C Stallmann

Datenlinkage von Primär- und Sekundärdaten

In Germany, research on health-care services addresses many topics within a regional context, and it predominantly uses a single (typically secondary) data sources for this purpose. The specific disadvantages and methodological challenges associated with these data sources may limit analysis. Various data sources break the data down by region and may be of interest in regional health-care research. Linking multiple data sources (data linkage) could therefore expand analysis options in this area. Researchers in this field are currently discussing various approaches for using data linkage to overcome the respective weaknesses of primary and secondary data. This contribution covers the various types of data linkage (on an aggregate or individual level) and their potentials and limitations in small area health services research. The focus lies on individual data linkage, which requires written informed consent. Taking into account methodological and particularly data protection challenges, conclusions are drawn regarding future application areas and options of small area health services research and specific examples are provided.ZusammenfassungDie Versorgungsforschung in Deutschland behandelt eine Vielzahl von Themen im regionalen Kontext und nutzt dafür überwiegend eine (in der Regel: Sekundär-)Datenquelle. Deren spezifische Nachteile und methodischen Einschränkungen können sich limitierend auf eine Analyse auswirken. Zur regionalen Aufgliederung existieren vielfältige Datenquellen, die für die regionale Versorgungsforschung von Interesse sein könnten. Eine Verknüpfung verschiedener Datenquellen (Datenlinkage) könnte somit die Analysemöglichkeiten erweitern. In der Versorgungsforschung selbst werden derzeit verschiedene Ansätze diskutiert, um die jeweiligen Schwächen von Primär- und Sekundärdaten über ein Datenlinkage zu überwinden. Der vorliegende Beitrag thematisiert die verschiedenen Formen des Datenlinkage (auf aggregiertem bzw. individuellem Niveau) sowie deren Potenziale und Restriktionen für die kleinräumige Versorgungsforschung. Der Fokus liegt auf dem individuellen Datenlinkage, das ein schriftliches Einverständnis voraussetzt (informed consent). Unter Berücksichtigung der methodischen und insbesondere datenschutzrechtlichen Herausforderungen werden Schlussfolgerungen über zukünftige Anwendungsfelder und -möglichkeiten der kleinräumigen Versorgungsforschung gezogen und an Beispielen konkretisiert.AbstractIn Germany, research on health-care services addresses many topics within a regional context, and it predominantly uses a single (typically secondary) data sources for this purpose. The specific disadvantages and methodological challenges associated with these data sources may limit analysis. Various data sources break the data down by region and may be of interest in regional health-care research. Linking multiple data sources (data linkage) could therefore expand analysis options in this area. Researchers in this field are currently discussing various approaches for using data linkage to overcome the respective weaknesses of primary and secondary data. This contribution covers the various types of data linkage (on an aggregate or individual level) and their potentials and limitations in small area health services research. The focus lies on individual data linkage, which requires written informed consent. Taking into account methodological and particularly data protection challenges, conclusions are drawn regarding future application areas and options of small area health services research and specific examples are provided.

[Data linkage of primary and secondary data: a gain for small-area health-care analysis?].

