite its effectiveness, CRT is underused, in part owing to concerns of tolerability and the need for integrated multidisciplinary care. We investigated factors associated with and the impact of treatment discontinuation in patients with MIBC treated with CRT. Methods and Materials: In the US Veterans Affairs’ national database, we identified patients with urothelial histology, MIBC (T2-4a/N0-3/M0) diagnosed between 2000 and 2018 and treated with definitive-intent CRT. The primary endpoint of discontinued radiation was evaluated in a multivariable logistic regression. Secondary endpoints of 30-day and 90-day mortality, overall mortality, and nonbladder cancer mortality were evaluated in multivariable models. Results: Of 369 veterans with MIBC who underwent CRT, 30 patients (8.1%) did not complete radiation. The most common reasons for treatment discontinuation included comorbidities or infections necessitating hospital admission (63.3%) and treatment intolerance or declining performance status (26.7%). In multivariable logistic regression, variables associated with radiation discontinuation were creatinine clearance ≤ 50 (odds ratio [OR], 3.93; 95% CI, 1.63-9.50; P = .002), incomplete transurethral resection of bladder tumor (TURBT) (OR, 3.16; 95% CI, 1.15-8.63; P = .02), and nonpreferred chemotherapy (OR, 3.31; 95% CI, 1.31-8.36; P = .01). In the cohort that discontinued radiation, 30-day mortality was 33.3% and 90-day mortality was 50.0%, with the majority of deaths attributed to nonbladder cancer causes. No patient or tumor variables were associated with either endpoint. In the cohort that completed radiation, 30-day mortality was 2.7% and 90-day mortality was 6.8%. In multivariable analysis, radiation discontinuation was associated with worse overall mortality (hazard ratio [HR], 2.48; 95% CI, 1.36-4.50; P = .003) and worse nonbladder cancer mortality (HR, 2.32; 95% CI, 1.24-4.34; P = .008). Conclusions: With a low rate of treatment discontinuation, CRT is an effective and feasible treatment option for the typically elderly and comorbid population of patients with MIBC. In addition to identified predictors of treatment discontinuation (poor renal function, incomplete TURBT, etc.), further research is required to develop evidence-based guidelines for optimal patient selection.

OBJECTIVES Transoral laser microsurgery (TLM) is commonly utilized for early glottic cancer and offers favorable oncologic and functional outcomes. However, the survival implications of salvage therapy for recurrent or persistent disease have not been definitively characterized. STUDY DESIGN Retrospective, national database cohort study. METHODS Data were extracted from Veterans Health Affairs (VHA) Informatics and Computing Infrastructure (VINCI) concerning the TLM-based management of T1-T2 glottic squamous cell carcinoma patients between 2000 and 2017. Patients were characterized as either requiring TLM-only, or in cases of persistent or recurrent local disease, TLM plus change in treatment modality (radiotherapy, chemoradiotherapy, or open surgery). Predictors of overall survival (OS), cancer-specific survival (CSS), and salvage-free survival were evaluated via Cox and Fine-Gray models. RESULTS About 553 patients (70.9% T1a, 13.4% T1b, 15.7% T2) were included, with a median follow-up time of 74.5 months. The need for non-TLM salvage increased along with more advanced disease (11.7% T1a, 29.7% T1b, 32.2% T2). Compared to patients with T1a disease, those with T1b and T2 tumors initially treated with TLM had a significantly higher probability of receiving non-TLM salvage (T1b: HR 2.70, 95% CI: 1.61-4.54; T2: HR 3.02, 95% CI: 1.88-4.84). In a multivariable model, receipt of non-TLM salvage was not a significant predictor of either OS (HR = 0.91, 95% CI: 0.62-1.33, P = .624) or CSS (HR 1.21 95% CI 0.51-2.86, P = .667). CONCLUSION The majority of patients with early glottic cancer that are managed with TLM do not require additional salvage therapy. When non-TLM salvage was required, there was no decrement in OS or CSS. LEVEL OF EVIDENCE 4 Laryngoscope, 131:2766-2772, 2021.

