Search

e19025 Background: Over the past two decades, there has been a tremendous increase in the chemotherapy options available to cancer patients. In terms of overall, progression-free survival, and temporary suppression of cancer-related symptoms, chemotherapy has shown beneficial effects. However, the side effects of chemotherapy are sometimes life threatening which affects an individual’s physical health, emotional state and quality of life. There is a considerable increase in the prevention, early identification and timely management of toxicities associated with chemotherapy; however, chemotherapy-related deaths still occur. Methods: We conducted a retrospective analysis of the National Inpatient Sample Database for the year 2017. Patients who were admitted for the administration of chemotherapy are identified using ICD- 10 codes. The epidemiology, the role of insurance providers in the treatment outcome were studied. Results: A total of 29,018 hospitalizations for the administration of chemotherapy were there in 2017. The median age of patients who received chemotherapy was 48. The overall mortality related to chemotherapy admissions was 0.80% (n = 233). The mortality of females who were admitted for chemotherapy did not vary much when compared to males admitted for chemotherapy (0.89% vs 0.73%, p = 0.132). It was found that admissions for chemotherapy during weekend had 85 % higher odds of dying as compared to admission during weekdays (1.6% vs 0.76%, OR = 1.85, p = 0.001, CI = 1.16 – 2.95). Patients who were admitted electively for chemotherapy were 74% less likely to die in hospital when compared to those who were admitted emergently for chemotherapy (1.4 % vs 0.49% OR = 0.36, p = 0.001, CI = 0.266 – 0.49). Interestingly, patients who had Medicare and Medicaid had higher mortality than those who had private insurance and self-pay when admitted for chemotherapy (2.08 % vs 0.58% vs 0.36%, p = 0.00). Those who had private insurance were 60% less likely to die in hospital while admitted for chemotherapy. The average length of stay for chemotherapy admissions were 5.92 ± 7.9%. Conclusions: Medicare and Medicaid patients, weekend admissions and emergent admissions were more likely to die in hospital while admitted for chemotherapy. Further studies are needed to reveal the disparities in the mortality of chemotherapy admissions, based on the socioeconomic status and the insurance payers.

e19097 Background: Cardiotoxicity resulting in heart failure (HF) is one of the most common complications of chemotherapy which significantly impacts patient care and results in poor clinical outcomes. Based on prior clinical studies, cancer patients have multiple cardiovascular comorbidities and higher prevalence of cardiovascular diseases which makes them more vulnerable to cardiovascular injuries during chemotherapy. It may increase their risk of developing cardiomyopathy or worsening of the already existing cardiac conditions which result in HF and death. Methods: We conducted a retrospective analysis of the National Inpatient Sample Database for the year 2017. Patients who were admitted in the hospital for chemotherapy administration and those who have HF were identified using ICD-10 codes. The mortality of patients with HF who receive chemotherapy and the factors affecting the mortality were studied. The two groups that were studied are chemotherapy patients with and without HF. Results: A total of 29,018 admissions for chemotherapy were identified during the year 2017. Out of these, 2.64% of patients were found to have HF (n = 767). The median age for the patients who have an associated diagnosis of HF was 63. The overall mortality of chemotherapy-related admission was 0.80%. The mortality of patients with HF who received chemotherapy was higher than the patients without HF who received chemotherapy (4.3% vs 0.7%, OR = 1.82, p = 0.001). Patients with HF who were admitted for chemotherapy were 82% more likely to die while in hospital than those without HF. Females who have HF had a higher mortality rate and had 64% more odds of dying while admitted for chemotherapy than males with HF (7.92% vs 1.93%, OR = 1.64, P = 0.00, CI = 1.31- 1.82). Also, interestingly Asians who have HF had higher mortality while admitted for chemotherapy than all the other races who had HF who received chemotherapy (Asians vs Caucasians vs African Americans 13.63% vs 4.64% vs 2.85%, p = 0.041). Conclusions: The mortality of patients with heart failure who receives chemotherapy is higher than the overall mortality from chemotherapy-related side effects. Females and Asians who have HF had higher mortality while admitted for chemotherapy. Patients who have prior cardiovascular diseases have to be given diligent care and special attention while we plan for chemotherapy to prevent the development or worsening of heart failure.

