Attachment Q - Estimated Relative Standard Errors

Attachment Q Estimated Relative Standard Errors NHDS Redesign

Estimated Relative Standard Errors of selected discharge level statistics, for sample size of 240 hospitals (100 discharges per hospital per year)

Discharge Variable	Percent Estimate*	Responding Sample Size	Estimated Number of Discharges	Relative Standard Errors (%)
Number of discharges by characteristic:				240 Hospitals
- % ED admission	41.36%	7544	14,340,000	3.78 2.18 3.75 0.09 10.62 7.08 6.82 10.70 7.56 9.06 17.49 0.52
- % Medicare	39.77%	7254	13,787,065
- Female aged 15-44 years	22.48%	4100	7,794,615
- % Non-metro hospital	17.12%	3122	5,934,990
- Black or African American	11.90%	2170	4,108,786
- Patient aged <15 years	7.0%	1276	2,430,926
- % Delivery with C-section	3.64%	663	1,261,878
- Depression & Bipolar	2.7%	492	944,599
- Acute myocardial infarction	2.0%	364	682,699
- Asthma	1.4%	255	488,594
- % C-difficile (hospital infection)	0.22%	40	76,267
- Newborn	N/A	960	3,999,112
-7 day readmission	2.1%	383	728,894
Data from respondent hospitals in the 2005 National Hospital Discharge Survey were used to calculated these estimates and RSEs.
Strata are defined by hospital service type, area type (Central City, Fringe, Small Metro, Non-Metro), and bed-size where feasible.
Total sample includes non respondents and out-of-scope hospitals but assumed rate for being both in scope and responding is 80% of total sample hospitals.
*Except for newborn variable, newborns are excluded from estimates.