Oxygen supply and oxygen demand are mismatched in cardiogenic shock (CS) as the most severe manifestation of acute heart failure due to an impaired cardiac output (CO). Besides contractile dysfunction of the left, right or both ventricles, a severe end-organ hypoperfusion results in multi-organ failure syndrome occasionally accompanied by a systemic inflammatory response [1].

Acute coronary syndrome (ACS) is the underlying major pathology of CS (up to 80% of the cases) [2]. Furthermore, up to 13% of acute myocardial infarction (AMI) manifests with CS [3]. Rare findings are mechanical complications of AMI, such as ventricular septum defect (VSD), rupture of the free wall or papillary muscle rupture with subsequent severe acute mitral regurgitation. Recent papers hint at present high numbers of non-ischemic shock scenarios (up to 52%)[4]. They may result from pulmonary embolism, pericardial tamponade, myocarditis, arrhythmia, valvular disease, decompensated congestive heart failure or other cardiomyopathies (peripartal, autoimmune, stress induced) [5, 6].

Resolving the underlying cause in acute myocardial infarction complicated by CS is crucial: a revascularization delay of only 10 min results in significantly higher death rates (3.3 additional deaths for every 10 min delay) [7, 8]. The advances in survival (from formerly 30% to nowadays about 60%) [9] can clearly be attributed to rapid percutaneous revascularization, enhanced revascularization techniques (culprit lesion only strategy [10, 11], radial access site [12]), and highly effective platelet therapy—conforming to current European Guidelines [13, 14].

Differential inotropic strategies (dobutamine, dopamine, adrenalin) were adopted to oppose refractory low CO, but resulted in higher mortality rates, presumably due to higher oxygen consumption [15].

The goal of short-term MCS is to assist or to take over CO, and to sufficiently sustain organ perfusion, ideally without further harming the heart. The damaged heart is set at rest until recovery or until a permanent solution is obtained (i.e., LVAD or heart transplantation).

Sophisticated technologies to maintain temporary cardiac salvage mainly consist of percutaneous vascular access and different intra- or extracorporeal pump systems, allowing rapid access and a broad use [16].

IABP was the first broadly used support system. With technical progress, VA-ECMO became usable outside of heart surgery settings in the intensive care units and recently as resuscitation tools for refractory out-of-hospital cardiac arrest.

MCS strategies have evolved over the past two decades to act as bridging instruments in acute heart failure, either as bridges to recovery/decision, bridges to LVAD or heart transplantation, and in rare cases after extracorporeal CPR bridge to organ donation.

Guidelines do not recommend IABP therapy in CS after AMI following two landmark studies discouraging a broader use of IABP in these patients [14, 17, 18].

A small prospective and a larger retrospective study comparing Impella (pVAD) with IABP-guided therapy in cardiogenic shock showed comparable mortality rates (50 vs. 46%; p = 0.92 and 48.5 vs. 46.4%, p = 0.64). A higher incidence of bleeding and vascular entry site lesions complicated Impella therapy in both studies [19, 20].

VA-ECMO is increasingly adopted in Germany despite the lack of evidence and neutral IABP studies [21].

Mortality data are derived from registries such as the Extracorporeal Life Support Organization (ELSO), always inheriting the risk of selective reporting. Long-term data on quality of life after CS and MCS are scarce but promising [22‐24].

In this manuscript, we present the incidence and mortality of CS in Germany over 10 years considering different treatment protocols and competing MCS systems in the face of poor prior evidence.

Since 2005, data on all hospitalizations in Germany have been available for scientific use via the diagnosis-related group (DRG) statistics collected by the Research Data Center of the Federal Bureau of Statistics (DESTATIS). The World Health Organization published the 10th revision of the international statistical classification of diseases and related-health problems (ICD) in January 1998. These hospitalization data, including diagnoses and procedures, are a valuable source of representative nationwide data on in-hospital treatment of patients.

This database represents a virtually complete collection of all hospitalizations in German hospitals that are reimbursed according to the DRG system. From this database, we extracted data on all patients that were hospitalized between 2007 and 2017 with documented cardiogenic shock (ICD-10 code R57.0 as main or secondary diagnosis). Utilization of mechanical circulatory support was identified using the German Procedure Classification/OPS code 8-83a3* (pVAD/Impella), 8-83a0* (IABP), and 8-8523 (VA-ECMO).

Our study did not involve direct access by the investigators to data on individual patients, but only access to summary results provided by the Research Data Center. Therefore, approval by an ethics committee and informed consent were determined not to be required in accordance with German law. All summary results were anonymized by DESTATIS. In practice, this means that any information used in reaching conclusions regarding a single patient or a specific hospital is censored by DESTATIS to guarantee data protection. Especially, the use of the anonymous, persistent ‘institute indicator of hospitals’ is highly restricted in order not to publish any information directly attributable to a single hospital.

The primary outcome was in-hospital mortality which is part of DESTATIS own set of variables. A number of patient characteristics (defined by Reinöhl et al. [25]) are shown in the supplemental appendix. Incidence of cardiogenic shock per 100,000 person-years was approximated by dividing the number of patients with CS by the number of inhabitants of Germany in the respective year according to the Federal Statistical Office of Germany´s estimate [26]. Differences in in-hospital mortality between groups were calculated with the Chi-square test. Trends in in-hospital mortality over time were estimated by means of logistic-regression analyses with time as the sole covariate. Analyses were carried out with the use of Stata software, version 16 (StataCorp, College Station, USA) and Prism, version 5 (GraphPad, San Diego, USA).

We provide epidemiological descriptive data on the incidence, real-life usage of different MCS systems and survival in CS over the last decade in a European high-income country. To note, survival data cannot be derived for evidence for superiority or minority of different MCS systems.

We present data derived from the German governmental coding system. By nature of this study, our data are descriptive. While data for reimbursement-relevant characteristics like MCS usage are highly accurate, non-reimbursement-relevant characteristics like accompanying baseline characteristics might be underreported. The superiority of one MCS device over another or over medical therapy cannot be derived from data presented. Also, we cannot compare patient groups (with and without different MCS devices) to another since a significant selection bias has to be presumed. An adjustment of potential confounders cannot be reasonably performed, since data on disease severity at the time of the treatment decision (like lactate; catecholamine treatment; time, duration and quality of CPR with respect to MCS implantation; experience of the physician and his team; patient wish; necessity of pulmonary support with respect to MCS implantation; heart rhythm; PCI success rate or complications; etc.) are not supplied by the DESTATIS dataset. In the dataset supplied used for this research, the cause of CS is not encoded. We therefore cannot differentiate between different subcategories of CS. Also, we have no information concerning the rationale for using a specific MCS. Thus, we can only speculate why different devices were implanted. Different protocols may have led to a selection bias for or against a given MCS. Maybe, only one method was available in some settings or operators were only trained in one method. As reimbursement differs between the support devices, we cannot exclude a bias by economic considerations. We, further, cannot identify combination therapies or subsequent use of different MCS, which may bias survival data. Results may differ in other high-income countries due to differences in the available intensive care resources or reimbursement situations. The etiology of cardiogenic shock and the supporting medical therapy like inotropic therapy cannot be characterized. The data do not allow us to draw any conclusions regarding long-term survival or quality of life.

The reported incidence of CS is significantly increasing. Hospital survival rate is low and stagnant. Only a minority of patients with CS is managed with MCS. The type of MCS device used in CS has shifted from the exclusive use of IABP to a variety of pVAD, VA-ECMO and IABP and combination therapies. Now, the use of various devices results in different survival rates. This should inspire further investigation to deliver the most suitable technical option with the best chances of survival for different etiologies of CS.