Census 2000 Population Estimates - How Do They Affect Visit Rates?

Data from the 2000 NAMCS and NHAMCS were published in NCHS Advance Data reports in April and June 2002. At the time of publication, population estimates from Census 2000 were not  available for use in calculating visit rates. Therefore, the three reports (AD 326, AD 327, and AD 328) presented rates for 2000 that were available from the U.S. Census Bureau at that time. These were monthly postcensal estimates of the civilian noninstitutional population of the United States, based on the 1990 Decennial Census, which were adjusted for net underenumeration.

Recently, population estimates based on Census 2000 have become available. If these newer population estimates of the 2000 population had been used with the 2000 NAMCS and NHAMCS data, would the resulting visit rates have been different?

Table 1 presents 2000 visit rates to physician offices and hospital outpatient and emergency departments using both sets of population estimates. The first column of numbers uses 1990-based estimates as denominators for rate calculation; they are the same as the rates presented in our 2000 Advance Data reports. The second column of numbers uses provisional population estimates from Census 2000 to calculate visit rates. (The visit rate is the number of visits divided by the appropriate age/sex/race group and then multipled by 100 to get a rate per 100 persons.) As can be seen, the visit rates look quite similar. In no case were the differences between corresponding rates found to be statistically significant.