Background: The completeness of cancer registration is a major validity index of any reported cancer incidence.The present study aimed to evaluate the esophageal cancer incidence registered in the Tehran MetropolitanArea Cancer Registry. Materials and methods: The data on esophageal cancer abstracted from three sourcesof 1) pathology departments, 2) medical records, and 3) death certificates during 2003 till 2007 were utilized.The completeness of the data sources were evaluated using coverage (defined as the proportion of a communitypopulation with esophageal cancer identified by the source) and density (defined as the proportion of non-emptyfields of the data by source).
Results: A total 1,404 cases of esophageal cancer were reported for the duration ofthe study. Pathology provided 771, medical records 432, and death certificates 609. The coverage was 0.55 forpathology, 0.31 for medical records, and 0.43 for death certificates. The respective density values were 0.82, 0.96and 0.98, respectively. Pathology (0.45) was the most complete source followed by medical records (0.42), anddeath certificates (0.29). Discussion: A low degree of completeness dictates putting more effort into case findingplus abstracting data more thoroughly.