233 Background: Prior reports from our institution demonstrated high rates of racial segregation, unfavorable social determinants of health (SDoH) in Milwaukee, WI, and statewide reports of inferior outcomes for cancer patients from minority communities. At the Medical College of Wisconsin's Cancer Center (Milwaukee, WI), during the first through last quarters of 2018-2019, cancer patients from the low socioeconomic status (SES) communities who were hospitalized to inpatient oncology units had an average length of stay (LOS) of 7.2 days compared to 5.6 days for high SES group. Under the auspices of the American Society of Clinical Oncology's Quality Training Program (QTP) initiative, we aimed to reduce the hospital LOS by 10% or less by May 2021 for inpatient oncology teams. Methods: A multidisciplinary team collaboration between the inpatient and outpatient providers was developed during this QI initiative. We examined LOS index data, payer types, and other diagnostic criteria for the oncology inpatient solid tumor service and two comparator services (bone marrow transplant, BMT; internal medicine). We generated workflow, a cause-and-effect diagram, and a Pareto diagram to determine the relevant factors associated with longer hospital LOS. Institution-wide implementation of the SDH screen project was launched to evaluate and address specific barriers to SDoH to expedite a safe discharge process during the pandemic. Results: Through one test of change (Plan-Do-Study-Act cycles 1, 2 &3), we identified the problem of extended LOS and patient-related barriers to discharge during this QI initiative. Compared to the baseline LOS, after the launch of the SDoH screen project, there was a 6.5% decrease in the inpatient average LOS for oncology patients (7.89 to 7.40days, p = 0.004),10.7% for BMT (15.96 days to14.26, p = 0.166), and 2.4% for Internal Medicine (4.61 to 4.50 to days, p = 0.131). There was a 10.0% decrease in LOS (8.07 to 7.26 days, p = < 0.001) for the three specialties combined. With collaboration from inpatient and outpatient providers, appropriate referrals were generated to address patient-specific SDH before discharge (i.e., transportation coordination, nutritional and physical therapy referrals, social worker assistance with food, and housing insecurities). Conclusions: In this pilot project, implementing SDoH screening-based-care delivery at the time of inpatient admission demonstrated a slight improvement in LOS for solid tumor oncology patients and provided timely referrals, opportunities to engage and explore the discharge facilities early on during the COVID 19 pandemic. With this preliminary data, we plan to continue to expand our efforts through a systemwide implementation of this SDoH survey both in the inpatient and outpatient settings to address cancer inequities.
Application of Cox and Parametric Survival Models to Assess Social Determinants of Health Affecting Three-Year Survival of Breast Cancer Patients