In Germany, research on health-care services addresses many topics within a regional context, and it predominantly uses a single (typically secondary) data sources for this purpose. The specific disadvantages and methodological challenges associated with these data sources may limit analysis. Various data sources break the data down by region and may be of interest in regional health-care research. Linking multiple data sources (data linkage) could therefore expand analysis options in this area. Researchers in this field are currently discussing various approaches for using data linkage to overcome the respective weaknesses of primary and secondary data. This contribution covers the various types of data linkage (on an aggregate or individual level) and their potentials and limitations in small area health services research. The focus lies on individual data linkage, which requires written informed consent. Taking into account methodological and particularly data protection challenges, conclusions are drawn regarding future application areas and options of small area health services research and specific examples are provided.ZusammenfassungDie Versorgungsforschung in Deutschland behandelt eine Vielzahl von Themen im regionalen Kontext und nutzt dafür überwiegend eine (in der Regel: Sekundär-)Datenquelle. Deren spezifische Nachteile und methodischen Einschränkungen können sich limitierend auf eine Analyse auswirken. Zur regionalen Aufgliederung existieren vielfältige Datenquellen, die für die regionale Versorgungsforschung von Interesse sein könnten. Eine Verknüpfung verschiedener Datenquellen (Datenlinkage) könnte somit die Analysemöglichkeiten erweitern. In der Versorgungsforschung selbst werden derzeit verschiedene Ansätze diskutiert, um die jeweiligen Schwächen von Primär- und Sekundärdaten über ein Datenlinkage zu überwinden. Der vorliegende Beitrag thematisiert die verschiedenen Formen des Datenlinkage (auf aggregiertem bzw. individuellem Niveau) sowie deren Potenziale und Restriktionen für die kleinräumige Versorgungsforschung. Der Fokus liegt auf dem individuellen Datenlinkage, das ein schriftliches Einverständnis voraussetzt (informed consent). Unter Berücksichtigung der methodischen und insbesondere datenschutzrechtlichen Herausforderungen werden Schlussfolgerungen über zukünftige Anwendungsfelder und -möglichkeiten der kleinräumigen Versorgungsforschung gezogen und an Beispielen konkretisiert.AbstractIn Germany, research on health-care services addresses many topics within a regional context, and it predominantly uses a single (typically secondary) data sources for this purpose. The specific disadvantages and methodological challenges associated with these data sources may limit analysis. Various data sources break the data down by region and may be of interest in regional health-care research. Linking multiple data sources (data linkage) could therefore expand analysis options in this area. Researchers in this field are currently discussing various approaches for using data linkage to overcome the respective weaknesses of primary and secondary data. This contribution covers the various types of data linkage (on an aggregate or individual level) and their potentials and limitations in small area health services research. The focus lies on individual data linkage, which requires written informed consent. Taking into account methodological and particularly data protection challenges, conclusions are drawn regarding future application areas and options of small area health services research and specific examples are provided.

Verknüpfung großer Sekundär- und Registerdatenquellen mit Daten aus Kohortenstudien

Cohort studies provide the best evidence of all epidemiological observational studies for the identification of causal relationships between risk factors and diseases. However, this design may lead to drawbacks that may affect the validity and reliability of the results. This follows in particular from systematic errors, such as selection bias or recall bias. One possibility to avoid or counteract some of these drawbacks is to link primary data from cohort studies with secondary and register data. The linkage of these data may also be used for mutual validations. Data that were previously linked with primary data within the context of cohort studies in Germany were obtained from statutory health insurances and pensions as well as data from the Federal Employment Agency and cancer registries. All these data have two features in common: First, they all cover detailed information about a large population and over a long period of time. Second, all sources are in principle able to provide data on an individual level such that an individual data linkage, e.g. with primary data, is possible. However, use and linkage of each of these data sources are restricted by several limitations. These have to be accounted for as well as numerous legal restrictions that exist in Germany to especially prevent the misuse of social data.

Verknüpfung großer Sekundär- und Registerdatenquellen mit Daten aus Kohortenstudien@@@Linkage of large secondary and registry data sources with data of cohort studies: Doppeltes Potenzial nutzen@@@Usage of a dual potential

Cohort studies provide the best evidence of all epidemiological observational studies for the identification of causal relationships between risk factors and diseases. However, this design may lead to drawbacks that may affect the validity and reliability of the results. This follows in particular from systematic errors, such as selection bias or recall bias. One possibility to avoid or counteract some of these drawbacks is to link primary data from cohort studies with secondary and register data. The linkage of these data may also be used for mutual validations. Data that were previously linked with primary data within the context of cohort studies in Germany were obtained from statutory health insurances and pensions as well as data from the Federal Employment Agency and cancer registries. All these data have two features in common: First, they all cover detailed information about a large population and over a long period of time. Second, all sources are in principle able to provide data on an individual level such that an individual data linkage, e.g. with primary data, is possible. However, use and linkage of each of these data sources are restricted by several limitations. These have to be accounted for as well as numerous legal restrictions that exist in Germany to especially prevent the misuse of social data.

[Linkage of large secondary and registry data sources with data of cohort studies : usage of a dual potential].

