National Cardiovascular Disease Database Research Articles

Risk stratification of Asian patients after ST-elevation myocardial infarction using machine learning methods

Abstract Background Risk stratification in ST-elevation myocardial infarction (STEMI) that is population-specific is essential. Conventional risk stratification methods such Thrombolysis in Myocardial Infarction (TIMI) score is used to evaluate the risk associated with the acute coronary syndrome (ACS) which are derived from Western Caucasian cohort with a limited participant from the Asian region. In Malaysia, multi-ethnic developing country, patients presenting with STEMI are younger, have a much higher prevalence of diabetes, hypertension and renal failure, and present later to medical care than their western counterparts. Purpose We aim to investigate the predictors, predict mortality and develop a risk stratification tool for short and long term mortality in multi-ethnic STEMI patients using machine learning (ML) method. Methods We created three separate mortality prediction models using support vector machine (SVM) to identify predictors and predict mortality for in-hospital, 30-days and 1-year for STEMI patients. We used registry data from the National Cardiovascular Disease Database of 6299 patient's data for in-hospital, 3130 for 30-days and 2939 for 1-year for ML model development. Fifty parameters including demographics, cardiovascular risk, medications and clinical variables were utilised for training the models. The Area under the curve (AUC) was used as the primary performance evaluation metric. All models were validated against conventional method TIMI and tested using testing data. SVM variable importance method were used to select and rank important variables. We converted the final algorithm into an online tool with a database for continuous algorithm validation. We implemented the online calculator in selected hospitals for further testing using prospective patients data. Results The calculator is available at http://myheartstemi.uitm.edu.my. The calculator outperforms TIMI on testing data for in-hospital (15 predictors) (AUC=0.88 vs 0.81), 30 days (12 predictors) (AUC=0.90 vs 0.80) and 1-year (13 predictors) (AUC=0.84 vs 0.76). Common predictors for in-hospital, 30 days and 1-year mortality model identified in this study are; age, heart rate, Killip class, fasting blood glucose and diuretics. Invasive and less invasive treatments such as PCI pharmacotherapy drugs are also selected as important variables that improve mortality prediction. Our results also suggest that TIMI score underestimates patients risk of mortality. 90% of non-survival patients are classified as high risk (>30%) by the calculator compared 10–30% non-survival patients by TIMI. Conclusions In the multi-ethnicity population, patients with STEMI are better classified using ML method compared to the TIMI score. ML allows identification of distinct factors in unique ASIAN population for better mortality prediction. Availability of population-specific calculator and continuous testing and validation allows better risk stratification. Machine learning and TIMI performance Funding Acknowledgement Type of funding source: Public Institution(s). Main funding source(s): University of Malaya Grant

Predicting Sudden Deaths Following Myocardial Infarction in Malaysia Using Machine Learning Classifiers

Myocardial infarction (MI) is among the top causes of death in Malaysia. The mortality rate following MI was high, especially within the first 30 days after the onset. This paper study the ability of k-Nearest Neighbors (kNN) and Naïve Bayes algorithms to predict the 30-day mortality of MI patients, using. The dataset used for this study is provided by National Cardiovascular Disease Database (NCVD) which consist of 2840 MI patients from hospitals in Malaysia. The sudden death predictions made by the machine learning are based on the age, gender, year of onset, smoking habit, BMI, diabetes, hypertension and cholesterol level. The result suggests that kNN algorithm has better performance in predicting the sudden death compared to Naïve Bayes. The number of independent variables plays an important role in mortality prediction, and removing insignificant variables improve the performance.

