Going beyond summed stress scores: Correlating global and territorial coronary flow reserve by single photon emission tomography with routine myocardial perfusion imaging.

a Background: Technological advancement in the recent years has enabled the application of single photon emission tomography (SPECT) to evaluate myocardial blood flow (MBF). This method offers increased sensitivity in the assessment of coronary health, quantifiable through non-invasive imaging beyond the more conventional methods such as with myocardial perfusion imaging (MPI). b Aims: To correlate MBF, derived by dynamic SPECT, both global and by coronary territories to the summed stress scores (SSS) on conventional MPI. c Methods: Images obtained from dipyridamole-gated SPECT MPI stress and rest studies performed on recruited subjects were examined. We calculated the global and regional coronary flow reserve (CFR) via a standard software package, taken as the ratio of stress MBF to rest MBF, using CFR<2.5 as the cut off. d Results: Amongst the 90 recruited subjects (mean age 67 ± 8 years; of which 76% were males), 49% had MPI within normal limits (summed stress score (SSS) 0-3; Left ventricular ejection fraction (LVEF) > 50%). We observed a progressive reduction in global and regional CFR across the normal SSS category to that of severely abnormal (SSS >13). Reduced global CFR with correspondent lower CFR across the regional arteries were detected in scans within normal limits of MPI scans in subjects who were older (69 ± 7 vs. 62 ± 9 years, p = 0.034). Decreasing CFR was significantly associated with increasing age across the regional arteries. e Conclusion: In our study we depict the global and regional MBF values obtained via SPECT MPI in correlation to the respective SSS categories. Our data proposes that dynamic SPECT has a part in refining cardiac risk stratification, particularly in the older adult population, who are at greater risk.

Longitudinal analysis of thoracic aortic expansion in non-syndromic real-world patients.

Remodeling of the thoracic aorta is commonly seen and viewed as a precursor to an aortic aneurysm. However, while aneurysms have been shown to expand at a rate of approximately 1 mm annually, the expansion of the pre-aneurysmal aorta is poorly characterized, especially in relation to age, gender, and aortic size per se. We identified patients that had undergone echocardiography at least twice at a large university medical center. Diagnosis codes, medications, and blood test results were obtained from hospital records. Syndromic patients were excluded (e.g., Marfan's syndrome, bicuspid aortic valve). Final population comprised n = 24,928 patients (median age 61.2 years (inter-quartile range (IQR): 50.6-71.5); 55.8% males) that had undergone a median of 3 echocardiograms (2-4; range 2-27) during a median of 4.0 years (IQR: 2.3-6.2). Hypertension was present in 39.6% of patients and diabetes in 20.7%, median LV ejection fraction was 56.0% (IQR: 41.0-62.0). Aortic size measurements were analyzed in mixed models while clustering on individual patients. Mean expansion was determined for sinus of Valsalva as 1.93 (95% confidence interval; CI95: 1.87-1.99) mm per decade, and for ascending aorta as 1.76 (CI95: 1.70-1.82) mm per decade. Faster expansion was found in males, with larger aortic size, and younger age (p for interaction <0.05 for all). In conclusion, expansion of the thoracic aorta, in real world, non-syndromic patients, is slow and averages <2 mm per decade. This will help to inform management of this large patient group.

Machine learning versus classical electrocardiographic criteria for echocardiographic left ventricular hypertrophy in a pre-participation cohort.

BACKGROUND: Classical electrocardiographic (ECG) criteria for left ventricular hypertrophy (LVH) are well studied in older populations and patients with hypertension. Their utility in young pre-participation cohorts is unclear. AIMS: We aimed to develop machine learning models for detection of echocardiogram-diagnosed LVH from ECG, and compare these models with classical criteria. METHODS: Between November 2009 and December 2014, pre-participation screening ECG and subsequent echocardiographic data was collected from 17 310 males aged 16 to 23, who reported for medical screening prior to military conscription. A final diagnosis of LVH was made during echocardiography, defined by a left ventricular mass index >115 g/m2. The continuous and threshold forms of classical ECG criteria (Sokolow-Lyon, Romhilt-Estes, Modified Cornell, Cornell Product, and Cornell) were compared against machine learning models (Logistic Regression, GLMNet, Random Forests, Gradient Boosting Machines) using receiver-operating characteristics curve analysis. We also compared the important variables identified by machine learning models with the input variables of classical criteria. RESULTS: Prevalence of echocardiographic LVH in this population was 0.82% (143/17310). Classical ECG criteria had poor performance in predicting LVH. Machine learning methods achieved superior performance: Logistic Regression (area under the curve [AUC], 0.811; 95% confidence interval [CI], 0.738-0.884), GLMNet (AUC, 0.873; 95% CI, 0.817-0.929), Random Forest (AUC, 0.824; 95% CI, 0.749-0.898), Gradient Boosting Machines (AUC, 0.800; 95% CI, 0.738-0.862). CONCLUSIONS: Machine learning methods are superior to classical ECG criteria in diagnosing echocardiographic LVH in the context of pre-participation screening.

