To the Editor The excellent work of Navar et al1 demonstrates the unreliability of using the US Social Security Death Master File (SSDMF) for clinical research. It is important for readers to understand why this is the case. The SSDMF has been incomplete since November 1, 2011, when 4.2 million records were removed from the historical file of 89 million.2 Since that time, the file has lost approximately 40% of deaths per year. In 2011, the US Social Security Administration concluded that it could not release state-owned data (ie, information from the death certificate) to the SSDMF.3 The consequences of this decision for researchers were profound, as there is no alternative to the SSDMF, which contained the only up-to-date, publicly available death records for the United States. The National Death Index is now the only reliable source of nationally available, identifiable death information for researchers. However, as the authors state,1 the National Death Index can be expensive, and there is a 2-year lag. Researchers should heed the study’s conclusion that the SSDMF is not reliable and should not be used alone to estimate mortality rates.