Probabilistic Record Linkage: Relationships between File Sizes, Identifiers, and Match Weights

Abstract:

This study investigates relationships between file sizes, amounts of information contained in commonly used record linkage variables, and the amount of information needed for a successful probabilistic linkage project. We present an equation predicting the amount of information needed for a successful linkage project. Match weights for variables commonly used in record linkage are measured using artificially created databases. Linkage algorithms were successful when the sum of minimum weights for variables used in a linkage exceeded the predicted cutoff. Linkage results were acceptable when this sum was near the predicted cutoff. This technique enables researchers to determine if enough information exists to perform a successful probabilistic linkage.

Keywords:

Probabilistic Record Linkage - Match Weights - Confidentiality

• REFERENCES

• 1 Weiss HB, Dill SM, Garrison HG. et al. The potential of using billing data for emergency department injury surveillance. Acad Emerg Med 1997; 4 (Suppl. 04) 282-7.
• 2 Jamieson E, Roberts J, Browne G. The feasibility and accuracy of anonymized record linkage to estimate shared clientele among three health and social service agencies. Methods of Information in Medicine 1995; 34 (Suppl. 04) 371-7.
• 3 Jaro MA. Probabilistic linkage of large public health data files. Stat Med 1995; 14 5-7 491-8.
• 4 Hogg RS, Whitehead J, Ricketts M. et al. Patterns of geographic mobility of persons with AIDS in Canada from time of AIDS index diagnosis to death. Clin Invest Med 1997; 20 (Suppl. 02) 77-83.
• 5 Clark DE, Hahn DR. Hospital trauma registries linked with population-based data. J Trauma 1999; 47 (Suppl. 03) 448-54.
• 6 Alsop JC, Langley JD. Determining first admissions in a hospital discharge file via record linkage. Methods of Information in Medicine 1998; 37 (Suppl. 01) 32-7.
• 7 Cook LJ, Knight S, Olson LM. et al. Motor vehicle crash characteristics and medical outcomes among older drivers in Utah, 1992-1995. Ann Emerg Med 2000; 35 (Suppl. 06) 585-91.
• 8 Johnson SW, Walker J. The Crash Outcome Data Evaluation System (CODES). Washington DC: National Highway Traffic Safety Administration; 1996
• 9 Computerised record linkage: compared with traditional patient follow-up methods in clinical trials and illustrated in a prospective epidemiological study. The West of Scotland Coronary Prevention Study Group. J Clin Epidemiol 1995; 48 (Suppl. 02) 1441-52.
• 10 Waien SA. Linking large administrative databases: a method for conducting emergency medical services cohort studies using existing data. Acad Emerg Med 1997; 4 (Suppl. 11) 1087-95.
• 11 Wiklund K, Eklund G. Reliability of record linkage in the Swedish Cancer-Environment Register. Acta Radiol Oncol 1986; 25 (Suppl. 01) 11-4.
• 12 Newcombe HB. Handbook of Record Linkage: methods for health and statistical studies, administration and business. New York City: Oxford University Press; 1988
• 13 Brenner H, Schmidtmann I, Stegmaier C. Effects of record linkage errors on registry-based follow-up studies. Stat Med 1997; 16 (Suppl. 23) 2633-43.
• 14 Brenner H, Schmidtmann I. Effects of record linkage errors on disease registration. Methods of Information in Medicine 1998; 37 (Suppl. 01) 69-74.
• 15 Brenner H, Schmidtmann I. Determinants of homonym and synonym rates of record linkage in disease registration. Methods of Information in Medicine 1996; 35 (Suppl. 01) 19-24.
• 16 Health Insurance Portability and Accountability Act of 1996. Available at: www.hcfa.gov/hipaa/hipaahm.htm accessed on Dec. 12, 2000.
• 17 Jaro MA. Advances in record linkage methodology as applied to matching the 1985 census of Tampa, Florida. J Am Statist Assoc 1989; 84 (Suppl. 406) 414-9.
• 18 McGlincey M. Probabilistic Linkage Issues. CODES Technical Assistance Meeting. Las Vegas, Nevada: 1998
• 19 Roos LL, Wajda A. Record linkage strategies. Part I: Estimating information and evaluating approaches. Methods of Information in Medicine 1991; 30 (Suppl. 02) 117-23.