Disparities in the management of cardiogenic shock complicating acute myocardial infarction: a questionnaire survey in Beijing

Abstract

Objective

To investigate disparities in the management of cardiogenic shock complicating acute myocardial infarction (AMI-CS) across different hospital levels in Beijing.

Methods

A standardized questionnaire was electronically distributed to 61 hospitals through the Beijing Cardiovascular Internal Medicine Quality Control and Improvement Center and the Beijing Extracorporeal Life Support Quality Control and Improvement Center. A multistage sampling method was also used to survey hospitals in eight Beijing districts that were not included in these quality control networks. Hospitals were categorized as Level 1 hospitals, Level 2 hospitals, or Level 3 hospitals according to international standards and the domestic context.

Results

A total of 142 hospitals were analyzed. All participating hospitals were capable of managing cardiovascular diseases. 33.1% of hospitals did not have percutaneous coronary intervention (PCI) capability; 40.8% provided only intra-aortic balloon pump and PCI, whereas 26.0% offered temporary or durable mechanical circulatory support. Level 3 hospitals often lacked established protocols for AMI-CS management. Level 1 hospitals (59.4%) and Level 2 hospitals (56.5%) were more likely to maintain a minimum mean arterial pressure target of 65 mmHg, whereas Level 3 hospitals more commonly accepted a lower target (45.5%). Pulmonary artery catheter monitoring was infrequently used across all hospital levels. Continuous lactate monitoring also differed significantly (Level 3, 39.1% vs. Level 2, 74.6% vs. Level 1, 84.4%; overall p < 0.001). Among the 61 hospitals participating in the quality control network, only 3.6% adopted the Society for Cardiovascular Angiography and Interventions shock stage classification. Shock teams were available in 71.9% of Level 1 hospitals and 44.4% of Level 2 hospitals. Furthermore, only 31.9% of Level 3 hospitals had established referral networks.

Conclusion

Significant disparities exist in the diagnosis and management of AMI-CS across different hospital levels in Beijing. There is an urgent need to enhance the understanding and management of AMI-CS, especially in Level 2 and Level 3 hospitals.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12872-026-06287-6.

Introduction

Cardiogenic shock (CS) is a clinical syndrome characterized by impaired tissue perfusion caused by cardiac dysfunction and inadequate cardiac output [1]. CS has various etiologies, the most common being acute myocardial infarction complicated by CS (AMI-CS), which carries a mortality rate of 40%–50% [2]. Other causes include acute decompensated chronic heart failure, fulminant myocarditis, severe valvular heart disease, and stress-induced cardiomyopathy [3]. The management of CS should be individualized according to its underlying etiology [4, 5].

The 2025 ACC expert consensus statement on CS evaluation and management is based on evidence-based medicine and emphasizes multidisciplinary collaboration. It provides standardized treatment pathways, a tiered care framework, and a regional transfer system. The consensus addresses evidence gaps in the heterogeneous management of CS and offers specific diagnostic and therapeutic recommendations [6]. International cross-sectional surveys have demonstrated substantial variation in CS management among countries, hospitals, and clinicians, identifying challenges including unequal resource distribution, transfer barriers, and insufficient multidisciplinary teams and standardized care pathways, thereby emphasizing the importance of regional care network [7–9]. These studies provide practical evidence for optimizing diagnostic and therapeutic strategies for AMI-CS.

However, no studies have evaluated the distribution of medical resources, quality control, or AMI-CS management capabilities in China, where the burden of cardiovascular diseases continues to increase. Beijing’s advanced healthcare system provides an important model for CS management and may help establish benchmarks for other regions. Therefore, investigating CS management in Beijing is an essential first step toward developing national treatment strategies and improving clinical outcomes. Questionnaire-based data provide an important foundation for future research. This study aimed to assess AMI-CS treatment capacity and management practices across different levels of medical institutions involved in CS care in Beijing.

Methods

Design of the survey

From September to November 2024, the questionnaire was developed based on a previously published international survey on CS management [10], considering local clinical practice and study objectives (Supplementary Material 1). The questionnaire was reviewed and revised by an expert panel, comprising specialists in critical care and emergency medicine. The survey contained 44 questions across five sections: respondent characteristics, hospital characteristics, shock management, coronary revascularization, and mechanical circulatory support (MCS). The questionnaire also addressed diagnostic practices, emergency referral and transfer pathways, as well as multidisciplinary team collaboration. These data were collected to provide a comprehensive evaluation of the current status and inter-hospital differences in AMI-CS management across Beijing.

Hospital sampling and survey distribution

A cluster sampling approach was used to collect data from 61 hospitals participating in the Beijing Cardiovascular Internal Medicine Quality Control and Improvement Center and the Beijing Extracorporeal Life Support Quality Control and Improvement Center. These institutions represent most hospitals in Beijing with capabilities in emergency revascularization, MCS, and multidisciplinary collaboration. The questionnaire was distributed online via Wenjuanxing (www.wjx.cn; Changsha Ranxing Information Technology Co., Ltd., China) to hospitals participating in the quality control network.