Ethnic Differences and Trends in ST-Segment Elevation Myocardial Infarction Incidence and Mortality in a Multi-Ethnic Population.

INTRODUCTION: This study aimed to compare the incidence and mortality of ST-segment elevation myocardial infarction (STEMI) across the 3 main ethnic groups in Singapore, determine if there is any improvement in trends over the years and postulate the reasons underlying the ethnic disparity. MATERIALS AND METHODS: This study consisted of 16,983 consecutive STEMI patients who sought treatment from all public hospitals in Singapore from 2007 to 2014. RESULTS: Compared to the Chinese (58 per 100,000 population in 2014), higher STEMI incidence rate was consistently observed in the Malays (114 per 100,000 population) and Indians (126 per 100,000 population). While the incidence rate for the Chinese and Indians remained relatively stable over the years, the incidence rate for the Malays rose slightly. Relative to the Indians (30-day and 1-year all-cause mortality at 9% and 13%, respectively, in 2014), higher 30-day and 1-year all-cause mortality rates were observed in the Chinese (15% and 21%) and Malays (13% and 18%). Besides the Malays having higher adjusted 1-year all-cause mortality, all other ethnic disparities in 30-day and 1-year mortality risk were attenuated after adjusting for demographics, comorbidities and primary percutaneous coronary intervention. CONCLUSION: It is important to continuously evaluate the effectiveness of existing programmes and practices as the aetiology of STEMI evolves with time, and to strike a balance between prevention and management efforts as well as between improving the outcome of "poorer" and "better" STEMI survivors with finite resources.

Allometric Relationships for Cardiac Size and Longitudinal Function in Healthy Chinese Adults　- Normal Ranges and Clinical Correlates.

BACKGROUND: Cardiac size measurements require indexing to body size. Allometric indexing has been investigated in Caucasian populations but a range of different values for the so-called allometric power exponent (b) have been proposed, with uncertainty as to whether allometry offers clinical utility above body surface area (BSA)-based indexing. We derived optimal values for b in normal echocardiograms and validated them externally in cardiac patients. METHODS AND RESULTS: Values for b were derived in healthy adult Chinese males (n=1,541), with optimal b for left ventricular mass (LVM) of 1.66 (95% confidence interval 1.41-1.92). LV hypertrophy (LVH) defined as indexed LVM >75 g/m1.66 was associated with adverse outcomes in an external validation cohort (n=738) of patients with acute coronary syndrome (odds ratio for reinfarction: 2.4 (1.1-5.4)). In contrast, LVH defined by BSA-based indexing or allometry using exponent 2.7 exhibited no significant association with outcomes (P=NS for both). Cardiac longitudinal function also varied with body size: septal and RV free wall s', TAPSE and lateral e' all scaled allometrically (b=0.3-0.9). CONCLUSIONS: An optimal b of 1.66 for LVM in healthy Chinese was found to validate well, with superior clinical utility both to that of BSA-based indexing and to b=2.7. The effect of allometric indexing of cardiac function requires further study.

Triaging Primary Care Patients Referred for Chest Pain to Specialist Cardiology Centres: Efficacy of an Optimised Protocol.

