387 Dispositions for Patients Ruled Out for Acute Myocardial Infarction Using a 1-Hour High Sensitivity Troponin I Algorithm: Home or Hospital?

Abstract

The multicenter High Sensitivity Cardiac Troponin I (hs-cTnI) in the United States (HIGH-US) study has reported a baseline/1-hour emergency department (ED) hs-cTnI algorithm with an acute myocardial infarction (AMI) rule-out rate of 50.4% (negative predictive value 99.7%, sensitivity 98.7%). In the ruled-out patients the 30-day AMI/death rate was 0.2%. Our objective was to describe any differential clinical characteristics of the AMI ruled-out patients placed in observation/inpatient beds (OBS/ADM) compared to those with an ED discharge (EDD). 2113 consenting adults with any symptoms suspicious for AMI were enrolled from 2015-2016 in 29 US medical centers. There were no exclusion criteria for enrollment. Baseline and 1-hour plasma samples were analyzed using the Siemens Atellica hs-cTnI assay (99th % = 47.0 ng/L). AMI diagnosis was independently adjudicated by cardiologists and ED physicians using local contemporary troponin assays and all following 30-day clinical information available. Patients were placed in OBS/ADM or EDD according to local practice during the study period. Step logistic regression modeling was used to determine which of 25 clinical patient characteristics were associated with OBS/ADM disposition decisions. 1020 (48.3%) individuals were ruled out for AMI at 1 hour. Of these, 584 (57.3%) were EDD and 436 (42.7%) were placed in OBS/ADM. No AMI/deaths occurred in the latter group while they were in the hospital. The frequency of patients having an EDD rather than being placed in OBS/ADM varied significantly (0.0% - 94.1%) amongst the participating sites. 30-day AMI/death rates in these 2 groups were not significantly different (0.0% v 0.5% {2 patients, days 7 and 28}, p=0.185). The highest Hs-cTnI values were very low in both groups (mean 3.1 and 2.3 ng/L in the OBSD/ADM compared to the EDD patients). The clinical variables significantly associated with a decision to place a patient in OBS/ADM (table) were generally risk factors for the development of coronary artery disease. Of the many potential factors that could influence the decision for an extended length of stay for patients at very low risk for AMI/death, those with a personal and/or family history of coronary artery disease, stroke, no prior heart failure hospitalizations, hypertension, or having an abnormal ECG were likely to be placed in OBS/ADM for further cardiac evaluations. Considering their excellent prognosis and recent evidence (ISCHEMIA Trial) showing revascularization in the absence of AMI does not reduce AMI/death compared to medical therapy, reliance on clinical variables should be judiciously considered and systems developed to reduce the frequent decisions for extended stay for rapidly ruled-out for AMI patients.Tabled 1Variables associated with OBS/ADM placementOBS/ADM versus EDD variablesp-valueOdds ratio estimate95% CI lower limit95% CI upper limitHistory of CAD<0.0012.9401.9994.325Previous stroke0.0022.7421.4605.148No prior HF hospitalization0.0192.20221.11213.647History of hypertension<0.0011.8951.3612.638Abnormal ECG<0.0011.7661.2922.413Family history of CAD0.0351.3991.0241.912 Open table in a new tab