CONDITION FILE INCLUSIONS The Condition file contains the following types of records: 1. Condition records, either acute or chronic, which are associated with disability days, doctor visits and/or hospitalizations in the two weeks preceding the interview (location 107=1 or 2 and 111-112=01-14 or 109=1). 2. Chronic condition records when the condition is reported to be the main or secondary cause of activity limitation (location 107=1 and 126=1 or 2) or limitation in ability to work. 3. Chronic condition records reported in response to a check list of conditions for each body system, e.g., digestive. The total sample was divided into six subsamples. Each subsample was asked the set of conditions corresponding to one of the six chronic disease check lists. Use of Condition Records NHIS condition records can be used to calculate aggregate estimates and rates of the incidence of acute conditions (Recode A), prevalence of selected chronic conditions (Recode C), disability days and utilization associated with acute or chronic conditions, and chronic conditions causing limitation of activity. Explanations of the procedures used to calculate these figures and rates are given below as well as restrictions due to content and reliability factors. Included below is a list of special codes in ICD format and tape location created by NHIS as well as a list of ICD codes not used by NHIS. NHIS Created Codes The codes listed below are ICD-like codes with meanings specific to NHIS. These codes and the Impairment X-codes (see Medical Coding Manual for complete list of X-codes) are found in the ICD column locations of the Condition and Doctor Visit Records. The X of the Impairment Codes (represented by a hyphen '-') is in the first of the four columns. NHIS Code Category Title 019 Tuberculosis, arrested or inactive Psychic factors associated with diseases classified elsewhere in specified system: 316.0 Musculoskeletal 316.1 Respiratory 316.2 Cardiovascular 316.3 Skin 316.4 Gastrointestinal 316.5 Genitourinary 316.6 Endocrine 316.7 Organs of special sense 316.8 Other 316.9 Unspecified 399 Rheumatic fever, inactive 477.7 Allergies with multiple causes 693.2 Skin allergies with multiple causes 799.5 Observation with no condition found Codes Not Used by NHIS These ICD codes are not specifically used by NHIS. They are usually included in the categories specified: Unused ICD Coded To: 009.1 558 009.3 558 042-044 799.9 137-139 Impairment, as appropriate 230-234 Not applicable (not reported in NHIS) 235-238 Benign, malignant or unspecified, by site 268.1 Not applicable (not reported in NHIS) 306.0,6,7,8,9 Code as if non-psychogenic 315.0-2 X14 317-319 X19 326 Code residual effect to appropriate impairment or ICD 320-325 331.3,4 X93 321 X41 343 X50 344.0,2-6,8,9 Less than 3 months, see Medical Coding Manual, Appendix III If greater than 3 months, code as impairment, by site and extent 353.5 X70-X79, by site 369 X00-X03, by severity and site Unused ICD Coded To: 388.2 X05-X09, by severity and site 389 X05-X09, by severity and site 391 Usually 398.9 (see Short Index) 410-412 414 429.6 414 436 437 438 Appropriate impairment usually (see Medical Coding Manual, Appendix III) 520.0-2,5 X92 521.6 X92 524 X92 718.6 X85 728.2 X70-X79, by site 728.4 Orthopedic impairment, by site 728.6 X74 734 X77 735 X76 736 Impairment, by type, site and cause 737 X70 738 Impairment, by type, site and cause 740 Not applicable (not reported in NHIS) 741 X71.9 Unused ICD Coded To: 742.0,1,3 X93.9 743 Code to equivalent acquired diseases of the eye 744.0-3388.9 and/or appropriate impairment (see Appendix III) 744.5 X79.9 744.8,9 of face, coded to X90.9 of neck, coded to X79.9 749 X91.9 750.0,1 X11.9 753.0 X31.9 754.2 X70.9 754.3 X75.9 754.4 X76.9 754.5-7 X78.9 except flatfoot coded to X77 754.8 Appropriate impairment, by type and site 755 Appropriate impairment, by type and site 756.0-3 Appropriate impairment, by type and site 758.0 X19.9 759.7.9 X99.9 761.6 Not applicable 779.6 Not applicable (for fetus) 781.5 X74 Unused ICD Coded To: 783.1 278.0 798 Not applicable 871.3 X00-X03 885-887 Appropriate impairment, by site 895-897 Appropriate impairment, by site 905.6-8 X80-89, by site 905.9 X20-29, X35, by site 908.6 Code to effect 909.0,1,3,9 Code to effect 958 Code injury only 960-979 977.9 980.987 989.9 In addition, the NHIS does not use certain ICD codes, which are conditions known to be caused by other conditions. In the Ninth Revision of the ICD, some causal conditions are marked by a dagger and the resulting condition by an asterisk. NHIS codes the dagger and not the asterisk. Acute Conditions Annual incidence estimates can be calculated by the following (using a single year of data): 1. Use acute condition records (107:2) and categories from Recode A (location 100-103). 