Identification of Long-Term Mortality Predictors and Risk Score Development in Left Ventricular Assist Device Recipients

Abstract

Objective: Long-term mortality remains a significant concern in patients receiving left ventricular assist devices (LVADs). Identifying reliable prognostic factors and developing a validated risk score could improve patient selection and long-term management.

Methods: In this retrospective single-center study, 35 patients who underwent LVAD implantation between August 2019 and May 2025 were evaluated. Patients who died during the index hospitalization were excluded. Baseline clinical, laboratory, echocardiographic, and hemodynamic parameters were collected. Long-term mortality predictors were identified using Cox regression analyses. A risk score (CACA score) was developed based on significant predictors.

Results: Four variables – age, prior coronary artery bypass grafting (CABG), creatinine, and albumin levels – were associated with long-term mortality in univariate analysis. Among these, only age remained an independent predictor in multivariate analysis (hazard ratio [HR]: 1.48; 95% CI: 1.03–2.12; p=0.032). The CACA risk score was derived using the β-coefficients from multivariate analysis. Patients were stratified into low, intermediate, and high-risk groups with corresponding mortality rates of 0%, 45.4%, and 75%, respectively. The CACA score demonstrated superior discriminative ability compared to individual variables (area under the curve: 0.88).

Conclusion: Age, prior CABG, renal function, and albumin levels are key predictors of long-term mortality in LVAD recipients. The proposed CACA score effectively stratifies mortality risk and may serve as a practical tool for clinical decision-making.

Introduction

Left ventricular assist devices (LVADs) have become an established therapy for patients with advanced heart failure who are not eligible for transplantation, offering improved survival and quality of life compared with optimal medical treatment.[1] Despite this knowledge and advancements in device technology, mortality rates remain still high in this population. Historical data, such as the REMATCH trial, reported 1- and 2-year survival rates of only 52% and 23%, respectively.[1] These patients often face complications, such as right ventricular failure, bleeding, stroke, arrhythmic events, and device-related issues, including infection, pump thrombosis, and suction events, all of which contribute to the multifactorial causes of death.

The presence of multifactorial causes of death makes it challenging to identify objective risk factors for mortality in LVAD recipients. The etiologies of short-term and long-term mortality differ among these patients. For instance, INTERMACS status has been shown to strongly predict early mortality after LVAD implantation;[2,3] however, its predictive value for long-term mortality diminishes in previous studies.[4–10] Similarly, early post-operative sepsis accounts for substantial early mortality, while later deaths are more often related to chronic comorbidities, bleeding, or device-related issues. Predictors of long-term mortality remain controversial and vary across studies.[4–11] Several risk scores incorporating laboratory, clinical, and hemodynamic parameters have been proposed,[4–11] but their discriminatory and predictive power for long-term mortality in LVAD recipients remains uncertain. The present study, therefore, sought to evaluate potential determinants of long-term mortality among LVAD patients and to develop a new risk score to support individualized risk stratification.

Materials and Methods

This retrospective, single-center study included patients who underwent implantation of a LVAD as destination therapy for advanced heart failure with reduced ejection fraction between August 1, 2019, and May 30, 2025. Patients who died during the early post-operative hospitalization period were excluded (n=9), resulting in a final cohort of 35 adult patients. All data were retrospectively retrieved from the hospital electronic medical records and individual patient charts, recorded at the time of admission.

Clinical information included demographic characteristics, existing comorbidities, ongoing medication regimens, and baseline laboratory results. In addition, comprehensive echocardiographic assessments and right and left heart catheterization measurements were obtained.

Transthoracic echocardiography was performed with a Vivid E95 ultrasound platform (General Electric Vingmed Ultrasound, Milwaukee, WI) using 3.5-MHz or M5S transducers. Parameters evaluated included left ventricular ejection fraction (LVEF) via the biplane Simpson method, left-sided chamber dimensions, and valvular pathologies. Pulmonary artery systolic pressure (PASP) was estimated from tricuspid regurgitant jet velocity using the Bernoulli equation with right atrial pressure estimation. Right ventricular function was assessed by tricuspid annular plane systolic excursion (TAPSE) and tissue Doppler-derived systolic velocity of the tricuspid annulus (RV-St). Measurements followed the standards of the American Society of Echocardiography and the European Association of Cardiovascular Imaging.[12]

All patients underwent right and left heart catheterization the day before LVAD implantation. Vascular access was obtained from the femoral vessels under ultrasound guidance. Pressure measurements were standardized with the mid-thoracic line as the zero reference. Aortic and intracardiac pressures were obtained using 6F pigtail catheters, while pulmonary capillary wedge pressure (PCWP) was recorded with a Swan-Ganz catheter. Cardiac output was determined by the Fick principle, and stroke volume, stroke volume index, and cardiac index were calculated according to body surface area and heart rate.

