WPC 2 ZB 0 X&n6X@&@Ы&n6X@&@; 9Z .Courier New Regular&6X@@; 9Z .Courier New Regular&n6X@&@; 9Z .Courier New Regular&Xx6X@X@; 9Z .Courier New RegularX&n6X@&@; 9Z .Courier New Regular&d6X@@; 9Z .Courier New Regular6X@@; 9Z .Courier New RegularXx6X@X@; 9Z .Courier New RegularX&n6X@&@; 9Z .Courier New Regular&d6X@ @; 9Z .Courier New Regular6X@ @; 9Z .Courier New RegularXx6X@ X@; 9Z .Courier New RegularX&n6X@ &@; 9Z .Courier New Regular&Xx6X@ X@; 9Z .Courier New RegularX&n6X@&@; 9Z .Courier New Regular&6X@@; 9Z .Courier New Regular&n6X@&@; 9Z .Courier New Regular&d6X@@; 9Z .Courier New Regular6X@@; 9Z .Courier New RegularXx6X@X@; 9Z .Courier New RegularX&n6X@&@; 9Z .Courier New Regular&d6X@@; 9Z .Courier New Regular6X@@; 9Z .Courier New RegularXx6X@X@; 9Z .Courier New RegularX&n6X@&@; 9Z .Courier New Regular&Xx6X@X@; 9Z .Courier New RegularX&n6X@&@; 9Z .Courier New Regular&6X@@; 9Z .Courier New RegularXx6X@X@; 9Z .Courier New RegularX&n6X@&@; 9Z .Courier New Regular&Xx6X@X@; 9Z .Courier New RegularX&n6X@&@; 9Z .Courier New Regular&d6X@ @; 9Z .Courier New Regular23|x@Page @    #6X@@#1994 National Health Interview Survey #&n6X@&@# ,"#Xx6X@X@#State Data Files #&n6X@&@# #d6X@@#` ` ` Introduction.........................................xxx 2 ` ` ` NCHS contact person.................................xxx 3 ` ` ` NCHS Data Use Agreement..............................xxx 4 ` ` ` Description of files included on this disk...........xxx 5 ` ` ` Estimates............................................xxx 8 ` ` ` Weighting............................................xxx 9 ` ` ` Standard errors......................................xxx 10 ` ` ` Usage notes..........................................xxx 11 l  #6X@@#1994 NHIS State Data Files #Xx6X@X@#у #&n6X@&@#  Introduction  The National Health Interview Survey (NHIS) State Data Files have been constructed to allow for the production of direct state estimates (note: see Estimates and Weighting sections) or to use state as an ecologic variable by merging state level data such as reimbursement policies or laws onto the files. These files differ from the NHIS public use data in that statistical noise at both the variable level and record level have been added to allow for the protection of respondent confidentiality and, at the same time, allow for release of the files with state identifiers. In addition to statistical noise, a number of variables have been removed and continuous and pseudocontinuous variables have been recoded into categorical variables. There are a number of states for which the sample size is too small to allow for reasonable estimates. These states have been combined into two groups: a.small states in the western part of the United States have been given the numeric state code (FSTATE) of '91' and the alpha state code (ASTATE) of 'WE'. Included states are AK, ID, NM, NV, and WY.; b.small states in the remainder of the country have been given the numeric state code of '92' and the alpha state code of 'OT'. Included states are NH, SD, and WV. Additionally note that the 19851994 NHIS sample did not have samples in the states of ND or NE. We realize that this file structure may not be suitable for all analytic purposes. Should these files not provide sufficient information for your work, please contact the NCHS contact person to explore other approaches. The NHIS state data files are in SAS version 6.06 format. A format library is provided to give the code structure for all variables. Instructions for the use of these files is in the Usage notes section. This document is not intended to provide complete file and survey documentation. Complete survey documentation for the 1994 NHIS Public Use data files may be found on the CDROM (NCHS CDROM Series 10 No. 6, Issued January, 1995). #d6X@ @#l  #6X@ @#1994 NHIS State Data Files #Xx6X@ X@#у #&n6X@ &@#  NCHS Contact Person  John Horm National Center for Health Statistics 6525 Belcrest Road, Room 850 Hyattsville, Md. 20782 V (301) 4367085 X123 F (301) 4363484 Email JDH3@CDC.GOV #Xx6X@ X@# 1994 NHIS State Data Files ă #&n6X@&@#  NCHS Data Use Agreement  The user is subject to the NCHS data use agreement reproduced below: The Public Health Service Act (Section 308) (d) provides that the data collected by the National Center for Health Statistics (NCHS), Centers for Disease Control and Prevention (CDC), may be used only for the purpose of health statistical reporting and analysis. Any effort to determine the identity of any reported case is prohibited by this law. NCHS does all it can to assure that the identity of data subjects cannot be disclosed. All direct identifiers, as well as any characteristics that might lead to identification, are omitted from the dataset. Any intentional identification or disclosure of a person or establishment violates the assurances of confidentiality given to the providers of the information. Therefore, users will: 1.Use the data in this dataset for statistical reporting and analysis only. 2.Make no use of the identity of any person or establishment discovered inadvertently and advise the Director, NCHS, of any such discovery. 