Zero risk is not possible in emergency medicine

An impossible paradigm has been proposed by recent medico-legal cases in Australia. These have the potential to cause patient harm, inhibit good clinical practice and place a heavy burden on ED clinicians and the health system. The most fundamental principle in emergency medicine is identifying patients with critical illness who need urgent intervention; this frequently involves rapid clinical assessment with limited information. Not all serious illness can be easily identified by clinical assessment and basic investigations available at all times in EDs. Every test or intervention carries both a potential benefit and a risk, which varies with each patient. Knowing when to apply an intervention should be based, when possible, on strong research evidence. The article by Kelly et al. published in this issue demonstrates the conundrum faced by clinicians when seeing a patient with a rare condition that has potentially fatal consequences if missed.1 Aortic dissection (AD) usually (but not always) presents with chest or abdominal pain; the vast majority of patients with chest pain or abdominal pain do not have AD. The distinguishing features that make the diagnosis more likely are often absent and there are no validated clinical decision rules. The Australian medico-legal experience highlighted by Kelly et al. appears to be following that experienced in the USA. In a 12-year period, there were 150 medical malpractice cases involving AD, most for failure to diagnose and treat.2 Although the majority of cases were ruled in favour of the defendants, increased use of CT scans is suggested as a potential solution to reducing the incidence of missed ADs in the ED.3 AD is not alone, with similar conundrums faced in patients with ischaemic bowel presenting as abdominal pain, spinal cord compression presenting as back pain and subarachnoid haemorrhage presenting as headache. Liberal use of advanced imaging such as CT scans with contrast as a screening tool is fraught with risks, including malignancy,4 anaphylaxis and the physical and psychological trauma associated with incidental diagnoses of benign conditions resulting in further diagnostic tests. To chase all rare diagnoses that present with common symptoms will tip the risk–benefit ledger towards nett harm for most patients. Apart from individual patient harm, the systemic impact will be reduced equity of medical care, and many patients with stroke, trauma and other critical illnesses will have less access to CT scanning. It will also cause worse patient flow and bed access, particularly after hours. Also, each scan will generate hundreds of images, with and without contrast, that will demand expert reporting. If emergency clinicians are expected to have negative predictive values of 100% for all serious diagnoses among patients presenting to the ED, the impact on the health system, from the perspective of service provision and cost, will not be sustainable. The only logical solutions are the use of better screening tools and acceptance of some level of evidence-based risk. In the future, biomarkers for screening may offer a solution. For AD assessment, widespread use of the d-dimer levels is limited by low specificity, but newer biomarkers such as Aggrecan, IL1R1 and soluble ST2 have shown some potential in pre-clinical studies.5, 6 Other potential targets include biomarkers from the aortic extracellular matrix such as metalloproteinases, TGF-β, elastin fragments and those associated with vascular smooth muscle.7 Similarly, micro RNAs (non-coding, small RNA sequences that are stable and detectable in circulating blood) show promise, but all are a long way from demonstrated diagnostic utility in humans.8 As patient advocates, we cannot allow the current paradigm to continue. Our specialist colleges, the Australian Medical Association and patient consumer groups should advocate for common sense. Clinical pathways, based on available evidence should be agreed and followed across Australia, and adherence to such guidelines should define reasonable practice. Evidence-based guidelines for most serious conditions are available, and most accept that a small risk of delayed or missed diagnosis exists and may cause harm. Such adverse events are understandably distressing for the individual patient and their family, but the system should not be influenced into causing more harm. Clinicians following these evidence-based pathways should not be subject to inappropriate censure by courts, coroners or other external bodies who do not, or cannot, appreciate the impossibility of always being right and the impacts of chasing rare diagnoses. In this current climate, as described by Kelly et al., there is an even greater need for regular optimisation of clinical pathways, research to improve diagnostics and benchmarking of practice, funded and governed at a federal level. PAC, BM and AMK are section editors for Emergency Medicine Australasia and were excluded from the peer-review process and all editorial decisions related to the acceptance and publication of this article. Peer review was handled independently by members of the Editorial Board to minimise bias.