Population description
In the project proposal, you must describe the population shortly and precisely (without involving technical details), and document who creates the population. You do so under the population description in Denmark's Data Portal.
Private institutions are able to create the population themselves and get a full register extraction if the project is surveying a major group of entities. To get a full register extraction, private institutions must give reasons for this need based on the size of the population.
When Research Services must create the population
If Research Services is going to create the population for your project, this is done on the basis of a framework agreement. Under the population description in Denmark's Data Portal, you describe the population shortly and precisely (without involving technical details) and add that Statistics Denmark is going to create the population. When Research Services have received the project proposal, they will contact you about the creation of the population.
Examples of population descriptions:
`The population consists of all persons who have been hospitalised with asthma, which is matched with five controls on sex and birth year per case. The controls must be alive and be residing in Denmark on the index data of the case. Statistics Denmark creates the population.'
`The population consists of persons who have had residence permits as refugees, and family members reunited with refugees. Statistics Denmark creates the population.'
Framework agreement for extraction description and population creation
Research Services prepares a framework agreement, which covers counselling regarding the extraction description as well as the subsequent population creation. Based on the framework agreement, we prepare a detailed extraction description in collaboration with the relevant institution. Research Services uses the extraction description for the final population creation. Based on the institution's criteria and needs regarding the population, we give advice on which registers, variables and variable values that are necessary to create the wanted population. The final extraction description is attached as an appendix to the project proposal. When the extraction description is ready, Research Services creates the population for the project.
The following elements must be uncovered for the extraction description:
- Registers or additional data to be used
- Periods, including if you want to use registers that are updated annually, quarterly or monthly (for example, BEF (population) is updated quarterly)
- Conditions based on specific variables and delimitation on specific variable values (for example, if the population must be delimited by age from 15-76 years)
- How registers must be linked (if several registers are applied), including linking based on specific variables and, if relevant, key register
Especially about case control populations
Research Services uses the term `case control populations' for analyses where cases (e.g. exposed) are compared with a reference group (controls). The term is used regardless of the type of study. Under the population description in Denmark's Data Portal, enter a short and precise description of the criteria for cases and controls in the case control population, without involving technical details (including registers and variables).
In collaboration with Research Services, we prepare a detailed extraction description of the case control population. The final extraction description is attached as an appendix to the project proposal. Please note that Research Services only creates case control populations based on date and register criteria, not based on more complicated statistical methods such as for example Propensity Score Matching.
The following elements must be uncovered for the extraction description:
- What characterises cases:
- Registers, periods, conditions, and how registers are linked (see description below)
- If relevant, index date (for example date of first completed vocational education, first hospital discharge date)
- What characterises the pool of possible controls:
- Registers to be used for creating the pool of possible controls
- Inclusion and exclusion criteria based on specific variables and variable values (for example sex = 2 (women), municipality = 607 (Fredericia), residence in the period 01-01-2020 until 31-12-2023)
- Specific criteria for the case control population including:
- How many controls are extracted per case?
- Whether cases are allowed to be controls of other cases
- If controls are allowed to change status in the inclusion period
- Extraction with or without replacement:
- is a control allowed to be used as a control for more than one case (replacement)?
- or can a control only be a control for a specific case (without replacement)?