| Mortality rates and binary variable [message #19823] |
Tue, 18 August 2020 10:39  |
Chery87
Messages: 6
Registered: November 2019
|
Member |
|
|
I have been working on U5M by assigning value 1 for children who are alive and 0 otherwise. The stata commands I am using are as follows.
I am using Ethiopia DHS of 2016: ETKR71DT
gen wt= v005/10^6
svyset psu[pweight=wt], strata(strata)
gen ch_alive=b5
label var ch_alive "child is alive"
gen ch_died=1-b5
tab ch_alive ch_died
gen ch_age=hw1
replace ch_age=b7 if hw1==.
gen ch_59=0
recode ch_59 0=1 if ch_age<60
replace ch_59= ch_age if ch_age==.
svy: regress ch_died (this gives the mean of the variable, which I am considering it as the rate)
However, the coefficient is not even comparable to the u5mr calculated in the DHS report. It does work for the malnutrition indicators but not for my estimation of u5mr.
Any technical support to deal with this, the soonest reply the best will be.
Thank you very much in advance.
Chery
|
|
|
|
|
|
|
|
| Re: Mortality rates and binary variable [message #19855 is a reply to message #19830] |
Thu, 20 August 2020 13:56  |
Bridgette-DHS
Messages: 3214
Registered: February 2013
|
Senior Member |
|
|
Following is a response from DHS Research & Data Analysis Director, Tom Pullum:
U5M and the other child mortality rates are calculated in a very complex way. The code is available on our GitHub site (see the new landing page for the DHS forum). What you are doing would only be a rough estimate, perhaps for individual-level analysis.
You are doing a linear probability model. You should use logit regression, especially for probabilities outside the range .3 to .7. If you just replace "regress" with "logit" you will get an intercept, b0, which is the log odds that your outcome is 1. U5MR would be analogous to 1000*[exp(b0)/(1+exp(b0))]. However, this estimate is affected by censoring. Children born in the past 5 years have not had full exposure to the risk of dying before age 5. There are various ways to deal with this. The coarsest approach would be to reduce the sample to children born more than 5 years ago (i.e. children with v008-b3>60), even if their death occurred in the past 5 years.
You could look at publications on the topic and at methods such as hazard models. You only need to go down this path if you want to do multivariate analysis of under-five mortality.
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