To increase sample diversity across hospital levels, a multistage sampling method was further applied. Eight administrative districts in Beijing were randomly selected, and 8–12 hospitals with cardiovascular care capacity that were not included in the quality control network were randomly selected from each district. Questionnaire data from 89 hospitals were collected during this stage.

For hospitals within the quality control network, questionnaires were completed by personnel responsible for cardiovascular disease quality control. For hospitals outside the network, questionnaires were distributed to emergency departments and completed by physicians with experience in cardiovascular emergency care, attending physician rank or higher, and familiarity with departmental CS management.

The exclusion criteria were as follows:

1. Hospitals exclusively providing pediatric care;

2. Hospitals lacking laboratory tests necessary for AMI diagnosis;

3. Hospitals without an emergency department;

4. Hospitals with questionnaires containing substantial missing data.

Hospital classification and definitions

According to international standards and national practice [11, 12], hospitals were categorized as Level 1 (hub) hospitals (n = 32), Level 2 (spoke) hospitals (n = 63) or Level 3 (spoke) hospitals (n = 47) based on institutional capacity and their role within the regional CS care network (Supplemental Table S1).

Level 1 hospitals were defined as institutions with a cardiac catheterization laboratory, emergency cardiac surgery operating room with surgeon, cardiac intensive care unit or cardiac care unit, temporary MCS (tMCS) capability, referral and transfer support, and the capacity to provide comprehensive care from resuscitation and active treatment to recovery or end-of-life/palliative care.

Level 2 hospitals were defined as institutions capable of performing percutaneous coronary intervention (PCI) and intra-aortic balloon pump (IABP) implantation but without advanced MCS devices. Level 3 hospitals were defined as institutions without PCI or MCS capability but with basic cardiovascular care resources.

Sample and administration

The sample size was calculated using the standard formula for cross-sectional surveys, n = [Z² × p × (1 − p)] / e², where Z = 1.96 (corresponding to a 95% confidence level), p = 0.75 based on previous reports that 75% of clinicians may not accurately identify CS, and e = 0.05 as the acceptable margin of error. The initial sample size was 288. The estimated sample size was further adjusted using the finite population correction formula, m = nN / (N + n − 1), where N = 202 represented the total number of hospitals in Beijing with cardiovascular care capacity according to the 2023 Beijing Cardiovascular Internal Medicine Quality Control Expert Working Meeting. After finite population correction, the adjusted sample size was 119. After accounting for an anticipated response rate of 90%, the final target sample size was set at 133 hospitals.

Statistical analysis

Categorical variables were presented as frequencies and percentages, and the Wilson score method was used to calculate 95% confidence intervals for proportions. Differences were analyzed using the chi-square test or Fisher’s exact test, with Bonferroni correction for multiple comparisons. Ordinal variables were compared using the Kruskal–Wallis H test. Multiple-response analysis was applied to multiple-choice questions. Differences in CS management were evaluated among Level 1 hospitals, Level 2 hospitals, and Level 3 hospitals. Statistical analyses were performed using SPSS version 26.0 (IBM Corp, Armonk, NY, USA), GraphPad Prism version 10.1.2 (GraphPad Software, Boston, MA, USA), and R version 4.5.1 (http://www.R-project.org). A p value < 0.05 was considered statistically significant.

Result

Respondents’ specialties and the MCS capabilities of their affiliated hospitals

A total of 150 questionnaires were distributed and returned, yielding a response rate of 100%. After excluding eight questionnaires with > 20% missing required items, 142 questionnaires were included in the final analysis. Of the affiliated hospitals, 61 were part of the aforementioned quality control network, whereas 81 were not. All included hospitals had emergency departments and were capable of providing initial assessment and stabilization for patients with suspected AMI-CS.

Among respondents, 35.2% specialized in general cardiology, 21.1% in critical care medicine, 15.4% in emergency medicine, 9.2% in interventional cardiology, 3.5% in heart failure, 1.4% in cardiothoracic surgery, and 14.1% in other specialties (Fig. 1A). Physicians from these specialties were familiar with emergency care processes and participated in the management of patients with suspected CS through urgent consultations, multidisciplinary collaboration, and fast-track pathways when specialized evaluation or treatment was required. Regarding MCS capabilities, 33.1% of hospitals did not have PCI capability, 40.8% had only IABP and PCI capability, 11.3% had multiple tMCS devices and PCI capability, 7.7% could provide all forms of MCS (both durable and temporary MCS devices), but could not perform heart transplantation, and 7.0% possessed all the above capabilities, including heart transplantation (Fig. 1B).

