The under-five mortality rates are compound rates that can only be calculated for aggregates. To proceed with individual-level date, you might try pooling the BR files from successive rounds and working with the child's date of birth (b3), survived / died (b5), age at death (b7) and covariates. You could try a survival / hazard model.]]>

This has really took my time and your help is highly appreciated.

Kind regards,

Amanuel]]>

With the approach you describe, you cannot distinguish between the possible effects of the policy and other influences. However, if you can localize the intervention geographically, you can then compare trends in intervention areas and other (control) areas. Here is a link to a report that did this sort of thing in Uganda: https://dhsprogram.com/pubs/pdf/WP142/WP142.pdf to other examples. Hope you can apply this strategy.]]>

Amanuel

]]>

Information about antenatal care is only collected for the youngest child under 5. You can see this if you enter "tab m14 bidx". Then enter "tab m14 neonatal_death if bidx==1". The children in this table are the only ones for whom you have information about the relationship between antenatal visits and neonatal mortality / survival.

If you calculate the mean number of visits with "summarize m14 if bidx==1 & neonatal_death==0 & m14<=20" and "summarize m14 if bidx==1 & neonatal_death==1 & m14<=20" you will find a slightly greater mean number of visits for the children who died--although the difference is probably not significant.

This is a difficult topic to analyze because women who have problematic pregnancies are more likely to be referred for antenatal care AND are more likely to have a neonatal death. For this reason it's difficult to show that better care results in fewer deaths.

]]>

replace neonatal_death=1 if b7==0

tab neonata_death[iweight=v005/1000000]

....... then....

tab m14 neonatal_death,

Information on 4+ ANC was available in only 66 of the 167 diseased neonates, of whom only 1 has no ANC visits, 2 have one-time ANC visits, none have 2 ANC visits, and 3 have 3 ANC visits. Therefore, the overall <4 ANC visit among the diseased neonates was 6(9%).

My questions:

1. Am I using the right data set to address my objectives? If not which data set should I use?

2. Do I need to merge two or more data sets? If yes, which one should I merge?

3. Could you please share with me the STATA commands on how to filter the above variables of any data set sources?

Looking forward to your reply.

My kind regards

]]>

Just a silly question, perhaps.

Am I able to estimate the effect of X policy on infant mortality rate or on under-five mortality rate relying on pooled KR files across surveys (4 rounds)?

2 rounds are before the policy and 2 are after the policy, so I just want to look at if that reduces such outcomes. infant or under-five mortality rate is just a single number in each survey and I am not sure if that is possible to conduct differences in difference estimation "reg rate Xolicy [iw=weight], robust.

I would be grateful if anyone of reply!

regards

]]>

On your second question this depends on what the variables are that you are interested in. If they are variables that already exist in the individual recode (IR) file then you just need to carry them through in your program, by including those variables on the /keep parameter of the varstocases command. You shouldn't need to merge files if the variables of interest are already in the IR file.]]>

*Set up the period length and related variables.

is it the same has this?

* Define the period to use, the upper limit (months preceding the interview), and the lower limit.

define periodlen ().

* To use 5 years durations.

compute period = 5.

compute kmax = 1.

compute kmin = 12*period.

compute totexp = kmin-kmax+1.

!enddefine.

And secondly, my research is on some of the factors affecting maternal mortality in Nigeria 2013 and 2018. please how will merge the data?

thanks!

]]>

No, DHS surveys do not produce data on mortality that is sufficiently complete across ages to allow the calculation of life expectancy. The infant death rate, child death rate, and under-5 death rate, when divided by 1000, are equivalent to 1q0, 4q1, and 5q0, respectively. The chapter on adult and maternal mortality (when included) usually gives an estimate of 30q15 or 45q15. With these numbers you could work with model life tables, but the reference time intervals for child mortality and adult mortality are different, so even the use of model life tables would be difficult to justify.

]]>