Identifying deaths before 1979 using the Social Security Administration Death Master File

Area of Science:

• Public Health
• Epidemiology
• Biostatistics

Background:

• The Social Security Administration (SSA) was a primary source for death data before 1979.
• The SSA replaced its death accounting system with the Death Master File in 1988.
• The completeness of the Death Master File is crucial for accurate cohort study mortality ascertainment.

Purpose of the Study:

• To evaluate the completeness of the SSA Death Master File for cohort studies.
• To assess the impact of using the Death Master File on mortality analyses.
• To identify potential biases in epidemiological research due to incomplete death records.

Main Methods:

• Comparison of SSA Death Master File data with U.S. Vital Statistics records.
• Verification of known decedents from seven cohorts against the SSA Death Master File.
• Re-analysis of cohort data excluding deaths not found in the SSA Death Master File.

Main Results:

• Only 53% of reported U.S. deaths and 75% of known cohort deaths were found in the SSA Death Master File.
• Ascertainment improved significantly for deaths after 1975 (89-95%).
• Excluding pre-1979 deaths not in the file led to 20-35% decreases in mortality ratios and dose-response trends.

Conclusions:

• The SSA Death Master File is inadequate for determining vital status, particularly for deaths before 1979.
• Incomplete death ascertainment can significantly bias mortality metrics in cohort studies.
• Research using cohorts with substantial pre-1979 deaths may be severely affected, especially before the National Death Index inception.