This was a retrospective trial of chest pain patients presenting to the ED in 14 hospitals in the Asia-Pacific region. The goal of the investigation was to externally validate the HEART score against a large, independently acquired prospective study dataset concerning chest pain symptoms. The HEART score risk stratifies patients with the following 5 components: history, electrocardiogram, age, risk factors, and troponin. Three risk groups were established: < 3, low risk; 7-10, high risk. The primary endpoint was to assess the predictive value of the occurrence of a 30-day major adverse coronary event (MACE). MACE included: AMI, emergency PCI, CABG or death. The secondary enpoint also included elective PCI. The HEART's discriminative power was compared with that of the TIMI score using the C-statistics on the same data. Of the 2906 patients, 28.2% were categorized low risk. The low risk category was shown to have a MACE of 1.7% (the percentage of false negatives) with a mortality rate of 0%. Of the total patients, 16% were high risk and were documented as having a MACE of 43.1%. In comparing the predictive accuracy of HEART vs TIMI, their respective C-statistics were 0.83 and 0.75 (p < 0.01). The authors of the study concluded that this was strong external validation of the HEART score as a powerful clinical tool, which can help identify large proportions of low risk patients that can be discharged early without additional testing. The group discussed that more implementation of the HEART score in daily practice seems an appropriate step in patient management and risk stratification. However, as this scoring system has not be widely adopted at our current facility, there was much apprehension in relying on a low risk stratification to warrant discharge to home with outpatient follow-up rather than the current practice of overnight observation with a chest pain rule out protocol.
Air versus oxygen in ST-segment elevation myocardial infarction
This was a prospective, randomized controlled trial comparing the use of oxygen versus no supplemental oxygen in the setting of a STEMI without hypoxia (Spo2 > 94%) at 9 metropolitan hospitals in Melbourne, Australia. The study evaluated 441 patients with confirmed STEMI. The primary end point was MI size as determined by cardiac enzymes (troponin and creatine kinase). Secondary endpoints included: recurrent MI, arrhythmia, and infarct size at 6 months based on cardiac MRI. In terms of primary endpoints, the study saw no difference in mean peak troponin (p = 0.18), but did note increased creatine kinase in the oxygen group (1948 vs 1543, p = 0.01). In terms of secondary endpoints, the oxygen group had significant increase in rate of recurrent MI (5.5% vs 0.9%, p = 0.006), with less significant increases in arrhythmia (40.4% vs 31.4%, p = 0.05) and infarct size on MRI at 6 months (20.3 vs 13.1, p = 0.04). The authors concluded that supplemental oxygen for STEMI patients without concurrent hypoxia may increase early myocardial injury and result in larger infarct sizes at 6 months. Discussion of the article revealed an interest in more appropriately using supplemental oxygen, however, a large concern was that the trial was not powered for clinical end points. The results did not give a true sense of the subsequent morbidity and mortality when deciding upon supplemental oxygen, and that is an endpoint that holds must more weight when deciding to change the standard practice. Considering that supplemental oxygen is a drug with its own side effect profile, there was a general interest in withholding oxygen for the non-hypoxic STEMI patient. However, more data is needed on patient-centered outcomes.
The incidence and significance of bacteremia in out of hospital cardiac arrest
This was a prospective observational convenient sampling study assessing for bacteremia in out of hospital cardiac arrests (OHCA) seen at one urban academic ED in Detroit, MI. Bacteremia was assessed by collecting two blood cultures in the ED during ACLS or ROSC. Patients were bacteremic if they had two cultures growing skin flora pathogens, or one culture growing non-skin flora pathogens. The primary objective was identifying the incidence of bacteremia in OHCA patients. The secondary objective was assessing the characteristics, hospital course, and mortality of those bacteremic patients compared to the non-bacteremic in the patients who survived the ED and were able to transfer to the ICU. Of the 173 patients evaluated, 38% had bacteremia. Although bacteremia had no significant difference at 28-day mortality (93.8% vs 92.6%, p > 0.05), ED survival was significantly lower for bacteremic patients (25% vs. 40%, p < 0.042). The order of most common rhythm for both bacteremic and non-bacteremic: asystole, PEA, then VF. The most commonly identified gram-positives were S. epidermidis and Strep non-pneumoniae (combined 35.5%). The most commonly identified gram-negatives were E. coli, Klebsiella pneumoniae, and Proteus mirabilis (combined 12.2%). The study showed some significance for clinical discrepancies in bacteremic patients such as lower arterial pH, higher lactate, higher pCO2, and higher base excess (p < 0.05). Other significant lab discrepancies for bacteremic patients included higher potassium, BUN, creatinine, magnesium, and phosphate (p < 0.05). The article goes on to describe four pathways of the proposed association with bacterial infections and sudden cardiac arrest. Three of the pathways begin with a bacteremic infection that leads to sudden cardiac arrest via SIRS or acute myocardial inflammation or stress with underlying coronary artery disease. The fourth scenario is incidental bacteremia in the setting of an acute cardiac event. The authors of the study conclude that further research is needed to identify the causal relationship (bacteremia as byproduct of cardiac arrest vs. contributing factor of sepsis leading to cardiac arrest). However, they do note laboratory and survival changes that are more in line with severe sepsis and septic shock patients. Our group discussed how the findings of this study would change the management for our cardiac arrest patients. Although the data comes from a single urban ED, the results shed light on the importance of keeping differentials broad when caring for the acutely ill patient. The biggest take away was the importance of obtaining pertinent history of recent infectious symptoms and being cognizant of the metabolic derangements that can present in such patients, which may require empiric antibiotics early in the course of treatment.
Submitted by Dr. Shawn Joseph, PGY-1