17 Background: The UVB-vitamin D-breast cancer (BC) hypothesis is supported by ecological studies demonstrating an inverse correlation between sunlight exposure and BC incidence and mortality. Observational studies also favor an inverse association between vitamin D status and BC risk, recurrence and mortality. Yet controversies remain regarding the role of vitamin D in BC. We examined associations between vitamin D levels, geographic location, clinical, and pathologic characteristics of BC patients (pts). Methods: This was a retrospective analysis from the electronic health record database of pts diagnosed with BC between 1/2007 and 5/2013 across US Oncology Network practices categorized based on their geographic location into: northern (> 40° N), central (35 to 40°N) and southern (< 35°N) latitude. We collected age at diagnosis, BMI, smoking history, stage, estrogen receptor (ER), progesterone receptor (PR), HER2 status, and the first documented serum 25-hydroxyvitamin D (25-(OH)D) level, categorized as 30 optimal. Statistical comparison was performed using Chi-squared tests for categorical variables and Kruskal-Wallis tests for continuous variables. Logistic regression was used to predict the likelihood of vitamin D deficiency. Results: 20,338 BC pts with a documented vitamin D level were identified. Mean age at diagnosis was 58. Stage and receptor status distribution were: 8%, 41%, 32%, 11% and 4% for stage 0, I, II,III, and IV respectively; 63%, 13%, and 10% for ER+/HER-2-, HER-2+ and TNBC, respectively. 17.6% and 27.8% of pts had deficient or suboptimal vitamin D levels. The covariates of age < 60 years (OR 1.24), advanced stage (OR 1.32 stage II, OR 1.51 stage III, and OR 1.81 stage IV), TNBC (OR 1.45), BMI ≥ 25 (OR 2.02), current smoker (OR 2.11), and lower latitudes (OR 1.36 and 1.19 for central and southern latitude respectively) were independent predictors of a first documented vitamin D deficiency in a multivariate model. Conclusions: Vitamin D deficiency may be associated with TNBC and central and southern latitudes. The possible influences of differential vitamin D supplementation and timing of testing require further investigation.

65 Background: Several organizations have created oncology pathways (pws) to support evidence-based medicine, and both standardize and improve the quality of care. Information is scarce; however, regarding the requirements for successful implementation of a pws program and monitoring of adherence. Methods: We assessed the efficacy of a quality assurance initiative called Practice Pathways Improvement (PPI) program to increase pws adherence to The US Oncology Network’s physician-developed Level I Pathways. We queried our electronic health record (EHR) iKnowMed to collect data before and after program initiation. Education was provided to address policies and procedures; leadership roles and responsibilities; integration of pws into the EHR and workflow; and the importance of correct data entry into the EHR. We then measured metrics including levels of assessable data (AD) (complete data, no missing or conflicting elements); On-Pathway adherence (ON-PW); and documented rationale for exceptions (EX) of Off-Pathway treatments. Metric performance reports were distributed monthly. Results: Since 2010, 10 practices of varying size (6-342 medical oncologists) and sites (2-47) voluntarily participated. Program duration ranged from 6-12 months (mos). There were 42,266 regimens evaluated during the program, covering 19 disease pws. Baseline averages, results at program end, and current averages (avg) for AD, ON-PW, and EX are shown in the Table. As of June 2012, average follow up was 19 mos (range 13-25). A normal approximation shows significance of p < 0.001. Conclusions: Implementation of a quality assurance initiative containing operational, leadership, and workflow elements along with continued education is important for successful pws adherence. Tailored program components are necessary to meet individual practice needs. Establishing a foundation for a pws process is important to maintain performance over time. As the reimbursement landscape in oncology evolves, the ability to demonstrate quality of care is essential. [Table: see text]

1521 Background: Hospice improves the quality of life and care for cancer patients and reduces the likelihood of unwanted death in the hospital. Advance Care Planning (ACP) allows physicians to proactively initiate hospice and end-of-life discussions with identified patients, promoting timely hospice care enrollment. We developed a machine learning (ML) model to predict 90-day mortality risk for patients with metastatic cancer. The tool was designed to enable earlier ACP discussions leading to increased hospice enrollment. This study assesses the ML tool usage on ACP documentation in a community oncology setting. Methods: Twelve practices across the nation were included in the study, all participating in the Oncology Care Model. Five practices implemented the ML tool during 10/26/2020-9/30/2021, with patients scored every two weeks to provide insights on mortality risk. Patients identified as high-risk were evaluated for ACP utilization, obtained from timely EMR data and historical claims. Seven practices did not implement the ML tool and served as the control for the study. Results: A total of 1,663 patients were predicted to have a high risk of mortality at the 12 practices during the timeframe. The median age was 74 years. 53% of patients were males, and 47% were females. ACP documentation varied among the practices. The range was 19.4%-55.8% among ML tool participating practices and 7.4%-31.0% among non-participating practices. The weighted mean of ACP utilization was 34.4% for participating practices and 14.0% for non-participating practices. Compared with non-participating practices, the ACP rate increased significantly by 2.5-fold for participating practices (p-value = 0.03, two-sided T-test). Conclusions: This initial outcome study showed improved ACP documentation after deploying a mortality prediction tool in a community oncology setting. We are currently working on propensity score matching and regression analysis to reduce the effect of confounding factors such as practices, patient demographics, diagnosis, and treatment. Future studies will evaluate the impact of mortality tool use on other outcomes, including hospice enrollment, emergency department visits, and hospital admission. Implementing the mortality prediction tool is an ongoing effort with more practices planned to adopt.