Record Linkage in the Cancer Registry of Tyrol, Austria

W. Oberaigner; W. Stühlinger

doi:10.1055/s-0038-1634018

Methods of Information in Medicine, Table of Contents

Methods Inf Med 2005; 44(05): 626-630
DOI: 10.1055/s-0038-1634018

Original Article

Schattauer GmbH

Record Linkage in the Cancer Registry of Tyrol, Austria

W. Oberaigner

¹Cancer Registry of Tyrol, Department of Clinical Epidemiology of the Tyrolean State Hospitals Ltd., Innsbruck, Austria

²Department of Quality Science, Medical Planning and Information Management, University for Health Informatics and Technology Tyrol, Innsbruck, Austria

,

W. Stühlinger

²Department of Quality Science, Medical Planning and Information Management, University for Health Informatics and Technology Tyrol, Innsbruck, Austria

› Author Affiliations

Abstract

PDF Download

Summary

Objective: Record linkage of patient data originating from various data sources and record linkage for checking uniqueness of patient registration are common tasks for every cancer registry. In Austria, there is no unique person identifier in use in the medical system. Hence, it was necessary and the goal of this work to develop an efficient means of record linkage for use in cancer registries in Austria.

Methods: We adapted the method of probabilistic record linkage to the situation of cancer registries in Austria. In addition to the customary components of this method, we also took into consideration typing errors commonly occurring in names and dates of birth. The method was implemented in a program written in DELPHI^TM with interfaces optimised for cancer registries.

Results: Applying our record linkage method to 130,509 linkages results in 105,272 (80.7%) identical pairs. For these identical pairs, 88.9% of decisions were performed automatically and 11.1% semi-automatically. For results decided automatically, 6.9% did not have simultaneous identity of last name, first name and date of birth. For results decided semi-automatically, 48.4% did not have an identical last name, 25.6% did not have an identical date of birth and 83.1% did not have simultaneous identity of last name and date of birth and first name.

Conclusions: The method implemented in our cancer registry solves all record linkage problems in Austria with sufficient precision.

Keywords

Probabilistic record linkage - cancer registry - homonym rate - synonym rate

PDF (318 kb)

References

References
1 Parkin DM, Whelan SL, Ferlay J, Raymond L, Yuen J. Cancer Incidence in Five Continents. Volume VII. IARC Scientific Publications No. 143. Lyon: IARC. 1997
2 Parkin DM, Whelan SL, Ferlay J, Teppo L, Thomas B. Cancer Incidence in Five Continents. Volume VIII. IARC Scientific Publications No. 155. Lyon. 2002
3 Jensen OM, Parkin DM, MacLennan R, Muir CS, Skeet RG. Cancer Registration. Principles and Methods. Lyon: IARC; 1991
4 Schouten LJ, Hoppener P, van den Brandt P, Knottnerus JA, Jager JJ. Completeness of Cancer Registration in Limburg, The Netherlands. Int J Epidemiol 1993; 22 (03) 369-76.
5 Ziegler H, Stegmaier C. Bevölkerungsbezogene Krebsregistrierung in Deutschland. Onkologie 1996; 19: 268-77.
6 Calle EE, Terrell DD. Utility of the National Death Index for Ascertainment of Mortality among Cancer Prevention Study II Participants. Am J Epid 1993; 137 (02) 235-41.
7 Fellegi IP, Sunter AB. A Theory for Record Linkage. JASA 1969; 64: 1183-1210.
8 Jaro MA. Probabilistic Linkage of Large Public Health Data Files. Stat Med 1995; 14: 491-8.
9 Leitner H. Probleme und Lösungen beim Zusammenführen von Patientendaten unterschiedlicher Quellen. Innsbruck 1995
10 Stegmaier C, Ziegler H. Zusammenführung personenbezogener Informationen (Record-Linkage)– Probleme und Möglichkeiten am Beispiel des Krebsregisters Saarland. Krebsregister Saarland, Vierteljahresheft 1992; 4: 1-7.
11 Schmidtmann I, Michaelis J. Untersuchungen zum Record Linkage für das Krebsregister Rheinland- Pfalz. 1994
12 Brenner H, Schmidtmann I. Determinants of Homonym and Synonym Rates of Record Linkage in Disease Registration. Methods Inf Med 1996; 35 (01) 19-24.
13 Brenner H, Schmidtmann I, Stegmaier C. Effects of Record Linkage Errors on Registry-based Follow- up Studies. Stat Med 1997; 16 (23) 2633-43.
14 Newcombe HB, Smith ME, Howe GR, Mingay J, Strugnell A, Abbatt JD. Reliability of Computerized versus Manual Death Searches in a Study of Health of Eldorado Uranium Workers. Comput Biol Med 1983; 13 (03) 157-69.
15 van den Brandt P, Schouten LJ. Development of a Record Linkage Protocol for Use in the Dutch Cancer Registry for Epidemiological Research. Int J Epid 1990; 19 (03) 553-8.
16 MacLeod MC, Bray CA, Kendrick SW, Cobe S. Enhancing the Power of Record Linkage involving low quality Personal Identifiers: Use of the Best Link Principle and Cause of Death prior Likelihoods. Comput Biomed Res 1998; 31 (04) 257-70.
17 Gomatam S, Carter R, Ariet M, Mitchell G. An empirical Comparison of Record Linkage Procedures. Stat Med 2002; 21 (10) 1485-96.
18 Thomas B, Newman M. Use of Commercial Record Linkage Software and Vital Statistics to Identify Patient Deaths. J Am Med Inform Assoc 1997; 4 (03) 233-7.
19 Kendrick SW, Douglas MM, Gardner D, Hucker D. Best-link Matching of Scottish Health Data Sets. Methods Inf Med 1998; 37 (01) 64-8.
20 Oberaigner W, Kreienbrock L, Schaffrath-Rosario A, Kreuzer M, Wellmann J, Keller G, Gerken M, Langer B, Wichmann HE. Radon und Lungenkrebs im Bezirk Imst/Österreich. In Wichmann HE, Schlipkoeter HW, Fuelgraff G. (eds) Fortschritte in der Umweltmedizin. Landsberg/Lech: Ecomed Verlag; 2002
21 Pommering K, Miller M, Schmidtmann I, Michaelis J. Pseudonyms for Cancer Registries. Methods Inf Med 1996; 35 (02) 112-21.
22 Meux E. Encrypting Personal Identifiers. Health Serv Res 1994; 29 (02) 247-256.
23 Rabinowitz J. A Method for Preserving Confidentiality when Linking Computerized Registries. Am J Public Health 1998; 88 (05) 836
24 Kelman C, Smith L. It’s time: Record Linkage – The Vision and the Reality. Aust N Z J Public Health 2000; 24 (01) 100-1.