Key points are not available for this paper at this time.
Design
Editorial
Answers are grounded in this paper and cite their sources.
Extrapolating risk scores like EuroSCORE II to fundamentally different patient cohorts, such as those with acute type A aortic dissection, is misleading and potentially harmful compared to using disease-specific tools like the GERAADA score.
NežIć et al. present a manuscript in which they apply 2 risk stratification systems to the prediction of outcomes of a single disease. Only one of these risk stratification systems is specifically designed for this disease. This system was generated from data of patients who had the disease, namely from the German Registry of Acute Aortic Dissection Type A (GERAADA) score, a unique tool for predicting 30-day mortality in patients undergoing surgery for acute type A aortic dissection [1, 2]. The score has already been externally validated and works exceedingly well [3]. The other tool, the EuroSCORE II, was designed for patients undergoing adult cardiac surgery, mainly for structural heart valve disease and/or coronary artery disease [4]. The authors adapt a different score to make it fit, scotomizing the background being generation out of data from patients who did not have the disease they want to address. The relationships presented are questionable and might even be dangerous: 1 example is the equation for extracardiac arteriopathy and cerebral malperfusion, which goes against the fundamental nature of the disease. Acute aortic dissection that causes cerebral malperfusion (usually highly symptomatic) and chronic obliterative arteriopathy (usually completely asymptomatic) are almost always mutually exclusive. Fabricating such a connection indicates a fundamental ignorance of the pathophysiology of the underlying disease processes. However, to show how you can make things fit, for reasons of curiosity, we took a very different score, the CHA2DS2-VASc-Score and simply extrapolated the given parameters using our own interpretation. Our goal was to make the score fit to predict mortality after acute type A aortic dissection. To achieve this end, we exchanged the given parameters with the desired parameters, e.g. stroke/transient ischaemic attack/thromboembolism by malperfusion (which was analogous to what the authors did for extracardiac arteriopathy) and the CHA2DS2-VASc-Score quickly turned into a tool to predict 30-day mortality for patients undergoing surgery for acute type A aortic dissection. There is only one catch: The score was not designed for these patients nor was it developed from data obtained from a group of patients who had the disease [5]. Summarizing, every effort to improve the accuracy of currently available risk prediction tools is of utmost importance, needed and welcome. However, extrapolation of a score that was developed for different reasons using a fundamentally different patient cohort is counterproductive, misleading and eventually even harmful to those who need our attention most, namely our patients. Our patients trust us to correctly interpret the pathophysiological process, accurately determine the treatment strategy and anticipate the risk that is mirrored by a maximum of informed consent that is uniquely substantiated by the prediction of 30-day mortality by the GERAADA score—generated from data of patients who actually had the disease. Conflict of interest: none declared.
No takes yet. Share an insight, caveat, or question.
Martin Czerny (2021) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: