Original Research

Is computed tomography of the head justified in patients with minor head trauma presenting with Glasgow Coma Scale 15/15?

Chuma Singata, Sally Candy
South African Journal of Radiology | Vol 22, No 1 | a1329 | DOI: https://doi.org/10.4102/sajr.v22i1.1329 | © 2018 Chuma Singata, Sally Candy | This work is licensed under CC Attribution 4.0
Submitted: 20 February 2018 | Published: 27 September 2018

About the author(s)

Chuma Singata, Department of Radiology, Faculty of Health Sciences, University of Cape Town, South Africa
Sally Candy, Department of Radiology, Groote Schuur Hospital, University of Cape Town, South Africa, South Africa


Share this article

Bookmark and Share

Abstract

Background: In keeping with radiology departments in tertiary referral hospitals in developing countries offering computed tomography (CT) head scan services, the radiology department at Groote Schuur Hospital (GSH) in the Western Cape of South Africa undertakes several such scans annually. Of these scans, many are undertaken for post-trauma patients with minor head injury (MHI). While there is agreement that MHI patients with Glasgow Coma Scale (GCS) scores of 13–14/15 may well benefit, there is doubt as to the clinical utility of routine CT head scanning in MHI patients with GCS scores of 15/15.

Objectives: This retrospective descriptive study of patient records was undertaken to determine the frequency and clinical significance of any abnormalities found on CT head scans of 460 patients with MHI and GCS scores of 15/15, scanned at GSH between 2012 and 2014.

Method: Ethical clearance was obtained and the records of 460 MHI patients with GCS scores of 15/15, loss of consciousness (LOC) and amnesia who underwent CT head scanning at GSH between 2012 and 2014 were then retrieved from the Philips picture archiving and communication system (PACS). Patient records, containing illegible referral forms or technically inadequate CT head scans, were excluded from the study. Patients’ biographical, clinical and CT head scan data were entered into sequentially numbered data collection forms. These data were tabulated and summed as percentage distributions. Patients’ CT head scan findings were reviewed and classified as either showing normal or abnormal features. Abnormalities detected on CT head scans were classified as being either clinically significant or clinically non-significant.

Results: Referral forms and CT scan reports were obtained for 460 MHI patients from a sample of 497 patients, calculated by using the equation for estimating a single proportion from a large sample (precision 1.5%). The sample obtained yielded an acceptable response rate of 460/497 (92.6%). Of 460 (100%) scan reports, 320 (69.6%) showed no abnormality, while 140 (30.4%) showed abnormality. Of the 140 abnormal scans, 107 (23.3%) showed clinically non-significant abnormality, while 33 (7.2%) revealed clinically significant abnormality. Twenty-two (4.8%) of these clinically significant scans showed brain contusion and 11(2.4%) showed skull fracture. No subdural or extradural haematoma, shift or herniation were reported and none of the 33 patients whose CT scans showed clinically significant abnormality underwent urgent neurosurgical intervention.

Conclusion: Of the 460 CT head scans performed at GSH for MHI with LOC but normal GCS between 2012 and 2014, none required urgent neurosurgical intervention. This is in accordance with the 2012 Kimberley Hospital Rule (KHR), a management protocol which indicates that CT head scanning in patients with MHI and GCS scores of 15/15 can safely be delayed for 8 h. An audit of the records of patients excluded from this study coupled with an analysis of data from other Western Cape hospital CT head scan databases could help ensure that this scarce resource is used cost-beneficially for all head-injured patients in the Western Cape catchment area.


Keywords

CT head scans; computed tomography; minor head trauma

Metrics

Total abstract views: 125
Total article views: 560


Crossref Citations

No related citations found.