Investigator University of Pennsylvania, United States
Introduction: Extracorporeal membrane oxygenation, also known as ECMO, has seen increased use in Critical Care medicine in patients with refractory cardiogenic shock. ECMO provides advanced life support to these patients allowing us to provide them time for recovery and/or eventual organ transplantation. However in the last 5 years, according to the Extracorporeal Life Support organization registry, only 48% of patients placed on ECMO for cardiac dysfunction survived for eventual discharge or transfer. Due to the poor outcomes for patients on ECMO due to cardiogenic shock, our study aimed to evaluate the presence of effects secondary to geographic location/insurance coverage on outcomes for these patients.
Methods: The National Inpatient Sample (NIS) database (2016-2020) was analyzed to identify adults admitted with the ICD-10 diagnostic/procedural codes for Cardiogenic Shock and ECMO. Multivariate logistic regression analysis was used to obtain effects of geographic location and insurance payer in mortality outcomes. Weighted analysis using Stata 17 MP was performed.
Results: A total of 4,589 adult patients were identified with Cardiogenic Shock that underwent ECMO in our analysis, of which 2,214 (48%) died.
We found that West North Central (OR 0.67, p< 0.05, CI 0.45 - 0.995), South Atlantic (OR 0.65, p< 0.05, CI 0.045 - 0.94) and West South Central (OR 0.63, p< 0.05, 0.43 - 0.91) areas were less likely to die in comparison to the New England area. There was no significant difference in all other areas
Conclusions: When comparing the mortality rates of patients in different regions of the US on ECMO secondary to cardiogenic shock, our studies showed that the West North Central region, South Atlantic, and West South Central regions had significantly decreased mortality ratios when compared to the New England division when controlling for age, race, and Charlson Comorbidity Index. When comparing other regions to the New England division, no significant difference was noted.
These findings warrant further research in evaluating the factors that are causing the decreased odds ratio in the West North Central region, South Atlantic, and West South Central region to allow us to implement it to other regions of the US.
