Problems in a statistical study of disease based on death certificates

Area of Science:

• Epidemiology
• Public Health
• Cancer Research

Background:

• Death certificates are crucial for vital statistics and disease frequency.
• Cancer data collection methods have evolved over time.
• Accurate cause-of-death reporting is essential for epidemiological studies.

Purpose of the Study:

• To evaluate the scientific utility of historical death certificate data for cancer frequency analysis.
• To assess the reliability of cancer diagnoses recorded on death certificates in Morgan County, Georgia.

Main Methods:

• Retrospective review of 5,787 death certificates registered between 1927 and 1979 in Morgan County, Georgia.
• Data extraction focused on age, sex, race, and cause of death, with emphasis on neoplasms.
• Analysis of coding challenges, including unspecified tumor sites and unrecorded tumors.

Main Results:

• Significant limitations were identified in the cancer data recorded on death certificates.
• Inconsistent certificate formats and diagnostic coding issues compromised data accuracy.
• The recorded cancer information lacked sufficient detail for reliable scientific analysis.

Conclusions:

• Historical death certificates from Morgan County, Georgia, between 1927-1979, are not scientifically informative for cancer frequency studies.
• Data quality issues, including diagnostic ambiguity and coding inconsistencies, render the information unreliable for epidemiological research.
• Improvements in death certificate formats did not fully resolve data quality concerns during the study period.