Fig. 1.

Respondents’ specialties and the MCS capabilities of their affiliated hospitals. A Distribution of respondents by clinical specialty. B Availability of mechanical circulatory support for AMI-CS at the respondents’ hospitals. AMI-CS, acute myocardial infarction complicated by cardiogenic shock; PCI, percutaneous coronary intervention; IABP, intra-aortic balloon pump; tMCS, temporary mechanical circulatory support; MCS, mechanical circulatory support

Diagnostic approach for CS

Level 3 hospitals were more likely to diagnose CS based solely on symptoms and physical signs (38.3%). In contrast, Level 2 and Level 1 hospitals more frequently used comprehensive diagnostic strategies incorporating symptoms, physical signs, metabolic indicators, echocardiography, and invasive measurements, with rates of 49.2% and 77.4%, respectively. Diagnostic approaches differed significantly among hospital levels (overall P < 0.001; Fig. 2). Bonferroni-adjusted pairwise comparisons demonstrated significant differences in diagnostic strategies between Level 3 and Level 2 hospitals and between Level 3 and Level 1 hospitals (Supplemental Table S5).

Fig. 2.

Diagnostic approaches for cardiogenic shock. A Diagnostic strategies used for cardiogenic shock across all hospitals. B Differences in diagnostic strategies according to hospital level

Clinical management of CS

Dopamine was the most commonly selected vasoactive/inotropic agent in Level 3 hospitals (46.8%), followed by norepinephrine (29.8%) and epinephrine (23.4%). In Level 2 hospitals, norepinephrine was the preferred agent (66.7%), followed by dopamine (20.6%) and epinephrine (12.7%). Norepinephrine was also the most commonly selected agent among Level 1 hospitals (59.4%). Selection of vasoactive/inotropic agents differed significantly across hospital levels (overall p = 0.002; Fig. 3A). Bonferroni-corrected pairwise comparisons showed that the difference remained significant between Level 3 and Level 2 hospitals (adjusted p = 0.002; Supplemental Table S5).

Fig. 3.

Differences in hemodynamic management strategies for cardiogenic shock across hospital levels. A Primary vasopressor agents used for cardiogenic shock. B Target range of minimum mean arterial pressure maintained during management. C Lactate monitoring during cardiogenic shock management

Significant differences were also observed in monitoring and management strategies among hospital levels. The minimum mean arterial pressure (MAP) target varied significantly (overall p = 0.017; Fig. 3B). A MAP target of 65 mmHg was adopted by 59.4% of Level 1 hospitals and 56.5% of Level 2 hospitals, compared with only 29.5% of Level 3 hospitals. The difference between Level 3 and Level 2 hospitals remained statistically significant after Bonferroni adjustment (adjusted p = 0.030; Supplemental Table S5). In contrast, 45.5% of Level 3 hospitals targeted a lower MAP of 60 mmHg.

Continuous lactate monitoring to guide patient management also differed significantly among hospital levels (overall p < 0.001; Fig. 3C), with lower utilization in Level 3 hospitals than in Level 2 and Level 1 hospitals (39.1% vs. 74.6% and 84.4%, respectively; both adjusted p < 0.001; Supplemental Table S5). Additionally, 23.9% of Level 3 hospitals did not routinely assess right ventricular function, whereas Level 2 and Level 1 hospitals performed continuous monitoring more consistently. Pulmonary artery catheters (PACs) were not routinely used, and only a few Level 1 hospitals reported occasional or frequent use (Table 1).

Table 1.