PURPOSE/OBJECTIVE(S) Despite recent guideline recommendations, the utilization of low-dose CT (LDCT) for lung cancer screening has remained low. Little is known about the driving factors behind these low rates or the effect of real-world LDCT utilization on lung cancer outcomes. We hypothesize that LDCT screening will be associated with less advanced disease at diagnosis and improved cancer survival. Furthermore, we utilize a machine learning approach to investigate predictors for receiving LDCT screening. MATERIALS/METHODS We identified patients with non-small cell lung cancer (NSCLC) within the Veterans Health Administration (VHA) diagnosed from 2015 to 2017. LDCT screening was identified using procedural codes. We calculated LDCT rates using the pre-diagnostic observation period that included up to 3 years prior to diagnosis. Multivariable logistic regression assessed the influence of LDCT usage on metastatic disease at diagnosis. Lead-time correction using published LDCT lead-times was performed. Cancer-specific mortality (CSM) was evaluated using Fine-Gray regression and non-cancer death as a competing risk. A lasso machine learning model was used to identify important predictors for receiving LDCT screening. The data were split 75%/25% into training/testing datasets. We constructed the model with training data and evaluated performance within the test data using area under the curve (AUC). An AUC of 1.0 indicated perfect prediction. RESULTS The cohort included 4,664 patients diagnosed with NSCLC at a mean age of 67.8. The pre-diagnostic period ranged from 1-3 years, median of 1 year. Median follow-up was 12 months. Overall LDCT screening rate was 1.3%. From 2015-2017, this rate increased from 0.1% to 6.6%. Compared with no screening, Patients with ≥ 1 LDCT were more likely to present with stage I disease and less likely to present with stage IV disease (Stage I: 64% vs 42%, P < 0.01; Stage IV: 10% vs 22%, P < 0.01). In multivariable models, screened patients had lower risk of metastatic disease at diagnosis (odds ratio = 0.40, 95% confidence interval (CI) = [0.22-0.74], P = 0.03) and CSM (sub-distribution hazard ratio = 0.17, 95% CI = [0.04-0.63], P < 0.01). Our lasso model achieved an AUC of 0.82 and identified year of diagnosis and regional VA site of care as the most important predictors associated with receiving LDCT screening. Non-Hispanic White patients were more likely to undergo LDCT screening. CONCLUSION Within the VHA, utilization of LDCT screening is increasing though the usage remains low. Patients undergoing LDCT were diagnosed at earlier stages and were less likely to die from cancer. While overdiagnosis and lead-time bias need to be considered, these findings are consistent with randomized data and support the need for increased utilization of LDCT screening. Availability of LDCT appeared to be the main predictor of utilization. Providing access to more patients, including those in diverse racial and ethnic groups, should be a priority to reduce lung cancer mortality.

Purpose/objective(s) Outcomes for muscle-invasive bladder cancer (MIBC) treated with definitive chemoradiation (CRT) have largely remained stagnant in recent decades. To understand how best to improve outcomes for this population, we investigated patterns of failure after CRT for MIBC within the national Veterans Affairs' (VA) database. Materials/methods In the VA database, we identified patients diagnosed between 2000 and 2018 with urothelial histology, MIBC (T2-4a/N0-3/M0) who underwent CRT (defined as at least 50% of an intended definitive radiation regimen in conjunction with at least 1 cycle of chemotherapy). All patients' electronic health records were chart reviewed to obtain detailed clinical, pathologic, treatment (radiation dose, type of chemotherapy), and outcomes (dates and locations of recurrences, cause-specific death) data. Median times to recurrence are reported along with 25th-75th percentiles. Endpoints of local recurrence (LR, defined as recurrence in the bladder or upper tract), distant recurrence (DR, defined as recurrence beyond regional lymph nodes), overall mortality (OM), and bladder cancer-specific mortality (BCM) were evaluated in multivariable Cox and Fine-Gray models. Results Overall, 356 patients were included: median age 76, T2 disease 85%, lymph node positive (LN+) disease 5%. 12% of patients had carcinoma in situ, 32% had multifocal disease, and 33% had pretreatment hydronephrosis. Median follow up time was 77 months. 5-year OM was 69% (95% CI 64-74%) and 5-year BCM was 39% (95% CI 34-44%). 155 (44%) patients had no recurrence. In the 135 (38%) patients that developed DR, the median time to DR was 12.1 mo (8.0-23.4 mo). 106 (79%) patients of the DR group developed DR as a first recurrence with either DR only (n = 73) or DR and LR simultaneously (n = 33). In the 123 (35%) patients that developed LR, median time to LR was 11.1 mo (7.3-20.9 mo). Of the 92 patients who developed LR first without a simultaneous DR, 26 (28%) went on to develop DR. Median time from LR to DR in this group was 12.4 mo (3.0-27.0 mo). In multivariable models, LN+ disease, pretreatment hydronephrosis, and creatinine clearance (CrCl) ≤ 50 were associated with higher rates of DR, inferior OM, and inferior BCM. No patient, tumor, or treatment variables were associated with LR. Conclusion 44% of patients with MIBC treated with CRT had no recurrence while 35% developed LR and 38% developed DR. These results highlight the need for continued local surveillance after CRT as over a third of patients developed LR. Given the high proportion of DR and the advent of novel agents (antibody drug conjugates, immunotherapy, etc.) for advanced urothelial carcinoma, these data emphasize the need for clinical trials exploring such agents to improve outcomes for patients with MIBC undergoing definitive intent CRT. Our identified clinical markers (LN+ disease, hydronephrosis, decreased CrCl) as well as novel biomarkers (e.g., circulating tumor DNA) may help delineate populations most likely to benefit from additional therapy.

