Priority levels in cancer patients admitted to medical intensive care at a safety net hospital.

Abstract

e18599 Background: Admissions of cancer patients to intensive care units (ICU) are increasing with improved mortality. While ICU care can be lifesaving, its higher cost does not always result in reduced mortality. Moreover, timely goals of care (GOC) discussions correlate with less ICU use in those with certain cancers. We investigate if hospital mortality and disposition outcomes for cancer patients correlate to triage by ICU providers. Methods: This subgroup analysis of a prospective cohort of 808 patients admitted to the ICU from 1 July 2015- 15 June 2016 at an academic safety net hospital included 106 patients diagnosed with cancer. Medical records were reviewed by ICU physicians, who assigned priority ranks using Society of Critical Care Medicine guidelines: 1: critically ill, needing treatment/monitoring not provided outside of ICU, 2: not critically ill, but requiring close monitoring/potentially immediate intervention, 3: critically ill patients with reduced likelihood of recovery, 4: not appropriate for ICU, equivalent outcomes achieved with non-ICU care. We did a chart review for factors like prior therapy and documentation of GOC discussions. Statistical tests were conducted to examine if priority levels correlate with disposition, mortality, and length of stay (LOS). Results: χ2-tests revealed priority rank correlated with disposition after hospitalization (p<.05) with group 3 having the highest proportion of deaths and lowest proportion of discharges home. It revealed that mortality rate differed by group (p<.05) with logistic regression showing that priority 3 predicted increased mortality (p<.05). ANCOVA indicated ICU LOS differed by priority group (p<.05), with priority 3 averaging the longest LOS. While priority 3 had the most in-hospital GOC discussions, relatively few were documented pre-hospitalization. Conclusions: Overall, our patients were accurately triaged, with worse mortality and discharge outcomes among priority 3 and a dearth of pre-hospitalization GOC documentation for all groups. Our data show the importance of triaging patients and having early, frequent GOC discussions to minimize ICU admission given increasing demand and costs. GOC discussions are associated with less aggressive medical care near death and better patient quality of life. Thus, holding these talks with our sickest patients prior to potential ICU admission is an area to improve cost-effective high quality care.[Table: see text]