Kid

Kids Impatient db

Last Updated Aug 20, 2008 01:57 PM

Geographic Region - Northeast, Midwest, West, and South
Control â?? public, private not-for-profit, and proprietary
Location â?? urban or rural
Teaching Status â?? teaching or non-teaching
Bed Size â?? small, medium, and large
Hospital Type â?? childrenâ??s or other hospital.
If there were fewer than two frame hospitals, 30 uncomplicated births, 30 complicated births, and 30 non-birth pediatric discharges sampled in a stratum, we merged that stratum with an "adjacent" stratum containing hospitals with similar characteristics. We created the discharge weights by stratum in proportion to the number of AHA newborns for newborns and in proportion to the total number of (non-newborn) AHA discharges for non-newborns.

Weight Data Elements
In addition to the regular discharge weight data element, DISCWT, the 2000 KID contains a new discharge weight data element named DISCWTCHARGE. Texas discharges were not included in the calculation of DISCWTCHARGE. This data element was set to zero for all Texas discharges because total charges were not available for the first half of the year from that state. Consequently, DISCWTCHARGE differs from DISCWT for hospitals in the South.

To produce national estimates, use DISCWT or DISCWTCHARGE to weight sampled discharges in the Core file to the discharges from all U.S. community, non-rehabilitation hospitals. For the 2000 KID, DISCWT should be used to create national estimates for all analyses except those that involve total charges, and DISCWTCHARGE should be used to create national estimates of total charges. In the 1997 KID, DISCWTCHARGE is not available, and DISCWT should be used to create all national estimates.

The 2000 KID Sample
The Agency for Healthcare Research and Quality (AHRQ) obtained agreements with 27 HCUP State Partners to participate in the 2000 KID. Over 90% of the hospital universe is included in the sampling frame for all but six of these states. Five State Partners â?? Georgia, Hawaii, Missouri, South Carolina and Virginia â?? imposed sampling restrictions that limited the percentage of state hospitals included in the frame to between 50 and 82 percent. (Restrictions from other states did not have an appreciable effect on the percentage of hospitals in the sampling frame.) One State Partner, Texas, supplied data from only 70% of the stateâ??s hospitals. This is because certain Texas state-licensed hospitals, primarily the smaller hospitals, are exempt from statutory reporting requirements. As a result, small Texas hospitals are substantially less likely to be included in the sampling frame, while larger hospitals are more likely to be included. Although the number of hospitals omitted appears sizable, these missing hospitals contain only 6% of Texas discharges.

Although pediatric discharges from hospitals from each region are selected for the KID, the comprehensiveness of the sampling frame varies by region. The percentage of hospitals included in the sampling frame is highest in the Northeast (90%) and in the West (77%), while figures are lower for the South (63%) and the Midwest (30%).

Because the KID sampling frame has a disproportionate representation of the more populous states and contains hospitals with more annual discharges, its comprehensiveness in terms of discharges is higher. The proportion of the regional population in the KID states ranges from 95% in the Northeast to 26% in the Midwest. The five Southern states added for 2000 have substantially increased the percentage of the Southern population represented, from 38.7% in the 1997 KID to 80.8% in the 2000 KID.

There were 2,788 hospitals in the 2000 sampling frame, a 10% increase from the 1997 KID. The final 2000 KID sample included 2,516,833 discharges of children from 2,784 hospitals drawn from 27 frame states representing each region of the United States. The 2000 KID is larger than the 1997 KID across several dimensions:

The number of states included increased from 22 to 27.
The number of hospitals included increased from 2,521 to 2,784.
The number of discharges increased from 1.9 million to 2.5 million.
Variance Calculations
It may be important for researchers to calculate a measure of precision for some estimates based on the KID sample data. Variance estimates must take into account both the sampling design and the form of the statistic. If hospitals inside the frame were similar to hospitals outside the frame, the sample hospitals can be treated as if they were randomly selected from the entire universe of hospitals within each stratum. Discharges were randomly selected from within each hospital. Standard formulas for stratified, two-stage cluster sample without replacement may be used to calculate statistics and their variances in most applications.

The KID database includes a Hospital weights file with variables required by statistical software to calculate finite population statistics. In addition to the sample weights described earlier, hospital identifiers (Primary Sampling Units, or PSUs), stratification variables, and stratum-specific totals for the numbers of discharges and hospitals are included so that finite-population corrections (FPCs) can be applied to variance estimates.

See also Kid 680 1 - 3