Derived indices of right ventricular–pulmonary arterial coupling were computed as follows:

• Right ventricular stroke work index (RVSWI): (mean pulmonary artery pressure – mean right atrial pressure) × stroke volume index,

• Pulmonary artery pulsatility index (PAPi): (systolic pulmonary artery pressure – diastolic pulmonary artery pressure) / mean right atrial pressure,

• RAP/PCWP ratio: Mean right atrial pressure/PCWP

This study was approved by the İstanbul Medipol University (Date: 11.07.2025, Decision no: 810). This study was conducted in accordance with the Declaration of Helsinki. All procedures complied with the Declaration of Helsinki. Written informed consent for scientific use of clinical data was obtained from each participant at the time of hospital admission and before invasive procedures.

Statistical Analysis

Normality of variable distribution was evaluated using the Kolmogorov–Smirnov test and inspection of histograms. Continuous data were expressed as mean±standard deviation for normally distributed variables or as median with interquartile ranges for skewed variables. Categorical data were presented as counts and percentages.

Patients were grouped according to survival status at long-term follow-up. Comparisons between groups were performed using the Chi-square test for categorical variables and the Mann– Whitney U test for continuous variables. The primary outcome was all-cause long-term mortality.

Univariate Cox proportional hazards regression was applied to identify potential predictors. Variables reaching significance were subsequently included in a multivariate Cox regression model. Forest plots were used to visualize the results of multivariate analysis.

A composite risk score was created from the β-coefficients of the multivariate model. Patients were divided into tertiles of risk, and mortality rates were compared among these categories using bar charts and Kaplan–Meier survival curves. Discriminative performance of the risk score was assessed with receiver operating characteristic (ROC) analysis and comparison against individual predictors.

Analyses were conducted using Python version 3.11 (Python Software Foundation, Wilmington, DE, USA). Statistical significance was defined as a two-tailed p<0.05.

Results

A total of 35 patients were analyzed, with 21 alive and 14 deceased at the end of follow-up. Median follow-up time for the entire cohort was 550 days (IQR: 169.5–1469.5).

Patients who died during follow-up were significantly older than survivors (68.2±8.2 vs. 55.9±9.4 years, p<0.001). Sex distribution was similar between groups (male: 85.7% vs. 76.2%, p=0.796). A greater proportion of deceased patients had a history of coronary artery bypass grafting (CABG) (35.7% vs. 4.7%, p=0.05). Other baseline clinical characteristics, comorbidities, medication use, and LVAD type were comparable (Appendix 1).

The deceased group showed worse baseline renal function, as reflected by higher creatinine values (1.88±0.93 vs. 1.40±0.79 mg/dL, p=0.030), along with lower serum albumin (3.43±0.62 vs. 4.04±0.50 g/dL, p=0.007), total cholesterol (126.0±40.2 vs. 169.3±39.9 mg/dL, p=0.005), and thyroid-stimulating hormone (TSH) levels (1.64±1.69 vs. 4.10±3.24 µIU/mL, p=0.007). Inflammatory ratios also differed, with a higher neutrophil-to-lymphocyte ratio (NLR) (11.00±7.43 vs. 7.47±9.48, p=0.021) and lower lymphocyte-to-monocyte ratio (LMR) (1.16±0.71 vs. 2.21±1.42, p=0.018) in the deceased group. Other laboratory markers were not significantly different between groups (Appendix 1).

Interestingly, patients who died exhibited higher baseline LVEF values compared with survivors (20.4±4.1% vs. 18.3±2.5%, p=0.039). Other echocardiographic indices (LV size, TAPSE/PASP, RV dimensions, RV-St) and invasive hemodynamic measurements (including PCWP, PAPi, RAP/PCWP, cardiac output and index, PVR, and RVSWI) were similar across groups (Appendix 2).

