Novel low-delta value troponin algorithms have good precision for rule-out of NSTE-ACS

Abstract

Abstract Background Rapid rule-out algorithms for non-ST-elevation infarction (NSTEMI) may be beneficial for logistics in the emergency room. Current algorithms are designed to rule-out NSTEMI, but do not differentiate between unstable angina (UAP) in need of revascularization and non-cardiac chest pain patients (NCCP) who could be discharged. Recent improvements in analytical precision of high sensitivity troponin (cTn) assays allow for trialing algorithms with very small delta values. Purpose Could the use of lower delta values produce rule-out algorithms for NSTE-ACS with a false negative rate of ≤5%, and a sufficient high rule-out rate of patients with NCCP. Method 927 patients with suspected NSTE-ACS were consecutively included. Serum samples were collected at 0, 3 and 8–12 hours. The final diagnosis was adjudicated by two independent cardiologists based on all clinical data including routine cTnT. The 0- and 3-hour samples were additionally measured for cTnIand cTnI from Singulex Clarity System (cTnI(sgx)). The diagnostic performance to rule-out NSTE-ACS was compared between one low-delta value algorithm from each assay (cTnT, cTnI and cTnT). Results The prevalence of NSTEMI was 13.4%, UAP 11.4% and NCCP 60%. Median age was 63 years, 60% males. Fig. 1 shows baseline and 3-hour delta cTn values for the UAP and NCCP patients for the three different assays. The baseline cTn value differed significantly between UAP and NCCP for all assays, p value <0.001. The novel low-delta cTnT algorithm (Table 1) ruled out 8 NSTE-ACS patients (3.5%), the cTnI algorithm and cTnI (sgx) algorithm ruled out 11 (4.8%) and 12 (5.2%) patients with NSTE-ACS, respectively. Moreover, the cTnT algorithm allocated 35.3% of the NCCP patients to discharge. Respective numbers for the cTnI(sgx) and cTnI algorithm were 30.6% and 33.5%. Comparing the ROC curves, the cTnT algorithm had significantly higher AUC compared to the cTnI(sgx) algorithm (p value =0.005, DeLong test). Conclusion The low-delta value algorithms correctly ruled in ≥95% of the NSTE-ACS patients whilst >30% of NCCP patients were ruled out. The cTnT algorithm had the best performance with a significant higher AUC compared to the cTnI(sgx) algorithm. Figure 1 Funding Acknowledgement Type of funding source: Public Institution(s). Main funding source(s): Western Norway Regional Health Authority, Haukeland and Stavanger University hospital