PURPOSE/OBJECTIVE(S) Hospital readmissions are associated with poor patient outcomes including increased morbidity and mortality, as well as increased healthcare costs. The Hospital Frailty Risk Score (HFRS) represents a validated risk stratification tool used to predict readmission and death among non-cancer patients, however this tool has not been evaluated among cancer patients, which represent a unique cohort of patients. This study sought to evaluate the utility of the HFRS among cancer patients to predict the risk of 30-day readmission and death. MATERIALS/METHODS We analyzed cancer patients within the National Readmission Database, a nationally representative longitudinal database, between January and November 2017. HFRS was calculated from ICD-10 diagnosis codes associated with the initial hospitalization. Study endpoints included readmission 30-days after the initial admission, length of readmission, hospital mortality, and hospital charges. A mixed effects multivariate logistic regression model was used to determine the association of HFRS on study outcomes while controlling for confounding variables. RESULTS We identified 1,685,099 weighted cancer admissions, of which 924,095 (54.8%) were deemed frail (HFRS > 5). Most common causes of readmission included sepsis, acute kidney injury, respiratory failure, and neutropenia. On unadjusted analysis frailty was associated with a lower 30-day readmission rate (23.4% vs. 23.6%, P = 0.04), but a significantly higher 30-day mortality rate (4.1% vs. 3.2%, P < 0.001). Among those readmitted, frail patients had longer hospital stays (7.0 days vs. 6.4 days, P < 0.001), and incurred higher charges across their hospitalizations ($20,644 vs. $18,321, P < 0.001). On multivariable analysis increased HFRS was associated with a decreased risk of readmission (odds ratio [OR] = 0.91, P = 0.04), and an increased risk of mortality (OR = 1.50, P < 0.0001). CONCLUSION An increased Hospital Frailty Risk Score was paradoxically associated with a decreased risk of readmission, though was associated with 30-day mortality among cancer patients. This research demonstrates that commonly used risk prediction scores may not generalize to cancer populations. More research is needed to develop and test risk-prediction algorithms specifically on cohorts of cancer patients to help personalize their supportive care.

PURPOSE Pancreatic cancer is an aggressive malignancy with patients often experiencing nonspecific symptoms before diagnosis. This study evaluates a machine learning approach to help identify patients with early-stage pancreatic cancer from clinical data within electronic health records (EHRs). MATERIALS AND METHODS From the Optum deidentified EHR data set, we identified early-stage (n = 3,322) and late-stage (n = 25,908) pancreatic cancer cases over 40 years of age diagnosed between 2009 and 2017. Patients with early-stage pancreatic cancer were matched to noncancer controls (1:16 match). We constructed a prediction model using eXtreme Gradient Boosting (XGBoost) to identify early-stage patients on the basis of 18,220 features within the EHR including diagnoses, procedures, information within clinical notes, and medications. Model accuracy was assessed with sensitivity, specificity, positive predictive value, and the area under the curve. RESULTS The final predictive model included 582 predictive features from the EHR, including 248 (42.5%) physician note elements, 146 (25.0%) procedure codes, 91 (15.6%) diagnosis codes, 89 (15.3%) medications, and 9 (1.5%) demographic features. The final model area under the curve was 0.84. Choosing a model cut point with a sensitivity of 60% and specificity of 90% would enable early detection of 58% late-stage patients with a median of 24 months before their actual diagnosis. CONCLUSION Prediction models using EHR data show promise in the early detection of pancreatic cancer. Although widespread use of this approach on an unselected population would produce high rates of false-positive tests, this technique may be rapidly impactful if deployed among high-risk patients or paired with other imaging or biomarker screening tools.