Management of cardiogenic shock across different hospital levels

Variables	All responders (n = 142)	Level 3 (n = 47)	Level 2 (n = 63)	Level 1 (n = 32)	χ²	p value
Established protocol %(95%CI)					41.089	<0.001
Yes	43.0 (35.1–51.2)	14.9 (7.4–27.7)	41.3 (30.0–53.6)	87.5 (71.9–95.0)
No	57.0 (48.8–64.9)	85.1 (72.3–92.6)	58.7 (46.4–70.0)	12.5 (5.0–28.1)
Right ventricular function* %(95%CI)					20.195	<0.001
Yes	91.3 (84.5 -95.0)	76.1 (62.1–86.1)	98.4 (91.3–99.7)	100 (89.0–100.0)
No	8.7 (5.0–14.6)	23.9 (13.9–37.9)	1.6 (0.3–8.7)	0.0 (0.0 -11.0)
hemodynamic monitoring
PAC%(95%CI)					53.130	<0.001
Always / Routinely	1.4 (0.4–5.1)	0.0 (0.0–7.7)	1.6 (0.3–8.6)	3.1 (0.6–15.7)
Often / Frequently	9.3 (5.5–15.2)	0.0 (0.0–7.7)	9.7 (4.5–19.5)	21.9 (11.0–38.8)
Sometimes	18.6 (13.0–25.8)	8.7 (3.4–20.3)	14.5 (7.8–25.3)	40.6 (25.5–57.7)
Rarely	34.3 (26.9–42.5)	21.7 (12.3–35.6)	51.6 (39.4–63.6)	18.8 (8.9–35.3)
Never	36.4 (28.9–44.7)	69.6 (55.2–80.9)	22.6 (14.0–34.4)	15.6 (6.9–31.8)
Cardiac output (or cardiac index) %(95%CI)					8.764	0.073
Swan-Ganz catheter	39.6 (31.0–48.9)	45.0 (25.8–65.8)	45.0 (33.1–57.5)	25.8 (13.7–43.2)
Arterial catheter	36.0 (27.7–45.3)	30.0 (14.5–51.9)	40.0 (28.6–52.6)	32.3 (18.6–49.9)
Use both methods	24.3 (17.3–33.1)	25.0 (11.2–46.9)	15.0 (8.1 -26.1)	41.9 (26.4–59.2)
pharmacologic therapy† %(95%CI)
β-blockers or ARBsǂ	63.4 (55.2–70.9)	55.3 (41.2–68.6)	69.8 (57.6–79.8)	62.5 (45.3–77.1)	2.460	0.292
Statins	83.8 (76.9–89.0)	85.1 (72.3–92.6)	84.1 (73.2–91.1%)	81.3 (64.7–91.1)	0.217	0.897
Antiplatelet therapy	89.0 (82.6–93.2)	84.1 (70.6–92.1)	88.3 (77.8–94.2)	96.9 (84.3–99.4)	3.130	0.177
Neurological monitoring§ %(95%CI)					24.832	<0.001
Yes	52.1 (43.9–60.2)	30.4 (19.1–44.8)	50.0 (37.9–62.1)	87.5 (71.9–95.0))
No	47.9 (39.8–56.1)	69.6 (55.2–80.9)	50.0 (37.9–62.1)	12.5 (5.0–28.1)
Database for cardiogenic shock %(95%CI)					4.267	0.116
Yes	14.8 (9.9–21.6)	6.4 (2.2–17.2)	17.5 (10.0–28.6)	21.9 (11.0–38.8)
No	85.2 (78.4–90.1)	93.6 (82.8–97.8)	82.5 (71.4–90.0)	78.1 (61.2–89.0)

Categorical variables were described using %(95%CI) and significant differences were compared using the chi-square test or Fisher’s exact test

PAC pulmonary artery catheter, ARB angiotensin receptor blocker

^*In managing cardiogenic shock, do you consistently monitor RV function?

^†Are they used within 24 h of diagnosis of cardiogenic shock?

^ǂIn the initial 24-hour period following resolution of cardiogenic shock, are β-blockers or ARBs administered?

^§Is neurological monitoring implemented in cardiac arrest patients receiving mechanical circulatory support?

Standardized management protocols for patients with CS remained suboptimal in Level 3 hospitals and Level 2 hospitals. Nearly half lacked defined resuscitation time thresholds and standardized neurologic monitoring practices when considering MCS for patients with cardiac arrest (Table 1). Moreover, 50.0% of Level 3 hospitals did not have a designated area for the admission and management of patients with CS. Notably, 78.1% of Level 1 hospitals had not established a dedicated clinical database for CS management.

A shock team was reported by 71.9% of Level 1 hospitals and 44.4% of Level 2 hospitals, whereas no Level 3 hospitals had such a team (overall p < 0.001; Fig. 4). All pairwise comparisons remained significant after Bonferroni correction (adjusted p < 0.05; Supplemental Table S5). The most commonly involved members were intensivists (94.1%), cardiologists (94.1%), and emergency physicians (80.4%), followed by nurses (74.5%), surgeons (43.1%), physiotherapists or rehabilitation specialists (33.3%), pharmacists (25.5%), nutritionists (23.5%), and respiratory therapists (11.8%), and other healthcare professionals. Additionally, among the 61 hospitals in the quality control network, 8.9% did not use any scoring system to evaluate the stage or severity of CS. Among hospitals that used scoring systems, the most frequently applied were the Acute Physiology and Chronic Health Evaluation II (APACHE II), Sequential Organ Failure Assessment (SOFA), Global Registry of Acute Coronary Events (GRACE), and Intra-Aortic Balloon Pump in Cardiogenic Shock II trial (IABP-SHOCK II) scores. Only 3.6% of hospitals used the Society for Cardiovascular Angiography and Interventions (SCAI) shock stage classification for outcome prediction (Fig. 5).

Fig. 4.

Availability of multidisciplinary shock teams across hospitals of different levels

Fig. 5.

