Arrest location, initial rhythm and outcome in intensive care unit-treated cardiac arrest patients.
 
Purpose of the study
To evaluate the effect of arrest location and initial rhythm on outcome of intensive care unit (ICU) treated cardiac arrest (CA) patients.
 
Materials and methods
We performed a registry-based study to determine associations between CA location and initial rhythm with one-year mortality in adult CA patients receiving post-arrest treatment in the ICUs of a tertiary hospital between 2005 and 2013. We used logistic regression analysis adjusting for age, comorbidity, pre-admission functional status, time to Return of Spontaneous Circulation, and Sequential Organ Failure Assessment score. Cerebral Performance Category score of 1 or 2 was defined as good neurological outcome in survivors.

Results
Of 1,139 CA patients, 692 patients (60%) had out-of-hospital CA (OHCA), 338 patients (30%) had non-ICU in-hospital CA (IHCA), and 109 patients (10%) had in-ICU CA (ICUCA). The initial rhythm was ventricular fibrillation or ventricular tachycardia (VF/VT) in 58% of patients, pulseless electrical activity (PEA) in 25% of patients, asystole in 11% of patients, and other/unknown in 6% of patients. Unadjusted mortality was highest among ICUCA patients (62%), followed by IHCA (57%) and OHCA patients (42%).  ICUCA had significantly higher risk of mortality (odds ratio [OR] 2.2 (95% confidence interval CI 1.2-4.1) compared to OHCA. An initial rhythm of PEA (OR 2.1, 95% CI 1.4-3.0) or asystole (OR 3.3, 95% CI 1.9-5.5) was associated with a higher risk for one-year mortality compared to VF/VT. Among one-year survivors, good neurological outcome was seen in 94% of patients with no significant differences between CA locations or initial rhythms (p>0.05 for both).

Conclusion 
In patients surviving to be admitted to the ICU, OHCA as location and VF/VT as initial rhythm were associated with better survival. Most one-year survivors had a good neurological outcome.