The calculation of the under-five rates is vastly more complex. DHS uses a synthetic (or life table) approach. Deaths and risk are assessed within an interval of time, such as the five years before the survey, and within eight sub-intervals of age. The method is described in the Guide to DHS Statistics https:// www.dhsprogram.com/pubs/pdf/DHSG1/Guide_to_DHS_Statistics_29 Oct2012_DHSG1.pdf.

It happens that within the next couple of weeks I will be putting a Stata program onto the DHS website that calculates these rates. I'm afraid you will have to wait until then.

The table you are referring to (Table 8.2 on page 107 of the report on the Kenya 2008-09 survey) is correct. I just re-calculated the numbers. However, the following observations may be relevant.

First, did you notice in the title of the table that it refers to the TEN years before the survey?

Second, the table itself does not include an estimate for the total (that is, all Kenya) for the ten years before the survey. I calculate that number to be 84 (83.72). That number is not mentioned anywhere, in the text or a table or a figure.

Third, I'm sorry to see that Figure 8.2, beneath Table 8.2, has an error in it. For all Kenya it gives an estimate of 74. That's obviously wrong because it is not in between the urban and rural estimates of 74 and 86, respectively. The value of 74 is for the FIVE years before the survey, given in table 8.1. The correct value, should be 84 (as I said). I don't know whether you are trying to match the 74 for the five years before the survey or the 84 for the ten years before the survey.

Fourth, and this will affect your calculations, the DHS estimate for the five years before the survey is not based just on children born in the past five years. It is based on any exposure to any of the eight component age intervals in the past five years. Similarly for the past ten years.

I can't tell what it is that you want to do. Do you want under-five rates for some covariates that are not in table 8.2? Do you want to repeat table 8.2 for the past five years, rather than the past ten years? Or something else? I would like to be able to help but I don't really know what you are trying to do.

DHS staff can help on the forum with specific questions about DHS datasets, indicators, tables in reports, and that sort of thing. We cannot help you set up your analysis. Maybe someone else on the forum will be willing to help.