INTRODUCTION: Patients referred for chest pain from primary care have increased, along with demand for outpatient cardiology consultations. We evaluated 'Triage Protocol' that implements standardised diagnostic testing prior to patients' first cardiology consultation. MATERIALS AND METHODS: Under the 'Triage Protocol', patients referred for chest pain were pretriaged using a standardised algorithm and subsequently referred for relevant functional diagnostic cardiology tests before their initial cardiology consultation. At the initial cardiology consultation scheduled by the primary care provider, test results were reviewed. A total of 522 triage patients (mean age 55 ± 13, male 53%) were frequency-matched by age, gender and risk cohort to 289 control patients (mean age: 56 ± 11, male: 52%). Pretest risk of coronary artery disease was defined according to a Modified Duke Clinical Score (MDCS) as low (<10), intermediate (10-20) and high (>20). The primary outcome was time from referral to diagnosis (days). Secondary outcomes were total visits, discharge rate at first consultation, patient cost and adverse cardiac outcomes. RESULTS: The 'Triage Protocol' resulted in shorter times from referral to diagnosis (46 vs 131 days; P <0.0001) and fewer total visits (2.4 vs 3.0; P <0.0001). However, triage patients in low-risk groups experienced higher costs due to increased testing (S$421 vs S$357, P = 0.003). Adverse cardiac event rates under the 'Triage Protocol' indicated no compromise to patient safety (triage vs control: 0.57% vs 0.35%; P = 1.000). CONCLUSION: By implementing diagnostic cardiac testing prior to patients' first specialist consultation, the 'Triage Protocol' expedited diagnosis and reduced subsequent visits across all risk groups in ambulatory chest pain patients.

HEART score performance in Asian and Caucasian patients presenting to the emergency department with suspected acute coronary syndrome.

BACKGROUND: The HEART score is a simple and effective tool to predict short-term major adverse cardiovascular events in patients suspected of acute coronary syndrome. Patients are assigned to three risk categories using History, ECG, Age, Risk factors and Troponin (HEART). The purpose is early rule out and discharge is considered safe for patients in the low risk category. Its performance in patients of Asian ethnicity is unknown. We evaluated the performance of the HEART score in patients of Caucasian, Chinese, Indian and Malay ethnicity. METHODS: The HEART score was assessed retrospectively in 3456 patients presenting to the emergency department with suspected acute coronary syndrome (1791 Caucasians, 1059 Chinese, 344 Indians, 262 Malays), assigning them into three risk categories. RESULTS: The incidence of major adverse cardiovascular events within six weeks after presentation was similar between the ethnic groups. A smaller proportion of Caucasians was in the low risk category compared with Asians (Caucasians 35.8%, Chinese 43.5%, Indians 45.3%, Malays 44.7%, p<0.001). The negative predictive value of a low HEART score was comparable across the ethnic groups, but lower than previously reported (Caucasians 95.3%, Chinese 95.0%, Indians 96.2%, Malays 96.6%). Also the c-statistic for the HEART score was not significantly different between the groups. CONCLUSIONS: These results show that the overall performance of the HEART score is equal among Caucasian and Asian ethnic groups. The event rate in the low risk group, however, was higher than reported in previous studies, which queries the safety of early discharge of patients in the low risk category.

Impact of Direct Cardiovascular Laboratory Activation by Emergency Physicians on False-Positive Activation Rates.

INTRODUCTION: Door-to-balloon (DTB) time is critical to ST elevation myocardial infarction (STEMI) patients' survival. Although DTB time is reduced with direct cardiovascular laboratory (CVL) activation by emergency physicians, concerns regarding false-positive activation remain. We evaluate false-positive rates before and after direct CVL activation and factors associated with false-positive activations. MATERIALS AND METHODS: This is a retrospective single centre study of all emergency CVL activation 3 years before and after introduction of direct activation in July 2007. False-positive activation is defined as either: 1) absence of culprit vessel with coronary artery thrombus or ulceration, or 2) presence of chronic total occlusion of culprit vessel, with no cardiac biomarker elevations and no regional wall abnormalities. All false-positive cases were verified by reviewing their coronary angiograms and patient records. RESULTS: A total of 1809 subjects were recruited; 84 (4.64%) identified as false-positives. Incidence of false-positive before and after direct activation was 4.1% and 5.1% respectively, which was not significant (P = 0.315). In multivariate logistic regression analysis, factors associated with false-positive were: female (odds ratio (OR): 2.104 [1.247-3.548], P = 0.005), absence of chest pain (OR: 5.369 [3.024-9.531], P <0.0001) and presence of only left bundle branch block (LBBB) as indication for activation (OR: 65.691 [19.870-217.179], P <0.0001). CONCLUSION: Improvement in DTB time with direct CVL activation by emergency physicians is not associated with increased false-positive activations. Factors associated with false-positive, especially lack of chest pain or LBBB, can be taken into account to optimise STEMI management.