Use of scoring systems for outcome prediction in patients. APACHE II, Acute Physiology and Chronic Health Evaluation II; SOFA, Sequential Organ Failure Assessment; IABP-SHOCK II, Intra-Aortic Balloon Pump in Cardiogenic Shock II trial; GRACE, Global Registry of Acute Coronary Events; SAVE, Survival After Veno-arterial ECMO; SCAI, Society for Cardiovascular Angiography and Interventions; INTERMACS, Interagency Registry for Mechanically Assisted Circulatory Support

MCS

The optimal timing of MCS initiation remained heterogeneous. Among Level 1 hospitals (n = 32), 56.3% considered initiating MCS when patients did not improve despite treatment with two or more vasopressors or inotropic agents. IABP was the most commonly used device (71.9%), followed by veno-arterial extracorporeal membrane oxygenation (VA-ECMO) (12.5%) (Supplemental Fig. S1). For patients with pre-existing or newly developed CS in the catheterization laboratory, MCS initiation was determined by a shock team in 50.0% of Level 1 hospitals and by interventional cardiologists in 31.3% (Supplemental Fig. S2). VA-ECMO implantation was primarily performed by intensivists (34.4%), surgeons (21.9%), or interventional cardiologists (18.8%) (Supplemental Fig. S3). Among Level 1 hospitals providing VA-ECMO, IABP was the preferred left ventricular unloading strategy (56.3%) (Supplemental Fig. S4). In 62.5% of Level 1 hospitals, the interval from the decision to initiate VA-ECMO to deployment was less than one hour for patients with CS.

Referral for CS

Among all hospitals included in the analysis (n = 142), Level 1 hospitals provided extensive medical transport services, ranging from the local region to nationwide levels. Level 3 hospitals frequently transferred patients because of the need for escalation of care. However, only 31.9% of Level 3 hospitals had established referral networks, whereas 44.7% were uncertain about their hierarchical referral pathways (Table 2). In this survey, 51 of the 142 participating hospitals had on-site cardiac surgery capability.

Table 2.

Comparison of CS transfers among hospitals of different levels

Variables	All responders (n = 142)	Level 3 (n = 47)	Level 2 (n = 63)	Level 1 (n = 32)	χ²	p value
Accept referred patients %(95%CI)					53.913	< 0.001
Always	12.1 (7.7–18.6)	4.3 (1.2–14.5)	3.2 (0.9–11.0)	40.6 (25.5–57.7)
Selectively, with administrative approval	37.1 (29.6–45.4)	15.2 (7.6–28.2)	48.4 (36.4–60.6)	46.9 (30.9–63.6)
Never	50.7 (42.5–58.9)	80.4 (66.8–89.3)	48.4 (36.4–60.6)	12.5 (5.0–28.1)
Scope of referral acceptance %(95%CI)					14.928	0.038
Nationwide	26.1 (17.2–37.5)	11.1 (2.0–43.5)	15.6 (6.9–31.8)	42.9 (26.5–60.9)
Beijing and the surrounding regions	33.3 (23.4–45.1)	33.3 (12.1–64.6)	37.5 (22.9–54.7)	28.6 (15.3–47.1)
District-level or nearby areas	27.5 (18.4–39.0)	44.4 (18.9–73.3)	37.5 (22.9–54.7)	10.7 (3.7–27.2)
Affiliated branch or consortium hospitals	10.1 (5.0–19.5)	0.0 (0.0–29.9)	9.4 (3.2–24.2)	14.3 (5.7–31.5)
Others	2.9 (0.8–10.0)	11.1 (2.0–43.5)	0.0 (0.0–10.7)	3.6 (0.6–17.7)
Transfer for escalation or de-escalation %(95%CI)					27.061	0.002
Always / Routinely	22.7 (16.3–30.6)	37.0 (24.5–51.4)	18.9 (10.6–31.4)	6.9 (1.9–22.0)
Often / Frequently	20.3 (14.3–28.1)	19.6 (10.7–33.2)	22.6 (13.5–35.5)	17.2 (7.6–34.5)
Sometimes	39.1 (31.0–47.7)	34.8 (22.7–49.2)	47.2 (34.4–60.3)	31.0 (17.3–49.2)
Rarely	16.4 (11.0–23.8)	8.7 (3.4–20.3)	11.3 (5.3–22.6)	37.9 (22.7–56.0)
Never	1.6 (0.4–5.5)	0.0 (0.0–7.7)	0.0 (0.0–6.8)	6.9 (1.9–22.0)
Referral pathways %(95%CI)					12.070	0.010
Fixed referral network	32.4 (25.2–40.5)	31.9 (20.4–46.2)	39.7 (28.5–52.0)	18.8 (8.9–35.3)
Assigned by administrators	15.5 (10.5–22.3)	23.4 (13.6–37.2)	4.8 (1.6–13.1)	25.0 (13.3–42.1)
Uncertain	52.1 (43.9–60.2)	44.7 (31.4–58.8)	55.6 (43.3–67.2)	56.3 (39.3–71.8)
CS admission for non-cardiac surgery %(95%CI)					14.086	0.001
All PCI centers	37.0 (29.4–45.3)	15.9 (7.9–29.4)	51.6 (39.4–63.6)	37.5 (22.9–54.7)
Conditional admission with transfer to the nearest advanced shock center	63.0 (54.7–70.6)	84.1 (70.6–92.1)	48.4 (36.4–60.6)	62.5 (45.3–77.1)

