Time to computed tomography for patients with severe acute neurological symptoms: a quality assurance study

Design and tuning

This was a retrospective, register-based quality assurance study conducted at Rigshospitalet Copenhagen, a tertiary teaching hospital with a level 1 trauma center and emergency department.

The study was approved by the management of Rigshospitalet hospital. According to Danish law, no ethics committee approval or informed consent is required for quality assurance studies. The study was conducted in accordance with the Declaration of Helsinki.

Patient selection

Activation of the acute medical team in our emergency department is based on prehospital findings, including resuscitated cardiac arrest, suspected ruptured aortic aneurysm, major bleeding, and suspected neurological conditions or non-trauma emergency neurosurgery. These time-critical neurological or neurosurgical conditions are classified as “Acute Serious Neurological Symptoms” in the ED Database. We included patients admitted to the Rigshospitalet emergency department with severe acute neurological symptoms between April 1st2016 and September 30e2020. We excluded patients without a valid social security number and patients transferred after primary evaluation to another hospital.

Data gathering

Data were extracted from the ED database containing patient record data of all patients admitted to the ED. Data is prospectively collected by the treating clinician using a data collection sheet and then stored in the approved research web application, Research Electronic Data Capture (REDCap).

Data extracted included age, gender, comorbidities (neurological comorbidity, obesity, hypertension), and medications (antihypertensives, anticoagulants). We defined neurological comorbidity as previously diagnosed SAH, ICH, ischemic stroke, epilepsy and sequelae of head trauma. Prehospital information data included symptoms and outcomes (paresis, seizures, vomiting, unequal pupil size, GCS, cardiac arrest, and vital signs) and interventions (hypertonic saline, hemostatic agents, and airway management). Admission data to our emergency department included time of arrival, time of CT scan, medical specialties present in the emergency department, and time of neurosurgical procedure (defined as time of incision). In addition, we collected information on International Classification of Diseases (ICD-10) discharge diagnoses, admission mortality, and 30-day mortality. For deceased patients, electronic patient records were reviewed for consideration of organ donation and for completed organ procurement. ICD-10 discharge diagnoses were subgrouped into SAH, HCI, ischemic stroke, brain tumor, seizure, headache, subdural hematoma (SDH), intoxication, infection, and “other” (Appendix 1).

If a discrepancy was found between the registry information and the electronic patient record information, the data was corrected based on the electronic patient record.


Continuous variables were reported as medians with interquartile ranges (IQR). Categorical data were reported as numbers (percentage) with 95% confidence interval (CI) where applicable. We have set a quality target that the time from hospital admission to brain scan should not exceed 30 minutes in more than 10% of patients. With four or fewer out of 100 patients exceeding 30 minutes before having a CT scan, we would be 95% confident that we had met our quality target.

In the analyses, we stratified patients by age: children (

Ethical approval and consent to participate

No informed consent or ethical approval is required for quality assurance projects.

Comments are closed.