Data for validation and adjustment of APACHE II score in cardiogenic shock patients treated with a percutaneous left ventricular assist device

A precise prognosis is of imminent importance in intensive care medicine. This article provides data showing the overestimation of intrahospital mortality by APACHE II score in various subgroups of cardiogenic shock patients treated with a percutaneous left ventricular assist device. The data set includes additional baseline characteristics regarding age, pre-existing diseases, characteristics of coronary artery disease, characteristics of cardiopulmonary resuscitation, and hemodynamic parameter not included in the APACHE II score. Further data were provided which characterize derivation and validation group. Both groups were used for adjustment of the APACHE II approach. The data are supplemental to our original research article titled “Predictive value of the APACHE II score in cardiogenic shock patients treated with a percutaneous left ventricular assist device” (Mierke et al., IJC Heart & Vasculature. 40 (2022) 101013. https://doi.org/10.1016/j.ijcha.2022.101013).


a b s t r a c t
A precise prognosis is of imminent importance in intensive care medicine. This article provides data showing the overestimation of intrahospital mortality by APACHE II score in various subgroups of cardiogenic shock patients treated with a percutaneous left ventricular assist device. The data set includes additional baseline characteristics regarding age, preexisting diseases, characteristics of coronary artery disease, characteristics of cardiopulmonary resuscitation, and hemodynamic parameter not included in the APACHE II score. Further data were provided which characterize derivation and validation group. Both groups were used for adjustment of the APACHE II approach. The data are supplemental to our original research article titled "Predictive value of the APACHE II score in cardiogenic shock patients treated with a percutaneous left ventricular assist device" (Mierke et

Value of the Data
• The database offers baseline characteristics, different clinical parameters and outcome data of cardiogenic shock patients receiving left ventricular unloading with a micro-axial left ventricular assist device (pLVAD). • The database is useful for exact prediction of outcome in different subgroups of cardiogenic shock patients treated with pLVAD. • The dataset enables the validation of the adjusted Acute Physiology and Chronic Health Evaluation (APACHE) II score in other cohorts of cardiogenic shock patients treated with pLVAD. • Researchers with interest in pLVAD in cardiogenic shock can utilize database, combine it with others' datasets, and analyze them for further insights. • The dataset can be used for comparison with other cardiogenic shock cohorts treated with another pLVAD than the Impella CP®.

Data Description
We present data of 180 patients of Dresden Impella Registry with severe CS, who received left ventricular (LV) unloading with a microaxial percutaneous left ventricular assist device (pLVAD). We compared real-world mortality with the predicted mortality estimated by the Acute Physiology and Chronic Health Evaluation (APACHE) II score [1] . Table 1 shows the baseline characteristics of the patients displaying a typical distribution of cardiovascular risk factors found in developed countries. Acute myocardial infarction was the Atrial fibrillation / % (n) 26.1 (46) Valvular disease / % (n) 5.0 (9) PAD / % (n) 6.8 (12) Interventional complication / % (n) 3.9 (7) History of stroke / % (n) 8.0 (14) Heart rhythm disturbances / % (n) 2.2 (4) History of AMI / % (n) 17.6 (31) Post cardiac surgery / % (n) 1.1 (2) History of PCI / % (n) 25.0 (44) Takotsubo-CMP % (n) 1.1 (2) History of CABG / % (n) 4.0 (7) Other 1) / % (n) 1.1 (2) 1) Aortic dissection type ABMI, body mass index; PAD, peripheral arterial disease; AMI, acute myocardial infarction; PCI, percutaneous coronary intervention; CABG, coronary artery bypass graft; CPR, cardiopulmonary resuscitation; ICM, ischemic cardiomyopathy; CMP, cardiomyopathy. prevailing cause of cardiogenic shock and around 50% of patients received cardiopulmonary resuscitation. Table 2 presents the characteristics of coronary artery disease of the patients with cardiogenic shock complicating acute myocardial infarction (n = 120). Coronary three-vessel disease was prevailing in these patients, whereby most frequently culprit lesions of the left coronary system caused an acute myocardial infarction. Percutaneous coronary intervention was the predominant treatment.  Table 3 shows clinical parameters, that are not included in the APACHE II score but are known to influence outcome . Patients of the Dresden Impella Registry had high concentration of serum lactate and required intensive ionotropic and vasopressor support in the first 24 hours. The left ventricular ejection fraction was severely impaired before pLVAD. Table 4 compares APACHE II Score, length of hospital stays of survivors, observed mortality, and predicted mortality in different subgroups. The APACHE II score overestimated intrahospital mortality in nearly all sub-categories. The comparisons within the dichotomous or trichotomous subgroups showed no significant difference in observed mortality. The last column of the table displays the adjusted Diagnostic Category Weight, a specific constant for calculation of predicted mortality by the APACHE II score. The approach is described in detail by Mierke et al. [2] . Table 5 presents the baseline characteristics of derivation and validation group, which were obtained by random division o thef total study cohort. These groups were used for adjustment of Diagnostic Category Weight and its internal validation. The derivation and validation group showed well balanced baseline characteristics. Differences were only observed between body mass index and occurrence of peripheral arterial disease.
Individual raw data on outcome and on all measured parameters are listed in a supplementary Excel sheet.

Experimental Design, Materials and Methods
Clinical data were collected from the prospective Dresden Impella Registry during the period from February 2014 to May 2018. The predicted intrahospital mortality estimated by APACHE II score was calculated as described by Knaus et al. [1] and compared with the registry mortality. The comparison was performed conservatively by using Kaplan-Meier estimator at survivors' length of hospital stay. Patients who died intrahospital were excluded from the calculation of length of hospital stay. Receiver operating characteristics (ROC) analysis was performed to prove the accuracy of APACHE II score. In case of an overestimation of predicted mortality and an acceptable accuracy of APACHE II score, a specific constant (Diagnostic Category Weight) for calculation of predicted mortality was adjusted. For this purpose, the total study cohort was randomly divided into a derivation and a validation group. The derivation group was used for the calculation of the adjusted Diagnostic Category Weight (DCW). The observed mortality of the validation group was compared with predicted mortality calculated by adjusted Diagnostic Category Weight by using two approaches. First, goodness of fit was proved by the Hosmer-Lemeshow statistics. Second, differences between observed and predicted mortality by using either original DCW or adjusted DCW were compared. Finally, the adjusted DCW was cal- culated for every subgroup which showed a significant difference between observed and predicted mortality.

Ethics Statement
The study was performed in accordance with the Helsinki Declaration and local law and was approved by the institutional ethics committee of the Technische Universität Dresden (EK 457-122-014). All patients were adequately informed about the objective of the study and presented data are anonymized.