Outcomes of STEMI patients with chronic kidney disease treated with percutaneous coronary intervention: the Malaysian National Cardiovascular Disease Database \u2013 Percutaneous Coronary Intervention (NCVD-PCI) registry data from 2007 to 2014

BackgroundPatients with renal impairment often left out from most major clinical trials assessing the optimal treatment for ST-elevation myocardial infarction (STEMI). Large body of evidence from various cardiovascular registries reflecting more ‘real-world’ experience might contribute to the knowledge on how best to treat this special cohort. We aim to analyze the outcomes of Malaysian STEMI patients with renal impairment treated with coronary angioplasty.MethodsUtilizing the Malaysian National Cardiovascular Disease Database-Percutaneous Coronary Intervention (NCVD-PCI) registry data from 2007 to 2014, STEMI patients treated with percutaneous coronary intervention (PCI) were stratified into presence (GFR < 60 mls/min/1.73m2) or absence (GFR ≥ 60 mls/min/1.73m2) of chronic kidney disease (CKD). Patient’s demographics, extent of coronary artery disease, procedural data, discharge medications, short (in-hospital) and long (1 year) term outcomes were critically assessed.ResultsA total of 6563 patients were included in the final analysis. STEMI CKD cohort was predominantly male (80%) with mean age of 61.02 ± 9.95 years. They had higher cardiovascular risk factors namely diabetes mellitus (54.6%), hypertension (79.2%) and dyslipidemia (68.8%) in contrast to those without CKD. There were notably higher percentage of CKD patients presented with Killip class 3 and 4; 24.9 vs 8.7%. Thrombolytic therapy remained the most commonly instituted treatment regardless the status of kidney function. Furthermore, our STEMI CKD cohort also was more likely to receive less of evidence-based treatment upon discharge. In terms of outcomes, patients with CKD were more likely to develop in-hospital death (OR: 4.55, 95% CI 3.11–6.65), MACE (OR: 3.42, 95% CI 2.39–4.90) and vascular complications (OR: 1.88, 95% CI 0.95–3.7) compared to the non-CKD patients. The risk of death at 1-year post PCI in STEMI CKD patients was also reported to be high (HR: 3.79, 95% CI 2.84–5.07).ConclusionSTEMI and CKD is a deadly combination, proven in our cohort, adding on to the current evidence in the literature. We noted that our STEMI CKD patients tend to be younger than the Caucasian with extremely high prevalence of diabetes mellitus. The poor outcome mainly driven by immediate or short term adverse events peri-procedural, therefore suggesting that more efficient treatment in this special group is imperative.

Mortality and lipid trends in ACS populations: A study from the Malaysian national cardiovascular disease database - Acute coronary syndrome (NCVD-ACS) registry

Mortality and lipid trends in ACS populations: A study from the Malaysian national cardiovascular disease database - Acute coronary syndrome (NCVD-ACS) registry

Association between body mass index and outcomes after percutaneous coronary intervention in multiethnic South East Asian population: a retrospective analysis of the Malaysian National Cardiovascular Disease Database—Percutaneous Coronary Intervention (NCVD-PCI) registry

ObjectiveTo examine the relationship between body mass index (BMI) and outcomes after percutaneous coronary intervention (PCI) in a multiethnic South East Asian population.SettingFifteen participating cardiology centres contributed to the Malaysian...

Sex Differences in Acute Coronary Syndrome in a Multiethnic Asian Population: Results of the Malaysian National Cardiovascular Disease Database—Acute Coronary Syndrome (NCVD-ACS) Registry

