Evaluating risk characteristics for patients who die within 30 days of hospitalization.

Abstract

161 Background: Advanced cancer patients often receive aggressive end of life care despite questionable benefit. As a result, there are ongoing efforts to improve end of life care and coordinate palliative care supportive services. This study is an expansion of a previous study performed examining descriptive data in a population of oncology patients who died within 30 days of admission. Here, we compare patients who died within 30 days of admission against those who survived to evaluate differences in patient characteristics and healthcare utilization. Methods: Adult oncology patients who were admitted from 10/1/2018-3/30/2019 at an academic medical center were evaluated. Two groups of patients were studied–oncology patients who died within thirty days of admission and those who survived. The patients were selected using ICD-10 codes, EMR systems support, and manual chart review. Additionally, we examined demographic (i.e. gender, ethnicity, cancer diagnosis) and clinical characteristics (i.e. level of care, code status, previous palliative care consult, palliative care consult in the hospital, nutrition status, clinical trial status, advance care planning, hospice enrollment). Statistical analysis included chi-squared and ANOVA tests, and logistic regression models. Results: A total of 267 patients were included in the analysis. For all patients in the study, 38% had a change in code status during their admission. 26% of patients had palliative care involvement and 23% were known to palliative care prior to admission. Twenty three percent spent the duration of their admission in the ICU for their end of life care. Significant mortality-level variation was found compared to overall mean number of admissions for the past 6 months (ANOVA F=25.3, p<0.0001). We conducted a logistic regression and adjusted for ethnicity, number of admits in the last 6 months, and length of stay to identify the outcome of patients who died within 30 days of admission vs. those who did not. Factors associated with increased odds of mortality included the number of admits in the last 6 months (OR 1.753, 95% CI: 1.397-2.200). Length of stay did not increase one’s odds for mortality (OR 0.989, 95% CI: 0.965-1.014). Conclusions: Low utilization of palliative care and advanced care planning was seen widely in both populations. Previous hospitalization in the last 6 months was a predictor of mortality in this patient population.