High-sensitivity cardiac troponin T in the emergency department : admissions, resource utilization and outcomes

Sammanfattning: Background Patients presenting with chest pain in the emergency department (ED) may have myocardial infarction (MI) requiring immediate treatment. High-sensitivity cardiac troponin T (hs-cTnT) was recently introduced as a biomarker that aids in determining whether the patient requires hospital admission or can be safely discharged home. The aim of this thesis was to evaluate the implementation of hs-cTnT in the ED, with respect to hospital admission, resource utilization and patient outcomes. Methods and Results Two separate datasets were created by combining administrative information from the ED at Karolinska University Hospital with laboratory data and linking several national health care registers through the National Board of Health and Welfare. The first dataset was used for Studies I and II, while the second dataset was used for Studies III and IV. Cox regression was used to calculate hazard ratios (HRs) with 95% confidence intervals (CIs). Study I: In total, 14,636 patients with chest pain who presented to the EDs at Karolinska University Hospital, Solna and Huddinge, during 2011 and 2012 were included to evaluate whether a first undetectable (<5 ng/L) hs-cTnT level and an electrocardiogram (ECG) without signs of ischaemia could be used to safely rule out MI in the ED. We identified 15 patients with an undetectable hs-cTnT level and non-ischaemic ECG who were diagnosed with MI within 30 days. The negative predictive value for MI using this strategy was 99.8%, and for death 100%. Study II: We included 13,046 patients with chest pain who visited the ED at Karolinska University Hospital, Solna and Huddinge, during 2011 and 2012. We calculated HRs at different hs-cTnT levels for the risk of revisits to the ED, readmissions to hospital and resource utilization in terms of whether the patient was discharged or admitted. In patients with a hs-cTnT level of <5 ng/L who were admitted to the hospital compared with discharged home, we observed a 24% increased risk (adjusted HR 1.24, 95% CI 1.05–1.46) of revisiting the ED within 30 days and a three-fold increased risk of coronary angiography or revascularization during follow-up. Study III: We evaluated trends in admission rates among 15,472 patients with chest pain who presented to the ED at Karolinska University Hospital, Huddinge from 2011 to 2014. Proportions of admitted patients were calculated using each year of the study period (2012, 2013 and 2014) as exposure with year 2011 as reference. We found a 36% relative reduction in hospital admissions. All-cause mortality increased (adjusted HR 1.51, 95% CI 1.18–1.92), but for non-cardiovascular causes only. Coronary angiography significantly increased, but revascularizations remained stable. Study IV: Survival and resource utilization in 31,904 patients with chest pain were compared during the initial 3 years (2011 -2013) when the hs-cTnT assay was implemented to the preceding 2 years (2009-2010) when the conventional troponin (cTnT) assay was in use at Karolinska University Hospital, Solna and Huddinge. Patients who were tested with hs-cTnT had a 15% increase in all-cause mortality (adjusted HR 1.15, 95% CI 1.02–1.29), 13% increase in coronary angiography (adjusted HR 1.13, 95% CI 1.00–1.28) and 18% increase in revascularizations (adjusted HR 1.18, 95% CI 1.01 – 1.37). Conclusions [I] Patients presenting with chest pain, a first undetectable hs-cTnT level and a normal ECG may be safely discharged from the ED because the risk of MI or death is minimal. [II] When patients with chest pain and an undetectable hs-cTnT level are admitted to the hospital instead of discharged home, they have an increased risk of revisits to the ED, recurrent hospital stays, coronary angiography and revascularization. [III] Admissions for chest pain were reduced by 36% during the first 4 years of hscTnT use. All-cause mortality increased, but for non-cardiovascular causes only. [IV] After the introduction of hs-cTnT testing in the ED, an increase in mortality, coronary angiography and revascularizations was observed.