11014 Background: While physician sex has been shown to impact salary even after adjusting for productivity, gender-based differences in clinical activity and reimbursement for oncologists and hematologists are not completely understood. We evaluated the differences in Medicare reimbursement for male and female physicians in medical oncology and hematology. Methods: A retrospective analysis using Centers for Medicare & Medicaid Services Physician and Other Supplier Public Use File (POSPUF) for the year 2017 was performed to identify charges and total payment information for individual oncologists and hematologists. Mean values were compared using two-sample t-test, and the medians were compared by Wilcoxon rank-sum test. Results: A total of 8553 oncologists and hematologists (2710 women and 5843 men) were included in the POSPUF in 2017. Female physicians submitted a mean of 16,754 fewer charges (95% CI, −19,696 to −13,812; P < .0001), collected a mean of $173,632 less in revenue (95% CI, −201,184 to −146,080; P < .0001), and submitted a mean of 5.65 fewer unique billing codes (95% CI, - 6.69 to - 4.61; P

669 Background: Palliative care provides support and improves the quality of life of patients who has terminal illness such as cancer. Studies have shown that integrating palliative care into cancer care soon after the diagnosis of metastatic cancer can improve the quality of life and also prolong survival. The objective of the study was to analyze the acceptance of palliative care and to determine the factors affecting the utilization of palliative care in patients with genitourinary (GU) cancers. Methods: The National Inpatient Sample Database for the year 2016 was queried for the data. Patients who were admitted with a principal diagnosis of Malignant Neoplasm of Prostate (MNP), Renal Cell Carcinoma (RCC), Malignant Neoplasm of Ureter (MNU) were identified using ICD-10 codes and those who had Encounter for Palliative Care (PC) was also identified. Results: A total of 58765 hospitalizations were identified with MNP (n=39853), RCC (n=17786), MNU (n= 1126) during the study period. The total PC utilization for the above patients were 6.4% (n= 3785). Among those, 6.7% (n=1186), 6.3% (n=2531), 6.03% (n=68) from the groups RCC, MNP, MNU respectively received PC. The mean age for the PC utilization was 73 years (MNP= 72 yrs, RCA= 63 yrs, UCC= 73 yrs). Females received more PC when compared to males (6.7% vs 6.4%). Among those who had PC 27.4% (n=1029) died in the hospital. Interestingly, patients who had Medicare and Medicaid had more PC encounters when compared to those with private insurance (OR= 1.21, P= 0.001). Patients admitted on the weekend received more PC when compared to those who admitted during weekdays (OR= 1.12, P= 0.001). Patients admitted to hospitals in the West received more PC than other regions (OR= 1.42, P=0.03). Impact of race, teaching vs non-teaching hospital admissions were not found to be statistically significant in the utilization of PC in the above GU cancers. Conclusions: Medicare and Medicaid patients, weekend admissions, admissions to hospitals in the West received more palliative care. Further studies are needed to reveal the role of socioeconomic status and insurance in the utilization of palliative care in GU malignancies.

69 Background: Prostate cancer is one of the most common cancers in men and one of the leading causes of death among men of all races worldwide. Prostate cancer prevalence and mortality vary substantially by race, ethnicity and geography. The reason behind the disparity is unclear, however, access to screening and treatment, variation in exposure to risk factors, genetic susceptibility, lifestyle practices are the suggested factors that affect the disparity. Methods: We conducted a retrospective analysis of the National Inpatient Sample Database for the year 2016. Patients who were admitted with a principal diagnosis of Malignant Neoplasm of Prostate (MNP) were identified using ICD-10 codes. The epidemiology, racial disparities and regional variations in the mortality of prostate cancer, trend in the cost and length of stay were studied. Results: A total of 39,853 hospitalizations were identified with a principal diagnosis of MNP. The prevalence of MNP was found to be 0.56% with mean age of diagnosis 71.36 ± 0.55. The prevalence of prostate cancer was found to be highest in African Americans (0.65%), followed by Caucasians (0.60%). The disease-specific mortality rate for MNP was 3.45% (n=1,375). African Americans (AA) had 32% more risk of dying with prostate cancer when compared to all the other groups (OR= 1.322, P= 0.001, 95% CI= 1.14-1.52). Caucasians had 25% less chance of dying and Hispanic had 28% less chance of dying with prostate cancer when compared to AA (OR= 0.755, P= 0.001, CI= 0.65-0.87 vs OR= 0.72, P= 0.016, CI= 0.56- 0.94). Those who admitted to West were 22% more likely to die and those admitted to Northeast were 18% more likely to die when compared to Midwest and south regions (OR= 1.22, P= 0.037, CI= 1.01- 1.47 vs OR= 1.18, P= 0.062, CI= 0.91-1.4). The length of stay and also cost of stay were found to be highest for African American population. Conclusions: The prevalence, mortality of prostate cancer, cost and length of inpatient stay were highest among African Americans. There was a statistically significant difference in the mortality rate of prostate cancer based on the hospital regions in the United States. More studies are needed to better understand the reasons behind the regional difference.

