Lab #5: Direct and Indirect Standardization of Mortality Rates

We have been discussing mortality as it has changed over time in developed and developing regions. Even within the same country, mortality can change due to differences in environmental or cultural factors. Also, a population’s age and gender composition can cause variations in mortality rates. Geographers interested in mortality or disease often want to be able to compare rates in two locations or at two scales in order to identify whether there are spatial factors such as the local natural or built environment or cultural factors that influence a rate. In order to make such a comparison, one needs to adjust for differences in age composition. Since the highest age-specific mortality rates occur at the youngest and oldest age cohorts, populations with large child and elderly populations will have higher mortality rates. In order to eliminate the influence of age composition, one can compute a standardized death rate.

A standardized death rate is a crude death rate that has been adjusted for differences in age composition between the region under study and a standard population. Standardization allows for comparisons when the population structures differ and is key in assessing the potential influence of environmental or cultural factors on death rates in a region.

There are two ways of computing standardized death rates – direct and indirect standardization. The results will be a little bit different. The one you would use varies based on the data available to you.

Direct Standardization (SDR1) calculates a weighted average of the region’s age-specific mortality rates where the weights represent the age-specific sizes of the standard population.

Indirect Standardization (SDR2) uses age-specific mortality rates from the standard population to derive expected deaths in the region’s population.

Direct Standardization:

SDR1 = [Sum age groups (Mar Pas)]/Ps x 1000

Mar is the age-specific mortality rate for the region.
Pas is the number of people in the age group in the standard population.
Ps is the total standard population.

To compute the direct standardized mortality rate:

1. For each age group, you need to multiply Mar by Pas.

2. Add them up.

3. Divide the sum by the total standard population.

4. Multiply by 1000, or give the rate in terms of "per thousand population."

Indirect Standardization:

SDR2 = Dr/[Sum age groups (Mas Par)] x CDRs

Mas is the age-specific mortality rate for the standard population.
Par is the number of people in the age group in the region’s population.
Dr is the number of deaths in the region.
CDRs is the crude death rate for the standard population.

To compute the indirect standardized mortality rate:

1. For each age group, you need to multiply Mas by Par.

2. Add them up.

3. Divide the sum into the number of deaths in the region.

4. Multiply by the crude death rate.

Choosing which formula to use will depend on what data you have access to. More typically, the data you have will be the components for indirect standardization. You are more likely to be able to find age-specific mortality rates for a standard population than for a specific region such as a state. In this case, since I have given you state population data in thousands, you will need to divide your final result by 1,000.

For this lab, you will be computing standardized death rates of New York and Arizona using the following data sets and the US as the standard population.

The fastest way to compute the direct and indirect rates will be to enter the measures you need into Excel, remembering to compute formulas from the inside out. See if you can figure out how to enter this into Excel. (These tables are taken from D. Plane and P. Rogerson. 1994. The Geographic Analysis of Population with Applications to Planning and Business. New York: John Wiley & Sons, Inc.)  Be careful!  Think about what data you need and whether you need all the data I have given you!  You will need to type these into Excel.  These tables will not directly import into Excel.

Table 1. US Population in thousands and State Age-Specific Mortality Rates  
Age Population New York Arizona        
0-4 18,456 0.00272 0.00269        
5-24 72,053 0.00066 0.00076        
25-34 43,675 0.00192 0.00146        
35-44 35,264 0.00321 0.00224        
45-54 24,163 0.0054 0.00484        
55-64 21,830 0.01212 0.01102        
65-74 17,897 0.02793 0.02408        
75+ 12,470 0.08753 0.07359        

Source: U.S. Bureau of the Census, 1990; U.S. Dept. of Health and Human Services, 1989.

Table 2. U.S. Age-Specific Mortality and State Populations in Thousands 
Age US Death  New York  Arizona      
  Rate Population Population      
0-4 0.00251 1,275 299      
5-24 0.00065 5,023 1,028      
25-34 0.00135 3,018 614      
35-44 0.0022 2,572 479      
45-54 0.00486 1,909 315      
55-64 0.01236 1,784 307      
65-74 0.0273 1,347 277      
75+ 0.08513 981 170      

Source: US Bureau of the Census (1989). US Dept. of Health and Human Services (1990).

Total deaths in New York: 176,289
Total deaths in Arizona: 27,646
US Crude Death Rate: 8.818

 

In this lab, you will be selecting the right data from these tables in order to compute a direct and an indirect standardization of the death rate of these two states.

SDR1 for New York =

SDR2 for New York =

SDR1 for Arizona =

SDR2 for Arizona =