Outcomes with mechanical circulatory support devices among patients with mechanical complications of acute myocardial infarction

Abstract

Aims

The aim of this study is to examine the trends and outcomes of mechanical circulatory support (MCS) device use among patients with mechanical complications of acute myocardial infarction (AMI).

Methods and results

Using data from the National Inpatient Sample (a large admirative database in the USA) years 2016–20, we identified AMI admissions (ST-elevation and non-ST-elevation myocardial infarction) with mechanical complications (ventricular septal defect, free wall rupture, or papillary muscle rupture). Among 4 450 219 AMI patients, 7025 (0.2%) had a mechanical complication of which 3115 patients (44.3%) received at least one MCS device. There was a rising trend in MCS use (39.3% in 2016 to 48.9% in 2020, P_trend = 0.02), but there was no corresponding reduction in the incidence of in-hospital mortality (36.9% in 2016 vs. 43.4% in 2020, P_trend = 0.75). There was no significant difference in in-hospital mortality between those who received MCS vs. those who did not (48.4 vs. 34.5%, respectively).

Conclusion

In this large observational analysis of AMI hospitalizations, mechanical complications were rare and associated with very high in-hospital mortality. Although the use of MCS has increased, in-hospital mortality rates remain high even among patients who received MCS. Further investigations are needed to clarify the role of MCS devices among patients with mechanical complications of AMI.

Graphical Abstract

Graphical Abstract.

(A) There was a rise in MCS device use among patients with mechanical complications of acute myocardial infarction from 2016 to 2020. (B) The rates of in-hospital mortality did not change. (C) In-hospital mortality rates. This figure was created by using Biorender.com. ECMO, extracorporeal membrane oxygenation; IABP, intra-aortic balloon pump; MCS, mechanical circulatory support.

Introduction

Mechanical complications of acute myocardial infarction (AMI) are rare and associated with considerably high mortality.¹ Mechanical complications include ventricular septal defect (VSD), papillary muscle rupture (PMR), and ventricular free wall rupture (FWR). Although there has been a substantial reduction in the incidence of mechanical complications with the widespread adoption of primary percutaneous coronary interventions (PCIs) compared with the thrombolytic era, data indicate that the mortality rates associated with these complications have not changed in the past decade.^2–4 Surgical repair remains the mainstay treatment for these complications, but the timing remains controversial.¹ Early surgery carries a high mortality (>50%) due to the challenges of operating on necrotic tissue and managing patients with multi-organ failure. Delayed surgery (i.e. >2 weeks) allows for healing of the necrotic tissue and has been linked with lower mortality (∼30–40%),^5,6 though this might also be related to selection bias as relatively healthier patients with lower comorbidity burden have better chances of surviving long enough to have surgery.

Mechanical circulatory support (MCS) devices play a role in stabilizing the haemodynamics and can be used as a bridge to surgery or recovery.⁷ The MCS devices can potentially improve end-organ perfusion while reducing cardiac filling pressures. The MCS devices can also help unload the myocardium, ameliorate ischaemia, and support revascularization.^8,9 The primary types of MCS include intra-aortic balloon pumps (IABPs), Impella (Abiomed, Danvers, MA), and extracorporeal membrane oxygenation (ECMO).^10,11 Few single-centre experiences and case series with a limited number of patients have suggested that these devices may improve in-hospital mortality among AMI patients with mechanical complications.^12–14 To better understand the potential role of MCS devices in this high-risk population, we aimed to examine the trends in MCS device use in the management of mechanical complications of AMI and investigate the association between these modalities on in-hospital mortality using a nationally representative sample.

Methods

Data source, study population, and outcomes

This is an observational study using the National Inpatient Sample (NIS) years 2016–20. The NIS is an all-payer, nationally reflective data set capturing hospital admissions across the USA, curated by the Center for Health Data Excellence (CHDE) under the Healthcare Research and Quality (AHRQ). As the most extensive publicly accessible resource for inpatient care data, it aggregates millions of discharge records each year, detailing patient profiles, medical conditions, treatments, facility attributes, and clinical results. The NIS utilizes a layered sampling framework, applying statistical weights to each entry to produce findings that mirror national trends.