Purpose As cancer burden has risen worldwide, physicians, patients, and their advocates have become aware that the clinical cancer trial research paradigm is not ubiquitous. Furthermore, the number and characteristics of trials that are registered in low- and middle-income countries (LMICs) compared with that in high-income countries (HICs) are unknown. Methods We collected retrospective data on trials for breast, lung, and cervical cancer registered in ClinicalTrials.gov or with the WHO International Clinical Trial Registry Platform between 2010 and 2017. The data were then classified as trials within LMICs or HICs using definitions from the World Bank. Results Included in these analyses were 6,710 trials, of which 3,164 (47%) were breast cancer trials, 3,283 (49%) were lung cancer trials, and 263 (4%) were cervical cancer trials. There were 1,951 (29%) trials from LMICs and 4,759 (71%) trials from HICs (P < .001). Although the proportion of phase III trials in HICs versus LMICs was similar (18% v 17%; P = .66), the number of phase I trials in LMICs was significantly lower than that of HICs (20% v 2%; P < .001). For several LMICs with the highest mortality-to-incidence ratios for breast, lung, or cervical cancer, there were no cancer trials registered in the registration data bases searched for this work. Conclusion There are differences in access to cancer clinical trials in LMICs compared with HICs. Several factors, such as excessive cost and a lack of infrastructure and expertise, may explain these differences.

Introduction Disseminated Intravascular Coagulation (DIC) is a systemic coagulopathy which leads to widespread thrombosis and hemorrhage and ultimately results in multiorgan dysfunction. DIC usually occurs as a complication of illnesses like severe sepsis, malignancies, trauma, acute pancreatitis, burns, and obstetrical complications. The prognosis and mortality of DIC depend on the etiology, however, the mortality of DIC is known to be on the higher side. The aim of the study is to analyze if gender, race, regional differences have any association with the mortality of hospitalized patients with DIC. Method The National Inpatient Sample database from the Healthcare Cost and Utilization Project (HCUP) for the year 2016 was queried for data. We identified hospital admissions for DIC with the International Classification of Diseases, Tenth Revision, Clinical Modification diagnosis code D65. The data was analyzed with STATA 16.0 version and univariate and multivariate analysis were performed. We studied the characteristics of all such hospitalizations for the year 2016 and the factors associated with the in-hospital mortality rate (MR) of DIC. We used length of stay, cost of stay as an outcome to determine if gender, race, and location play a role in the mortality. Results A total of 8704 admissions were identified with a diagnosis of DIC during the year 2016. The mean age for admission was found to be 56.48± 0.22. The percentage of admissions in females and males did not have a notable difference (50.57% vs 49.43%). The disease specific MR for DIC was 47.7%. Admission during weekend vs weekdays did not carry a statistically significant difference in terms of MR. Females with DIC were less likely to die in the hospital when compared to males with DIC (OR= 0.906, CI 0.82 - 0.99, p= 0.031). Interestingly, African Americans (AA) with DIC admissions were found to have 24% more risk of dying when compared to Caucasians admitted with DIC (OR= 1.24, CI 1.10 - 1.39, P= 0.00), Native Americans (NA) has 67% more risk of dying when compared to Caucasians (OR= 1.67, CI 1.03 - 2.69, p= 0.035). The mortality rate of NA, AA, Caucasians with DIC was found to be 57%, 52%, 47% respectively. The MR was found to be highest in hospitals of the northeast region (52%), then hospitals in the south (47%), followed by west and mid-west (46%), p= 0.000. Patients admitted to west and mid-west were 24% less likely to die when compared to patients admitted to northeast region hospitals (OR= 0.76, p= 0.001). The average length of stay and cost of stay were also less in west and mid-west regions when compared to north east. The difference in outcomes persisted after adjusting for age, gender, race, hospital division, co-morbid conditions. Conclusion Our study demonstrated that African Americans and Native Americans with DIC have high risk of dying in the hospital. Also, there exists a difference between the mortality rate, length and cost of stay among different regions in the United States. More research is needed to elucidate the factors that might be impacting the location-based variation in mortality. Disclosures No relevant conflicts of interest to declare.

