Denmark has launched one of the most comprehensive injury surveillance resources in the world, a nationwide register-based cohort that captures more than 7.2 million hospital contacts linked to injuries and nearly 27,000 injury-related deaths over a thirteen-year period. Known as the National Danish Injury Cohort, or NDIC, the resource was described in detail in a cohort profile published in the European Journal of Epidemiology by researchers at the National Institute of Public Health, University of Southern Denmark. The cohort reconstructs, at the level of the individual, the burden of both unintentional injuries and intentional events such as violence, suicide attempts and self-harm across the entire Danish population from 2010 onwards, filling a gap left by the closure of the former Danish Injury Register in 2009 and restoring Denmark’s national capacity for injury surveillance.
The technical architecture of NDIC rests on Denmark’s long-standing tradition of nationwide administrative and health registries, all linked through the unique Civil Personal Register number assigned to every resident at birth or upon immigration. This personal identifier functions as the key that allows individual-level records to be joined across central registries with near-complete accuracy. The cohort draws its primary injury data from the Danish National Patient Register, which records all inpatient admissions, outpatient visits and emergency department contacts across both public and private hospitals, spanning somatic and psychiatric care. Injury-related deaths are identified through the Danish Register of Causes of Death, where each fatality is categorised by manner of death as natural, accident, suicide, violence or homicide, or undetermined following autopsy. Because Danish healthcare is tax-funded and free at the point of use, socioeconomic barriers to seeking care are reduced, which supports the completeness of the register-based capture of injury events.
The classification machinery underlying the cohort is deliberately standardised. Diagnoses are recorded using the Danish Medical Classification System, known as SKS, which incorporates the World Health Organization’s tenth revision of the International Classification of Diseases. Injury codes from ICD-10 Chapter XIX, spanning DS00 to DT98, cover physical harm to every major body region as well as burns, corrosions, frostbite, poisonings, complications of medical procedures and long-term sequelae of injuries. Chapter XX codes, DX60 through Y09, classify intentional self-harm, assaults and other external causes. External circumstances are recorded using the NOMESCO Classification of External Causes of Injuries, a Nordic standard translated into SKS as the EUA to EUZ code range. These codes specify the activity the injured person was engaged in, such as sport or paid work, the place of occurrence, such as home, school or road, and the mode of injury, such as a fall or transport accident. For traffic incidents, reporting rules also mandate the transport modes of the injured person and any counterpart, distinguishing for example a cyclist from a passenger car occupant or pedestrian.
Identifying a genuine injury incident within routine administrative data is far from trivial, and the cohort profile lays out the operational logic in detail. Each hospital contact in the National Patient Register carries a reason for contact, which since 2010 has been mandatory for acute somatic encounters and must be coded as an accident, an act of violence, or a suicide attempt when applicable. From early 2019, intentional self-harm was introduced as a distinct reason for contact, whereas before that date self-harm was recorded under illness and cannot be captured through the reason-for-contact field, a structural discontinuity the researchers flag explicitly. A contact qualifies for inclusion in NDIC if it carries any injury reporting code, or a relevant reason for contact, or an action diagnosis in the ranges DX60 to DX99 or DY0 to DY09. Because a single injury episode often generates multiple registrations as patients move through emergency departments, admissions and outpatient clinics, the cohort designates a single key contact per incident and links subsequent contacts into individual incident trajectories, while allowing individuals to accumulate multiple incidents over time. Deaths classified as accident, suicide, or violence or homicide complete the picture, although fatalities can only be identified for registered residents.
The analytical yield of the resource is already substantial. Between 2010 and 2022 the researchers identified 7,217,351 primary injury contacts and 26,871 injury-related deaths. Incidence rates were computed as the number of new events divided by the mid-year population, an accepted approximation of person-time in large dynamic populations, with exact 95 percent confidence intervals derived from the Poisson distribution using chi-square limits. The headline finding is a gradual decline in overall injury incidence from 113.3 per 1,000 person-years in 2010 to 94.4 in 2019, followed by a sharp drop in 2020 coinciding with the COVID-19 pandemic and a partial rebound in subsequent years. Men consistently sustained more injuries than women, with incidence rates between 16.7 and 24.2 percent higher across the period. Mortality told a starker story: men were between 49.7 and 66.2 percent more likely than women to die from an injury, while the overall mortality rate remained relatively stable at 34.2 to 38.0 deaths per 100,000 person-years throughout the study window.
Stratification by age and sex reveals patterns with direct preventive implications. Among women, the highest incidence rates occurred in girls aged 0 to 15 years, ranging from 116.3 to 145.9 per 1,000 person-years, followed by women aged 75 and older at 106.7 to 141.0, and young women aged 16 to 29 at 88.1 to 131.3, all three groups showing declining trends over time and a marked pandemic-era drop among the young. Women aged 30 to 74 sat at much lower and more stable rates of roughly 57 to 65. Among men, boys aged 0 to 15 recorded rates of 139.7 to 171.7 and carried the highest rate in 2022 at 151.5, while men aged 16 to 29 peaked at 195.5 in 2010, declined 26.8 percent to 143.2 by 2015, and remained comparatively stable thereafter aside from the pandemic dip. In a notable reversal of the overall sex pattern, women aged 75 and older consistently exhibited higher injury incidence than men of the same age, while rates were similar between the sexes at 65 to 74 years, underscoring the particular vulnerability of the oldest women, most plausibly linked to falls and fragility.
