Comparison of in-hospital mortality and 30-day mortality in cardiac surgery

I read with great interest the article by Santarpino et al. [1] comparing the age, creatinine and ejection fraction II risk score (ACEF II) and the European System for Cardiac Operative Risk Evaluation II (EuroSCORE II) to predict 30-day mortality in cardiac surgery (cohort of 14804 patients operated on at 8 Italian cardiac surgery centres in the period 2009–2019). The ACEF II [2] has recently been introduced as an updated version of original ACEF score [3] including risk prediction for patients undergoing emergency surgery and with preoperative anaemia. First of all, I absolutely disagree with the author's [1] claim that the previous version of ACEF [3] score has been included in the latest (2018) guidelines of the European Society of Cardiology and the European Association for Cardio-Thoracic Surgery (2018 ESC/EACTS Guidelines on myocardial revascularization) as a risk stratification tool for surgical and percutaneous myocardial revascularization [4]. The ACEF score [3] has not been mentioned anywhere in the text of 2018 ESC/EACTS Guidelines on myocardial revascularization [4], either was quoted between basic 786 references [4] or in 2018 ESC/EACTS Guidelines on myocardial revascularization—Supplementary Material, including Supplementary Material, Table S4, and references for material in the Supplementary Material, Appendix [4]. The EuroSCORE II risk scoring system predicts in-hospital mortality in cardiac surgery. Although authors [1] defined operative mortality as death occurring up to 30 postoperative days (but they also wrote that for the generation of receiver operating characteristic curves, observed in-hospital mortality was used as end point for testing the discrimination of the score [1]), there is a theoretical possibility that comparing data of in-hospital mortality only, for both scores, could offer a more accurate comparison between scores. Namely, Siregar et al. [5] reported that the in-hospital and the 30-day mortalities do not refer to the same patients, as 20% of the patients counted in each measure are not included in the other measure. Thus, they concluded that in-hospital mortality and 30-day mortality measure two different types of end point and are not interchangeable [5]. Hickey et al. [6] wrote: ‘Assessing a risk prediction model “off label” should be explicitly stated, and conclusions should not mislead readers that the performance was for the intended outcome. For example, the EuroSCORE II was developed to predict in-hospital mortality. “Validating” the model in a sample of patients with the outcome of 30-day mortality would be “off label”, and so caution should be taken’. Even the authors [1] mentioned that problem in the Limitation section of their manuscript. The outcome variable in ACEF II study [2] was the operative mortality, defined as in-hospital mortality at any time or mortality after discharge but within 30 days from surgery. I assume that authors [1] have separate data for in-hospital and for 30-day mortality. Therefore, it would be of great interest if authors [1] are able to supplement their results comparing and validating EuroSCORE II and ACEF II testing only sample containing data for in-hospital mortality. Conflict of interest: none declared.