Introduction The overall impact of morbidity related to sickle cell disease (SCD) is enormous due to decreased quality of life, high health care utilization and immense financial strain on the patients and health care system. Patients with SCD have also been found have high hospital readmission rates compared to other medical conditions, further leading to increased health care burden. The objective of this study was to explore the common reasons, healthcare utilization and identify modifiable factors associated with 30-day readmission in patients with SCD using the most recently available national data in the United States. Methods Cohort selection. 2016 Nationwide Readmission Database (NRD) from the Healthcare Cost and Utilization Project (HCUP) by the Agency for Healthcare Research and Quality (AHRQ) was queried for analysis. NRD captures discharge data from 22 states, representing about 50% of all hospitalizations in the United States. National estimates can be produced by using sampling weights provided by the NRD. Patients with SCD were identified by using the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) diagnosis code D57. Patients younger than 18 years were excluded. Readmission was defined as any admission within 30 days of index hospitalization discharge. Statistical Analysis. Sampling weights were used throughout all calculations, facilitating appropriate national projections. Percentages in all figures reflect national estimates. Chi-square test and Student's t-test were used for univariate analysis. Multivariable logistic regression analysis was done to determine independent predictors of 30-day readmission in patients with SCD. Data analyses was performed using SAS v9.4 (SAS Institute, Cary, NC). Results In a total of 83,692 hospitalizations for SCD in 2016, 15,880 (18.9%) had at least one 30-day readmission and 40% of the readmissions occurred within the first two weeks of discharge. The most common reason for 30-day readmission was due to complications of SCD (29.8%) with sickle-cell pain crisis (18.4%) being the most frequent one. The other common causes for readmission were sepsis (10.3%), cardiac related (9.7%), respiratory failure (6.2%), renal failure (4.3%) and mood related (4.04%). Female sex (P Conclusion This study showed that the 30-day readmission rate in patients with SCD in 2016 has decreased to less than 20% compared to previously published rates (31.9%, Elixhauser A, 2013). While this decrease in readmission rate is encouraging, further studies are needed to investigate the reasons for this trend. Several modifiable risk factors were identified in this study such as alcohol, tobacco, opioid, substance abuse and depression which can be addressed to potentially bring down the readmission rate further and improve patient care. Disclosures No relevant conflicts of interest to declare.

