Drug overdose deaths plague the human condition within the United States, the following dashboard and data analysis is an attempt to understand this problem. Additionally, this blog will conduct an analysis of the trends discovered. The dashboard comes from a data set collected from the state of Connecticut for 2018. The following blog is an instructional guide on how to navigate the dashboard to analyze the data further.

The dataset was mined via Tableau and the Dashboard can be accessed at:

Drug OD Dashboard 

The dashboard consists of three visualizations: the top compares age and gender, the middle compares overdose deaths overtime, the final being drug types and gender. There are two filters one for sex and one for age. These filters are linked to the datasets and as you manipulate them all data and visualizations will change accordingly. Additionally, holding the cursor over any of the X&Y points in the visualizations will populate the death count.

In the first visualization, Accidental Drug Deaths by Gender and Age, the X-axis holds the ages of the deceased, and the Y-axis holds the counts.

For an example baseline, select both female and male boxes in the sex filter on the right side of the dashboard and below it select “All” in the Age (bin) filter. Take note of the visualizations that populate. In the first visualization the distributions are multi-modal, with two distinct modes; One mode of 25 and another mode with an average age of around 55. From this visualization, it appears that people with in their 40’s are not using hard drugs and overdosing; At the same time, they hit around 55 they're using much harder. Therefore, 55 year-olds and 25 year-olds are approximately equal in risk with 55 year-old being at a slightly higher overdose rate. Moreover, the male death count is almost four to one to the female.

The second visualization, Overdose Deaths Overtime, is a time series with year on the X-axis and counts on the Y-axis. Two trendlines (blue for male) start at the lower left and rise to the upper right.  

 The second visualization shows a steep rise in male overdose deaths at year 2014. However, females have slightly risen.

In the final visualization, Drug types and Gender, is a bar graph representing the number of drugs in the bodies of the deceased. The drug names are on the X axis and instance counts on the Y-axis. Bear in mind, that one overdose death may have multiple drugs associated.  

Interestingly, when female is the only gender selected, the difference in rate of overdose deaths between 25 year old woman in a 55 year old woman is significantly marked. A 50 year-old woman is at a much higher rate of overdose death then a 25 year-old. Now go to the Age(bin) filter on the right-hand side of the dashboard and slide it around to the different ages; Notice what happens to the visualizations. The visualizations change in relation to the data in the filter. Concurrently, the time series changes as well as the preferred drug of choice found in the bodies of the deceased. Undeniably, fentanyl is the leading drug of overdose choice and heroin being a close second. Third, is the high trend of Cocaine among those under 65. However, notice the interesting trends in the drug types as they relate to age. It would appear that 65 and up females prefer alcohol along with fentanyl in contrast to the broader array of drugs used by younger women. One could speculate that 65+ are being prescribed fentanyl due to the absence of other drugs in their system.

In final analysis, a recommended course of action would be to gather more data and monitor the trends. Granted this is only a sample of Connecticut’s overdose deaths and should not be taken as a generalization from the sample.