859

Abstract

Introduction: With limited intensive care unit (ICU) beds at many hospitals, patients with acute myocardial infarction (MI) may be triaged into ICUs with less experience managing MI. To our knowledge, volume-outcome relationships have not been studied according to an ICU’s annual MI volume. Hypothesis: We hypothesized that MI patients admitted to ICUs with low annual MI volume would have lower rates of MI quality indicators and worse clinical outcomes when compared to ICUs with high MI volume. Methods: We used data from the eICU— a telemedicine system for ICU patients at >400 US hospitals— to examine MI management in ICUs with low vs high annual MI volume in 2008. In order to focus on ICU triage between specialty ICUs, hospitals with <250 beds were excluded. Low MI volume was defined as =50th (=16 pts/yr) and high MI volume was defined as =90th (=96 pts/yr) percentile of ICUs managed by eICU programs. Performance measures were compared and modified Poisson regression was used to assess the association between mortality and admission to an ICU with low MI volume. Results: Of 7482 patients with a primary ICU admission diagnosis of MI, 5313 were admitted to ICUs with high and 288 to ICUs with low MI volume. Patients in the low MI volume group were slightly older, more likely to be female, and had more comorbidities and worse vital signs and labs at ICU admission. These patients had less frequent coronary revascularization (70% vs 77%), less early aspirin (74% vs 86%) and?blocker use (81% vs 90% among pts without contraindications), and longer length-of-stay in the ICU (2.8 vs 2.4 days) and overall (6.3 vs 5.2 days) when compared to MI patients in high MI volume ICUs (all p=0.05). ICU mortality (6% vs 3%, p<0.001) and hospital mortality (10% vs 4%, p<0.001) were higher in the low MI volume group. After adjustment for 24 patient-level characteristics, admission to an ICU with low annual MI volume was independently associated with higher hospital mortality (RR 1.46, 95% CI 1.02-2.09, p=0.038). Conclusions: Among unselected patients with acute MI, admission to an ICU with less experience treating MI is associated with less application of evidence-based therapies and higher rates of in-hospital mortality.