It's basically "Average VS Median", they are 2 different measures.
Quick example with 10 people:
1 death with 0 comorbidity, 2 with 1, 3 with 4, 3 with 6, 1 with 7
The average would be 3.9 comorbidities, all death included, and 90% of those deaths had at least 1. However, 70% had at least 4 comorbidities.
While the average is showing the general trend, this new stat (similarly to a median) is showing how skewed (or not) it is.
The average being significantly different from the median could also show the presence of outliers. You would need to look at the curve to see if it's skewed.
Average =/= At least [X]
It's basically "Average VS Median", they are 2 different measures.
Quick example with 10 people:
1 death with 0 comorbidity, 2 with 1, 3 with 4, 3 with 6, 1 with 7 The average would be 3.9 comorbidities, all death included, and 90% of those deaths had at least 1. However, 70% had at least 4 comorbidities.
While the average is showing the general trend, this new stat (similarly to a median) is showing how skewed (or not) it is.
Thank you for that breakdown.
The average being significantly different from the median could also show the presence of outliers. You would need to look at the curve to see if it's skewed.