nhamcs design on survey package (R)

Question

I am trying to analyze nhamcs http://www.nber.org/data/national-hospital-ambulatory-medical-care-survey.html using Lumley's survey package but getting the wrong standard errors when comparing them against http://www.cdc.gov/nchs/data/ahcd/nhamcs_emergency/2011_ed_web_tables.pdf

current design object is defined as:

data.design <- survey::svydesign(id = ~CPSUM, strata = ~CSTRATM, weight = ~PATWT, data = data, nest = TRUE)

but this gives me the wrong s.e. for example, for the total population number of visits I get:

    total      SE

[1,] 136296400 7370965

while the pdf above reports (in thousands for both totals and s.e.):

136,296 (6,413)

would appreciate any thoughts on where I might be going wrong