and Publication Committee, North American Association of Central Cancer Registries OBJECTIVES

The Data Evaluation and Publication Committee of NAACCR used data submitted by member registries to examine the effect of death certificate follow-back on cancer registration in the U.S. METHODS For this analysis, we used cancer registration data in electronic format from 35 U.S. population-based cancer registries that participated in the 1994 NAACCR Call for Data (Table 1). Registries that submitted data also consented to have their data reproduced in an electronic format for distribution with Cancer Incidence in North America, 1988-1991.1 We defined three categories of registries based on extent of death certificate follow-back. Registries that participate in the Surveillance, Epidemiology, and End Results (SEER) program of the National Cancer Institute composed the category with the most comprehensive resources for conducting case follow-back. SEER registries routinely abstract all death certificates in which cancer is listed as a cause of death, and conduct extensive searches for medical records pertaining to the individuals thus identified, abstracting all relevant information. The remaining categories of follow-back protocols included other population-based registries in the U.S. that perform death certificate follow-back, and others that do not perform death certificate follow-back. We examined four measures of tumor registry data quality in our analysis: the percent of cases that wer