BackgroundSex differences in acute coronary syndrome (ACS) have been well studied in major registries and clinical trials in Western populations. Limited studies have examined the sex differences in ACS using a large number of Asian women as the subjects. ObjectivesThe aim was to study the sex differences in ACS using the NCVD-ACS (National Cardiovascular Disease Database—Acute Coronary Syndrome) registry. MethodsWe analyzed 13,591 ACS patients, of which 75.8% were men and 24.2% were women, from March 2006 to February 2010. Data were collected on demographic characteristics, risk factors, anthropometrics, treatments, procedures, mortalities, and complications. The results were compared among 3 cohorts of ACS (ST-segment elevation myocardial infarction [STEMI], non–STEMI, and unstable angina). ResultsWomen were older and more likely to have diabetes, hypertension, previous heart failure, and cerebral vascular accidents than men were. Women were less likely to receive in-hospital administration of aspirin, beta-blockers, angiotensin-converting enzyme inhibitors, or angiotensin receptor blockers, and they were less likely to undergo angiography and percutaneous coronary intervention. In STEMI, a significantly lower proportion of women than men received primary percutaneous coronary intervention (6.2% vs. 6.7%, respectively, p = 0.000) and fibrinolysis (64.4% vs. 74.6%, respectively, p = 0.000). In addition, with regard to STEMI, women had a significantly higher unadjusted in-hospital mortality rate than men did (15.0% vs. 8.1%, respectively, p < 0.000). There was no statistically significant in-hospital mortality difference between sexes for non-STEMI and unstable angina. After adjustment for age and other covariates, a multivariate analysis showed no sex differences in the in-hospital mortality in all spectrums of ACS. ConclusionsOur study showed significant sex differences in the demographic characteristics, risk factors, treatments, and outcomes of ACS. More importantly, in ACS patients, we found evidence of suboptimal treatments and interventions in women versus men. Our findings provide an opportunity to narrow the sex gap in the care of women with ACS in Malaysia.

Ethnic differences in the occurrence of acute coronary syndrome: results of the Malaysian National Cardiovascular Disease (NCVD) Database Registry (March 2006 - February 2010).

BackgroundThe National Cardiovascular Disease (NCVD) Database Registry represents one of the first prospective, multi-center registries to treat and prevent coronary artery disease (CAD) in Malaysia. Since ethnicity is an important consideration in the occurrence of acute coronary syndrome (ACS) globally, therefore, we aimed to identify the role of ethnicity in the occurrence of ACS among high-risk groups in the Malaysian population.MethodsThe NCVD involves more than 15 Ministry of Health (MOH) hospitals nationwide, universities and the National Heart Institute and enrolls patients presenting with ACS [ST-elevation myocardial infarction (STEMI), non-ST elevation myocardial infarction (NSTEMI) and unstable angina (UA)]. We analyzed ethnic differences across socio-demographic characteristics, hospital medications and invasive therapeutic procedures, treatment of STEMI and in-hospital clinical outcomes.ResultsWe enrolled 13,591 patients. The distribution of the NCVD population was as follows: 49.0% Malays, 22.5% Chinese, 23.1% Indians and 5.3% Others (representing other indigenous groups and non-Malaysian nationals). The mean age (SD) of ACS patients at presentation was 59.1 (12.0) years. More than 70% were males. A higher proportion of patients within each ethnic group had more than two coronary risk factors. Malays had higher body mass index (BMI). Chinese had highest rate of hypertension and hyperlipidemia. Indians had higher rate of diabetes mellitus (DM) and family history of premature CAD. Overall, more patients had STEMI than NSTEMI or UA among all ethnic groups. The use of aspirin was more than 94% among all ethnic groups. Utilization rates for elective and emergency percutaneous coronary intervention (PCI) and coronary artery bypass graft (CABG) were low among all ethnic groups. In STEMI, fibrinolysis (streptokinase) appeared to be the dominant treatment options (>70%) for all ethnic groups. In-hospital mortality rates for STEMI across ethnicity ranges from 8.1% to 10.1% (p = 0.35). Among NSTEMI/UA patients, the rate of in-hospital mortality ranges from 3.7% to 6.5% and Malays recorded the highest in-hospital mortality rate compared to other ethnic groups (p = 0.000). In binary multiple logistic regression analysis, differences across ethnicity in the age and sex-adjusted ORs for in-hospital mortality among STEMI patients was not significant; for NSTEMI/UA patients, Chinese [OR 0.71 (95% CI 0.55, 0.91)] and Indians [OR 0.57 (95% CI 0.43, 0.76)] showed significantly lower risk of in-hospital mortality compared to Malays (reference group).ConclusionsRisk factor profiles and ACS stratum were significantly different across ethnicity. Despite disparities in risk factors, clinical presentation, medical treatment and invasive management, ethnic differences in the risk of in-hospital mortality was not significant among STEMI patients. However, Chinese and Indians showed significantly lower risk of in-hospital mortality compared to Malays among NSTEMI and UA patients.