Categorical variables were described using %(95%CI) and significant differences were compared using the chi-square test or Fisher’s exact test

CS cardiogenic shock, PCI percutaneous coronary intervention

Among the 61 hospitals participating in the quality control network, 31 hospitals (50.8%) believed that centers without cardiac surgery programs should admit patients with CS only when a nearby shock center (Level 1) was available; otherwise, patients should be transferred directly to facilities capable of advanced shock management. Conversely, the remaining 30 hospitals (49.2%) considered that all centers with PCI capability, regardless of cardiac surgery capability, should be able to admit patients with CS.

Discussion

To our knowledge, this is the first multicenter survey in China investigating AMI-CS management, although it was limited to hospitals in Beijing. The findings provide important insights into the current status and identify areas requiring improvement. The main findings were as follows: (1) substantial heterogeneity existed among hospital levels in diagnostic strategies, with Level 3 hospitals showing less frequent use of comprehensive assessment and invasive hemodynamic monitoring; (2) most hospitals did not routinely use the SCAI classification for risk stratification and outcome prediction; (3) IABP remained the most commonly used MCS device, followed by VA-ECMO, and was the preferred strategy for left ventricular unloading; and (4) the inter-hospital referral networks remained insufficiently established.

Early and comprehensive hemodynamic monitoring is essential when CS is suspected because it guides treatment, enables timely recognition of deterioration, and supports decisions regarding vasoactive agents and MCS initiation [13, 14]. Current guidelines recommend combining invasive hemodynamic assessment with non-invasive cardiac imaging to better characterize the hemodynamic phenotype of patients with CS [6]. However, only 49.2% of Level 2 hospitals and 21.3% of Level 3 hospitals in this survey diagnosed CS using a comprehensive approach incorporating symptoms, physical findings, metabolic parameters, echocardiography, and invasive monitoring. This approach was most frequently adopted by Level 1 hospitals, whereas most Level 3 hospitals primarily relied on clinical presentation. The marked differences among hospital levels likely reflect limited diagnostic and management capabilities in Level 3 hospitals, as well as unequal resource distribution and equipment availability.

Level 1 hospitals and Level 2 hospitals were more likely to maintain a minimum MAP target of 65 mmHg during CS management, whereas Level 3 hospitals more commonly accepted lower blood pressure targets. Moreover, 60.9% of Level 3 hospitals did not use continuous lactate monitoring to guide hemodynamic management, possibly because limited point-of-care testing (POCT) availability and early inter-hospital transfer make repeated lactate assessment less feasible. These findings suggest that some institutions rely mainly on arterial pressure while underutilizing lactate assessment. However, elevated lactate levels may indicate persistent tissue hypoperfusion despite apparently adequate MAP. Both lactate concentration and lactate clearance are established prognostic biomarkers in CS [15–18]. Recent guidelines recommend reassessing lactate and other perfusion markers every 2–8 h to ensure adequate organ perfusion while maintaining a MAP of 60–65 mmHg [6]. A retrospective study of patients with CS admitted to cardiac intensive care units showed that persistent MAP < 65 mmHg was associated with increased mortality risk [19]. MAP targets should be individualized according to patients’ baseline cardiovascular status and clinical response. In patients with septic shock, an initial MAP target of 65–70 mmHg is generally recommended. However, higher targets may be appropriate for patients with a history of chronic arterial hypertension, if they are accompanied by improved tissue perfusion or clinical status at higher blood pressure levels [20]. Therefore, both MAP and lactate levels should be monitored simultaneously to optimize CS management.

PAC monitoring was infrequently used across all hospital levels, including Level 1 hospitals. PAC has several important roles in CS, including confirming diagnosis and hemodynamic classification, providing continuous monitoring in patients receiving MCS to guide management and weaning, and determining suitability for advanced therapies such as left ventricular assist devices or heart transplantation [21]. In addition, PAC enables calculation of cardiac power output/index (CPO/CPI) and the pulmonary artery pulsatility index (PAPi), which provide prognostic and therapeutic information in CS [21–23]. However, the clinical value of PAC in CS remains controversial. Several studies have associated PAC monitoring with improved outcomes [24, 25], whereas a meta-analysis showed no overall reduction in in-hospital mortality among patients with CS, although patients with acute decompensated heart failure-related CS may benefit from PAC monitoring [26]. We believe that hemodynamic data obtained through PAC may facilitate dynamic treatment adjustment and potentially improve outcomes, rather than PAC itself serving as a therapeutic intervention [27, 28]. Consistent with these conflicting findings, the 2025 ACC expert consensus recommends selective PAC use in patients with refractory CS or uncertain hemodynamic phenotypes, rather than routine use in all patients with CS [6]. Future prospective studies are required to clarify the optimal indications and clinical utility of PAC in this population.

