Skip to contents

introduction.Rmd

Skåne Emergency Medicine (SEM) Cohort Data Overview

Introduction

The Skåne Emergency Medicine (SEM) cohort is a comprehensive data platform established to support research and development of clinical decision support systems (CDSS) in emergency medicine. The cohort consists of 325,539 unique patients with 630,275 emergency department (ED) visits from January 1st, 2017 to December 31st, 2018 at eight EDs in the Skåne region of southern Sweden.

The SEM cohort provides a unique resource for emergency medicine research, with its combination of:

  • Rich clinical data from multiple health information systems

  • Comprehensive sociodemographic information from national registries

  • Longitudinal follow-up data on outcomes and healthcare utilization

  • Linkage to radiological, laboratory, and intensive care data

This vignette provides an overview of the SEM cohort data structure, sources, and potential applications.

Cohort Characteristics

The SEM cohort comprises all patients presenting at eight general EDs in the Skåne region during 2017-2018, with nearly 100% complete coverage (fewer than five patients declined participation). The cohort includes:

  • Population: 325,539 unique patients with 630,275 ED visits

  • Demographics: Mean age 55 years, 49% male

  • Clinical profile: 23.5% arrived by ambulance, most common triage category was Yellow (Priority 3)

  • Comorbidities: 11% had previous diagnoses of diabetes, 10% of cancer, 8% of pulmonary disease

  • Chief complaints: Most common were abdominal pain (10.7%), chest pain (8.1%), and dyspnea (6.1%)

  • Outcomes: 24.4% admitted to inpatient care, 0.9% mortality within 7 days, 2.2% within 30 days

The cohort is representative of a general emergency department population in Sweden, though with limited representation of psychiatric, obstetric/gynecologic, ophthalmologic, and pediatric patients (without orthopedic problems) as these conditions are typically treated at specialized EDs in the region.

Data Organization

The SEM cohort data is organized into three temporal categories that provide a comprehensive view of patients’ healthcare journey:

1. Pre-SEM Data

Data collected before patient inclusion in the SEM cohort, providing context for their clinical presentation:

  • Historical diagnoses: Diagnoses recorded within 5 years prior to SEM contact

  • Previous medications: Medication prescriptions before SEM contact

  • Prior healthcare utilization: Previous hospital admissions and specialized care

2. SEM Cohort Data

Data collected during the index SEM cohort contact/hospitalization period:

  • ED visit information: Chief complaints, triage levels, timing metrics

  • Clinical data: Diagnoses, procedures, laboratory tests

  • Medications: Prescriptions and administrations during the visit

  • ED outcomes: Disposition decisions, admissions, diagnoses

3. Post-SEM Data

Data collected after discharge from the index SEM cohort contact, capturing outcomes and follow-up:

  • Return visits: Subsequent ED presentations and hospital admissions

  • Post-discharge diagnoses: New diagnoses recorded after the index visit

  • Interventions: Procedures performed after discharge

  • Mortality: Death dates and causes

  • Healthcare utilization: Inpatient stays, hours of care, costs

Data Sources

The SEM cohort integrates data from multiple sources to provide a comprehensive view of patients’ healthcare journeys:

Hospital Information Systems (Region Skåne)

  • Melior: Electronic health record system capturing diagnoses, interventions, and clinical documentation

  • Patientliggaren: ED patient registry tracking patient flow, triage, and disposition

  • SVAR: Swedish Emergency Care Registry with vital signs and additional clinical parameters

  • SECTRA: Radiology information system with examination details and reports

  • PASIVA: Intensive care unit registry with ICU admission details

  • MUSE: ECG database containing electrocardiogram recordings and analyses

  • Laboratory systems: Clinical chemistry and microbiology test results

National Registries (Sweden)

  • National Patient Register: Inpatient and specialized outpatient care contacts

  • Prescribed Drug Register: Dispensed medications from pharmacies

  • Cause of Death Register: Mortality data and causes

  • Statistics Sweden (SCB): Socioeconomic and demographic information

    • LISA Database: Education, income, employment status

    • DeSO: Geographic residence data

    • Birth Country Data: Migration background information

Data Linkage and Integration

All datasets in the SEM cohort are linked through common identifiers:

  • patient_id: Pseudonymized identifier for linking all patient-level data

  • contact_id: Identifier for specific healthcare encounters

This linkage enables researchers to:

  1. Follow individual patients across different care settings

  2. Connect pre-hospital factors with in-hospital processes and post-discharge outcomes

  3. Integrate clinical data with socioeconomic context

  4. Create comprehensive patient profiles incorporating multiple data dimensions

For more detailed information about specific data components, refer to:

Conclusion

The SEM cohort represents a comprehensive platform for emergency medicine research, connecting patient characteristics, clinical presentations, care processes, and outcomes across multiple data sources. By providing detailed information on over 630,000 ED visits, the SEM cohort enables researchers to develop improved decision support tools and gain insights into emergency care practices and outcomes.

The cohort’s main strengths include its population-based nature, rich clinical detail, socioeconomic context, and longitudinal follow-up. These characteristics make it a valuable resource for diverse research objectives, from algorithm development to quality improvement and health policy evaluation.

References

Ekelund, U. et al. The skåne emergency medicine (SEM) cohort. Scand. J. Trauma Resusc. Emerg. Med. 32, 37 (2024).