This CD-ROM contains information for users of the National Hospital
Discharge Survey Multi-year Data File for 1979-1997.  Included here
are the data file in ASCII format, the data tape documentation, and 
LOTUS files to accompany Appendices D and E of the file documentation.  

Appendix D contains detailed instructions on how to calculate standard 
errors for NHDS estimates.  To accomplish this, tables of approximate 
relative standard errors for 1979 through 1997 are presented in LOTUS 
files with the designation RSExx.WK1, where xx can be replaced by the 
last two-digits of a year between 1979 and 1997 (e.g. 79).
  
Appendix E provides U. S. Census population estimates, which are
used to calculate hospital utilization rates for NHDS estimates. 
The files to accompany Appendix E are named either 1YRxx.WK1 or
or REGxx.WK1 and contain U.S. Census Bureau estimates of the 
civilian resident population, as of July 1, for the years 1979 through 1997.
Again, the xx can be replaced by the last two-digits of a year between
1979 and 1997.   

NOTE: The population estimates provided here for 1997 have been 
adjusted for net underenumeration using the 1990 National Population
Adjustment Matrix.  PRIOR TO THE 1997 DATA YEAR, CENSUS 
ESTIMATES OF THE CIVILIAN POPULATION PROVIDED WITH
THE NHDS PUBLIC USE DATA TAPE DOCUMENTATION WERE
NOT ADJUSTED FOR THE UNDERCOUNT.  HENCE, THE
POPULATION ESTIMATES PROVIDED HERE FOR 1996 AND
EARLIER ARE ALSO UNADJUSTED.

For the years 1980 through 1997, the files named 1YRxx.WK1 contain 
estimates disaggregated by sex, race, and single-year age groupings.  
The files named REGxx.WK1 contain estimates disaggregated by sex, 
region, and single-year age groupings.  These population estimates have been
adjusted based on the 1990 decennial census.  There is only one
spreadsheet for 1979 because data disaggregated by region are not
available for this year.  The 1979 data were adjusted based on
the 1980 Census.  Note that the figures in the 1979 spreadsheet
have been rounded to the nearest thousand, whereas those for all
the other years are unrounded.  

It should be noted that rates calculated with these estimates may
differ slightly from those in published NCHS reports or those
calculated from population estimates disseminated with the NHDS
annual data tape documentation.  

Before 1981, estimates of rates of discharges and days of care
which appeared in published reports from NCHS were calculated
using the civilian noninstitutional population.  However, beginning 
in 1981, the civilian resident population estimates were used.  
The civilian resident population was determined to be more appropriate 
because persons in institutions, for example nursing home patients, 
are hospitalized when necessary.  A report has been published which 
discusses differences in discharge rates based on the different denominators 
(see Reference 9 in documentation).