2. Select those with onset in last two weeks (108:1) 3. Use 6.5 weight (228-236) for all four quarters of data. Denominators for subdomain rates should be computed by summing the weights for the entire domain of interest. When associated disability days such as annual bed days are needed, the weighted number of bed days in location 246-254 should be substituted for the 6.5 weight in 3 above). For additional examples of calculating estimates associated with acute conditions, see the back of this document. Chronic Conditions Recode C is the list of selected chronic conditions for which NHIS calculates prevalence and associated disability and use data. We feel that the collection and coding procedures used do not justify any further breakdown of categories for purposes of estimating prevalence from our survey. These categories correspond to questions asked on the six chronic condition lists and therefore produce reliable estimates except where the numbers are small. In order to obtain NHIS prevalence estimates, use the Recode C categories following the Tape Layouts. For chronic conditions not specified in the six lists, estimates do not reflect prevalence but may be useful in epidemiological investigations. For most of the non-list conditions only those resulting in limitation of activity, recent disability days, or doctor visits will be included. Because chronic conditions-related estimates are made based on a one-sixth subsample, many estimates, especially for population subgroups, may not be statistically reliable. The weights for the six subsamples have been adjusted to the Census population controls using the same age-sex-race cells as were used for the entire sample. The most accurate chronic condition aggregate estimates for population variables other than age, sex and race are achieved by multiplying rates calculated for the one-sixth subsample by the population estimate based on the total sample. For example, to estimate the prevalence of hypertension in the population with income $25,000+, the rate calculated using the entire circulatory system subsample and those in the sample with hypertensive disease should be multiplied times the population estimate from the entire NHIS sample in the $25,000 income category. In order to calculate annual prevalence estimates (using a single year of data): 1. Use chronic condition records (107:1) 2. Select on the appropriate list number in tape Location 92 3. Select on specified codes in Location 104-106 (Recode C). 4. Multiply the chronic condition prevalence factor (Location 190-200 divided by 1,000,000,000 to adjust for the implied decimal) times the annual basic weight in location 219-227. Rates should be calculated using the population estimates from the subsample as the denominator. Population estimates from the total sample also may be used as denominators; however, if rates are to be calculated for groups other than those defined by age, sex, and race, it may be preferable to use denominators based on the subsample. In order to calculate associated disability days, doctor visits, etc., the specifications above should be used, substituting the appropriate weight for the basic weight factor in 4. For example, to calculate annual bed day estimates based on a two-week recall period, use the chronic condition factor multiplied by the weighted bed days in location 246-254. Limitation of Activity due to Chronic Condition To calculate limitation of activity due to chronic conditions on a condition basis, use the entire NHIS sample, rather than the one-sixth subsamples. NHIS recommends that Recode B (Location 97-99) be used to categorize the causal conditions. Use of Recodes Recode A - Recode categories for acute conditions. It contains the smallest categories considered reliable by NHIS for estimating acute conditions. Recode B - The recommended recode for limitation of activity. It corresponds directly to Ninth Revision ICD categories. Recode C - Recode categories for chronic conditions. It is used by NHIS for estimating prevalence of chronic conditions.