Introduction Sickle cell disease (SCD) associated pain is a significant health care issue in the United States which prompts physicians to prescribe opioids to help treat and prevent the recurrent acute painful episodes. Despite nationwide efforts to reduce narcotic pain medication usage, opioids still remain as the mainstay of pain management in SCD. Many SCD patients are using marijuana to help with their pain, anxiety, appetite, mood and sleep as per recent studies. Cannabinoids in marijuana interact with the body's endocannabinoid system which has receptors in almost every major bodily system. The effect of cannabinoids on these receptors reduces the signaling of inflammatory responses and also reduce cytokine production. Very few states have approved SCD as a qualifying condition for medical marijuana. But we are still unsure about the medical benefits of marijuana in SCD patients as there are very limited studies done so far. In our study, we sought to examine the characteristics and complications of marijuana usage in sickle cell patients. Methods The National Inpatient Sample database for the year 2016 was used to identify admissions with a primary diagnosis of SCD and we grouped patients into those who have a diagnosis of cannabis related disorders (CRD) and those who do not have the diagnosis. ICD- 10 codes are used for identifying the SCD patients and also for CRD. Statistical analysis was performed using STATA and univariate and multivariate analysis were performed. The outcomes that are studied included mortality, length and cost of stay, hospital regions and the association of marijuana use with anxiety, mood disorders. We also studied the association of marijuana with the complications of SCD such as sickle cell pain crisis, vaso occlusive crisis, acute chest syndrome, splenic sequestration, avascular necrosis. Results A total of 37,307 admissions with a principal diagnosis of SCD were identified, out of which 4.09% (N= 1526) had cannabis use disorders. The median age of patients with CRD was found to be 31.21 ± 0.3 when compared to 30.67 ± 0.09 in patients without CRD. Even though SCD admissions were more commonly seen in females when compared to males (61.78% vs 38.22%), cannabis use was seen more associated with males (57.97% vs 42.03%). The in-hospital mortality of SCD was less (0.56%) as compared to the mortality rates of other hematological malignancies. The association of cannabis use with in-hospital mortality was found to be not statistically significant. Also, the median length of stay was less in patients with CRD when compared to patients without CRD (4.88 ± 0.2 vs 5.11 ± 0.03) and also likewise cost of stay. Based on the hospital regions in the US, Cannabis use in SCD was seen more prevalent in South region (44%), then Midwest or north-central (26%), northeast (19%), west (10%) and the result was statistically significant (p= 0.003). The association of cannabis use was not found to be statistically significant with acute chest syndrome and splenic sequestration. Cannabis use was, however, found to be associated with the vaso occlusive crisis and avascular necrosis (OR=1.02, p=0.003 and OR= 1.14, 0.022 respectively) even though we cannot say that cannabis use could be a risk factor as there are other confounding factors like coagulopathy, chronic debilitating conditions. Interestingly, SCD patients with CRD have more risk of developing anxiety (OR= 2.32, p=0.000) and also mood disorders (OR= 2.5, p= 0.001) when compared to SCD patients without CRD. The difference persisted after adjusting for age, gender, race, co-morbidities. Conclusion Marijuana use is more seen in the southern and north-central regions in patients with SCD. Marijuana use was not found to be associated with in-hospital mortality in sickle cell patients. SCD patients are using marijuana mainly for alleviating their pain and sometimes for its euphoria effect. Our study showed that it can cause anxiety and mood disorders. The main limitation of our study was the moderate sample size for SCD patients with CRD. The impact and interaction between CRD and SCD complications need to be evaluated separately in a larger study to get accurate values. Large randomized control trials have to be done to assess if SCD qualifies for prescription of medical marijuana as it possesses benefits as well as risks. Disclosures No relevant conflicts of interest to declare.

72 Background: Intractable pain is one of the invincible entities in cancer and about 75% of cancer patients require inpatient hospital admission at least once in their treatment course. Palliative care service has a significant role in managing cancer pain and thus reducing the recurrent hospital admissions. Methods: We analyzed the association of Mortality Rate (MR) with age, sex, race, emergency admission, hospital regions among patients admitted for Neoplasm Related Pain (NRP). Data from the National Inpatient Sample (NIS) database for the year 2016 was reviewed. Patients admitted with a principal diagnosis of NRP were identified using ICD-10 code. The prevalence of each cancer and characteristics of admissions were also studied. Statistical analysis was performed using STATA and the association was depicted with Chi-Square test. Results: A total of 28609 admissions with a primary diagnosis of NRP was identified. The mean age of admission was found to be 59.38 ± 0.08, with 53.19% patients ≥ 60 years of age. 51.39% were females, 65.58% were Caucasians,16.25% were African Americans. The overall inpatient hospital MR for NRP was 8.16%. There was a significant difference in the MR between inpatients ≥ 60 years of age and

