Cardiac Health Metrics
Investigating your Heart Health
This analysis is done to understand the leading causes and targets of heart disease. The analysis is an in-depth study of metrics associated with heart disease. The data is taken from IntelliMed healthcare and analytics vendor which acts as a medical decision support system for various hospitals across Victoria. In accord with their focus, I have helped them enhance their patient care by providing them with clinical insights using data analytics. With their interest to understand and helping predict cardiac arrest, which is among the leading causes of death across Australia, I continue my analytical and visual assessment to predict the likelihood of a heart attack.
What leads to Myocardial Infarction?
A 'Myocardial Infarction', commonly known as 'Heart attack' occurs when one or more areas of the heart muscles do not receive enough oxygen. This could be due to low or no blood flow to the heart muscles.
There may be numerous factors involved in causing a Heart Attack. They may be influential like cholesterol and high blood pressure, or general factors involving heart functioning itself. With our data analysis, we intend to detect such factors and help our doctors approach their patients with more care. As said- "Prevention is always better than Cure", nonetheless, with today's new normal, "Prevention is the Cure".
Step-wise Approach:
When skimming through the sample set, binary information on patients having heart disease was noted. This helped in the choice of making a binomial regression model, namely, the Logistic Regression Model for the predictive analysis of the data set. To start with our analysis, it was pivotal to understand the terms involved in the causal of Heart Disease. For this, a dig into the domain knowledge was important, which was possible in a few meetings with some GPs and Cardiologists. This further research, brought into the light, that data here, denotes cases that could have an inclination towards myocardial infarction and the generic and influential factors involved with it. We carried out our data analysis using R, discovering important variables impacting heart disease, visualizing reports from data, and then created a dashboard for easy tracking and reporting through PowerBI.
To apply a modeling technique to predict a Heart Attack, we needed to understand the patient's lifestyle and the characteristics behind their heartache. We categorized the patients into groups of 5 based on their age as this sample distribution gives clear insight into their general vitals. We also noticed the maximum heart rate the patients have and their resting blood pressure, to understand their normal heart functioning every day. Resting ECG with the slope of their ST segment and the type of chest pain the patient endured tells the investigative clinical diagnosis on the type of heart disease.
Electrocardiogram Waves (P-QRS-T)
Waveforms
P-wave: Atrial depolarization
QRS-complex: Ventricular depolarization
T-wave: Ventricular repolarization
U-wave: Unknown wave
Some Insights:
Upon intense analysis, we found that there are different types of heart attacks suffered by patients here. Almost 44% of the cases suffer from heart disease.
Approximately 10% of the patients showed ST-segment elevated but did not have heart disease. Almost 8% of the patients with heart disease were found to suffering from Left Ventricular Hypertrophy and High Blood Pressure.
Nearly 64% of these cases are suffering from a Silent Heart-Attack, which means their ECG shows no signs of any symptom but still has suffered through a heart attack. Almost 78% of the total cases of heart disease belong. to the Silent Myocardial Infarction (SMI), which may result in premature deaths due to heart failure or strokes.
Almost 51% of cases of heart disease have suffered from asymptomatic chest pain with exercise angina (a cholesterol-clogged coronary artery disease that causes chest pain while exercising) and are found prominent with high cholesterol and borderline-to-high blood pressure.
Males above 50 years of age are more prone to heart disease whereas women of 45 years and above have a likelihood of heart disease.
Of which nearly 20% of Males suffer from diabetes and exercise angina while only 6 in 38 women suffer from the likewise agony.
Results and Reports:
Recommendations
Maintain a healthy lifestyle, i.e. no smoking, no alcohol, no cannabis, and low stress.
Keep blood sugar levels and blood pressure under control by maintaining a healthy lifestyle.
Encourage patients to achieve a healthy weight, acquire a nutritional diet, and exercise regularly.
Ask patients to monitor their cholesterol, especially if they have a family history of high or borderline cholesterol.
Must maintain regular checks on their ECG and old peaks or history of heart ailments.
Patients with a family history of heart disease must maintain a healthy lifestyle.
Suggest healthy habits to patients, especially patients above 45 years of age, as they are more prone to coronary risks.