Finding the holy grail is not a short-term project.

C. Atzema,M. Schull
DOI: https://doi.org/10.1161/CIRCOUTCOMES.115.001611
2015-03-01
Abstract:Clinical decision instruments exist for a myriad of disease states as they present to the emergency department (ED), including cervical spine imaging rules,1,2 instruments that estimate the risk of venous thromboembolism,3 and the mortality risk associated with pneumonia,4 to name a few. As the second most common reason to visit an ED,5 chest pain has been the target of decades of clinical decision instrument research efforts. Article see p 195 Given that less than a quarter of these patients are suffering from an acute coronary syndrome,6 multiple groups have attempted to derive a highly sensitive decision instrument that identifies essentially all the patients with an acute coronary syndrome, while at the same time facilitating the safe, early discharge of a cohort of low-risk patients.7–10 Emergency physicians have identified 1% as the maximum miss rate that would acceptable to them in order to actually use such a tool in the clinical setting.11,12 A tool that has a sensitivity with a lower confidence interval of 99%, but still has a high enough specificity to identify a substantial proportion of chest pain patients for ED discharge would be the holy grail of clinical decision instruments. Early instruments had poor clinical uptake because of unacceptably low sensitivity: these include the Goldman Risk score, acute cardiac ischemia time–insensitive predictive instrument (ACI-TIPI), the Thrombolysis in Myocardial Infarction (TIMI) risk score, and Global Registry of Acute Coronary Events (GRACE).13–16 More recently, the North American Chest Pain Rule, the Accelerated Diagnostic Protocol to Assess Patients With Chest Pain Symptoms Using Contemporary Troponins (ADAPT) study, and the HEART (History, Electrocardiogram, Age, Risk factors, Troponin) Pathway, among others, have incorporated conventional troponins into their clinical decision instruments.7,8,10 The North …
What problem does this paper attempt to address?