Comparability

Comparability - geographical

In other – especially European – countries, registers exist of the same type as the Danish National Hospital Discharge Register. The coding on diagnoses will typically be by international classification. For international comparison, it is recommended that you look at data from Eurostat and the OECD which make comparable data collections and publish data that are comparable to a certain extent in this field. There are a number of organizational and institutional conditions that we must keep in mind when analysing possible differences.

Comparability over time

The statistics have been compiled on the same basis from the beginning in 1990 and until 2018. From 2019, the statistics will be revitalized on the basis of LPR3 since it contains a new data register structure. The Danish Health Data Agency describes the modernization of the National Patient Register (the transition from LPR2 to LPR3) and the substantive implications thereof in more detail Sundhedsdatastyrelsen. (Link in Danish). The Capital Region (Hovedstaden), Region Zealand and Region Midt transferred to LPR3 on the 2nd and 3rd. February 2019. Region South and Region North Jutland switched to LPR3 on 2-3 March 2019. 2019 cannot therefore be compared directly with 2020 and 2021, as 2019 does not contain LPR3 data regarding every 12 calendar months. The figures between the regions must also be viewed with caution for 2019.

From 2018 to 2019, a data break in the volume of admissions is expected. The same applies to a certain extent to the count of out-patient treatments. This has a background in structural conditions, such as the transition to a new register data structure (from LPR2 to LPR3) where admissions are formed in output data rather than in input data. Validation for raw admission figures from 2019 alone is therefore done against the National Health Data Agency's calculations Sundhedsdatastyrelsen.

Until 2019: Statistics Denmark compares the received data with data from the previous year and any major variations are investigated. If we observe any apparent error, we contact the Danish Health Data Agency for the purpose of clarification. We also assess the internal data, and in cases of doubt we contact the person in charge of the statistics for further explanation.

Before 2019: Information is applied about admissions to, outpatient treatments at and emergency room visits to public somatic hospital wards during the calendar year. If a person is transferred during a hospital stay from one hospital ward to another, this will count as two admissions. The number of bed days in connection with admissions is applied. Accordingly, the units in the statistics are persons, admissions and bed days, outpatient treatments and emergency room visits. The statistics are broken down by sex, age, diagnosis, region of residence and a number of background variables: family type, occupational group, education, type of accommodation, ancestry and job function. (Up to and including 2008, the statistics were also broken down by predominant social security benefit).

Furthermore, in the statistics from before 2019, classifications from other sets of statistics are applied: · Socio-economic status (from AKM) (self-employed persons; assisting spouses; senior executives; high-level employees; mid-level employees; ground-level employees; other employees; unemployed persons; persons temporarily outside the labour force; students; retired persons etc.; recipients of cash benefit; other persons outside the labour force) · Education (basic general education or N/A; upper secondary education; basic vocational training and education; short-cycle education; medium-cycle education; bachelor; long-cycle education) · Family type (single; married/registered; cohabiting couples; cohabiting couples who have had children together) · Type of accommodation (single-family houses; terraced houses, linked houses and semi-detached houses; flats; other types of accommodation) · Inherit (persons with Danish parents; immigrants from the Western part of the world; immigrants from outside the Western part of the world; children of persons from the Western part of the world; children of persons from outside the Western part of the world).

In previous table series, the following developments have taken place in the area up to the year of 2018: - From 1994, a number of new information at family level have been supplemented. - From and including 1999, the statistics are supplemented with information on ancestry. - From and including 2006, there is an inventory of outpatient treatments and emergency room visits.

As at 1 January 1994, the new classification of diseases (ICD10) was employed in Denmark. This replaced the former ICD8 classification. This means that we must be cautious when comparing the diagnosis pattern across this point in time. The development in the diagnosis pattern can further be influenced by changes in the registration practice. E.g. the number of admissions with diagnoses in the group Symptoms and insufficiently defined states has increased significantly. This is due to an enhanced tendency to register symptoms and less use of actual disease diagnoses in the examination phase or in case of uncertainty as to the nature of the disease. Consequently, the number of admissions e.g. with diagnoses in the group Tumours has declined. The National Hospital Discharge Register’s data about admissions caused by road traffic accidents is estimated to be insufficient up to and including 1994. From 1995, the data is considered to be sufficient.

In an evaluation of the Hospital Utilization for the years 1995 and 2008 and 2021 respectively, nurse disputes in these years must be taken into consideration. For the years 2020 and 2021, the Covid19 pandemic must be taken into account.

In an evaluation of the number of outpatient treatments and the number of emergency room visits, a data break between 2013 and 2014 must be taken into consideration. There may be a related effect of the data break in 2015, especially in the Capital Region of Denmark.

The statistics’ use of background information has continuously been extended. Accordingly, it will not be possible to retrieve certain statements for all years back in time, nor from 2019 forwards.

In 2017, there is a large drop in data regarding 'Diseases in pregnancy and during childbirth'. The decrease is due to changed registration practices for contacts during pregnancy, seen in LPR 2017 for outpatient visits with obstetric codes (DO00-DO99) as main diagnosis.

Coherence - cross domain

The Danish Health Data Agency (previously SSI and the Danish Health Authority respectively) publish information in eSundhed (eHealth) from the National Hospital Discharge Register, moreover, they publish key figures for the health sector on a quarterly basis Sundhedsdatastyrelsen. Deviations in key figures on the number of admissions and outpatient treatments are due to the fact that the Danish Health Data Agency makes publications based on non-cleansed versions of the National Hospital Discharge Register or based on “Det Grupperede Landspatientregister” (the Grouped National Hospital Discharge Register), where DRG is included. Furthermore, there may also be differences in the delimitation, e.g. inclusion of publicly financed treatments in private hospitals in the Danish Health Data Agency’s key figures. The development from one period to the next is generally consistent between the Hospitalisation rate and Key figures from the Danish Health Data Agency.

When counting the number of hospital admissions and outpatient treatments from 2019 onwards, a calculation method named by the Danish Health Data Agency has been applied Sundhedsdatastyrelsen. The link (in Danish) is accessed from Sundhedsdatastyrelsen

The Danish Health Authority published an annual set of statistics until 2005 – also based on the National Hospital Discharge Register – about the activity in the hospitals (Hospital Statistics). The hospitalisation rate is comparable to these statistics, except from the fact that the hospitalisation rate in most statements only include persons who were in the population as at 1 January (and consequently not persons born or immigrated during the year) and that the hospitalisation statistics in the geographical statements group the persons with their residence as at 1 January, whereas the Danish Health Authority’s statements in the Statistics relating to hospitals group the persons with their municipality of residence at the time of admission.

Coherence - internal

Data are highly internally consistent.

There is an extremely limited number of gender reassignment. In such rare cases, data on gender is set to undisclosed.