Geography matters as much as demography. When the researchers stratified incidence by the five Danish regions, they found striking disparities that persisted even after national reporting guidelines were applied uniformly. The Region of Southern Denmark recorded the highest rate in 2010 at 136.4 per 1,000 person-years, falling to around 119.4 by 2014, while the Capital Region declined from 113.8 to 97.0 and Central Denmark from 107.7 to 70.8. The North Denmark Region moved against the trend, rising from 51.0 to 66.0 over the same period, and Region Zealand dipped and then returned to roughly its 2010 level. From 2015 onwards all five regions stabilised with minor fluctuations, and by 2022 the lowest rates were found in Central and North Denmark at 60.5 to 73.1, while Region Zealand reached 113.8 and the Southern and Capital Regions hovered just above 103. The authors are careful to note that such regional gradients may partly reflect differences in healthcare organisation rather than true differences in injury occurrence, since residents of areas far from emergency departments are less likely to attend hospital for the same injury.
That caveat points to the cohort’s central methodological tension, which the profile addresses candidly. NDIC captures only injuries that result in hospital contact or death, meaning minor events treated solely by general practitioners, out-of-hours services, or on-site prehospital care are invisible, and previous self-reported data suggest only about half of all injuries in Denmark reach a hospital setting. Care-seeking behaviour therefore introduces residual selection that may vary geographically with distance to acute care. Misclassification risks also arise from variation in diagnostic and coding practices across clinicians, hospitals and time, with some institutions applying structured routines that generate rich contextual detail while others record only the mandatory minimum. The coding framework itself imposes structural limits, since coders cannot capture circumstances for which no adequate category exists. Yet the compensating strengths are considerable: nationwide population-based coverage, freedom from the selection biases of restricted cohorts or surveys, standardised international coding that enables cross-national comparison through the European Injury Data Base, and statistical power that renders confidence intervals almost invisible at plotted scale for incidence analyses.
The design also anticipates sophisticated downstream research. Through linkage on the Civil Personal Register number, cohort members can be connected to education registers, labour market and income registers, population statistics, prescribed medication records and healthcare utilisation data, enabling sociodemographic analyses of inequalities alongside studies of immigration status, family structure and housing. The longitudinal structure supports at least one year of follow-up after each incident, opening the door to survival analysis of time-to-event outcomes, causal inference methods for evaluating preventive interventions, and prediction modelling to identify high-risk populations. The researchers envision applications across the life course, from childhood injury prevention and workplace accidents among working-age adults to falls among the elderly, as well as mechanism-specific studies of traffic accidents, poisonings, drownings, burns, sports and occupational injuries. Data are held pseudonymously on a secure Statistics Denmark server, all analyses were conducted in R version 4.5.2 using the tidyverse, and the cohort is currently updated through 2022 with plans for annual or biennial updates, positioning NDIC as a durable foundation for evidence-based injury prevention in Denmark and a benchmark for register-based injury epidemiology internationally.
Subject of Research: A nationwide register-based cohort, the National Danish Injury Cohort (NDIC), established to monitor the incidence, causes and consequences of injuries and injury-related deaths in Denmark from 2010 onwards.
Subject of Research: Medicine
Article Title: Design and characteristics of the national Danish injury cohort (NDIC)
Article References: Rasmussen, M., Kruckow, S., Mairey, I. P., & Christensen, A. I. (2026). Design and characteristics of the national Danish injury cohort (NDIC). European Journal of Epidemiology. https://doi.org/10.1007/s10654-026-01413-2
Image Credits: AI Generated
DOI: 10.1007/s10654-026-01413-2
Keywords: Injury, Accident, Register-based, Cohort, Unnatural death, Denmark, Injury surveillance, Incidence rates, ICD-10, NCECI codes, National Danish Injury Cohort, Public health
Cite Scienmag News
APA MLA Chicago
Phoebe Ingram. (September 4, 2026). Denmark launches national injury cohort with detailed design features. Scienmag. https://scienmag.com/denmark-launches-national-injury-cohort-with-detailed-design-features/
Phoebe Ingram. “Denmark launches national injury cohort with detailed design features.” Scienmag, 4 September 2026, https://scienmag.com/denmark-launches-national-injury-cohort-with-detailed-design-features/. Accessed 4 September 2026.
Phoebe Ingram. “Denmark launches national injury cohort with detailed design features.” Scienmag. September 4, 2026. https://scienmag.com/denmark-launches-national-injury-cohort-with-detailed-design-features/
Copy citation Download RIS
Tags: administrative health data linkagecohort profile European Journal of EpidemiologyDenmark injury cohortindividual-level injury datainjury and violence surveillanceinjury burden assessmentinjury data collection methodsinjury epidemiologyinjury epidemiology Denmarkinjury mortality datainjury prevention researchinjury registry data integrationinjury surveillanceinjury-related hospital contactsinjury-related mortality datanationwide injury cohort Denmarknationwide injury registerpublic health injury monitoring
