PUBH 1242 HW #1

Introduction

This dataset contains 46 patients reported with traumatic brain injury. Variables in this dataset include age, gender, Glasgow Coma Scale scores & seizure indicators.

The Glasgow Coma Scale is a tool used to assess how severe brain injury is. It categorizes patents based off their visual, verbal, and motor responses and ranges from a 3 (severe injury) to 15 (no injury).

For this assignment, we will use various visualization techniques to explore the how GCS scores vary based on the mechanism of injury, their relationship with seizures, and whether they can predict long-term patient outcomes to better understand the factors influencing TBI recovery and diagnosis.

Histogram of Field GCS

The histogram shows the distribution of Field GCS scores for TBI patients. It should be noted that there is a concentration of scores around the lower end of scale, implying patients had severe brain injuries during their initial assessment. The drop-off in frequency as the patient GCS scores increase suggest fewer patients had mild injuries.

Density Plot of Age

This density plot highlights the central tendency and spread of ages in the dataset, with the mean age (37 years old) indicated. The plot reveals that a majority of patients are younger than the average age.

Pie Chart of Gender Distribution

This pie chart displays the distribution of males and females in the TBI dataset. The chart shows a higher percentage of males vs females, emphasizing a gender imbalance in the sample which can be relevant when trying to understand gender-specific trends.

Heatmap of correlation matrix between numerical variables

The correlation matrix displays the relationship between all of the numerical variables in the dataset. The heatmap reveals positive correlations between the various seziure levels and injurys indicating a connection betwen the two variables.

Bar plot for mechanism of injury

The bar plots show the count of different injuries sustained by TBI patients. By ordering the variables in orer of frequency we can see that falls and motor vehicle accidents are the most frequent causes for TBI.

Paired correlation

Each pairwise scatterplot and correlation coefficient offers insights into the relationship between different GCS scores and the scores at 6 months. For example, it is likely to show that the field.gcs, er.gcs, icu.gcs, and worst.gcs scores are strongly correlated with each other, indicating consistency in GCS measurements across different stages. Additionally, the 6m.gose score’s relationship with the GCS scores can reveal how early assessments of coma severity might predict long-term outcomes.