View Abstract

A6841 - Risk Factors for Unplanned Intensive Care Unit Transfer After Inter-Hospital Transfer of Medical Patients
Author Block: M. Montalvo1, R. V. Patel1, A. M. Sheth2, E. J. Yoo3, K. F. D'Mello3; 1Hahnemann University Hospital, Drexel University, Philadelphia, PA, United States, 2Lehigh Valley Health Network, Allentown, PA, United States, 3Drexel University College of Medicine, Philadelphia, PA, United States.
Rationale: Early unplanned ICU transfer (UIT) after hospitalization has a well-established association with increased mortality. It is also known that patients admitted to the ICU from general medical floors have worse outcomes than those admitted directly from the emergency room. Less well known are the risk factors associated with UIT after inter-facility medical transfer. Appropriate initial placement of patients directly admitted from outside facilities is an important safety goal to help mitigate poor outcomes. The objective of this study is to identify the risk factors associated with early (within 24 hours of arrival) unplanned ICU transfers for inter-hospital medical patients transferred to an urban tertiary care center. Methods: A retrospective chart analysis of all medical adult patients directly admitted to either general medical floors or the progressive care unit at Hahnemann University Hospital (HUH) during 2015 and 2016 was performed. The data consisted of the variables sex, age at time of admission, admitting diagnosis, code status, time between call to transfer center and arrival at HUH, transfer to ICU, time from admission to transfer to ICU, time to ICU transfer (as a grouped variable), diagnosis for ICU transfer, ICU length of stay (LOS), ICU mortality, hospital LOS, and Combined Age Charleston Comorbidity Index (CCI) score, and Mortality Probability Model III (MPMIII score) for patients transferred to ICU. A multivariate logistic regression was used to identify risk factors for ICU upgrade. Results: During 2015 and 2016, 156 (21.2%) of inter-facility medicine transfers were transferred to the ICU. 42 patients (5.7 %) were unplanned ICU transfers within the first 24 hours. Admitting diagnoses of pancreatic or liver disease (p=0.006) and ""other"" diagnoses (including anaphylaxis, acute renal failure, etc.) were most associated with early UIT as well as eventual ICU transfer. Significant diagnoses at time of UIT included pulmonary (p=0.021), pancreatic and liver (p=0.011), and sepsis (p=0.045). There was a significant correlation with UIT and delays between call to transfer center and arrival at the receiving hospital (p=0.031) and CCI score (p=0.002). MPMIII score was a significant predictor (p=0.037) for ICU mortality. Conclusions: Interfacility transfers with a pancreatic and liver related diagnosis as well as those with greater delays between the call to transfer center and arrival on the general medical floors were at increased risk of both early and eventual UIT. These patients would likely benefit from closer observation or improved triage methods upon arrival from receiving facilities.