In univariate Cox regression analysis, advanced age, prior CABG, elevated creatinine, and reduced albumin were significantly associated with long-term mortality (Table 1). These variables were entered into the multivariate model. Of these, only age remained an independent predictor of long-term mortality (HR: 1.48 per 5 years; 95% CI: 1.03–2.12; p=0.032) (Table 2 and Fig. 1).

Table 1. Univariate Cox regression analysis for the prediction of long-term mortality

Table 2. Multivariate Cox regression analysis for the prediction of long-term mortalitys

Figure 1. Forest plot of multivariate Cox regression analysis. CABG: Coronary artery bypass grafting.

The CACA score was constructed from β-coefficients of the multivariate regression:

CACA = (0.392×Age/5) + (1.087×CABG [yes=1]) + (0.474×Creatinine) – (0.258×Albumin). Patients were classified into three categories based on score percentiles: low-risk (0–33rd), intermediate-risk (34–66th), and high-risk (67–100th). Mean scores were 3.37±0.50, 4.58±0.43, and 6.23±0.73, respectively (Table 3). Corresponding mortality rates were 0% in the low-risk, 45.4% in the intermediate-risk, and 75% in the high-risk group (Fig. 2).

Table 3. Risk score values for risk groups and mortality rates

Figure 2. Long-term mortality rates according to risk groups.

In ROC analysis, the CACA score achieved the best discriminatory capacity for mortality (AUC: 0.88), outperforming each individual predictor included in the model (Fig. 3). Kaplan–Meier analysis demonstrated poorer survival in the high-risk group compared with the intermediate group, with borderline significance (log-rank p=0.055; Fig. 4).

Figure 3. Receiver operating characteristic curves of risk score and its individual components for the prediction of long-term mortality.

Figure 4. Kaplan–Meier survival curves of high- and intermediate risk groups.

Discussion

Different etiologies appear to contribute to mortality at different time points in LVAD recipients. Factors, such as INTERMACS status, multiple organ failure, sepsis, and right heart failure remain among the most significant predictors of early mortality following LVAD implantation.[2,3] However, in the long term, the causes of mortality tend to shift, and factors, such as age, prior surgeries, comorbidities, such as renal dysfunction, laboratory markers, and right ventricular function parameters gain greater prognostic importance.[4–11]

In the HeartMate 3 risk score, predictors of long-term mortality included age, prior CABG surgery, elevated blood urea nitrogen, low serum sodium, small left ventricular size, and the RAP/ PCWP ratio.[4] The J-MACS risk score identified age, history of cardiac surgery, serum creatinine, and the CVP/PCWP ratio as key components.[6] In the Penn-Columbia study, age, creatinine, total bilirubin, body mass index, right ventricular dysfunction, and aortic insufficiency were incorporated into the risk model. [7] Nayak et al.[11] reported that pre-implant blood urea nitrogen, prior CABG or valve surgery, post-implant ventricular arrhythmias, renal impairment, and hemocompatibility-related adverse events were key predictors for 5-year mortality. Inflammation surrogates, such as NLR and CRP have also been associated with long-term mortality.[9,10] In summary, although there are similarities among the components of these risk scores, no clear consensus exists. Furthermore, these risk scores may not accurately predict long-term mortality in other patient populations. For example, in the study by Moeller et al.,[13] the HeartMate 3 risk score was associated with post-implant survival, but it was not an independent predictor of long-term mortality after adjusting for ischemic etiology and severe diabetes. These discrepancies underscore the need for new, validated approaches.

Age appears to be a consistent predictor of long-term mortality following LVAD implantation.[4,6,7,14] In the study of Radhoe et al.,[14] older age was associated with higher mortality and an increased incidence of bleeding events, but not with heart failure hospitalization, right ventricular failure, ventricular arrhythmia, or device-related complications, such as pump thrombosis or device infection. In line with these findings, our cohort demonstrated that age was the only independent predictor of longterm mortality in the multivariate model.

Albumin is a marker of nutritional status, inflammation, hepatic function, and catabolism, making it a valuable global indicator of physiologic reserve. Previous reports linked low baseline albumin to worse long-term outcomes after LVAD implantation, although it has not been consistently incorporated into risk models.[15–17] Thus, the prognostic significance of pre-implantation albumin levels remains unclear.

