Our findings have implications for guideline developers, clinicians, service managers and research funders.
Guideline developers
The ADD-RS and D-dimer, alone and in combination, provide useful diagnostic information and could be included in guidelines, but specific guidance needs to take into account the role of clinical judgement in determining whether AAS is suspected in people presenting with possible symptoms. Using the ADD-RS and/or D-dimer to select patients for CTA from an unselected population is likely to lead to high rates or CTA.
If clinical judgement can identify patients with higher AAS prevalence, then a strategy based on the Canadian guideline or a strategy of CTA if ADD-RS is greater than zero or D-dimer > 500 ng/ml appears to be cost-effective and could be recommended in guidelines. These strategies could also be used in guidelines to reduce the number of CTA performed, albeit with a small increase in the risk of missed AAS. The role of clinical judgement emphasises the need for clinician training and awareness of AAS. Our findings suggest no role for clinical scores other than the ADD-RS or for biomarkers other than D-dimer in guidelines.
Clinicians
The ADD-RS and D-dimer, alone and in combination, provide useful diagnostic information. Clinicians can use the estimates of diagnostic accuracy from our meta-analysis in their diagnostic assessment of suspected AAS. These estimates are based on reasonably robust evidence for cohort studies.
The main problem with using the ADD-RS and D-dimer in diagnostic assessment is that they are likely to generate a high rate of FPs if used indiscriminately in patients with a very low risk of AAS, which will presumably then require potential unnecessary CTA. Clinicians should, therefore, not use the ADD-RS and D-dimer in diagnostic assessment unless they consider there to be a realistic possibility of AAS. In these circumstances, the ADD-RS and D-dimer can be used to select patients for CTA. Using the ADD-RS and D-dimer is also inappropriate when the diagnosis of AAS is obvious or the patient is critically ill and urgent definitive diagnostic assessment is required.
We have published a British Medical Journal ‘Easily missed’ paper that describes how clinicians can use the findings reported in this synopsis to improve their diagnostic assessment of suspected AAS.132
Service managers
The base-case analysis suggested that applying the strategies to an unselected population would miss more than half of AAS cases or would require a three- to fourfold increase in CTA capacity (from 298 to 1079 CTA per year), which is unlikely to be deliverable. To provide some context, in 2023, 17,491 patients received 20,378 CT scans at the Royal Infirmary of Edinburgh, while 15,399 CT scans were requested from the Northern General Hospital Emergency Department in Sheffield. Most of these were non-contrast CT scans of the brain, so unselective implementation of the strategies would markedly increase requirements for CT with contrast.
If clinicians can select a population with higher prevalence, then cost-effective strategies using ADD-RS and D-dimer would be potentially deliverable, albeit with a potential increase in demand for CTA. A strategy using ADD-RS > 1 or D-dimer > 500 ng/ml could be cost-effective and reduce CTA use among those currently receiving CTA, but would be difficult to implement without also being applied to patients who do not currently receive CTA.
Research funders
Diagnosis of AAS is currently the seventh top priority in the James Lind Alliance Emergency Medicine Priority Setting Partnership (https://rcem.ac.uk/research-priorities/) and third top priority in the James Lind Alliance Aortic Priority Setting Partnership (www.jla.nihr.ac.uk/priority-setting-partnerships/vascular-conditions/aortic-top-10-priorities.htm). Our evidence synthesis has identified areas of uncertainty in current knowledge and informed the research recommendations outlined in Research recommendations.