Re: Diabetic women at a higher risk of dying

glass318 wrote:
<snip>

Death rate for a population is not always 100%.
Unless you count those who rose from the dead, or possibly vampires it is
Its a rate that is a
ratio of total deaths in the specified population over time. Usually,
it is the number of deaths per 1000 (population) per year.
OFTEN, not usually - if so, SAY so.
The
percentage increase is determined by dividing the rate difference by
the initial rate and multiplying by 100. For example, if the death
rate goes up from 12 per 1000 per year to 18 per 1000 per year, there
is a 50% increase (difference of 6, divided by 12, then multiply
result by 100).


Worthless - an example of why they ought to teach math to physicians

The ONLY meaningful death rate is q[x], usually expressed as 1000q[x] (rate per thousand) it is D[x] (number dying during the year at a SPECIFIC AGE!) divided by L[x] (population living at the beginning of the year AT A SPECIFIC AGE)

you can not compare total populations - it is worthless, you have to compare them at specific points, Ideally you can plot them for all values of x, and then compare the areas under the resultant curves


The standard measure of overall mortality is Life Expectancy, This is also age specific, but is usually shown at x=0 or,in old tales, by convention, x=0.5 This is a little better to use, but still requires some work.


i
.

Relevant Pages