A history of prior CABG surgery increases the risk of a complicated intraoperative and post-operative course due to prolonged cardiopulmonary bypass and cross-clamping times, increased intra-operative bleeding, and a higher likelihood of post-operative right heart failure.[18] Consequently, previous CABG surgery has been included as a risk factor in several established risk scores.[4,6,11] Consistent with these studies, a history of CABG was associated with long-term mortality in our study based on univariate Cox regression analysis.

Baseline creatinine level is also a well-established predictor of long-term mortality, and renal function markers – such as creatinine, blood urea nitrogen, and glomerular filtration rate – have been incorporated into various risk prediction models.[4,6,7,11]

Right ventriculo-arterial coupling parameters, such as the RAP/ PCWP ratio and PAPi, have been shown to be associated with long-term survival after LVAD implantation in several studies. [3,4,19–22] An inadequate right ventricular response to elevated PCWP may indicate coexisting right ventricular failure in addition to left ventricular failure in patients undergoing LVAD implantation. After LVAD implantation, right ventricular failure may become more pronounced due to factors, such as increased RV preload, leftward septal shift, and loss of pericardial constraint.[23] Surrogates of right ventriculo-arterial coupling – particularly the RAP/PCWP ratio – have been included in some risk prediction models,[4,6] though not all studies have incorporated them.[7,11]

In our multivariable analysis, age emerged as the only independent predictor of long-term mortality. This finding is likely influenced by the limited sample size and event count, which may have reduced the statistical power needed to detect independent associations for other clinically relevant variables. Although prior CABG, creatinine, and serum albumin levels were all significant in univariate analysis – and have been consistently reported in the literature as prognostically important – they did not retain significance in the multivariable model, possibly due to the relatively small cohort and overlapping confidence intervals. Similarly, right ventricular–pulmonary arterial coupling parameters, such as RAP/PCWP and PAPi – known from prior studies to be strong predictors of right heart failure and adverse outcomes after LVAD implantation – did not reach statistical significance in our population, likely reflecting both cohort homogeneity and sample size limitations. In our study, several right ventricular–pulmonary arterial coupling parameters, including RAP/PCWP and PAPi, did not reach statistical significance despite their well-established prognostic importance in the LVAD literature. This absence of significance may be explained by several methodological characteristics of our cohort. First, the relatively small sample size and limited number of events inevitably reduce statistical power. Second, the values of these hemodynamic parameters were remarkably similar between survivors and non-survivors, which further attenuate the ability to detect meaningful differences. Third, early post-operative mortality related to acute right ventricular failure was intentionally excluded from the study population. As a result, the cohort represents a more homogeneous long-term LVAD population in whom baseline RV dysfunction may exert less pronounced prognostic impact compared with early post-implant phases. These factors likely contributed to the lack of statistical significance for RV parameters despite their recognized clinical relevance.

To avoid overlooking clinically meaningful predictors, we incorporated all univariate predictors into the CACA score using their β-coefficients. This strategy allowed us to integrate multiple physiologically relevant factors into a combined tool with superior discriminatory performance, as demonstrated by the high AUC value (0.88). Although age remained the only independent factor in the regression model, the composite score derived from these variables provided more robust risk stratification than any single parameter. The limited sample size may have led to an underestimation of the predictive value of certain parameters. Nonetheless, identifying strong predictors that maintain statistical significance even in small cohorts may offer valuable insights. Validation of the risk score developed in our study in larger patient populations through future research may help better define its prognostic value for long-term survival.

Study Limitations

This study has several limitations that should be acknowledged. First, it was conducted in a single-center with a relatively small sample size, which may limit the generalizability of the findings to broader populations. The small number of events, particularly in the low-risk group, may reduce the statistical power to detect associations with less prevalent variables. In addition, while the CACA risk score showed promising discriminatory performance within this cohort, it has not yet been externally validated. Therefore, prospective studies with larger, multicenter populations are necessary to confirm the predictive accuracy and clinical applicability of this risk model.

Conclusion

In this single-center experience, age, pre-operative serum albumin, renal function, and a history of CABG were identified as significant predictors of long-term mortality after LVAD implantation. From these variables, we derived the CACA score, which effectively stratified patients into low-, intermediate-, and high-risk groups with distinctly different survival outcomes. Although only age maintained independent significance in multivariate analysis, the overall score demonstrated superior discriminatory capacity compared with individual predictors (AUC 0.88). These findings emphasize the value of integrating simple clinical and laboratory parameters into a practical risk tool for long-term prognostication. While our findings highlight key prognostic factors, validation in larger, multi-center cohorts is necessary to confirm the clinical utility and generalizability of this risk stratification model.