3.Not link this dataset with individually identifiable data from other NCHS or non-NCHS datasets. l #6X@@# 1994 NHIS State Data Files #&n6X@&@#у  Description of files included on this disk:  README.W51` ` ` This file in Wordperfect version 5.1 format. A second version README.TXT is in text format for those without the ability to read Wordperfect files. Programs and documentation files (contained in the subdirectory ` ` ` hhh  \PROGRAMS on the CDROM): FORMAT.SAS` ` ` A SAS program to construct the format library used by the data files. This library contains the code structure for all variables. Instructions for constructing and using the format library are in the section of this document entitled Usage Notes. PRNTFORM.SAS A SAS program that produces a formatted listing of all variable formats. CONTENTS.SAS A SAS program that uses PROC CONTENTS on each data file. This gives the number of records, number of variables, associated variable labels and formats, variable names, and variable type for each variable on each data file. EXAMPLE1.SAS A SAS program that reads all ten data files, and does simple frequency distributions. This may be used as an example of how to access the data files and use the sampling weight variables. #d6X@@#l  #6X@@#1994 NHIS State Data Files #Xx6X@X@#у #&n6X@&@# Programs and documentation files (contained in the subdirectory ` ` ` hhh  \PROGRAMS on the CDROM): *.REC` ` ` There is one .REC file for each of the 11 data files for 1994. These files contain the recodes that were applied to the original data and are used in conjunction with the variable formats and the original variable formats (on the NHIS Public Use data CDROM) to fully decipher the code structure on these files. State data files (contained in the subdirectory \DATA on the CDROM): ` ` ` Number of File name records Description  PERSON.SD2 107,983hhhThe person level data file consisting of one record for each person in the 1994 NHIS. This corresponds to the Person Record from the NHIS public use data files. CONDITON.SD2 85,772hhhThe condition level data file consisting of one record for each chronic or acute condition reported. This corresponds to the Condition Record from the NHIS public use data files. DRVISIT.SD2 26,430hhhThe doctor visit data files consisting of one record for each doctor visit within the two week recall period. This corresponds to the Doctor Visit Record from the NHIS public use data files. HOSPITAL.SD2 12,131hhhThe hospitalization data file consisting of one record for each reported hospitalization. This corresponds to the Hospital Record from the NHIS public use data files. AIDSKNOW.SD2 19,766hhhAIDS knowledge and attitudes. This was administered to a randomly selected adult in each household. #d6X@@#l  #6X@@#1994 NHIS State Data Files #Xx6X@X@#у #&n6X@&@# ` ` ` Number of File name records Description  HLTHINSU.SD2 107,983hhhThis supplement collected detailed information on private and public health insurance plans for each family in the NHIS. IMMUNIZA.SD2 7,831hhhInformation on immunization history for children under the age of 6 years. Immunizations included are DTP, Polio, Measles or MMR, Hemophilus type B or HIB, and Hepatitus B. YEAR2000.SD2 20,386hhhBehaviors and knowledge of sample adults (ages 18+) in areas included in the year 2000 health goals such as environmental health, tobacco use, occupational health, clinical and preventive services and others. ACCESS94.SD2  107,983hhhSources of medical/health care, use of facilities and types of facilities. DPERS94.SD2  102,096hhhPerson records from Phase 1 of the Disability Survey. DCOND94.SD2  40,743hhhCondition records from Phase 1 of the Disability Survey. #Xx6X@X@# 1994 NHIS State Data Files ă #&n6X@&@#  Estimates  Estimates from this file may be slightly different than similar estimates from the NHIS public use data files because of the addition of statistical noise. National comparisons for state estimates should be constructed from this file as well as the NHIS public use files. This would give an approximation of the extent of any bias at the state level introduced by the methodology used for creating the data files. The sample design of the 19851994 National Health Interview Survey technically does not allow for producing state specific estimates. This shortcoming may be somewhat overcome by reweighting (poststratifying) the files controlling to state population estimates. See the next paragraph and the section on weighting. The user should assess the adequacy of these files for producing direct state estimates