Death rate calculations

 

Crude rates

The crude(unadjusted) rate is a quick look at the magnitude of mortality in the population. A crude or unadjusted death rate is simply the number of deaths divided by the population at risk, often multiplied by some constant. A population total death rate (crude rate or unadjusted rate) for the population is the crude death rate, calculated as stated earlier :

 

Total number of deaths for all causes
--------------------------- *1000

Total population

 

Cause-Specific death rate

When looking at death rates for "all causes of death"(the sum of all incidents of death), the common multiplier is 1,000. For death rates of a specific cause of death, for example, breast cancer, the multiplier is 100,000.

 

Total number of deaths for particular cause(e.g. cancer)

--------------------------------------------- * 100,000

Total population

 

Calculating Age-specific death rates

An age-specific or gender-specific death rate is simply the crude death rate for a specific age or gender, respectively. We use these specific rates to see if planning of health services needs to be designed for these groups. The age-specific death rate would be the death rate for the population of that age group. The numerator is the number of deaths in that age group. The denominator is the population of that age group:

 

Table 1. Age specific deaths rates for all causes of death

Hertford County, North Carolina Residents,1991-1995 Combined

Age group	Number of 1991-95 Deaths	Sum of 1991-95 population	Age-specific proportion who died	Death rates per 1,000 population
0-4	30	8150	.00368	3.68
5-14	2	17109	.00012	0.12
15-24	24	16601	.00145	1.45
25-34	34	14872	.00229	2.29
35-44	59	16199	.00364	3.64
45-54	85	12381	.00687	6.87
55-64	147	10277	.01430	14.30
65-74	305	9370	.03255	32.55
75-84	406	5631	.07210	72.10
85+	244	1829	.13341	133.41
TOTAL	1336	112419		11.9

 

 

 

Number of deaths in age group 1336

------------------------------------- *1000 --------- * 1000 =11.9 deaths per 1000

Number of population in age group 112,419

 

Age-Adjusted rates

The reason to adjust population rates is for doing comparisons of death rates in two or more populations or looking at a population at different time periods. Some populations, like those in Florida will have a higher proportion of older folks and a high death rate due to this age group. Thus it is important to use a standard population; this is called "standardizing for age" when comparing death rates. The age-adjusted rate, like the unadjusted(crude) rate describes the mortality in the whole population. To calculate the adjusted rate, the most common method is the "direct method" which is summarized below:

 

Table 2.

Age Adjustment by the direct method

Hertford County, North Carolina Residents, 1991-1995, All Causes

Age group	Death rates per 1,000 population from Hertford county, NC 1991-1995	Age-specific proportion who died	Standard population	Expected deaths
			(US 2000)
0-4	3.68	0.00368	19175798	70567
5-14	0.12	0.00012	41077577	4929
15-24	1.45	0.00145	39183891	56817
25-34	2.29	0.00229	39891724	91352
35-44	3.64	0.00364	45148527	164341
45-54	6.87	0.00687	37677952	258848
55-64	14.3	0.0143	24274684	347128
65-74	32.55	0.03255	18390986	598627
75-84	72.1	0.0721	12361180	891241
85+	133.41	0.13341	4239587	565603
TOTAL			281421906	3049452

 

 

 

• Choose a standard population to which all "study" populations of comparison would be applied.
• Start with the age-specific death rates in your study population which is usually reported per 1000 population.
• Divide each age-specific death rate by 1000 to get Proportion who Died in an age category
• Multiply the age-specific proportion who died to the standard population to get the expected deaths. Repeat for all age categories.
• Sum all the deaths together for all the age groups.
• Calculate the Age-Adjusted Death Rate.

 

For the example above, the age-adjusted death rate would be:

 

(3049452/ 281421906) *1000 =10.8 deaths per 1000 people

 

 

Rationale for using the US 2000 population

 

Public health experts use rates to compare population trends across populations and time. The most important aspect of the information that a user could get from a website is the ability to compare statistics from other states to New Hampshire(NH) and to compare the trends of death data from NH for many years. The standard population can be any population so the decision is now, which one. The best one to use is the population used in national agencies and by most of the other state agencies.

 

Constella compared the data on the NCHS website and the data on the Georgia website. For the data listed on both these websites, the 2000 US standard population is used for the population standardization in death rates. Constella also reviewed the proceedings from the CDC workshop on Age Adjustment (http://www.cdc.gov/nchs/data/series/sr_04/sr04_030.pdf) and the recommendations of the use of the 2000 US standard population. In looking at the types of comparisons that a user may make, whether to other states or to data within the state of NH, Constella makes the recommendation that the US 2000 standard population is used in NH's website.