Appendix 1. Comparison of alive and deceased patients in terms of baseline clinical characteristics, medications, and laboratory parameters

Appendix 1. Cont.

Appendix 2. Comparison of alive and deceased patients in terms of echocardiographic parameters and cardiac catheterization measurements

Cite This Article: Keskin B, Hakgör A, Demir İ, Erkanlı K, Çakal B, Yıldız Y, et al. Identification of Long-Term Mortality Predictors and Risk Score Development in Left Ventricular Assist Device Recipients. Koşuyolu Heart J 2026;29(1):1–7

The study was approved by the İstanbul Medipol University Ethics Committee (no: 810, date: 11/07/2025).

Informed consent was obtained.

Externally peer-reviewed.

Concept – B.K., A.H., B.B., İ.O.K.; Design – B.K., B.Ç., İ.O.K.; Supervision – K.E., B.Ç., Y.Y., B.B., İ.O.K.; Resource – İ.D., K.E., B.B.; Materials – İ.D., K.E., Y.Y., İ.Ö.K.; Data collection and/or processing – B.K., A.H., İ.D., Y.Y., B.Ç.; Analysis and/or interpretation – B.K., A.H., B.Ç.,Y.Y.; Literature review – B.K., A.H., İ.D., İ.O.K.; Writing – B.K., B.Ç., B.B., İ.O.K.; Critical review – A.H., K.E., B.Ç., Y.Y.

All authors declared no conflict of interest.

Artificial intelligence (AI)–assisted technologies were used only for language editing during the preparation of this manuscript.

The authors declared that this study received no financial support.

