Introduction

Heart failure (HF) is among the leading causes of hospitalization globally.¹ The American Heart Association reported that approximately 6 million Americans had HF between 2015 and 2018, and the population rate is expected to be 3.0% by 2030.² In Asian countries, its prevalence has been reported to be approximately 4.2 million people.³ The incidence of HF varies depending on many factors such as risk factors, etiology, region, and race.^1,2,4,5 The risk of encountering HF throughout life increases with the prolongation of lifespan and the addition of individual factors.^2,4,6

The 5-year mortality has been reported as 53% and 67% after diagnosis of HF.⁷ The deaths from HF have remained unchanged despite recent advances in evidence-based treatment modalities.¹ According to the Centers for Disease Control and Prevention data, HF is mentioned in 1 in 8 death certificates among the causes of death.² Previous hospital admission due to HF was found to be associated with increased readmission and mortality.^2,6,8 In addition, it is expected that HF expenditures, which were approximately 30 billion dollars in 2012 in the USA, will reach approximately 70 billion dollars in 2030 with an increase of 127%.²

Approximately, 80% of acute HF patients are admitted via emergency departments (EDs).⁹ Thus, EDs and emergency physicians play a crucial role in the management of patients with acute HF. However, it can be challenging for emergency physicians to decide on admission or discharge. Inappropriate admissions will increase hospital costs, and inappropriate discharges will increase the risk of adverse outcomes.⁹ Hence, it is crucial to discriminate between critical and noncritical patients. There is a need for reliable clinical decision support systems that are validated regionally and internationally.

In the literature, various derivation and validation studies of scoring systems have been carried out to discriminate critical HF patients.^3,10-12 However, the majority of studies were conducted among patients admitted to wards and/or intensive care units (ICUs). Studies including acute HF patients discharged from EDs after appropriate treatment and follow-up are limited.

Peterson et al¹³ developed and validated the Get With The Guidelines-HF (GWTG-HF) score using the GWTG program data. To the best of our knowledge, the GWTG-HF score has not been validated in the Turkish HF patient population.

This study aims to evaluate the prognostic accuracy of the GWTG-HF score and shock index (SI) alone and with lactate in patients with acute symptomatic HF.

Methods

Study Design

This retrospective cohort study was conducted in the ED of a tertiary training and research hospital between January 1, 2019, and December 31, 2019. The study hospital is a primary percutaneous coronary intervention, stroke, and level 3 trauma center. Approximately 350 000 patients present annually to the study ED. Local Ethics Committee approval was obtained before the study (approval ID: 2021/567; January 3, 2022). The study was conducted in line with the Declaration of Helsinki. Artificial intelligence technologies were not used in the production of the manuscript.

Patient Selection and Groups

Patients aged >18 years and diagnosed with acute symptomatic HF in the ED and patients who were Stage C and Stage D according to the American College of Cardiology and American Heart Association (ACC/AHA) stage definition were consecutively included in the study.¹⁴ Patients referred from another center and missing medical records were excluded. The informed consent was waived due to the retrospective design of the study.

Arrival type, vital parameters, demographic characteristics, consciousness status, comorbid diseases, laboratory results, ED length of stay, and outcomes (in-hospital, 24-hours, and 30-day mortality) of the patients were recorded. Patient data were obtained from electronic medical records and ED files. The values of the patients at the time of the initial examination were used in the calculation of scores and statistical analysis. Patients were divided into groups according to in-hospital mortality.

Acute HF was defined with Framingham criteria, clinical findings, X-ray, and echocardiographic evaluation.^15,16

Scores

The GWTG-HF score was developed and validated to predict in-hospital mortality in patients hospitalized with HF by Peterson et al¹³ in 2010. The GWTG-HF score is calculated with 7 variables (systolic blood pressure (SBP), blood urea nitrogen (BUN) value, sodium value, age, heart rate (HR), chronic obstructive pulmonary disease history, and race). Score points range from 0 to 100 and can be easily calculated using an online calculator (https://www.mdcalc.com/gwtg-heart-failure-risk-score).