We identified adult hospitalizations (>18 years) with any diagnosis of AMI defined as ST-elevation myocardial infarction and non-ST-elevation myocardial infarction who had a mechanical complication (i.e. VSD, PMR, or FWR) using the appropriate ICD-10 codes.¹⁵ Admissions were then classified according to the use of MCS devices (i.e. IABP, Impella, or ECMO). The ICD-10 codes were summarized in Supplementary material online, Table S1.

We identified the following patient-level characteristics: age, sex, race/ethnicity, primary insurance, and median income at the patient’s ZIP code and medical history including hypertension, diabetes mellitus, smoking, peripheral vascular disease, history of coronary artery disease (CAD) [defined as prior MI, prior PCI, and prior coronary artery bypass graft (CABG)], chronic kidney disease (CKD), and heart failure. We also identified other complications and procedures during admission including acute kidney injury (AKI), haemodialysis, PCI, CABG, surgical repair, mechanical ventilation, cardiogenic shock, and cardiac arrest. The hospital characteristics included bed size, teaching status, region, and transfer status.

The primary outcome was in-hospital mortality. We also assessed the trend of using MCS modalities during the study period.

Statistical analysis

All analyses employed weighted data to ensure national representativeness. Categorical variables were described as frequencies and percentages and compared using χ² tests. Continuous variables were described as means and standard deviations and compared with Student’s t-test or non-parametric tests as appropriate. Logistic regression was performed to evaluate the trend of MCS use while multivariable logistic regression and predictive margins were used to assess for the significance of the trend. Since it is not feasible to adjust for all relevant clinical variables such as lactate and the decision to pursue MCS, we reported only the unadjusted incidences of mortality for each group. A multilevel multivariable logistic regression model was used to determine the factors independently associated with in-hospital mortality. The model was adjusted for age, sex, primary insurance, hospital bed size, teaching status, hospital transfer status, peripheral vascular disease, history of MI, inpatient complications, and procedures including AKI, CABG, PCI, cardiac arrest, cardiogenic shock, mechanical ventilation, surgical repair, and haemodialysis. These covariates were selected after performing univariate analysis for each covariate individually against the primary outcome and significant covariates (P < 0.05) were included. Summary estimates were reported as odds ratio (OR) and corresponding 95% confidence intervals (CIs). A P < 0.05 was used as the threshold for statistical significance. According to the HCUP data use agreement, any variable with <11 observations was not reported. Analyses were conducted using STATA 18 (STATA Corp, College Station, Texas 77845, USA).

Results

Among 4 450 219 AMI patients from 2016 to 2020, 7025 (0.2%) had mechanical complications 5090 (0.1%) had VSD, 1255 (0.02%) had PMR, and 835 (0.01%) had FWR). Among AMI patients with mechanical complications, 3115 patients (44%) received at least one MCS modality [2410 (77.4%) received IABP, 850 (27.3%) received Impella, and 350 (11.2%) received ECMO] (Figure 1) and 845 patients (12%) underwent surgical repair during the hospitalization. IABP was the most commonly used MCS modality irrespective of the mechanical complication (27.4% of VSD patients, 22.8% of FWR patients, and 41.8% of PMR patients). Few patients received more than one MCS modality (244 patients received Impella and IABP, 135 received Impella and ECMO, and 145 received IABP and ECMO).

Figure 1.

Study flow chart.

The mean age was comparable in the two groups (68.8 years in both groups), and almost half of the patients were women (47% vs. 53%). The majority of the patients were White (81.7% vs. 78%). MCS use was more common in urban teaching hospitals (84.3% vs. 77.8%). Patients who received MCS were more likely to be transferred (49.1% vs. 43.1%). Regardless of the hospital location/teaching status, IABP was the most commonly used modality (64% in rural hospitals, 75% in urban non-teaching, and 64% in urban teaching hospitals). The baseline characteristics, comorbidities, and demographic features are summarized in Table 1.