BACKGROUND Black patients with laryngeal squamous cell carcinoma (LSCC) historically have inferior outcomes in comparison with White patients. The authors investigated these racial disparities within the Veterans Health Administration (VHA), an equal-access system, and within the Surveillance, Epidemiology, and End Results (SEER) program, which is representative of the US hybrid-payer system. METHODS Patients with invasive (T1 or greater) LSCC were included from SEER (2004-2015) and the VHA (2000-2017). The primary outcomes of overall survival (OS) and larynx cancer-specific survival (LCS) were evaluated in Cox and Fine-Gray models. RESULTS In the SEER cohort (7122 patients: 82.6% White and 17.4% Black), Black patients were more likely to present with advanced disease and had inferior OS (hazard ratio [HR], 1.37; 95% CI, 1.26-1.50; P < .0001) in a multivariable analysis. Black LCS was worse in a univariable analysis (HR, 1.42; 95% CI, 1.27-1.58; P < .0001), but this effect was attenuated by 83% when the authors controlled for the TNM category and was found to be insignificant in a multivariable analysis (HR, 1.05; 95% CI, 0.93-1.18; P = .42). In the VHA cohort (9248 patients: 79.7% White and 20.3% Black), the 2 racial cohorts presented with similar tumor characteristics and similar OS (HR, 0.95; 95% CI, 0.89-1.02; P = .14). Black LCS was similar in univariable (HR, 1.10; 95% CI, 1.00-1.22; P = .05) and multivariable analyses (HR, 1.02; 95% CI, 0.92-1.14; P = .67). CONCLUSIONS Black patients with LSCC had a tumor burden at diagnosis and survival outcomes comparable to those of White patients within the VHA; this was counter to what was observed in the SEER analysis and prior national trends. This study's findings point toward the notable role of health care access in contributing to racial health disparities in the realm of larynx cancer.

PURPOSE/OBJECTIVE(S) Radiation treatment with concurrent chemotherapy (CRT) is a definitive treatment option for muscle-invasive bladder cancer (MIBC). Despite its effectiveness, CRT is underutilized given perceptions regarding fitness of patients for treatment and subsequent utility. Within the national Veterans Affairs' (VA) database, we investigated the factors associated with and the impact of treatment discontinuation in patients with MIBC treated with CRT. MATERIALS/METHODS In the VA database, we identified patients with urothelial histology, MIBC (T2-4a/N0-3/M0) diagnosed between 2000 and 2018 and treated with definitive intent CRT. Patients' electronic health records were chart reviewed to document if they completed definitive radiation (defined as at least 55 Gy). Preferred chemotherapy regimen was defined as containing cisplatin or combination mitomycin/5-fluorouracil. Primary endpoint of discontinued radiation was evaluated in a multivariable logistic regression. Secondary endpoints of 30/90-day mortality, overall mortality (OM), bladder cancer-specific mortality (BCM), non-bladder cancer-specific mortality (NCM), locoregional failure (LRF), and distant failure (DF) were evaluated in multivariable Cox and Fine-Gray models. RESULTS Of an identified 369 veterans with MIBC who underwent CRT, 30 (8.1%) patients did not complete definitive radiation. The most common reasons for treatment discontinuation included comorbidities/infections necessitating hospital admission (63.3%) and treatment intolerance/declining performance status (26.7%). In multivariable logistic regression, the only significant sociodemographic or clinical variables associated with radiation discontinuation were creatinine clearance (CrCl) ≤ 50 (OR 3.61, 95% CI 1.51-8.66, P < 0.01) and non-preferred chemotherapy regimen (OR 3.43, 95% CI 1.36-8.67, P = 0.01). In the discontinued radiation cohort, 30-day mortality was 33.3% and 90-day mortality was 50.0% with the majority of deaths due to non-bladder cancer-specific causes. In this cohort, no patient or tumor variables were associated with either endpoint. In the completed radiation cohort, 30-day mortality was 2.7% and 90-day mortality was 6.8%. In multivariable analysis, radiation discontinuation was significantly associated with worse OM (HR 2.50, 95% CI 1.37-4.56, P < 0.01) and worse NCM (HR 2.29, 95% CI 1.21-4.32, P = 0.01). It was not associated with inferior BCM (HR 0.56, 95% CI 0.25-1.26, P = 0.16), LRF (HR 0.93, 95% CI 0.44-1.97, P = 0.85), or DF (HR 0.60, 95% CI 0.28-1.28, P = 0.19). CONCLUSION More than 90% of patients with MIBC who undergo definitive intent CRT are able to complete their radiation course. Decreased creatinine clearance and receipt of non-preferred chemotherapy notably increased the odds of not completing radiation. Further research is required to develop evidence-based guidelines on patient selection to ensure curative therapy is offered to appropriate patients.

