BACKGROUND: Very little evidence is available on the prevalence of serious spinal pathologies and the diagnostic accuracy of red flags in patients presenting to the emergency department (ED). This systematic review aims to investigate the prevalence of serious spinal pathologies and the diagnostic accuracy of red flags in patients presenting with low back pain to the ED.
METHODS: We systematically searched MEDLINE, PUBMED, EMBASE, Cochrane Library, and SCOPUS from inception to January 2019. Two reviewers independently reviewed the references and evaluated methodological quality.
RESULTS: We analyzed 22 studies with a total of 41,320 patients. The prevalence of any requiring immediate/urgent treatment was 2.5%-5.1% in prospective and 0.7%-7.4% in retrospective studies (0.0%-7.2% for vertebral fractures, 0.0%-2.1% for spinal cancer, 0.0%-1.9% for infectious disorders, 0.1%-1.9% for pathologies with spinal cord/cauda equina compression, 0.0%-0.9% for vascular pathologies). Examples of red flags which increased the likelihood for a serious condition were suspicion or history of cancer (spinal cancer); intravenous drug use, indwelling vascular catheter, and other infection site (epidural abscess).
CONCLUSION: We found a higher prevalence of serious spinal pathologies in the ED compared to the reported prevalence in primary care settings. As the diagnostic accuracy of most red flags was reported only by a single study, further validation in high-quality prospective studies is needed.
This review constitutes a good compendium of published studies bearing on the differential diagnosis in patients presenting to the ED with LBP. The inference regarding the importance of a number of 'red flag' elements of history or physical is clearly flawed in that the assumption is being made that these elements are to be taken in isolation without consideration of others, such as neurological symptoms or deficits. In fact, taken in context, any 'red flag' element with a LR + >1 should be taken very seriously.