The SI, modified SI (MSI), and age SI (Age-SI) are calculated using the parameters of HR, SBP, mean arterial pressure (MAP), and age. The formulas used to calculate the scores are given below.

SI = HR/SBP; MSI = HR/MAP; Age-SI = Age X SI

Outcomes

The primary endpoint of the study was the prognostic accuracy of the scores alone and with lactate in predicting in-hospital mortality in patients with acute HF.

The secondary endpoint is the predictive value of the scores alone and with lactate in predicting 24-hour and 30-day mortality of the patients and factors affecting the in-hospital mortality of the patients.

Statistical Analysis

The sample size was calculated using Kirkwood et al’s¹⁷ study as a reference (n = 105 388, mortality rate: 4%).¹⁷ The sample size was calculated as 234 acute HF patients (expected mortality = 8%, power = 80%, and alpha = 0.05). An online calculator was used to calculate the sample size (www.clincalc.com).

Categorical variables are presented as numbers and percentages. Numerical variables are presented as mean ± standard deviation or median (interquartile range). The compliance to the normal distribution of numerical variables of the groups tested by the Kolmogorov–Smirnov normality test. Categorical variables were evaluated with the chi-square test. According to normality test results, numerical variables were evaluated with the Student t-test or Mann–Whitney U-test. Variables with a P-value less than .05 and did not correlate strongly with each other in bivariate correlation analyses were included in the logistic regression model. The effects of prognostic factors on HF were analyzed with simple and multiple logistic regression analyses. Odds ratios (ORs) were presented with a 95% confidence interval. The area under the receiver-operating characteristic (AUROC) curve was used to evaluate the performance of the scores. DeLong’s test was used to compare ROC curves. The Youden index J point was used to determine the optimal cutoff values. The statistical significance level was determined as P < .05. SPSS® for Windows version 23.0 (IBM, Chicago, Ill, United States) was used for statistical analysis.

Results

A total of 378 consecutive patients were included in the study. Patients referred from another center (n = 6) and missing medical records (n = 4) were excluded. A total of 368 patients were included in the final analysis. The median age of the patients was 74 (65-81), and 197 (53.5%) were women. All patients had HF with reduced ejection fraction (EF < 40%). All patients received appropriate treatment according to the current guidelines. Mechanical ventilation (MV) support was applied to 145 (39.4%) patients in the ED. Two hundred fifty (67.9%) of them were discharged from the ED after initial treatment and 118 (32.1%) were hospitalized in the ward and ICU. The in-hospital mortality rate of the patients was 7.6% (n = 28). The 24-hour mortality and 30-day mortality rates of the patients were 1.9% and 12.5%, respectively. The baseline characteristics of the patients are shown in Table 1.

In univariate analyses, low blood pressure, peripheral oxygen saturation, and sodium values, and increased BUN, lactate dehydrogenase, C reactive protein, and lactate values were found to be associated with mortality (P < .05). Arrival by ambulance, receiving home oxygen therapy, new-onset altered mental status, presence of cerebrovascular disease (CVD), MV need, and ICU admission were found to be associated with mortality (P < .05). Gender and chronic obstructive pulmonary disease history were not associated with mortality (P > .05). In multivariate analyses, SBP (OR = 0.95, 95% CI: 0.93-0.98, P < .001), CVD history (OR = 7.69, 95% CI: 1.03-56.97, P = .046), ICU admission (OR = 86.9, 95% CI: 8.4-889.4, P < .001), sodium (OR = 0.82, 95% CI: 0.70-0.96, P = .017), BUN (OR = 1.08, 95% CI: 1.04-1.13, P < .001), and lactate (OR = 1.55, 95% CI: 1.14-2.10, P = .004) were found to be predictors of mortality. The factors associated with mortality are presented in Table 2 and Table 3.

