Direct Standardization

Direct standardization is a method of comparing population disease or death experiences, removing the effects of differences in population characteristics. Direct standardization is an important tool used in comparing the health of populations. Additionally, it can be used to monitor the health of the nation as is done with the Healthy People 2010 objectives. This entry uses an example of deaths from heart disease in the state of New York to illustrate the use of direct standardization.

Do men really have more heart disease deaths than women in the state of New York? In this case, two populations are compared, New York men and New York women. The heart disease death rate for New York men is 144.1 per 100,000 in 2000, and for New York women it is 160.5 per 100,000. Thus, the rate of death due to heart disease for women is higher than that for men. This finding is surprising but true, and is due largely to the fact that women tend to live longer than men, and thus live to an age where heart disease is a very common cause of death. In the terminology of epidemiology, the comparison of heart disease death rates is confounded by age. The death rates for men and women are really composed of two components: different heart disease death rates for each age group and different age distributions.

In this study, age is a confounder, a nuisance variable that affects our ability to clearly understand how gender affects the death rate for heart disease. Both the age distributions of men and women in the United States and the fact that heart disease deaths are more common as people get older are well known, but we need an analytical technique that incorporates these facts into our comparison of heart disease death rates by gender. Direct standardization provides that technique and allows us to calculate age-specific death rates, which compare heart disease death experiences by gender while controlling for age. Table 1 shows that for each age group, men have a higher heart disease mortality rate compared with women. The reason the overall, or crude, mortality rate is higher for women compared with men is because of the older age distribution for women in New York.

Both the crude mortality rates and the standardized mortality rates can be viewed as weighted averages of the age-specific mortality rates. In the crude mortality rate, the weighting reflects the number of men and women in each age category in New York State, while in the standardized rate, a reference or standard population is chosen. In Table 1, the crude rate for men is rewritten as weighted average.

Table 1 Crude Rates
			Men		Women
	Number of Cases	Number Population	Age Specific Rate∗∗	Number of Cases	Number Population	Age Specific Rate∗∗
< 45 years	774	15,304,246	5.1	336	15,959,393	2.1
45–54	1,684	1,226,925	137.3	613	1,326,011	46.2
55–64	2,977	786,187	378.7	1,492	901,800	165.4
65–74	5,540	561,262	987.1	3,772	714,784	527.7
Total	26,357	18,293,496	144.1	31,562	19,659,418	160.5
						Note: ∗∗Per 100,000 in 2000.
						Source: New York State Vital Statistics (2000).