Table 1.

Baseline characteristics

	MCS n = 3115	No MCS n = 3910	P value
Age, mean and SD	68.8 (±0.9)	68.8 (±1.1)	0.92
Female (%)	1140 (36)	1855 (47)	<0.001
Race (%)			0.25
White	2395 (81.7)	2885 (78)
African American	150 (5.1)	275 (7.4)
Hispanic	185 (6.3)	310 (8.4)
Asian or Pacific Islander	85 (2.9)	100 (2.7)
Native American/other	115 (3.9)	130 (3.5)
Primary insurance (%)			0.12
Medicare	1845 (59.3)	2390 (61.1)
Medicaid	215 (6.9)	385 (9.9)
Private insurance	805 (25.9)	835 (21.4)
Self-pay	165 (5.3)	180 (4.6)
No charge/other	80 (2.6)	120 (3.1)
Median household income for patient’s ZIP code (%)			0.46
0–25th percentile	820 (26.8)	1080 (28.2)
26th–50th percentile	770 (25.2)	1065 (27.8)
51st–75th percentile	830 (27.2)	920 (24)
76th–100th percentile	635 (20.8)	765 (19.9)
Hospital bed size (%)			<0.001
Small	190 (6.1)	480 (12.3)
Medium	675 (21.7)	985 (25.2)
Large	2250 (72.2)	2445 (62.5)
Hospital location and teaching status (%)			0.007
Rural	70 (2.3)	165 (4.2)
Urban non-teaching	420 (13.5)	705 (18)
Urban teaching	2625 (84.3)	3040 (77.8)
Hospital region (%)			0.77
Northeast	545 (17.5)	680 (17.4)
Midwest	860 (27.6)	995 (25.5)
South	1070 (34.4)	1435 (36.7)
West	640 (20.6)	780 (20.5)
Transferred (%)	1524 (49.1)	1674 (43.1)	0.03
Hypertension (%)	750 (24.1)	1135 (29)	0.04
Diabetes mellitus (%)	270 (8.7)	475 (12.2)	0.03
Tobacco smoking (%)	570 (18.3)	605 (15.5)	0.16
Peripheral vascular disease (%)	225 (7.2)	270 (6.9)	0.81
History of CAD (%)	560 (18)	850 (21.7)	0.08
CKD (%)	510 (16.4)	785 (20.1)	0.08
Heart failure (%)	2055 (66)	2360 (60.4)	0.03
AKI (%)	2130 (68.4)	1760 (45)	<0.001
PCI (%)	2180 (70)	1975 (50.5)	<0.001
CABG (%)	585 (18.8)	390 (10.0)	<0.001
Cardiac arrest (%)	230 (7.4)	340 (8.7)	0.37
Cardiogenic shock (%)	2530 (81.2)	1215 (31.1)	<0.001
Mechanical ventilation (%)	1155 (37)	620 (15.9)	<0.001
Haemodialysis (%)	255 (8.2)	115 (2.9)	<0.001
Surgical repair (%)	610 (19.6)	235 (6.0)	<0.001

SD, standard deviation; CAD, coronary artery disease; CKD, chronic kidney disease; AKI, acute kidney injury; PCI, percutaneous coronary intervention; CABG, coronary artery bypass graft.

There was an increase in the use of MCS devices among AMI patients with mechanical complications (39.2% in 2016 vs. 48.9% in 2020) (Graphical Abstract A). This trend was statistically significant after adjusting for covariates [adjusted odds ratio (aOR) 1.14, 95% CI 1.03–1.26, P = 0.02]. There was a rising trend for both IABP use (31.1% vs. 36.7%, P_trend = 0.66) and Impella use (11.3% vs. 12.9%, P_trend = 0.45) (Figure 2).

Figure 2.

Trend of using different mechanical circulatory support modalities. Extracorporeal membrane oxygenation was not included as there were no reported cases in some of the years during the study period. IABP, Intra-aortic balloon pump; MCS, mechanical circulatory support.