prior to analysis and interpretation. This assessment should include, at a minimum, sample size, and concordance of the age, race, sex distributions with those for the state(s) of interest. Estimates for several variables have been constructed from both these files and internal NCHS versions without the statistical noise. These comparisons are given in the attached table. These estimates were made to examine possible biases due to the addition of statistical noise. Estimates from the NHIS state files closely follow those from the internal NCHS files. A number of other comparisons were made but not included in the table since no deviations were found. Any suspicious findings should be reported to the NCHS contact person for resolution. l  #6X@@#1994 NHIS State Data Files #Xx6X@X@#у #&n6X@&@#  Weighting  All estimates made from NHIS data files must be weighted to yield unbiased estimates. The final basic annual weight included on these files was developed to yield national annual estimates. Adjustments for oversampling and nonresponse have been included and the annual basic weight has been poststratified to Census control totals by age, sex, and race (black, nonblack). However, this weight has not been adjusted to individual state population totals. Users of these files may want to ratio adjust the sampling weight to appropriate state population totals obtained locally. The procedure for ratio adjusting to state totals is reasonably straightforward:  ` ` ` Create a table of the sum of the weights by age, race, and sex. Denote each cell by W(i,j,k) where i is the indicator for age, j is the indicator for race, and k is the indicator for sex;  ` ` ` Create a similar table of local population controls. Denote each cell by P(i,j,k);  ` ` ` Form the ratio R(i,j,k) = P(i,j,k)/W(i,j,k);  ` ` ` Multiply each weight, at the record level, by the appropriate R(i,j,k). Include this adjusted weight in the analysis. This approach may be extrapolated to include adjustments for income, education, etc. if so desired.  Standard Errors  The statistical design of the NHIS is stratified, multiple stages with clustering and requires the use of sample design parameters such as the primary sampling units (PSU), strata, secondary sampling units (SSU), and substrata and specialized computer software for the estimation of standard errors. However, since these parameters pose confidentiality problems when combined with state identifiers, they have been removed from this file. Thus standard errors can only be estimated under the assumption of a simple random sample (SRS) such as is made in the common data analysis packages such as SAS, SPSS, BMDP, and others. Experience has indicated that standard errors estimated incorporating the sample design parameters are 2030 percent greater than those estimated assuming an SRS design. It is suggested that the user increase the standard errors by 30 percent for assessing statistical significance when using this file. The effects of a complex sample design such as that used for the NHIS are that designbased standard errors are always increased over those of an SRS of the same size. Thus nonsignificant findings assuming an SRS will also be nonsignificant for designbased estimates. #Xx6X@X@# 1994 NHIS State Data Files ă #&n6X@&@#  Usage Notes  These notes and example programs assume that the user will be working on a Windows (either 3.X or Windows 95) based PC in the directory C:\STATE94. It is further assumed that the user is using at least SAS Version 6.06 for Windows. The data files of interest and the program FORMAT.SAS must be copied into C:\STATE94. The EXAMPLE1.SAS program contains several examples of using the NHIS state files and may be copied into C:\STATE94 but is not necessary. The program FORMAT.SAS must be executed prior to using the NHIS statelevel data files. This creates the format library (FORMATS.SC2) with the variable value definitions required by data files. The format library will be placed into the directory 'C:\STATE94'. The programs 'PRNTFORM.SAS' and 'CONTENTS.SAS' are not necessary but should probably be run. These produce a formatted listing of the format library and execute SAS's PROC CONTENTS for each of the data files. Example program: This program is documented with comments and is meant only as an example of using these files. EXAMPLE1.SAS is configured to read all 11 data files from the directory C:\STATE94 and produce some simple crosstabulations while using the format library and the sampling weights for each file. This program is intended to get the user up and running with a minimum of floundering. Alternatively, the user may leave the data files *.SD2 on the CDROM and change the appropriate paths in the analytic programs. This will save space on the hard disk at the expense of some execution time. #d6X@ @#