84 Background: For a patient, receiving a diagnosis of cancer can be quite challenging and is often associated with depression and increased suicidal ideation. Previous studies have indicated that the risk of suicide in cancer patients is twice that of general population. Prostate cancer is the most common cancer in men. It has also been found to be associated with highest rates of suicide in genitourinary malignancies. We sought to evaluate the incidence and identify risk factors associated with suicide among men with prostate cancer. Methods: Patients with prostate cancer between the years 1973 and 2016 were identified in the Surveillance, Epidemiology, and End Results database (SEER). Multivariate logistic regression analysis was performed to identify independent risk factors for suicide death only on variables that were statistically significant in univariate statistical analysis. Results: In 1,307,625 patients with prostate cancer, 3,435 patients had committed suicide. Younger age (≤50 vs > 70: OR = 4.73 , P < 0.0001; 51-70 vs > 70: OR = 1.93, P < 0.04), Caucasian race (OR: 3.61, P < 0.0001) and not undergoing surgery (OR: 1.33, P < 0.0001) were significantly associated with increased risk of suicide. Furthermore, patients with < 1 year since the diagnosis of prostate cancer were also found to be associated with increased risk of suicide death ( < 1 vs > 5 years: OR = 1.46, P < 0.0001). Conclusions: Addressing mental health in patients with prostate cancer is an important issue. We found that several factors such as younger age, Caucasian race were significantly associated with risk of suicide. Patients who did not undergo surgery likely had advanced disease, which could explain their increased risk of suicide. Patients with < 1 year since the diagnosis of prostate cancer are more vulnerable and at increased risk of suicide which emphasizes the necessity of identifying and treating patients at risk of suicide as early as possible. It is vital that health care providers recognize these patients and offer them appropriate support.

1 Background: Advance care planning (ACP), which includes advance directive (AD) completion, ensures that patients receive care that is consistent with their beliefs and preferences. ACP is particularly important to patients with cancer as they have high rates of hospitalization and mortality. However, AD completion rates for oncology patients have been reported to be as low as 10%. We sought to evaluate barriers and improve AD completion rates in hospitalized oncology patients in our institution. Methods: 63 patients with cancer admitted to our inpatient oncology unit were surveyed. A multi-disciplinary team, consisting of resident physicians and social workers, was constituted to educate patients and provide assistance with AD completion. Barriers to AD completion were identified and AD completion and filing rates on patients’ discharge were calculated. Results: Most patients were female (66.6%) with mean age of 63.8 years. Majority of patients had stage IV cancer (71.4%) and were receiving palliative treatment (80.9%). Only 9.5% (6/63) of the patients had completed AD on file. However, 92% of the patients had previously heard about AD. After intervention, the AD filing rate improved to 41.2% (26/63). Some of the patient-reported barriers were needing more time to discuss with family (28.5%), wanting to file in clinic (23.8%), too sick to make decision (6.3%) and did not think they needed it (4.7%). Conclusions: AD completion rate in hospitalized oncology patients in our institute was lower than anticipated. We identified the barriers and improved the AD filing rate, however there is room for further improvement. While AD discussion and filing should ideally be done in clinic, hospitalization is an opportunity that should not be missed to educate and assist completion of AD. [Table: see text]

83 Background: Depression is a major ongoing problem with cancer patients and survivors. Depression can lead to functional impairment, decreased utilization of treatment, decreased quality of life and increased suicide risk. It also leads to an extended hospital stay and increased cost of treatment. Previous studies have shown that lung cancer has a high incidence for depression. Methods: We conducted a cross-sectional study using the National Inpatient Sample (NIS) database for the year 2016. Patients who were admitted with a principal diagnosis of Malignant Neoplasm of Bronchus and Lung (MNBL) and Major Depressive Disorder (MDD) were identified using ICD-10 codes. The prevalence of MDD and the risk factors contributing to MDD were studied. Statistical analysis was performed using STATA tool and tested with Pierson Chi-Square test. Results: A total of 79096 hospitalizations with a principal diagnosis of MNBL were identified. The prevalence of depression among lung cancer patients was found to be 12.49% (n = 9881). The mean age of lung cancer patients with depression was 62.27 ± 0.10. Females with lung cancer have more risk for MDD (15%, p < 0.001). Caucasian patients with lung cancer have a higher risk of MDD (13.6%, p < 0.001). It was also found that lung cancer patients who are ≤ 55 years have a higher risk of MDD (p < 0.001). The average length of stay for lung cancer patients with MDD was slightly higher than lung cancer patients without MDD (6.39 days and 5.99 days respectively). The average cost of stay also had a similar pattern. Severe morbidity was seen in a greater proportion of the MDD group. Conclusions: Younger patients who are ≤ 55 years of age, Caucasians or females with lung cancer have a higher risk of MDD. Particular attention should be given to them while being in the hospital which can decrease the morbidity, and will also help to improve their quality of life.

e18863Background: The Clinical Trials Registry of India (CTRI), launched in July 2007, was established to bring transparency and accountability to clinical trials performed in India. Mandatory regi...