I think the issue here is that the Kenya 1998 survey collected the maternal and child health indicators for births in the three years preceding the survey only and not for the standard 5 years.
The STATcompiler gives the option to show the data for this particular indicator both for children born in the three years preceding the survey and the 5 years preceding the survey.
For the three years preceding the survey it should show for this (and the other Kenya surveys), but for the five years preceding the survey there is no data because not collected in 1998.