While there are certainly life-or-death situations that I can only imagine as high-stress for everyone involved, the bulk of patients injure themselves in minor ways while doing everyday things. You can see this in injury data collected by the Consumer Product Safety Commission. They sample US hospitals and cull product-related injuries each year, which, from the CPSC documentation, includes the following:
- All poisonings and chemical burns to children under 5 years of age.
- All injuries where a consumer product, sport, or recreational activity is associated with the reason for the visit or related to a condition treated.
- Illnesses only if a consumer product or activity is associated with the onset of the illness.
So this data doesn’t include stuff like chronic conditions. That said, the 2014 sample includes about 367,000 records that represent 13.8 million visits. For each record, a main product is specified.
This chart shows the top 250 products, month-by-month.
At the top, you see general products: floors , stairs , and beds . Maybe someone slips on the floor and breaks a finger. Maybe someone was jumping on the bed and fell.
Keep mousing down, and you get to more specific products and activities.
For example, look at seasonal sports such as snow skiing , volleyball , baseball , or football . Naturally, you see peaks in the winter months and dips in the summer months, or vice versa.
Similarly, you can look at weather-related products like heaters or air conditioners . Or how about event-related? Like fireworks in July.
All in all, even though it’s data for emergency rooms, it’s a detailed view into the everyday, which kind of fascinates me.
- Data is from the CPSC’s National Electronic Injury Surveillance System, which is a probabilistic sample for 2014. Records are weighted to estimate the total population.
- Circle area represents number of visits for a product in a given month. It’s a redundant visual encoding meant to reinforce the vertical axis and lend some discrete anchors for an otherwise spaghetti mess.
- The vertical axis for visits is a logarithmic scale. I cheated slightly for the zero counts and show them at 1.
- I analyzed and formatted data in R, my thinking language of choice, and I made the interactive visualization with d3.js. I finally had an excuse to use the multi-line voronoi.