Young coronary artery disease in patients undergoing percutaneous coronary intervention

BACKGROUND AND OBJECTIVESUnderstanding the nature and pattern of young coronary artery disease (CAD) is important due to the tremendous impact on these patients’ socio-economic and physical aspect. Data on young CAD in the southeast Asian region is rather patchy and limited. Hence we utilized our National Cardiovascular Disease Database (NCVD)—Percutaneous Coronary Intervention (PCI) Registry to analyze young patients who underwent PCI in the year 2007 to 2009.DESIGN AND SETTINGSThis is a retrospective study of all patients who had undergone coronary angioplasty from 2007 to 2009 in 11 hospitals across Malaysia.METHODSData were obtained from the NCVD—PCI Registry, 2007 to 2009. Patients were categorized into 2 groups—young and old, where young was defined as less than 45 years for men and less than 55 years for women and old was defined as more than or equals to 45 years for men and more than or equals to 55 years for women. Patients’ baseline characteristics, risk factor profile, extent of coronary disease and outcome on discharge, and 30-day and 1-year follow-up were compared between the 2 groups.RESULTSWe analyzed 10 268 patients, and the prevalence of young CAD was 16% (1595 patients). There was a significantly low prevalence of Chinese patients compared to other major ethnic groups. Active smoking (30.2% vs 17.7%) and obesity (20.9% vs 17.3%) were the 2 risk factors more associated with young CAD. There is a preponderance toward single vessel disease in the young CAD group, and they had a favorable clinical outcome in terms of all-cause mortality at discharge (RR 0.49 [CI 0.26–0.94]) and 1-year follow-up (RR 0.47 [CI 0.19–1.15]).CONCLUSIONWe observed distinctive features of young CAD that would serve as a framework in the primary and secondary prevention of the early onset CAD.

Are There Gender Differences in Coronary Artery Disease? The Malaysian National Cardiovascular Disease Database – Percutaneous Coronary Intervention (NCVD-PCI) Registry

ObjectivesTo assess whether gender differences exist in the clinical presentation, angiographic severity, management and outcomes in patients with coronary artery disease (CAD).MethodsThe study comprised of 1,961 women and 8,593 men who underwent percutaneous coronary intervention (PCI) and were included in the Malaysian NCVD-PCI Registry from 2007–2009. Significant stenosis was defined as ≥70% stenosis in at least one of the epicardial vessels.ResultsWomen were significantly older and had significantly higher rates of diabetes mellitus, hypertension, chronic renal failure, new onset angina and prior history of heart failure whereas smokers and past history of myocardial infarction were higher in men. In the ST-elevation myocardial infarction (STEMI) cohort, more women were in Killip class III-IV, had longer door-to-balloon time (169.5 min. vs 127.3 min, p<0.052) and significantly longer transfer time (300.4 min vs 166.3 min, p<0.039). Overall, women had significantly more left main stem (LMS) disease (1.3% vs 0.6%, p<0.003) and smaller diameter vessels (<3.0 mm: 45.5% vs 34.8%, p<0.001). In-hospital mortality rates for all PCI, STEMI, Non-STEMI (NSTEMI) and unstable angina for women and men were 1.99% vs 0.98%, Odds ratio (OR): 2.06 (95% confidence interval (CI): 1.40 to 3.01), 6.19% vs 2.88%, OR: 2.23 (95% CI: 1.31 to 3.79), 2.90% vs 0.79%, OR: 3.75 (95% CI: 1.58 to 8.90) and 1.79% vs 0.29%, OR: 6.18 (95% CI: 0.56 to 68.83), respectively. Six-month adjusted OR for mortality for all PCI, STEMI and NSTEMI in women were 2.18 (95% CI: 0.97 to 4.90), 2.68 (95% CI: 0.37 to 19.61) and 2.66 (95% CI: 0.73 to 9.69), respectively.ConclusionsWomen who underwent PCI were older with more co-morbidities. In-hospital and six-month mortality for all PCI, STEMI and NSTEMI were higher due largely to significantly more LMS disease, smaller diameter vessels, longer door-to-balloon and transfer time in women.

