Background Chest pain accounts for approximately 6% of all Emergency Department (ED) attendances. We evaluated the diagnostic accuracy of a high-sensitivity cardiac troponin I assay (Siemens TNIH) on serial sampling for patients presenting to the ED with suspected cardiac chest pain. Specifically, we evaluated the accuracy of three previously reported accelerated diagnostic algorithms (figure 1), the 99th percentile upper reference limit and absolute and relative changes in TNIH over 3 hours.
This is a secondary analysis from a multi-centre prospective diagnostic accuracy study across 14 UK sites. Patients presenting to the ED with chest pain of suspected cardiac nature warranting investigation were included. The target diagnosis was acute myocardial infarction (AMI), which was adjudicated by two independent investigators. Serum blood samples were taken on ED arrival and 3 hours later. Stored frozen samples were subsequently analysed with the Siemens TNIH assay (ADVIA Centaur, 99th percentile upper reference limit: female 39.6 ng/L, male 58.0 ng/L; coefficient of variation 10% at 4.50 ng/L) and absolute (AΔ=|‘3h-TNIH’-‘0h-TNIH’|) and relative (RΔ%=(AΔ*100)/’0h-TNIH’) changes calculated.
Of 802 included patients, 13.8% had AMI. Absolute delta had superior accuracy to relative delta (C-statistic 0.94 vs 0.76, p<0.001). However, used alone no optimised delta could achieve sensitivity >95.5% for AMI. Simply ruling out AMI in patients with TNIH below the 99th percentile at 3h had only 88.3% (95% CI 80.8–93.6) sensitivity. The Basel algorithm had higher sensitivity (98.2%) than both High-STEACS (93.7%, p=0.03) and the ESC 0/3h algorithm (88.3%, p<0.001) (see figure 1). The algorithms ruled out 63%, 74% and 84% patients respectively.
With serial sampling over 3h, the Siemens TNIH assay should be used with a validated algorithm incorporating bespoke cut-offs and absolute delta changes. In our analysis, the Basel algorithm had greatest sensitivity. ‘Ruling out’ AMI using the 99th percentile upper reference limit of the assay cannot be recommended.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.