Purpose/Objective(s) A small fraction of cancer patients receiving definitive radiation experience unplanned hospitalizations related to treatment, which potentially leads to substantial morbidity or mortality. Identifying patients at high risk of unplanned hospitalizations could help with risk stratification, and potentially help inform future interventions aimed at improving patient outcomes. This study evaluated the predictive capacity of a machine learning prediction algorithm to identify unplanned hospitalizations and intensive care unit (ICU) visits among a large cohort of patients undergoing definitive radiation therapy. Materials/Methods We identified cancer patients undergoing definitive radiation therapy from the Veterans Health Administration (VHA) between 2000 and 2017. We used an extreme gradient boosting model to separately predict the risk of unplanned hospitalization and ICU admission within 30 days of completion of radiation treatment. Model covariates included patient demographics, cancer stage and subtype, and treatment information including previous chemotherapy and surgery. The data were split 75%/25% into training/testing datasets. We constructed the model with training data and evaluated performance within the test data using area under the curve (AUC), with an AUC of 1.0 indicating perfect prediction. Results We identified 60,804 cancer patients from 14 tumor subtypes receiving definitive radiation therapy, of which 936 (1.5%) experienced an unplanned hospitalization with 113 (0.19%) escalated to the ICU. Our final predictive model included 30 features, which contained 10 demographic variables and 20 tumor or treatment characteristics. The predictive models achieved an AUC of 0.82 and 0.79 for predicting unplanned hospitalization and ICU visit, respectively. Emergency department visits prior to radiation treatment was identified as the most important associated feature for predicting both unplanned hospitalizations and ICU visits. Other features including tumor subtype, Charlson comorbidity score, previous chemotherapy, and biologic effective dose were also important predictors in the final models. Conclusion This study demonstrates that a machine learning model can help provide a robust prediction of unplanned hospitalizations and ICU admission among cancer patients receiving definitive radiation. Machine learning prediction models show promise in helping to identify patients at risk of severe adverse outcomes, though additional validation is needed as well as research studying how to effectively implement these tools into practice.

6581 Background: Hospital readmission are associated with unfavorable patient outcomes and increased costs to the healthcare system. Devising interventions to reduce risks of readmission requires understanding patients at highest risk. Cancer patients represent a unique population with distinct risk factors. The purpose of this study was to define the impact of a cancer diagnosis on the risks of unplanned 30-day readmissions. Methods: We identified non-procedural hospital admissions between January through November 2017 from the National Readmission Database (NRD). We included patients with and without a cancer diagnosis who were admitted for non-procedural causes. We evaluated the impact of cancer on the risk of 30-day unplanned readmissions using multivariable mixed-effects logistic regression models. Results: Out of 18,996,625 weighted admissions, 1,685,099 (8.9%) had record of a cancer diagnosis. A cancer diagnosis was associated with an increased risk of readmission compared to non-cancer patients (23.5% vs. 13.6%, p < 0.001). However, among readmissions, cancer patients were less likely to have a preventable readmission (6.5% vs. 12.1%, p < 0.001). When considering the 10 most common causes of initial hospitalization, cancer was associated with an increased risk of readmission for each of these 10 causes (OR range 1.1-2.7, all p < 0.05) compared to non-cancer patients admitted for the same causes. Compared to patients aged 45-64, a younger age was associated with increased risk for cancer patients (OR 1.29, 95%CI [1.24-1.34]) but decreased risk for non-cancer patients (OR 0.65, 95%CI [0.64-0.66]). Among cancer patients, cancer site was the most robust individual predictor for readmission with liver (OR 1.47, 95%CI [1.39-1.55]), pancreas (OR 1.36, 95%CI [1.29-1.44]), and non-Hodgkin’s lymphoma (OR 1.35, 95%CI [1.29-1.42]) having the highest risk compared to the reference group of prostate cancer patients. Conclusions: Cancer patients have a higher risk of 30-day readmission, with increased risks among younger cancer patients, and with individual risks varying by cancer type. Future risk stratification approaches should consider cancer patients as an independent group with unique risks of readmission.