References

1. Rose EA, Gelijns AC, Moskowitz AJ, Heitjan DF, Stevenson LW, Dembitsky W, et al. Long-term use of a left ventricular assist device for end-stage heart failure. N Engl J Med 2001;345:1435–43.
2. Akin S, Soliman O, de By TMMH, Muslem R, Tijssen JGP, Schoenrath F, et al. Causes and predictors of early mortality in patients treated with left ventricular assist device implantation in the European Registry of Mechanical Circulatory Support (EUROMACS). Intensive Care Med 2020;46:1349–60.
3. Guglin M, Omar HR. Right atrial pressure predicts mortality among LVAD recipients: analysis of the INTERMACS database. Heart Lung Circ 2021;30:592–9.
4. Mehra MR, Nayak A, Morris AA, Lanfear DE, Nemeh H, Desai S, et al. Prediction of survival after implantation of a fully magnetically levitated left ventricular assist device. JACC Heart Fail 2022;10:948–59.
5. Bhat G, Yost GL, Ibrahim K, Pappas P, Tatooles A. Risk stratification with longitudinal neutrophil to lymphocyte ratio assessment after left ventricular assist device implantation. Int J Artif Organs 2018;41:445–51.
6. Imamura T, Kinugawa K, Nishimura T, Toda K, Saiki Y, Niinami H, et al. Novel scoring system to risk stratify patients receiving durable left ventricular assist device from j-macs registry data. Circ J 2023;87:1103–11.
7. Birati EY, Hanff TC, Maldonado D, Grandin EW, Kennel PJ, Mazurek JA, et al. Predicting long term outcome in patients treated with continuous flow left ventricular assist device: the penn-columbia risk score. J Am Heart Assoc 2018;7:e006408.
8. Adamo L, Nassif M, Tibrewala A, Novak E, Vader J, Silvestry SC, et al. The Heartmate Risk Score predicts morbidity and mortality in unselected left ventricular assist device recipients and risk stratifies INTERMACS class 1 patients. JACC Heart Fail 2015;3:283–90.
9. Tie H, Shi R, Welp H, Martens S, Li Z, Sindermann J, et al. C-reactive protein predicts early clinical outcomes and long-term mortality after left ventricular assisted device. Int J Artif Organs 2022;45:497–505.
10. Sundararajan S, Kiernan MS, Couper GS, Upshaw JN, DeNofrio D, Vest AR. The neutrophil-lymphocyte ratio and survival during left ventricular assist device support. J Card Fail 2019;25:188–94.
11. Nayak A, Hall SA, Uriel N, Goldstein DJ, Cleveland JC Jr, Cowger JA, et al. Predictors of 5-year mortality in patients managed with a magnetically levitated left ventricular assist device. J Am Coll Cardiol 2023;82:771–81.
12. Lang RM, Badano LP, Mor-Avi V, Afilalo J, Armstrong A, Ernande L, et al. Recommendations for cardiac chamber quantification by echocardiography in adults: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. J Am Soc Echocardiogr 2015;28:1–39.
13. Moeller CM, Rubinstein G, Oren D, Valledor AF, Lotan D, Raikhelkar JK, et al. Validation of the heartMate 3 survival risk score in a large left ventricular assist device center. J Thorac Cardiovasc Surg 2025;169:650–7.
14. Radhoe SP, Veenis JF, Jakus N, Timmermans P, Pouleur AC, Rubís P, et al. How does age affect outcomes after left ventricular assist device implantation: results from the PCHF-VAD registry. ESC Heart Fail 2023;10:884–94.
15. Imamura T, Combs P, Siddiqi U, Mirzai S, Stonebraker C, Bullard H, et al. Perioperative improvement in serum albumin level in patients with left ventricular assist device. J Card Surg 2020;35:3070–7.
16. Kato TS, Kitada S, Yang J, Wu C, Takayama H, Naka Y, et al. Relation of preoperative serum albumin levels to survival in patients undergoing left ventricular assist device implantation. Am J Cardiol 2013;112:1484–8.
17. Gopal DJ, Hanff TC, Mazurek JA, Grandin WE, Howard J, Forde-McLean R, et al. Prognostic implications of changes in albumin following left ventricular assist device implantation in patients with severe heart failure. Am J Cardiol 2017;120:2003–7.
18. Fitzpatrick JR, Frederick JR, Hsu VM, Kozin ED, O'Hara ML, Howell E, et al. Risk score derived from pre-operative data analysis predicts the need for biventricular mechanical circulatory support. J Heart Lung Transplant 2008;27:1286–92.
19. Beneyto M, Martins R, Galand V, Kindo M, Schneider C, Sebestyen A, et al. Right ventriculoarterial coupling surrogates and long-term survival in LVAD recipients: results of the ASSIST-ICD multicentric registry. J Card Fail 2025;31:388–96.
20. Gulati G, Grandin EW, DeNofrio D, Upshaw JN, Vest AR, Kiernan MS. Association between postoperative hemodynamic metrics of pulmonary hypertension and right ventricular dysfunction and clinical outcomes after left ventricular assist device implantation. J Heart Lung Transplant 2022;41:1459–69.
21. Akamkam A, Galand V, Jungling M, Delmas C, Dambrin C, Pernot M, et al. Association between pulmonary artery pulsatility and mortality after implantation of left ventricular assist device. ESC Heart Fail 2024;11:2100–12.
22. Ruiz-Cano MJ, Morshuis M, Koster A, Lauenroth V, Prashovikj E, Gummert J, et al. Risk factors of early right ventricular failure in patients undergoing LVAD implantation with intermediate Intermacs profile for advanced heart failure. J Card Surg 2020;35:1832–9.
23. Keskin B, Hakgor A, Yilmaz B, Erkanli K, Cakal B, Yazar A, et al. Right atrial pressure/pulmonary capillary wedge pressure ratio predicts in-hospital mortality in left ventricular assist device recipients. J Clin Med 2025;14:4784.

Figure and Tables

Figure 1. Forest plot of multivariate Cox regression analysis. CABG: Coronary artery bypass grafting.

Figure 2. Long-term mortality rates according to risk groups.

Figure 3. Receiver operating characteristic curves of risk score and its individual components for the prediction of long-term mortality.

Figure 4. Kaplan–Meier survival curves of high- and intermediate risk groups.

Table 1. Univariate Cox regression analysis for the prediction of long-term mortality

Table 2. Multivariate Cox regression analysis for the prediction of long-term mortalitys

Table 3. Risk score values for risk groups and mortality rates

Appendix 1. Comparison of alive and deceased patients in terms of baseline clinical characteristics, medications, and laboratory parameters

Appendix 1. Cont.

Appendix 2. Comparison of alive and deceased patients in terms of echocardiographic parameters and cardiac catheterization measurements