Another type of death rates frequently computed are the so-called specific rates. These are the rates for specific or limited groups of the population, such as those of age, sex, social condition, occupation, etc. They are stated as the proportion of the number of deaths per annum in the sub-group per i000 of the mean annual number of the population in that sub group. The groupings of age and occupation are most com monly employed (Fig. 131). The expression of births according to women of child bearing age may be considered a specific birth rate. They are the computations which should be em ployed for purposes of comparison or to secure a true insight into the prevalence or incidence of different diseases. Their computation can be readily understood from the following example. Divide the number of deaths of persons whose ages lie within the age groups employed, by the number of thou sands of persons alive in the same groups at the mid-year. Thus : A still better method to employ for comparative purposes is the standardized or corrected death rate. This is based upon a standard or stable population, not affected appreciably by migratory factors. The Swedish census of 1890 which had 5 age groups, or the standard million of England and Wales of 1901 with eleven age groups further subdivided according to sex, are most commonly used as a base. The standard Swed ish population is divided per too() as follows: Under I 25.5
1-19 398.o2o-39 269.6 192.3 6o and over 114.6 The method is employed as follows: Compute the specific death rates of the several age groups of the population whose standardized death rate it is desired to obtain (Table XII). Then take the corresponding age groups in the standard popula tion and compute the number of deaths that would have occurred in each age group at the specific death rate found to exist in the population for which the standard death rate is being computed. Add the number of deaths which it is thus found would have occurred in the age groups of the standard population. This gives the standardized rate per i,000,000. The rate per i,00c is secured by dividing this by i,000.
We may exemplify this by continuing the employment of the Boston data already given.
This is seen to be considerably higher than the crude death rate of 16.17 of our previous computation. Their value for comparative purposes is further shown in the following table: With the employment of only the crude rate erroneous im pressions might be drawn.
Death rates are influenced by the following factors: (a) Statistical methods employed in their computation. (b) The age, sex and social composition of the population. (c) The number and size of ,hospitals and institutions for the care of the sick and aged within the area. (d) Migration. (e) Hygienic and sanitary conditions. (f) The birth-rate.