Acute coronary syndrome in women of reproductive age

BackgroundThere is scarce or no data on prevalence and presentation of acute coronary syndrome (ACS) among women of reproductive age. Furthermore, whether women of reproductive age presenting with ACS have the same risk factors as men and older women is not known.ObjectiveTo analyze factors associated with ACS in women of reproductive age in comparison with older women and men of a similar age group.MethodologyA total of 9702 cases of acute coronary syndrome over a 3-year period ( 2006–2008) from the National Cardiovascular Disease database were analyzed, with focus on women of reproductive age (20–<40 years), looking into association with ethnicity, comorbid illness, and the ACS stratum. Comparison with older women (40–<60 years; Control 1) and men of similar age group (Control 2) was made and analyzed using Fisher’s exact test and chi-square test when necessary.ResultsFrom a total of 9702 cases, 2344 (24.2%) were women. Of these, 45 (1.9%) were women between 20 and <40 years, which is significantly lower than the two controls (older women 30.8%, and men of same age 6.2%, respectively; P < 0.0001). The distribution of ethnicity shows a similar pattern between the study group and the controls, but patients of Indian ethnicity were over-represented when compared with the Malaysian demographics of general population (31.3% versus 7.1%; P < 0.0001). ACS in women of reproductive age was associated with diabetes mellitus in 37.8%, hypertension in 40.0%, and dyslipidemia in 24.4% of cases, similar to men of the same age but significantly lower than the older women (P < 0.0001). Smoking is not a major risk factor in the study group, where only 6.7% ever smoked, similar to older women (6.8%, P = 1.000) and significantly much less compared with men of the same age (84.1%; P < 0.0001). Regarding the ACS stratum, a significantly higher percentage of women in the study group had ST-segment elevation myocardial infarction compared with older women (P = 0.0085) but less than that of men of similar age (P = 0.0187).ConclusionACS is rare in women of reproductive age. Diabetes, hypertension, and Indian ethnicity were identified as important contributors.

The journey of Malaysian NCVD—PCI (National Cardiovascular Disease Database—Percutaneous Coronary Intervention) Registry: A summary of three years report

IntroductionThe Malaysian National Cardiovascular Disease Database (NCVD) team presents Percutaneous Coronary Intervention (PCI) Registry report for the year 2007 to 2009. It provides comprehensive information regarding practice and outcome of PCI in Malaysia. MethodologyIt was a voluntary, multi-centered, observational, cohort study and included patients of 18years or above who underwent PCI at eleven participating centers in Malaysia from the year 2007 to 2009. ResultTen thousand six hundred and two patients underwent 11,498 PCI procedures with 18,116 stents for 15,538 lesions. Mean age of the patients was 57years and more than 98% of patients had at least one cardiovascular risk factor. A significant number of our patients were diabetic (50%) and had renal impairment (44.7% had≤stage 3 chronic kidney disease) at the time of procedure. Fifty eight percent of the lesions were type B2 or type C lesion. Twenty eight percent of the lesions had high risk characteristics. Procedural success rate was about 97% and post-procedural complications were low. Overall in-hospital, all cause mortality was 1%, of which 85% were cardiac related deaths. The poor prognostic factors for in-hospital mortality were acute coronary syndrome cases, higher Killip class and increasing age. ConclusionCompared to other registries, Malaysian patients undergoing PCI were much younger with high prevalence of risk factors. In spite of complex and high risk lesions, procedural success was high, with overall low mortality rate. NCVD—PCI Registry aims to improve over-all cardiac services in Malaysia through its ongoing journey.

Malaysian National Cardiovascular Disease Database (NCVD) – Acute Coronary Syndrome (ACS) registry: How are we different?

