Excess Deaths That May Be Related to COVID-19

Observed vs Expected Deaths over Time



Data visualization by Kevin Hartman
Source: National Center for Health Statistics
Estimates of excess deaths can provide insights about mortalities potentially related to COVID-19.

Excess deaths are defined as the difference between the observed deaths and the expected deaths per week based on historic norms. Counts include deaths from all causes, including deaths due to COVID-19. Because cause of death can often include co-morbidities, a view of all causes can provide insights into whether the deaths from COVID-19 are being overcounted or undercounted.



Problem Abstraction:

Use Tamara Munzner's framework to Abstract your problem (what, why, how)

Problem Abstraction
What Dataset type: Table
Availability: Dynamic
Attribute types: Categorical (States, Causes), Quantitative (Observed Deaths, COVID-19 Deaths, All Other Deaths, Expected Deaths), Temporal (Year, Week)
Source: National Center for Health Statistics
Why Who: Policy-makers, General Public
Action-Targets: {Discover: Trends}, {Compare: Similarities}, {Identify: Outliers}
How Mark: Area (Observed, COVID-19 Deaths, All Other Deaths), Lines (Expected Deaths), Points (COVID-19 Deaths, All Other Deaths, Expected Deaths)
Channels: Area->Position->Horizontal (Temporal), Area->Position->Vertical (Observed Deaths, COVID-19 Deaths, All Other Deaths), Line->Position->Both (Expected Deaths), Color Hue (COVID-19 Deaths, All Other Deaths)
Insight 1 An excess number of deaths occurred in the 2017-2018 Winter across most states.

For an example of this insight, zoom in on the main chart and pan over the various segments of time. Follow the dotted line as you move from right to left until you reach December of 2017.
Insight 2 The 2019-2020 Winter season saw an average number of deaths compared to historic norms.

For an example of this insight, zoom in on the main chart and pan over the various segments of time. Follow the dotted line as you move from left to the right side of the chart until you reach December of 2019.
Insight 3 An excess number of deaths occured in the Spring of 2020. The increase was more pronounced in some states versus others.

For an example of this insight, select New York as one of the states in the bottom drop down lists. Select Wisconsin for another. The vertical axis for the bottom 3 charts will adjust in scale to account for the largest state. This enables one to see the relative effect size across states.
Insight 4 States accrue excess deaths at different rates and stages.

For an example of this insight, select two or more states from the drop down selections. Mouse over the US overview chart at the top. A line mark with three figures for death counts will follow the mouse as it is moved across the timeline. The trajectories can be assessed visually by observing the three numeric counts in this marker as the line is moved. The line tracks to the same week for each state in each of the charts.
Insight 5 Excess deaths are higher than historic norms. Much of this can be attributed to COVID-19 as the cause of death. However, a large area above the expected death line remains shaded in beige where red is more probablistic. This effect is more pronounced in certain states than others. The effect may be attributed to varying standards by state for death record collection and classification.

For an example of this insight, adjust the zoom on the brush to bring 2020 into focus. Select Texas on any of the charts. Select California for another chart in comparison.