Decision-making in acute chest pain remains challenging despite normal (below ninety-ninth percentile) high-sensitivity troponin (hs-cTn). Some studies suggest that undetectable hs-cTn, far below the ninety-ninth percentile, might rule out acute coronary syndrome. We investigated clinical data in comparison to undetectable hs-cTnT. The study comprised 682 patients (November 2010 to September 2011) presenting at the emergency department with chest pain and normal hs-cTnT (<14 ng/l). The main end point was major adverse cardiac events (MACE: death, myocardial infarction, readmission for unstable angina, or revascularization) at a 4-year median follow-up; secondary end point was 30-day MACE. A clinical score was built by assigning points according to hazard ratios of the independent predictive variables: 1 point (male and effort-related pain) and 2 points (recurrent pain and prior ischemic heart disease). The negative predictive values of the clinical score and undetectable hs-cTnT (<5 ng/l), were tested. A total of 72 (10.6%) patients suffered long-term MACE. The C-statistics of the clinical score for long-term (0.75) and 30-day (0.88) MACE were higher than with the TIMI(Thrombolysis In Myocardial Infarction) risk (0.68, 0.77) or GRACE(Global Registry of Acute Coronary Events) (0.50, 0.47) scores. Likewise, the negative predictive values of score = 0 (97.5%, 100%) and ≤1 point (95.9%, 100%) were higher than using undetectable hs-cTnT (91.9%, 98.1%). Both clinical scores of 0 and ≤1 better classified patients at risk of MACE (p = 0.0001, log-rank test) than hs-cTnT <5 ng/l (p = 0.06). In conclusion, clinical data can guide decision-making and perform at least equally well as undetectable hs-cTnT, in patients presenting at the emergency department with chest pain and normal hs-cTnT.
Decision-making in acute chest pain with normal troponin levels (below ninety-ninth percentile) is challenging, despite the introduction of the high-sensitivity troponin (hs-cTn) assays. Clinical risk scores have been implemented as ancillary tools. Few of them, however, were derived from patients with normal troponin levels, and none was developed in the hs-cTn era. Recent studies have suggested that undetectable hs-cTn levels (far below the ninety-ninth percentile) might rule out acute coronary syndrome, meaning that such patients might safely be discharged from the emergency department. Therefore, if this were fully demonstrated, clinical evaluation could play a secondary role in patients with normal troponin levels. The present study involved patients presenting at the emergency department with acute chest pain and normal hs-cTnT levels (below the ninety-ninth percentile). The main goal was to compare clinical data to undetectable hs-cTnT for long-term outcomes.
Methods
The study comprised 682 consecutive patients presenting at the emergency department of the University Clinic Hospital in Valencia (Spain), with nontraumatic acute chest pain, without ischemic electrocardiographic (ECG) changes (defined by transient or persistent ST-segment elevation or ST-segment depression ≥0.5 mm) or hs-cTnT elevation. This was a retrospective cohort from November 2010 to September 2011. The information was obtained by reviewing all reports from the emergency department files with a diagnosis of nontraumatic acute chest pain at admission, during the mentioned study period. To be included in the study, all hs-cTnT tests performed had to show a negative result. The diagnosis of an obvious cardiac nonischemic or noncardiac cause after the initial assessment in the emergency department was exclusion criterion. According to the protocol in our institution, patients were first evaluated by a member of the emergency department staff. If the chest pain was considered to be of possible coronary origin, the patient was then evaluated by the cardiologist on duty according to a chest pain unit protocol consisting of early exercise testing followed by hospitalization and a coronary angiogram or discharge depending on the exercise testing result. The cohort consisted of all 682 patients initially evaluated by the emergency department physician. A total of 588 patients were directly discharged, whereas the remaining 94 were evaluated in the chest pain unit; 24 of them were discharged after a negative exercise test result, and the 70 remaining patients were hospitalized.
A large amount of clinical information was collected ( Table 1 ). The TIMI (Thrombolysis In Myocardial Infarction) risk score and the GRACE (Global Registry of Acute Coronary Events) score for 6-month death or myocardial infarction were calculated. The Elecsys assay (Roche Diagnostics, Switzerland) was used for hs-cTnT determination. This method has a detection limit of 5 ng/l, a limit of blank of 3 ng/l, and a ninety-ninth percentile cut-off point of 14 ng/l, which was considered to define hs-cTnT elevation and exclusion from the study. The physician treating each patient decided the number of troponin tests performed and their timing.