Cohort studies provide the best evidence of all epidemiological observational studies for the identification of causal relationships between risk factors and diseases. However, this design may lead to drawbacks that may affect the validity and reliability of the results. This follows in particular from systematic errors, such as selection bias or recall bias. One possibility to avoid or counteract some of these drawbacks is to link primary data from cohort studies with secondary and register data. The linkage of these data may also be used for mutual validations. Data that were previously linked with primary data within the context of cohort studies in Germany were obtained from statutory health insurances and pensions as well as data from the Federal Employment Agency and cancer registries. All these data have two features in common: First, they all cover detailed information about a large population and over a long period of time. Second, all sources are in principle able to provide data on an individual level such that an individual data linkage, e.g. with primary data, is possible. However, use and linkage of each of these data sources are restricted by several limitations. These have to be accounted for as well as numerous legal restrictions that exist in Germany to especially prevent the misuse of social data.

Datenlinkage von Primär- und Sekundärdaten@@@Data linkage of primary and secondary data: Ein Zugewinn auch für die kleinräumige Versorgungsforschung in Deutschland?@@@A gain for small-area health-care analysis?

In Germany, research on health-care services addresses many topics within a regional context, and it predominantly uses a single (typically secondary) data sources for this purpose. The specific disadvantages and methodological challenges associated with these data sources may limit analysis. Various data sources break the data down by region and may be of interest in regional health-care research. Linking multiple data sources (data linkage) could therefore expand analysis options in this area. Researchers in this field are currently discussing various approaches for using data linkage to overcome the respective weaknesses of primary and secondary data. This contribution covers the various types of data linkage (on an aggregate or individual level) and their potentials and limitations in small area health services research. The focus lies on individual data linkage, which requires written informed consent. Taking into account methodological and particularly data protection challenges, conclusions are drawn regarding future application areas and options of small area health services research and specific examples are provided.ZusammenfassungDie Versorgungsforschung in Deutschland behandelt eine Vielzahl von Themen im regionalen Kontext und nutzt dafür überwiegend eine (in der Regel: Sekundär-)Datenquelle. Deren spezifische Nachteile und methodischen Einschränkungen können sich limitierend auf eine Analyse auswirken. Zur regionalen Aufgliederung existieren vielfältige Datenquellen, die für die regionale Versorgungsforschung von Interesse sein könnten. Eine Verknüpfung verschiedener Datenquellen (Datenlinkage) könnte somit die Analysemöglichkeiten erweitern. In der Versorgungsforschung selbst werden derzeit verschiedene Ansätze diskutiert, um die jeweiligen Schwächen von Primär- und Sekundärdaten über ein Datenlinkage zu überwinden. Der vorliegende Beitrag thematisiert die verschiedenen Formen des Datenlinkage (auf aggregiertem bzw. individuellem Niveau) sowie deren Potenziale und Restriktionen für die kleinräumige Versorgungsforschung. Der Fokus liegt auf dem individuellen Datenlinkage, das ein schriftliches Einverständnis voraussetzt (informed consent). Unter Berücksichtigung der methodischen und insbesondere datenschutzrechtlichen Herausforderungen werden Schlussfolgerungen über zukünftige Anwendungsfelder und -möglichkeiten der kleinräumigen Versorgungsforschung gezogen und an Beispielen konkretisiert.AbstractIn Germany, research on health-care services addresses many topics within a regional context, and it predominantly uses a single (typically secondary) data sources for this purpose. The specific disadvantages and methodological challenges associated with these data sources may limit analysis. Various data sources break the data down by region and may be of interest in regional health-care research. Linking multiple data sources (data linkage) could therefore expand analysis options in this area. Researchers in this field are currently discussing various approaches for using data linkage to overcome the respective weaknesses of primary and secondary data. This contribution covers the various types of data linkage (on an aggregate or individual level) and their potentials and limitations in small area health services research. The focus lies on individual data linkage, which requires written informed consent. Taking into account methodological and particularly data protection challenges, conclusions are drawn regarding future application areas and options of small area health services research and specific examples are provided.