Specialized shock teams have been associated with reduced mortality among patients with CS [29, 30]. However, substantial differences in shock team development were observed across hospital levels. These findings reflect limitations in multidisciplinary collaboration, particularly in Level 3 and Level 2 hospitals, and emphasize the need for standardized criteria for shock team establishment to maximize their clinical benefit.

In 2019, the SCAI introduced a five-stage classification system (stages A-E) for CS [31]. The 2022 expert consensus further refined the criteria and proposed a three-axis model, emphasizing the SCAI classification as a key tool for assessing shock severity and predicting clinical outcomes [32]. However, among the 61 hospitals participating in the quality control network, only 3.6% applied the SCAI classification in the assessment of patients with CS. This finding indicates limited implementation of the SCAI staging, which may hinder accurate risk stratification and treatment optimization. The latest consensus emphasizes early recognition of CS, recommends the SUSPECT CS framework to facilitate diagnosis, and supports routine application of the SCAI classification to stage disease severity [6].

Although the IABP-SHOCK II trial showed that routine IABP use did not reduce 30-day mortality in patients with AMI-CS [33], our study found that IABP remains the most commonly used MCS device in Beijing. To date, no study has demonstrated a mortality benefit of VA-ECMO in patients with CS. However, several studies suggest that combined VA-ECMO and IABP therapy may reduce mortality, particularly in patients with AMI-CS [34–36]. Consistently, our findings also revealed that IABP remained the preferred device for left ventricular unloading. Current guidelines recommend considering IABP (Class IIa) in patients with AMI-CS who develop mechanical complications, including ventricular septal rupture or free-wall rupture [6, 37]. A recent study showed that early IABP implantation during SCAI shock stages A/B in patients with myocardial infarction–related ventricular septal defect (MI-VSD) was associated with reduced 30-day mortality and prevention of progression to more severe shock stages (C/D/E) [38]. However, further studies are required to clarify the role of IABP. The DanGer Shock trial was the first study to demonstrate that the microaxial flow pump (Impella CP) significantly reduced 180-day mortality in patients with AMI-CS, although it was associated with a higher risk of complications [39]. However, Impella use remains limited in Beijing, likely because of its high cost. The optimal selection of MCS devices and timing of implantation in different CS populations remain important areas for future research.

Dennis et al. surveyed 143 clinicians and reported that 75% believed CS remained underrecognized, whereas 67% considered its treatment inadequate [9]. These findings highlight gaps in clinical awareness, resource distribution, and access to advanced therapies. Consistent with this, our study identified deficiencies in CS management across hospitals in Beijing. Many Level 3 hospitals lacked standardized management protocols, dedicated treatment areas, and clear referral pathways. Previous studies have shown that patients transferred directly or indirectly to Level 1 hospitals had significantly lower in-hospital mortality [40]. These results emphasize the importance of establishing a structured tiered referral network to improve patient outcomes.

The crude incidence of AMI among Chinese adults was 87.6 per 100,000 population. Cardiogenic shock, the most severe complication of AMI, occurs in approximately 7.7% of patients with AMI and poses a substantial threat to patient survival [41]. A nationwide study conducted between 2019 and 2020 involving 253 chest pain centers found that Level 1 hospitals achieved better AMI outcomes than Level 2 hospitals, as reflected by lower fatality rates, likely because of superior equipment, technical expertise, and treatment capabilities [42]. Despite Beijing’s concentration of medical resources, substantial disparities in CS management persisted across hospital levels. Therefore, strengthening regional collaborative care networks and prioritizing professional training and system development in Level 3 hospitals are essential to reducing inter-hospital disparities.

These disparities may be attributed to unequal distribution of medical resources, differences in infrastructure and technical capacity, limited access to advanced hemodynamic monitoring and MCS, insufficient implementation of standardized AMI-CS management protocols, incomplete multidisciplinary shock teams, and unclear referral pathways.

