Saturday, August 29, 2015

Methods of Computing IMR

Posted by   on

Links to all courses
IMR = Infant Mortality Rate.
i) Conventional Infant Mortality Rate estimation:
In this method the rate is taken as numerator infant deaths irrespective of birth cohort from which they come.

Suppose IMR = 33.56 infant deaths per 1000 live births. The rate means that for 1000 babies recorded as having born alive in a year. About 34 were recorded as having died before reaching their first birthday.

ii) The exact matching of birth & death
For the use of this method knowledge is needed of the year of births of each infant who dies in the selected year & in the preceding year.
Additive Method:
Multiplicative method:
In this method the probability of an infant surviving to year 1 of age is the product of separate probabilities of its surviving to the end of the year in which it is born and having done so of its surviving to 1 year of age in the next year. 
Define, 
P1 = Probability of surviving of an infant to the end of year 1 who was born in year 1. 
P2 = Probability of surviving of an infant to age 1 who survived to the year 1. 
Then, P = P1 × P2
Here, 
Hence, IMR = 1- P

Factor Method:
Factor method consists of either separating the numerator or adjusting the denominator. 
Separative factor (f)
The Separative factor is the proportion of infant deaths in the current year in which the child involved was born in previous year.
Formula unclear? Click here to see bigger
Numerator separating factor: 
Let us assume that there were D1 (d2+d3) deaths in selected year and that there were b1 births in that year and bo births in the previous year. Then the adjusted IMR for the selected year –

Where f is the separation factor,
fD1 is the number of deaths in current year of infants born in previous year
(1-f)D1 is the number of deaths in the current year of infants born in the current year.
Denominator adjustment factor:
Where rb0 and (1-r)b1 is for each of two year the proportion of birth relating to D1. 
In practice the value of r & f are very close and therefore with very little error f can replace r in the method without much loss of accuracy and generality. 


No comments:
Write comments

Hey, we've a very popular astronomy portal in Bangla language. If You know Bangla, You'll like it - sky.bishwo.com
Join Our Newsletter