In-hospital mortality

During the study period, the rates of in-hospital mortality did not change significantly (36.9% in 2016 vs. 43.4% in 2020, P_trend = 0.75). In-hospital mortality was 48.4% in the MCS group and 34.5% in the no MCS group (Table 2). Mortality was comparable in early (<24 h) vs. late (>24 h) use of MCS (49.5% vs. 45.3%, respectively). In the subset of patients who underwent surgical repair (n = 845 patients), MCS was not associated with lower incidence of in-hospital mortality (56.6% vs. 42.6%). On multivariable logistic regression analysis, the following variables were independently associated with in-hospital mortality: older age (aOR 1.04, 95% CI 1.03–1.06, P < 0.001), female sex (aOR 1.36, 95% CI 1.04–1.79, P = 0.02), large hospital size (aOR 1.60, 95% CI 1.04–2.47, P = 0.03), AKI (aOR 1.55, 95% CI 1.17–2.04, P = 0.002), haemodialysis (aOR 6.10, 95% CI 3.09–12.04, P < 0.001), cardiogenic shock (aOR 2.43, 95% CI 1.72–3.43, P < 0.001), and cardiac arrest (aOR 9.34, 95% CI 5.18–16.85, P < 0.001). CABG during the hospitalization was associated with a lower incidence of mortality (aOR 0.56, 95% CI 0.38–0.84, P = 0.005), but surgical repair of the mechanical complication was not associated with lower mortality (aOR 1.45, 95% CI 0.99–2.14, P = 0.06) (Figure 3).

Table 2.

In-hospital mortality for overall MCS devices and individual MCS modalities

In-hospital mortality (%)	MCS n = 3115	No MCS n = 3910
	1510 (48.4)	1350 (34.5)
IABP (%)	IABP n = 2410	No MCS n = 3910
	1060 (43.9)	1350 (34.5)
Impella (%)	Impella n = 850	No MCS n = 3910
	565 (66.4)	1350 (34.5)
ECMO (%)	ECMO n = 350	No MCS n = 3910
	205 (58.5)	1350 (34.5)

IABP, intra-aortic balloon pump; ECMO, extracorporeal membrane oxygenation.

Figure 3.

Forest plot of select predictors associated with in-hospital mortality. AKI, acute kidney injury; CABG, coronary artery bypass graft; MCS, mechanical circulatory support.

In a subgroup analysis examining the different types of MCS devices, Impella, IABP, and ECMO were associated with a higher mortality compared with those who did not receive MCS (66.4% vs. 34.5%, 43.9% vs. 34.5%, and 58.5% vs. 34.5%, respectively).

Discussion

In this nationwide observational analysis of >4 million AMI hospitalizations from 2016 to 2020, we investigated the trends and in-hospital outcomes with MCS use among patients who developed mechanical complications of AMI. The salient findings of this study were as follows: (i) the rates of mechanical complications were low (∼0.2%) but associated with very high in-hospital mortality rates; (ii) there was an increase in MCS device utilization without a corresponding reduction in in-hospital mortality rates during the study period; and (iii) the rates of in-hospital mortality was high in the MCS and no MCS groups, even among the subset of patients who underwent surgical repair during the hospitalization.