Future efforts should focus on establishing a regionalized AMI-CS care network using a “hub-and-spoke” model [6]. This network should include standardized diagnostic and treatment protocols, regular training programs, and quality improvement initiatives. Advances in information technology, including artificial intelligence-assisted electronic medical records, multidisciplinary teleconsultation, and remote education, may enhance communication and coordination within the AMI-CS care network. Telemedicine-supported regional CS care models may help resource-limited hospitals obtain timely expert guidance for early recognition, stabilization, and transfer decisions, although their effectiveness requires further investigation [43]. Additionally, post-shock clinics may provide ongoing shock-related consultation and improve long-term outcomes [44]. “Excellent Chest Pain Centers” have been specifically designed recently in China to address inconsistent care and unacceptably high death rates in critical cardiovascular emergencies. The centers establish multidisciplinary teams and equip themselves with advanced devices. To guarantee true 24/7 readiness, these teams operate on fixed-schedule rotations with a rigid backup protocol. Through a “One-Network ECG” system, ECGs from primary facilities are transmitted instantly for remote expert interpretation [45]. For unstable patients, “Reverse Transport” is employed-a mobile ECMO team is dispatched to the referring hospital to establish cannulation and stabilize the patient on-site before transporting. These are implementation strategies for improving CS care systems in China and will play an essential role in the future. Furthermore, regional or national AMI-CS registries are also needed to support quality improvement, monitor treatment patterns, and evaluate patient outcomes.

Limitations

This study has several limitations. First, although a fully randomized sampling approach would have been preferable, it was not feasible because the limited number of Level 1 hospitals would have resulted in underrepresentation and prevented meaningful comparisons across hospital tiers.

Second, for hospitals outside the quality control network, although questionnaire respondents were selected based on their professional roles and familiarity with institutional CS management, physicians may have reported primarily based on their departmental practice and personal understanding. Therefore, potential information or reporting bias cannot be completely excluded. Nevertheless, this respondent selection strategy may have reduced such bias to some extent.

Third, this survey included only hospitals in Beijing and a relatively small sample size, which may have limited generalizability and introduced regional bias. Moreover, the absence of patient-level data prevented evaluation of direct associations between management strategies and clinical outcomes, as well as the differentiation of CS from mixed shock at the individual level. Future studies should include hospitals from more regions and incorporate patient-level investigations.

Fourth, this study did not include a formal pilot survey, psychometric validation, or test-retest reliability assessment, which may have limited the evaluation of the questionnaire’s validity and reliability. Nevertheless, the questionnaire was developed based on previous surveys, current relevant consensus statements, and guideline documents and was reviewed and revised by an expert panel. Therefore, the results may still offer valuable descriptive insights within the scope of this study.

Finally, although the overall response rate was 100%, eight questionnaires from the non-quality control network were excluded because of more than 20% missing data, whereas all questionnaires from the quality control network were included in the analysis. Therefore, potential non-responder bias or selection bias cannot be excluded because hospitals with better CS management may have provided more complete responses.

Conclusion

This study demonstrated substantial heterogeneity in CS management across different hospital levels in Beijing. Compared with Level 2 and Level 1 hospitals, Level 3 hospitals require improvement in diagnostic strategies, hemodynamic management, inter-hospital referral systems and the standardized implementation of the SCAI classification.

Abbreviations

AMI-CS

Acute myocardial infarction complicated by cardiogenic shock

IABP

Intra-aortic balloon pump

VA-ECMO

Veno-arterial extracorporeal membrane oxygenation

PCI

Percutaneous coronary intervention

tMCS

Temporary mechanical circulatory support

MCS

Mechanical circulatory support

APACHE II

Acute Physiology and Chronic Health Evaluation II

SOFA

Sequential Organ Failure Assessment

IABP-SHOCK II

Intra-Aortic Balloon Pump in Cardiogenic Shock II trial

GRACE

Global Registry of Acute Coronary Events

SAVE

Survival After Veno-arterial ECMO

SCAI

Society for Cardiovascular Angiography and Interventions

INTERMACS

Interagency Registry for Mechanically Assisted Circulatory Support

PAC

Pulmonary artery catheter

ARB

Angiotensin receptor blocker

CS

Cardiogenic shock

Authors' contributions

YR drafted the manuscript. YR, YL, YX, CZ, XT, LW, ZD, XH, and CL collected and analyzed the data, with contributions from the Beijing Cardiovascular Internal Medicine Quality Control and Improvement Center and the Beijing Extracorporeal Life Support Quality Control and Improvement Center. YD performed the statistical analysis. HW and XH designed the study. HW reviewed and revised the manuscript. All authors approved the final manuscript.

Funding

Capital’s Funds for Health Improvement and Research (2024-1-1051); Noncommunicable Chronic Diseases-National Science and Technology Major Project (2025ZD0547900 and 2025ZD0547904).

Data availability

The datasets analyzed during the current study are not publicly available but are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study was approved by the Institutional Ethics Committee of Beijing Anzhen Hospital (approval number: KS2024041). The study did not involve experiments on human participants, patient-level data, identifiable personal information, or human biological samples. Therefore, the requirement for written informed consent was waived by the Committee because this was hospital-level observational survey. All methods were conducted in accordance with the relevant guidelines and regulations.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.