e18618 Background: Cancer patients frequently utilize the Emergency Department (ED) for a variety of diagnoses, both related and unrelated to their cancer. Patients with cancer have unique risks related to their cancer and treatment which could influence ED-related outcomes. A better understanding of these risks could help improve risk-stratification for these patients and help inform future interventions. This study sought to define the increased risks cancer patients face for inpatient admission and hospital mortality among cancer patients presenting to the ED. Methods: From the National Emergency Department Sample (NEDS) we identified patients with and without a diagnosis of cancer presenting to the ED between 2016 and 2018. We used International Classification of Diseases, version 10 (ICD10-CM) codes to identify patients with cancer, and to identify patient’s presenting diagnosis. Multivariable mixed-effects logistic regression models assessed the influence of cancer diagnoses on two endpoints: hospital admission from the ED, and inpatient hospital mortality. Results: There were 340 million weighted ED visits, of which 8.3 million (2.3%) occurred in patients with a cancer diagnosis. Compared to non-cancer patients, patients with cancer had an increased risk of inpatient admission (64.7% vs. 14.8%; p < 0.0001) and hospital mortality (4.6% vs. 0.5%; p < 0.0001). Factors associated with both an increased risk of hospitalization and death included older age, male gender, lower income level, discharge quarter, and receipt of care in a teaching hospital. We identified the top 15 most common presenting diagnoses among cancer patients, and among each of these diagnoses, cancer patients had increased risks of hospitalization (odds ratio [OR] range 2.0-13.2; all p < 0.05) and death (OR range 2.1-14.4; all p < 0.05) compared to non-cancer patients with the same diagnosis. Within the cancer patient cohort, cancer site was the most robust individual predictor associated with risk of hospitalization or death, with highest risk among patients with metastatic cancer, liver and lung cancers compared to the reference group of prostate cancer patients. Conclusions: Cancer patients presenting to the ED have high risks for hospital admission and death when compared to patients without cancer. Cancer patients represent a distinct population and may benefit from cancer-specific risk stratification or focused interventions tailored to improve outcomes in the ED setting.

1512 Background: Elderly hospitalized cancer patients face high risks of inpatient hospital mortality. Identifying patients at high risk of hospital mortality could help with risk stratification, and potentially help inform future interventions aimed at improving outcomes. We evaluated the predictive capacity of a high-dimensional machine-learning prediction tool to predict inpatient mortality, and compared the performance of this new tool to existing prediction indices. Methods: We identified cancer patients 75 and over who presented to an emergency department (ED) and were subsequently hospitalized from the National Emergency Department Sample (NEDS) between 2016 and 2018. We used an extreme gradient boosting approach to predict the risk of death during hospitalization. Model covariates included patient demographics, hospital characteristics, and International Classification of Diseases, version 10 (ICD-10) diagnosis codes recorded during the ED visit. The data were split 75%/25% into training/testing datasets. We constructed the model with training data and evaluated performance within the test data using area under the curve (AUC), with an AUC of 1.0 indicating perfect prediction. We compared the performance of this risk prediction model to standard prediction indices including the Hospital Frailty Risk Score, modified 5-item frailty index, and Charlson comorbidity index. Results: We identified 1,892,690 weighted-hospitalizations among elderly cancer patients, of which 133,379 (7.0%) who died in the inpatient setting. Our final predictive model included 238 features, which contained 5 demographic variables, 3 hospital characteristics, and 230 ICD-10 diagnosis codes. The predictive model achieved an AUC of 0.92. Our comparator models including the Hospital Frailty Risk Score, modified 5-item frailty index, and Charlson comorbidity index achieved AUCs of 0.67, 0.56, and 0.60, respectively. Conclusions: Using a high-dimensional machine-learning model enabled a high level of precision in predicting hospital mortality among elderly cancer patients, substantially out-performing existing prediction indices. High-dimensional prediction models show promise in helping to identify patients at risk of severe adverse outcomes, though additional validation is needed as well as research studying how to implement these tools into practice.

