Cardiac risk assessment, morbidity prediction, and outcome in the vascular intensive care unit

• Published in: Vascular and endovascular surgery Vol. 47; no. 8; p. 585
• Main Authors: Dover, Mary, Tawfick, Wael, Hynes, Niamh, Sultan, Sherif
• Format: Journal Article
• Language: English
• Published: United States 01.11.2013
• Subjects: Aged; Area Under Curve; Coronary Artery Disease - complications; Coronary Artery Disease - therapy; Female; Heart Diseases - diagnosis; Heart Diseases - epidemiology; Heart Diseases - mortality; Heart Diseases - physiopathology; Heart Diseases - therapy; Humans; Incidence; Intensive Care Units; Ireland - epidemiology; Length of Stay; Male; Middle Aged; Myocardial Revascularization; Patient Admission; Predictive Value of Tests; Prognosis; Retrospective Studies; Risk Assessment; Risk Factors; ROC Curve; Time Factors; Vascular Diseases - diagnosis; Vascular Diseases - mortality; Vascular Diseases - physiopathology; Vascular Diseases - therapy; Ireland; coronary artery disease; ICU; Lee revised cardiac risk index; echocardiogram; risk prediction
• ISSN: 1938-9116
• DOI: 10.1177/1538574413502551

The aim of this study is to examine the predictive value of the Lee revised cardiac risk index (RCRI) for a standard vascular intensive care unit (ICU) population as well as assessing the utility of transthoracic echocardiography and the impact of prior coronary artery disease (CAD) and coronary revascularization on patient outcome. This is a retrospective review of prospectively maintained Vascubase and prospectively collected ICU data. Data from 363 consecutive vascular ICU admissions were collected. Findings were used to calculate the RCRI, which was then correlated with patient outcomes. All patients were on optimal medical therapy (OMT) in the form of cardioselective β-blocker, aspirin, statin, and folic acid. There was no relationship found between a reduced ejection fraction and patient outcome. Mortality was significantly increased for patients with left ventricular hypertrophy (LVH) as identified on echo (14.9% vs 6.5%, P = .028). The overall complication rates were significantly elevated for patients with valvular dysfunction. Discrimination for the RCRI on receiver-operating characteristic analysis was poor, with an area under the receiver-operating characteristic curve of .621. Model calibration was reasonable with an Hosmer-Lemeshow Ĉ statistic of 2.726 (P = .256). Of those with known CAD, 41.22% of the patients receiving best medical treatment developed acute myocardial infarction (AMI) compared to 35.3% of those who previously underwent percutaneous cardiac intervention and 23.5% of those who had undergone coronary artery bypass grafting. There was 3-fold increase in major adverse clinical events in patients with troponin rise and LVH. The RCRI's discriminatory capacity is low, and this raises difficulties in assessing cardiac risk in patients undergoing vascular intervention. The AMI is highest in the OMT group without prior cardiac intervention, which mandates protocols to identify patients requiring cardiac intervention prior to vascular procedures.