Objective The Malaysian NCVD-ACS (National Cardiovascular Disease Database-Acute Coronary Syndrome) registry attempts to determine the incidence and demographic profile of patients with ACS and to identify high risk group among them. It provides a comprehensive view to assess treatment strategies and adherence to existing guidelines for ACS patients; which can help in future development. It also aims to facilitate research using this database. Methods The study included patients with ST-segment elevation myocardial infarction (STEMI), non-ST-elevation myocardial infarction (NSTEMI) and unstable angina (UA) admitted to 11 participating sites in Malaysia from 1st January 2006 to 31st December 2006. The data were analyzed in terms of characteristics, clinical presentation, treatment, in-hospital outcome and 30-day outcome. Results A total of 3422 patients were enrolled, with male to female ratio of 3:1, mean age of 59 ± 12 years and mean BMI of 25.8 ± 4.4 kg/m 2. Among total study population, 96% had at least one established cardiovascular risk factor. In STEMI population, 70% received fibrinolytic therapy and 8% proceeded for primary percutaneous coronary intervention (PCI). Mean door-to-needle time for fibrinolytic therapy was 103 ± 143 min. Medical management was conducted for 86% of NSTEMI and 91% UA patients, with intervention for 14% and 9% respectively. The overall in-hospital mortality and 30-day mortality were 7% and 8% respectively. Conclusion In our NCVD-ACS registry, when compared with other registries, the subjects were much younger at presentation and had higher prevalence of established cardiovascular risk factors, indicating the importance of primary prevention programme. Majority of the patients were managed medically, with low rates of cardiac interventions, the factor being driven largely by availability of resources. Mean door-to-needle time was much higher than the recommended guidelines, an issue that needs attention. The results indicate many opportunities for improvement of in-hospital and 30-day mortality.

Highlights of the first Malaysian National Cardiovascular Disease Database (NCVD): Percutaneous Coronary Intervention (PCI) Registry

Summary Objective The Malaysian NCVD-PCI registry attempts to determine the number and to monitor the outcomes of Percutaneous Coronary Intervention (PCI), based on selected performance indicators. It provides a comprehensive view to determine the level of adherence to existing guidelines, to evaluate the cost-effectiveness of treatment and prevention programs and to facilitate quality improvement activities of the participants. It also aims to stimulate research and to act as a reference for future studies. Methods It was a voluntary, multi-centered, observational, cohort study and included patients of 18 years or above, with coronary artery disease who underwent PCI at eight participating centers in the year 2007. Results A total of 3677 patients underwent 3920 PCI procedures with 6299 stents for 5512 lesions. The mean age of patients was 56.7 ± 10.11 years. The mean BMI was 26.38 ± 4.21 kg/m2, while 80% of all subjects had a BMI above 23 kg/m2. Males constituted 81.2% of the total population and 98.4% of the total population had at least one cardiovascular risk factor. Regarding PCI status, 90.1% were elective cases and 94% of cases had a low TIMI risk index at the beginning of PCI. Femoral approach accounted for 59%, radial approach for 34% and brachial approach for 1% of all cases. The median fluoroscopy time was 15.7 min and the median door-to-balloon time for primary infarct PCI was 93.5 min. The commonest site of lesion was the left anterior descending artery, accounting for 48% of all lesions and 92.8% of all lesions were de novo. The mean lesion length was 24.4 ± 15.18 mm and about 28% of all lesions had high risk characteristics. Drug-eluting stents and bare metal stents were used in 53.6% and 42.5% of cases, respectively. After PCI, 91% of all lesions achieved TIMI grade 3 flow. Regarding pharmacotherapy, 99.5% of all patients received unfractionated heparin, 5% received LMWH prior to intervention, 96% received aspirin and 98% received clopidogrel. Over-all in-hospital mortality and 30 day mortality for the entire cohort was 1.1% and 1.8%, respectively. Conclusion A summary of the first nationwide PCI registry has been presented. The subjects were much younger with a high prevalence of cardiovascular risk factors. The majority of cases (90%) were elective procedures with a low TIMI risk index. Mean door-to-balloon time for primary PCI was higher than the recommended guidelines. There was good prescribing of antiplatelets and heparin. Over-all in-hospital and 30 day mortality were comparable to other registries.