e18006 Background: Both chemoradiotherapy (CRT) and total laryngectomy (TL) with adjuvant therapy are curative-intent treatment options for patients with T4a larynx cancer. Disease recurrence is a known negative prognosicator, but differences in recurrence patterns and the subsequent survival associations are not well characterized. To address this knowledge gap, we present long-term recurrence and survival outcomes from a novel longitudinal data source. Methods: Retrospective study of non-metastatic T4a larynx cancer patients diagnosed between 2000-2017 who underwent curative-intent treatment (TL with adjuvant therapy or primary CRT) from the VA Informatics and Computing Infrastructure database. Adjuvant therapy consisted of either postoperative radiotherapy (RT) or CRT. Fine-Gray and Cox models were used to evaluate primary outcomes – time to locoregional recurrence and distant recurrence. Secondary outcomes included overall survival (OS), cancer-specific survival (CSS), non-cancer specific survival (NCSS), and disease-free survival (DFS). These multivariable models accounted for age, race, alcohol history, smoking status, education and income, Charlson-Deyo score, N-classification, and tumor subsite. Results: The study included 1,114 patients with a median follow-up time of 63.3 months among those alive at last follow up. In the TL group, adjuvant RT was used in 69% and adjuvant CRT was used in 31%. Median time to first recurrence was 24.4 months with overall incidence of 28.5% locoregional and 9.5% distant recurrence. Primary CRT patients had higher rates of locoregional (37.2 vs. 22.9%) and distant recurrence (13.3 vs. 7.0%) (p < 0.0001). Median OS was 27.3 months for CRT (95% CI: 23.6-32.4 months) and 47.5 months (95% CI: 39.6-52.1 months) for TL. Median DFS was 14.1 months for CRT (95% CI:12.5-17.2 months) and 37.9 months (95% CI 31.2-47.5 months) for TL. On multivariable analysis compared to CRT, TL was associated with longer time to locoregional (HR 0.50, 95% CI:0.40-0.61) and distant recurrence (HR 0.50, 95% CI:0.34-0.73). Having N+ disease increased risk of distant recurrence (HR 2.20, 95% CI:1.42-3.41). TL was associated with improved OS (HR 0.78, 95% CI:0.67 – 0.91), CSS (HR 0.73, 95% CI:0.59 – 0.89), and DFS (HR 0.58, 95% CI 0.49-0.69) compared to CRT; NCSS was equivalent between groups (HR 1.09, 95% CI:0.88-1.35). Of the CRT patients with locoregional failures, 67/163 (41.1%) were salvaged with surgery. Conclusions: In this cohort of T4a larynx cancer patients, surgical management demonstrated favorable recurrence and survival results. TL with adjuvant therapy was associated with significantly lower incidence of both locoregional and distant recurrence and increased OS, CSS and DFS compared to CRT. Lower probability of disease recurrence, in addition to a survival advantage, should be considered as an important advantage to up-front surgery.

e22019 Background: Pediatric cancer patients represent a vulnerable cohort at risk of adverse outcomes after presenting to the emergency department (ED). Given the severity of cancer-related complications and uniqueness of this population, approaches to better risk stratify this cohort could potentially help define future interventions geared towards improving outcomes. We used a high-dimensional machine learning approach to help define the risk of hospitalization after an ED visit among pediatric patients with cancer. Methods: A cohort of cancer patients under 18 was identified from the Nationwide Emergency Department Sample (NEDS) between 2016-2018. We used a lasso regression model to predict inpatient admission after an ED visit. Model covariates included patient demographics, hospital characteristics, and International Classification of Diseases, version 10 (ICD-10) diagnosis codes. The data were split 75%/25% into training/testing data. The model was constructed with training data, and performance assessed on the test data using the area under the curve (AUC), with an AUC of 1.0 indicating perfect prediction. Results: We identified 129,631 pediatric cancer patients who visited the ED, of which 54.5% were subsequently admitted. The final predictive model included 150 variables, including 9 demographic, 6 hospital, and 135 unique ICD-10 codes. The model demonstrated excellent ability to predict hospitalization with an AUC of 0.96. The top 5 most important variables associated with inpatient admission were diagnoses of paralytic ileus/intestinal obstruction, neutropenia, sepsis, aplastic anemia/bone marrow failure, and bacterial infection. Conclusions: Pediatric cancer patients frequently present to the ED with complications of their cancer or their treatment, and over half of these patients are admitted. This study demonstrates the capacity of high-dimensional prediction models to help identify pediatric patients at risk of hospitalization. Additional research is needed to determine how to implement these predictive models in clinical practice.

e24097 Background: Radiation therapy plays an important role in symptom palliation for patients with incurable cancer. Currently limited data exists regarding the role of SBRT vs conformal radiation for the palliation of symptoms due to malignancy. We report the symptom management and local control of palliative SBRT for incurable thoracic malignancies at a single institution. Methods: We retrospectively identified patients who underwent palliative SBRT between Jan 1st, 2009 and March 26th, 2019. Patients all had thoracic tumors that were not candidates for curative radiation due to age, stage, comorbidity, and/or prior treatment. We identified courses with total doses between 25-50 Gy (median 40) and total fractions between 3-10 (median 5). Symptoms such as cough, chest pain, hemoptysis, and shortness of breath were assessed at time of consult and first follow-up between 1-6 months post treatment. We also reviewed follow-up CT imaging to evaluate for local control using RECIST criteria. Descriptive statistics were used to evaluate patients’ clinicopathologic data and symptom palliation. Local control was analyzed via Kaplan-Meier method. Results: Of the 76 patients who completed palliative SBRT to 92 total lung lesions, 45 patients reported symptoms at consult and completed 50 courses of radiation to 55 lesions. Within this symptomatic cohort, average age was 71 (range, 42-93), 32 were female (58%), and most were stage IV (n = 42, 76%). Most lesions treated were non-small cell lung cancer (n = 34, 62%) while the most common primary site of metastatic lesions was colorectal (n = 6, 11%). Additional primary sites included breast, renal, sarcoma, and others (n = 15, 27%). Of the 53 lesions treated with follow-up within 6 months, 21 (40%) showed relief of at least 1 symptom and 31 (58%) showed stable symptoms. Only 1 patient (2%) showed symptom progression. All patients with hemoptysis at presentation achieved hemostasis following SBRT. Among 48 treated lesions with follow-up CT imaging, 1 (2%) showed complete response, 28 (58%) showed partial response (PR), 15 (31%) showed stable disease (SD), and 4 (8%) showed progressive disease. With further follow-up (median 23 months), 30 of the 53 lesions with initial PR or SD demonstrated local control until death. Conclusions: There is conflicting literature regarding the ideal palliative radiation dose for thoracic tumors. SBRT has the advantage of allowing a higher biologic dose without protracted treatment courses in the setting of palliation of symptoms. Our symptomatic cohort showed good symptom palliation and long-term local control of treated lesions. Prospective studies are required to further confirm the role of palliative SBRT for symptomatic thoracic tumors.