Although mechanical complications of AMI are rare in the primary PCI era, they remain devastating challenges with very high mortality rates.¹⁴ Our study confirmed that in-hospital mortality rates for patients with mechanical complications remain considerably high, even among patients who received MCS devices. MCS devices are designed to provide haemodynamic support by augmenting cardiac output, ensuring systemic perfusion, reducing myocardial oxygen demand, improving coronary perfusion, particularly during diastole, and decreasing systemic vascular resistances.¹⁰ Despite these theoretical benefits, our findings contribute to a growing body of literature that questions the effectiveness of MCS in improving survival in these high-risk scenarios. Randomized trials have shown that MCS devices have not been shown to improve survival among a broad range of AMI patients with cardiogenic shock^16–18 and that only a certain subset of patients might drive benefit (i.e. patients with ST-elevation cardiogenic shock without risk of hypoxic brain injury).^18,19 We speculate that the lack of mortality reduction might be related to the severity of illness in patients with mechanical complications. The increased risk of bleeding and vascular complications associated with using MCS may play a role.¹¹ Although MCS may have been selectively offered to those who were initially considered to have a chance of survival, these patients remain critically ill, often with multi-organ failure that could not be fully mitigated by MCS. We could not ascertain these factors in our analysis since the data set does not provide granular data on multi-organ failure or hypoxic brain injury. Additionally, the heterogeneity of clinical presentations could have also been contributing factors. The decision to use MCS devices is likely dependent on the operator or centre experience, leading to variable practices that do not necessarily translate into better outcomes. We also observed that nearly 75% of MCS devices were inserted within 24 h of presentation but a comparison between early and late interventions was not performed to avoid confounding (as the time of cardiogenic shock was not available to account for). This raises questions about whether we are selecting the appropriate patients for MCS or if the devices themselves are not as effective as hoped in this context. It is possible that MCS is serving more as a temporary bridge to stabilize haemodynamics rather than as a definitive solution, which may explain the lack of observed mortality benefit.

Given the high mortality rates associated with mechanical complications of AMI and the unclear benefit of MCS, it is imperative to refine patient selection criteria and optimize the timing of interventions.^12,19 A multidisciplinary approach involving cardiologists, intensivists, and surgeons might play a role in improving outcomes. Additionally, continuous efforts in streamlining care processes, enhancing early recognition of ischaemic symptoms, and preventing delays in treatment are crucial.²⁰ While future research should ideally focus on large, randomized controlled trials to better understand the role of MCS devices, we recognize that such trials may be challenging to implement due to the rarity of these complications and the potential hesitance of physicians to randomize critically ill patients. Additionally, patients with mechanical complications have been excluded from randomized trials of MCS. Therefore, alternative research approaches, such as multicentre registries, may be necessary to provide insights into the optimal use of MCS in this high-risk population.^21,22 Additionally, new MCS devices are currently in development and could be tested among patients with AMI and mechanical complications.²³

In this study, we found that surgical repair of the mechanical complications during the hospitalization was performed in about 12% of the admissions and was not associated with lower incidence of in-hospital mortality. Indeed, small retrospective studies have demonstrated that early surgical repair of the mechanical complications is associated with a much higher mortality (∼50%), as compared with delayed surgical repair (∼30–40%).^1,5,24 Although delayed surgical repair allows time for maturation of friable necrotic tissue which in turn might facilitate surgical repair, the difference in the outcomes could also be explained by selection bias since patients who underwent early surgical repair are likely sicker.²⁴ This also could explain the lack of mortality benefit with surgical repair observed in our study. By contrast, CABG surgery was associated with lower in-hospital mortality. It is plausible that CABG surgery was offered to less sicker patients. This finding should be interpreted with caution given the wide CI.

Our study has several limitations. The observational nature of the study introduces potential for unmeasured confounding; therefore, we refrained from comparing the outcomes between MCS and no MCS. The NIS relies on administrative data which might be prone to coding errors.^25,26 The database also lacks data on haemodynamics, shock stage, time of presentation, and imaging, which are crucial for understanding the clinical rationale behind MCS use. MCS may have been offered to sicker patients with no other management options available which may represent selection bias by excluding healthier patients. It also may—at least partly—explain the absence of change in mortality despite increased use of MCS. Additionally, we were unable to assess the outcomes beyond hospitalization since the data set does not capture post-discharge outcomes. Despite these limitations, our study represents the largest study to date examining the role of MCS devices among this high-risk population.

Conclusion

In this large observational analysis of AMI hospitalizations, mechanical complications were rare and associated with very high in-hospital mortality. Although the use of MCS has increased, in-hospital mortality rates remain high even among patients who received MCS. Our study underscores the need for ongoing research and innovation in managing mechanical complications of AMI.

Supplementary material

Supplementary material is available at European Heart Journal: Acute Cardiovascular Care online.

Funding

None declared.

Data availability

The data underlying this article will be shared on reasonable request to the corresponding author.