A large proportion of endovascular aortic aneurysm repair (EVAR) patients are routinely admitted to the intensive care unit (ICU) for postoperative observation. In this study, we aimed to describe the factors associated with ICU admission after EVAR and to compare the outcomes and costs associated with ICU vs non-ICU observation.Methods:
All patients undergoing elective infrarenal EVAR in the Premier database (2009–2015) were included. Patients were stratified as ICU vs non-ICU admission according to location on postoperative day 0. Both patient-level (sociodemographics, comorbidities) and hospital-level (teaching status, hospital size, geographic location) factors were analyzed using univariate and multivariable logistic regression to determine factors associated with ICU vs non-ICU admission. Overall outcomes and hospital costs were compared between groups.Results:
Overall, 8359 patients underwent elective EVAR during the study period, including 4791 (57.3%) ICU and 3568 (42.7%) non-ICU admissions. Patients admitted to ICU were more frequently nonwhite and had more comorbidities, including congestive heart failure, coronary artery disease, chronic kidney disease, chronic obstructive pulmonary disease, diabetes, and hypertension, than non-ICU patients (all, P < .03). ICU admissions were more common in small (<300 beds), urban, and nonteaching hospitals and varied greatly depending on surgeon specialty and geographic region (P < .001). A pattern emerged when admission location was clustered by hospital; ICU patients were treated at hospitals where 96.7% (interquartile range, 84.5%-98.9%) of patients were admitted to ICU after EVAR, whereas non-ICU patients were treated at hospitals where only 7.5% (interquartile range, 4.9%-25.8%) were admitted to ICU after EVAR. A multivariable logistic regression model accounting for patient-, operative-, and hospital-level differences had a significantly lower area under the curve for predicting ICU admission after EVAR than a model accounting only for hospital factors (area under the curve, 0.76 vs 0.95; P < .001). The overall rate of adverse events was higher for ICU vs non-ICU patients (16.3% vs 13.7%; P < .001). Failure to rescue (2.9% vs 3.9%; P = .42) and in-hospital mortality (0.4% vs 0.4%; P = .81) were similar between groups. After adjusting for patient and hospital factors as well as for postoperative adverse events, ICU admission after EVAR cost $1475 (95% confidence interval, $768–2183) more than non-ICU admission (P < .001).Conclusions:
Among patients undergoing elective EVAR, postoperative ICU admission is more closely associated with hospital practice patterns than with individual patient risk. Routine ICU admission after EVAR adds significant cost without reducing failure to rescue or in-hospital mortality.