Abstract 214: Drivers of In-Hospital Cardiac Arrest Survival Rates: A Risk-Adjusted Model
Background Cardiac arrest in hospitalized patients is common and often fatal. Factors which predict which patients will survive arrest to discharge or die in the hospital are unknown. This gap in our understanding precludes hospitals from characterizing their relative in-hospital cardiac arrest (IHCA) outcome rates relative to other hospitals and identifying their individual needs for quality improvement.
Methods Using data from the Get With the Guidelines-Registry, years 2000-2010, we identified patients with IHCA. Excluded were patients <18 or >110 years of age, arrests in hospital visitors or employees, and events with missing variables. A regression model was used to identify predictors of survival rates across hospitals. The sequential model included demographics ( age, race, gender), then pre-arrest comorbidities, and then intra-arrest factors (e.g. initial rhythm, in-place interventions, arrest: etiology, location, time). The outcome measure was survival to discharge. Unadjusted and risk-adjusted survival rates per hospital were calculated.
Results We evaluated 136,663 IHCA at 588 hospitals. IHCA survival to discharge rates varied across hospitals; unadjusted: median: 0.19 (IQR 0.14-0.24); adjusted: median 0.18 (IQR 0.16-0.21). The regression model showed several predictors explained some but not all of the variability in outcomes (c-statistic =0.74). Several predictors of variability in survival outcomes included age 41-65 (OR 1.5; 95% CI 1.5-1.6, p<.01), female sex (OR 1.1; 95% CI 1.0-1.1 p<.01), black race (OR 0.83; 95% CI 0.79-0.87, p<.01), prior arrhythmia (OR 1.25; 95% CI 1.2-1.3, p<.01), cardiac arrest etiology (OR 1.2; 95% CI 1.2-1.3, p<.01), toxicologic arrest etiology (OR 2.4,95% CI 2.1-2.8, p<0.01), monitored bed status (OR 1.7; 95%CI 1.6-1.8, p<.01), and pre-arrest anti-arrhythmic drugs (OR 1.2; 95% CI 1.2-1.3, p<.01).
Conclusion There is significant variability in IHCA outcomes across hospitals although not all explained by patient level risk-adjustment of available variables. Further research is needed to investigate additional drives (e.g. hospital structure and process measures, patient level variables in other datasets) of this variability in order to develop valid comparisons of outcomes across hospitals.
- © 2012 by American Heart Association, Inc.