Abstract 17891: Underutilization of Palliative Care Consults and End-of-Life Care Planning After Cardiogenic Shock

Abstract

Background: Cardiogenic shock (CS) is associated with significant mortality. The 2022 ACC/AHA/HFSA guidelines discuss the importance of palliative care (PC) in advanced heart disease, but PC and end-of-life care (EOLC) integration after CS is not well characterized. We aimed to study the use of PC and EOLC planning after CS. Hypothesis: We assessed the hypothesis that PC consults and EOLC planning are underutilized after CS. Methods: This is a retrospective observational study of 63 consecutive patients at a quaternary care center in Boston diagnosed with CS between January and April 2022. CS onset and etiology were adjudicated by treating physicians. Results: Of 63 patients with CS (mean age 61±14 years, 30% Female, 73% White), 26 (41%) died during admission and 31 (49%) survived 1 year after CS onset. Only 6 patients (10%) received PC consults for EOLC considerations (consults placed on average 9 days from CS onset) (Table 1). Only 24 (38%) had documented advanced care plans after developing CS, of which 14 were documented by PC consultants. Of 44 patients with MCS or ICD placed, only 1 had a documented discussion about the option of discontinuing the therapy in the future. After CS, 28 (44%) had changes in resuscitation limits. Only 1 patient had their preferred location of death documented. Palliative inotropes were prescribed to 2 patients at discharge, and 3 patients were discharged to hospice. Readmission rates at 3 and 6 months were 24% and 36%, respectively. Conclusions: PC and EOLC planning are underutilized after CS, despite the significant mortality of CS. Secondary PC models may not be optimal for CS, given the acute presentation of CS and bimodal distribution of survival after onset. These data demonstrate a compelling need to address key considerations that impact EOLC after CS, such as resuscitation limits, duration of advanced therapies, and goal-concordant discharge locations. Further research is needed to explore optimal models of PC and EOLC integration after CS.