IMPORTANCE Accelerated diagnostic protocols (ADPs) for chest pain using high-sensitivity troponin (hsTn) levels have excellent sensitivity and negative predictive value for rapid risk stratification of patients with chest pain. However, little is known about the outcomes of patients who are discharged despite abnormal ADP results, ie, after "ruling-in" with a modest elevation of hsTn. OBJECTIVE To determine outcomes of patients discharged following ADP, including those who were ruled in with modestly elevated levels of hsTnT but discharged nonetheless. DESIGN, SETTING, AND PARTICIPANTSThis retrospective cohort study included patients with chest pain who presented to the emergency departments (EDs) of a large multisite health system ED between January 2017 to September 2019. Patients were assessed using an ADP, had a peak hsTnT level measured between the limit of quantitation and 52 ng/L, were discharged, and had follow-up in the electronic medical record. Data analysis was conducted from January 2017 to September 2019. EXPOSURES Application of an hsTnT ADP.MAIN OUTCOMES AND MEASURES Thirty-day major adverse cardiac events (MACE), including myocardial infarction, urgent coronary revascularization, and all-cause death, comparing patients who were discharged following ADP-concordant vs ADP-discordant results. RESULTSOf 10 342 patients with chest pain (mean [SD] age 51 [17] years; 5902 [57%] women) discharged following ADP, 29 (0.28%) had MACE. Patients with MACE were older (median [IQR] age, 66 [53-75] years vs 50 [38-62] years; P < .001) and more likely to have prior CAD (12 [41.4%] vs 1805 [17.5%]; P = .002) and hyperlipidemia (13 [44.8%] vs 2248 [21.8%]; P = .006). Additionally, patients with MACE were 5-fold more likely to have been discharged despite ADP discordance (16 [55.2%]vs 1145 [11.1%]; P < .001). A multivariable logistic regression analysis revealed only ADP discordance was independently associated with MACE (odds ratio, 6.42 [95% CI, 2.94-14.0]; P < .001). When stratified by peak hsTnT level, there were no differences in MACE between ADP-concordant and -discordant discharges provided the peak hsTnT measured was less than 12 ng/L. In contrast, patients with peak hsTnT level between 12 and 51 ng/L were significantly more likely to have MACE if they were discharged after ADP-discordant vs -concordant hsTnT series (14 of 609 [2.30%] vs 5 of 1047 [0.48%]; P < .002). Notably, a HEART (history, electrocardiogram, age, risk factors, troponin) score of 4 or greater retrospectively identified the most ADP-discordant discharges (13 of 16 [81.3%]) who had MACE. (continued) Key Points Question In patients presenting to the emergency department with chest pain, what are the outcomes following discharge despite modest elevations of high-sensitivity troponin? Findings In this cohort study of 10 342 patients with chest pain who were discharged following application of a high-sensitivity troponin-based advanced diagnostic protocol (ADP), patients with ADP-discordant discharges had significantly higher 30-day adverse car...

In order to choose from a list of functionally similar services, users often need to make their decisions based on multiple QoS criteria they require on the target service. In this process, different users may follow different decision making strategies, some are compensatory and some are non-compensatory. Most of the current QoS-based service selection systems do not consider these decision strategies in the ranking process, which we believe are crucial for generating accurate ranking results for individual users. In this thesis, we propose a decision strategy based service ranking model. Furthermore, considering that different users follow different strategies in different contexts at different times, we apply a learning to rank algorithm to learn a personalized ranking model for individual users based on how they select services in the past. Our experiment result shows the effectiveness of the proposed approach.

In order to choose from a list of functionally similar services, users often need to make their decisions based on multiple QoS criteria they require on the target service. In this process, different users may follow different decision making strategies, some are compensatory and some are non-compensatory. Most of the current QoS-based service selection systems do not consider these decision strategies in the ranking process, which we believe are crucial for generating accurate ranking results for individual users. In this thesis, we propose a decision strategy based service ranking model. Furthermore, considering that different users follow different strategies in different contexts at different times, we apply a learning to rank algorithm to learn a personalized ranking model for individual users based on how they select services in the past. Our experiment result shows the effectiveness of the proposed approach.