National Death Index
Death record information from state vital statistics offices. Includes date and cause of death.
All deaths in the US that are on file in state vital statistics’ offices
Access
Identified data are available to investigators for statistical purposes in medical and health research. Researchers must submit a data request application form to NCHS to access the National Death Index (NDI) data. As part of the application process, researchers must obtain approval from an Institutional Review Board (IRB), and sign a confidentiality agreement, which requires the researchers to specify how they will securely store and dispose of the data.
The request process, including links to necessary forms, is detailed in Chapter 1 of the NDI User’s Guide.
Timeline for Access
The review of an application for data usually takes 2-3 months. Once the application is approved, data are typically delivered 2 weeks after the researcher transmits finder files for matching.
After 5 years, any data that has identifiable information that came from the NDI match must be destroyed, and NCHS must be notified (unless the researcher specifies a justification for keeping the data longer in their initial application or applies for and receives an extension from NCHS).
Lag Time
Death records are added annually, typically 15 or more months after the end of the calendar year. NCHS has an "Early Release" program in which researchers can request data sooner. Early Release files are made available when more than 90 percent of the previous year’s death records have been processed, but no later than 6 months after the end of the calendar year. This is typically within a month of the end of a calendar year. Early Release files are updated at least twice before the release of final data. NCHS also tracks the percentage of records from each state that are available in the Early Release file.
Cost
The fees for routine NDI searches consist of a $350.00 service charge plus $0.15 per user record per year of death searched. For example, 1,000 records searched against 10 years would cost $350 + ($0.15 x 1,000 x 10) or $1,850. Fees for the “NDI Plus” service—which includes cause of death codes for each record—are $0.21 per user record per year searched. For subjects that are known to be deceased NDI charges $0.15 per subject for routine searches and $5.00 per subject for NDI Plus searches, regardless of how many years are searched. Volume discounts are available. More information on user fees is available on NCHS’ user fees worksheet.
Linking
Researchers must send a CD containing a file(s) of individuals to search for in the state death records to NCHS by overnight delivery or submit the files via SFTP. Further details regarding preparing records to be sent to NCHS can be found in Chapter 2 of the NDI User’s Guide.
Identifiers Available for Linking
- NCHS requires that the researcher's finder file has at least one of these three combinations of variables in order to search for matches in the NDI data set:
- First name, last name, Social Security number (SSN)
- First name, last name, month and year of birth
- SSN, date of birth, sex
- NCHS counts records as "possible matches" if they meet one of the following seven criteria:
- Exact match of SSN
- Exact month and +/- 1 year of birth, first and last name
- Exact month and +/- 1 year of birth, first and middle initials, last name
- Exact month and day of birth, first and last name
- Exact month and day of birth, first and middle initials, last name
- Exact month and year of birth, first name, father’s surname
- If the subject is female: Exact month and year of birth, first name, last name (on user’s record), and father’s surname (on NDI record)
Data Contents
Variable lists and data documentation for each file are available in the NDI User’s Guide.
Partial List of Variables
Date of death, state of death, death certificate number, cause of death codes in ICD-9 or ICD-10 format (NDI Plus only).
J-PAL Randomized Evaluations Using this Data Set
Jacob, Brian A., Jens Ludwig, and Douglas L. Miller. 2013. "The Effects of Housing and Neighborhood Conditions on Child Mortality." Journal of Health Economics 32(1): 195-206.
Finkelstein A, Zhou A, Taubman S, Doyle J. 2020. “Healthcare Hotspotting – A Randomized Controlled Trial.” The New England Journal of Medicine 382(2), 52-162.