Hospital Mortality SQL Prediction Analysis

To tackle the business problem of identifying the main causes of in-hospital mortality for admitted patients, I wanted to do a comprehensive analysis leveraging SQL and Tableau. My approach involved obtaining the dataset, importing it into Excel, cleaning the data, then importing it to MySQL and conducting insightful queries to extract valuable information. Using my SQL skills, I delved into the dataset, exploring various patient attributes such as age, ethnicity, gender, weight, BMI, heart rate, and comorbidities. After executing the queries with SQL, I uncovered insightful patterns, trends, and correlations within the data. To present my findings in a visually appealing manner, I used Tableau to design an interactive dashboard. The dashboard summarized the key insights derived from my SQL analysis, showcasing patterns and trends related to in-hospital mortality.

Age was the first big predictor of in-hospital mortality from this dataset. Consistently, advanced age has been associated with higher rates of medical conditions, complications, and death. According to the dataset, nearly 20% of the patients who were 70 years old or older experienced in-hospital mortality. This finding highlights the significant impact of age on mortality risk among older individuals. Another strong predictor of in-hospital mortality is comorbidities. Diabetes emerged as the most notable comorbidity, with a mortality rate of 24.45% among patients diagnosed with this condition. Heart rate can also be very telling to predict mortality as it reflects the cardiovascular status and overall physiological stability of patients. Abnormalities in heart rate, such as tachycardia (elevated heart rate) or bradycardia (low heart rate), often indicate underlying cardiovascular dysfunction or compromised perfusion. Research has consistently shown a strong association between abnormal heart rates and increased mortality risk. In this dataset, the average max heart rate for patients that died was 115.1. Generally, a resting heart rate is high if it is over 100 bpm . The length of stay at an ICU was also an important indicator of mortality, showing that patients who stayed for less than a day at the ICU had a much better chance for survival.