Comparison of two surveys of hospitalization: the National Hospital Discharge Survey and the NHANES I Epidemiologic Followup Study.

OBJECTIVES This report compares hospitalization data from the NHANES I Epidemiologic Followup Study (NHEFS) with data from the National Hospital Discharge Survey (NHDS), the benchmark for hospitalization in the United States, for men and women 35 years and older for the period 1971-87. The comparison is intended to help analysts evaluate the validity and generality of analyses based on the NHEFS. METHODS Hospital stays per 1,000 population and average lengths of stay are compared year by year for each age-sex group and for the entire period. Regression analyses test for differences between the two surveys by age and sex, and for differences in trends over time and the effect of the Medicare program's prospective hospital payment system. RESULTS Hospital stays per 1,000 population were lower in NHEFS than in NHDS in all age-sex groups at the beginning of the period, but the differences had almost disappeared by 1987. Lengths of stay, although somewhat longer in NHEFS, matched NHDS more closely. Differentials by age and sex were similar in the two surveys for both hospital stays per 1,000 population and length of hospital stay. With its extensive information on baseline risk factors, the NHEFS offers a unique opportunity to study determinants of hospitalization in a representative sample of U.S. adults. The evaluation presented here suggests two points for researchers who want to use the NHEFS. First, including age as a control should largely correct for differences in age distribution between NHEFS and NHDS. Second, a time trend should also be included to capture the effects of several factors that caused the count of stays to be low in the early years of NHEFS followup.