Background:Patients with advanced esophageal cancer often experience pain and dysphagia, yet the optimal palliative management remains unclear. This retrospective study evaluated outcomes and adverse effects of palliative radiotherapy (RT) compared with esophageal stenting among a cohort of U.S. veterans with metastatic esophageal cancer.Patients and Methods:We identified 1,957 veterans in the United States with metastatic esophageal cancer who received palliative RT to the esophagus or esophageal stenting, and assessed the risks of severe adverse effects, including esophageal fistula formation, perforation, obstruction, hemorrhage, and esophagitis. We determined palliative efficacy by evaluating pain and dysphagia scores before and after intervention. Multivariable analyses were used to control for potential confounding factors.Results:In our cohort, 1,593 patients underwent RT and 364 underwent esophageal stenting. The cumulative incidence of any severe adverse effect at 6 months was higher among patients who received stents compared with those who received RT (21.7% vs 12.4%;PPPP=.1029).Conclusions:Compared with esophageal stenting, RT was associated with a decreased risk of adverse effects, greater pain relief, and equivalent relief of moderate to severe dysphagia over time. Unmeasured patient- or tumor-related factors could have influenced the choice of intervention, thereby impacting our study outcomes. To our knowledge, this is the largest study to date analyzing the comparative risks and benefits of palliative RT and esophageal stenting among patients with metastatic esophageal cancer.

399 Background: Outcomes in bladder cancer are disproportionately worse for black patients compared to white patients. We hypothesize these disparities arise in part due to differences in access to healthcare and therefore may be mitigated in an equal access healthcare system, such as the Veterans Affairs’ (VA) system. Here, we examine outcomes by race for patients with bladder cancer within the VA system and then compare these outcomes to those in the Surveillance, Epidemiology, and End Results (SEER) database. Methods: We performed a retrospective cohort study using VA Informatics and Computing Infrastructure (VINCI) and SEER. We included all patients diagnosed with bladder cancer, American Joint Committee on Cancer (AJCC) stage 0-4 diagnosed between 2000 and 2018. Endpoints of overall survival (OS), bladder cancer-specific survival (BCS), and non-bladder cancer-specific survival (NCS) were evaluated in multivariable Cox and Fine-Gray models. Results: Using the VA dataset, we identified 36322 veterans (9.0% black, 91.0% white) with bladder cancer. Black veterans were more likely to have more comorbidities, reside in zip codes with lower median income and education levels, and present with higher stage disease (AJCC stages 2-4) than white veterans (23.3% vs 19%). In multivariable models accounting for disease stage among other covariables, there were no statistically significant differences in any survival endpoint (Table). Using the SEER dataset, we identified 130998 patients (5.9% black, 94.1% white) with bladder cancer. In similar multivariable models, SEER’s black patients had statistically significant inferior outcomes in all survival endpoints compared to SEER’s white patients (Table). Conclusions: While racial disparities for patients with bladder cancer in the SEER database were observed, no differences in survival outcomes between black and white patients were observed in the VA healthcare system. Of note, black veterans presented with more advanced stage, suggesting a delay in diagnosis or a more aggressive cancer phenotype compared to white patients. Our findings underscore the need to bridge healthcare disparities across diverse racial groups. Our study highlights the beneficial impact of an equal access healthcare system in reducing financial and social barriers to healthcare to counteract racial health disparities. Further research is required to delineate these disparities and guide appropriate screening strategies. [Table: see text]