Combining data from different sources

A few typical cases are presented below that illustrate how to combine data from different registers. The different procedures are mostly due to the fact that Statistics Finland and the social welfare registers of THL are under the jurisdiction of the Statistics Act, and thus data from them cannot be delivered to researchers in a format where individuals can be identified.

Data from Statistics Finland is used

Case 1. The registers used in the study are maintained by THL, Statistics Finland, and other register controllers; the study population has been defined based on data from THL.

It is assumed that, at THL, the study population is formed by sampling. The personal identification numbers (PINs) are then sent to other register controllers (such as the Social Insurance Institution (SII)), who conduct their own data sampling and then send the data including the PINs to Statistics Finland. THL also sends its data on the sample population (including PINs) to Statistics Finland. Statistics Finland samples the requested data from its registers and creates an identifier (ID) specific to the study and replaces all PINs with the IDs in all datasets, which are then sent to the researcher. It also takes into account other data protection considerations. The researcher can, by using the ID numbers that Statistic Finland has given, combine the data from the different register controllers.

Case 2. The registers used in the study are maintained by THL, Statistics Finland, and other register controllers; the study population has been defined based on data from Statistics Finland.

The situation is the same as in Case 1, with the exception that now Statistics Finland sends out the PINs to the different register controllers, who then send their data on the study population to Statistics Finland.

Case 3. Only the following variables are used from Statistics Finland: age, sex, education, occupation and cause of death.

In those cases where the researcher has a right, according to the Personal Data Act, to keep a personal data file, the researcher can obtain data including personal identifiers (PINs) from Statistics Finland. The researcher can combine the data with other data including PINs obtained from other register controllers.

Data from THL is used

Case 1. The registers used in the study are maintained by THL and other register controllers, the study population has been defined based on data from THL .

It is assumed that, at THL, the study population is formed by sampling, and an identifier (ID) specific to this study is created at THL. The ID and the personal identifiers (PIN) are then sent to other register controllers (such as the SII), who conduct their own data sampling and send the data to THL, where the data are combined and the PINs are removed. THL also takes into account other data protection considerations before sending the data to the researcher.

Case 2. The data from THL consists only of data from the health care registers

It is assumed that, at THL, the study population is formed by sampling. The researcher obtains the data from THL, including the PINs. The personal identifiers (PINs) of the study population are sent to other register controllers (such as the SII), who conduct their own data sampling and send the data to the researcher. The researcher receives the data and combines the data using the PINs.

Data from register controllers other than THL and Statistics Finland are used

A sample of the study population can be taken at one of the register controllers. The PINs are sent to the other register controllers, who send their data for the study population to the researcher. The researcher can combine the data using the PINs.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s