Underreporting of Race in the National Hospital Discharge Survey Advance Data 265. Race data from the National Hospital Discharge Survey (NHDS) has become increasingly incomplete in recent years. This report examines factors related to the underreporting of race and explores the effects of the underreporting on NHDS estimates of hospital use by race. Two main factors were found to be related to the increasing underreporting of race in the NHDS. First, a growing number of hospitals that participated in the NHDS in 1990-92 did not report race for any patients. Second, race was not reported for the majority of patients that were identified as Hispanic. Data Highlights: The number of hospitals in the NHDS that report race for less than 3 percent of discharges increased from 17 in 1989 to 63 in 1992. Most of the hospitals that did not report race in the 1990-92 period used their automated data collection methods. A specific race was reported for only 25-35 percent of Hispanic patients in the 1990-92 period; most Hispanic patients were probably white. In comparison with the number of discharges estimated from the 1990-92 National Health Interview Survey, NHDS estimates of discharges were significantly lower for white patients, but not significantly different for black patients. Proportional adjustment of NHDS race data may produce more accurate estimates of white discharges, but it does not improve comparison between racial groups.
This page last reviewed
January 11, 2007
|