The GWTG-HF score outperformed other scores in predicting in-hospital, 24-hour, and 30-day mortality. Adding lactate values to the GWTG-HF (GWTG-HF+L) score significantly improved the overall performance of the GWTG-HF score in predicting in-hospital (AUC = 0.807 vs. 0.872, P < .05, DeLong’s test), 24-hour (AUC = 0.868 vs. 0.936, P < .05, DeLong’s test), and 30-day mortality (AUC = 0.770 vs. 0.801, P < .05, DeLong’s test).

The overall performances of SI, MSI, and Age-SI were moderate to good in predicting in-hospital, 24-hour, and 30-day mortality. Adding lactate values to the SI, MSI, and Age-SI improved their overall performance for all 3 outcomes. The performances of the scores for each of the 3 outcomes are shown in Table 4 and Figure 1.

Discussion

In our study, the GWTG-HF score outperformed the SI, MSI, and Age-SI for all 3 outcomes. In addition, our study demonstrated that adding lactate value to the GWTG-HF, SI, MSI, and Age-SI scores improved the overall performance of each score.

The GWTG-HF score demonstrated moderate performance in predicting in-hospital mortality in studies conducted in the United States and Israel.^18,19 Similarly, in Lyle et al’s¹¹ study, the GWTG-HF score demonstrated moderate performance in cardiac ICU patients (with and without HF). In our study, the GWTG-HF score performed better than in other studies in the literature in predicting in-hospital mortality [AUC =0.807 (95% CI 0.763-0.846)]. The GWTG-HF score may be useful in discriminating critically ill patients with acute HF.

In another study, Shiraishi et al³ found moderate performance for the GWTG-HF score in the Japanese patient population and improved the overall performance of the GWTG-HF score by adding the BNP values. In our study, we added lactate value which is associated with mortality to the GWTG-HF score. The overall performance, sensitivity, and specificity of the GWTG-HF+L score were better than those of the original score. Thus, we can say that the GWTG-HF+L score is more successful than the original score in evaluating patients with acute HF.

After the initial stabilization of acute HF patients in the ED, it can be challenging to decide on hospitalization or discharge.⁹ In our study, both the GWTG-HF and GWTG-HF+L scores had high sensitivity (89.2% and 92.8%, respectively) in predicting in-hospital mortality. Therefore, both forms (original and with lactate) of the GWTG-HF score can be used safely in discriminating noncritical acute HF patients, especially in the ED.

Uyar et al²⁰ reported that high lactate levels were associated with cardiovascular death and hospitalization in patients with acute HF during a 6-month follow-up. In another study, SI and lactate demonstrated moderate-to-good overall performance in predicting 24-hour and 28-day mortality in patients with acute HF.¹⁰ In our study, SI, MSI, and Age-SI demonstrated moderate-to-good overall performance in predicting 24-hour, 30-day, and in-hospital mortality. Adding lactate to the SI, MSI, and Age-SI improved the overall performance of the scores for all 3 outcomes. Besides, adding lactate to SI and MSI produced 100% sensitivity for both and 88% and 90% specificity in predicting 24-hour mortality. The GWTG-HF+L score achieved 85% sensitivity and 91% specificity in predicting 24-hour mortality. The existing scores and lactate can be used together to predict early mortality in patients with acute HF.

In our study, the GWTG-HF+L score, SI+L, and MSI+L had high sensitivity in predicting in-hospital and 24-hour mortality (92%, 89%, 89%, and 85%, 100%, 100%, respectively). According to these results, the GWTG-HF+L score, SI+L, and MSI+L can be used safely to discriminate noncritical HF patients in the ED.

Age, SBP, HR, sodium, and BUN were found to be predictors of mortality in 2 cohort studies (ADHERE and GWTG-HF registry).^13,21 In the present study, SBP, sodium, BUN, and lactate were found to be predictors of mortality. Among these variables, lactate, which is not among the GWTG-HF score and SI variables, was added to the scores and improved the overall performance of the scores. Lactate, which is used in the evaluation of the severity of many diseases, can also be used in the evaluation of critical acute HF patients.

Study Limitations

The present study had several limitations. The first is that it is single-center and retrospective. Second, since there were no Black race patients in our study, no comment could be made on the effect of race. Third, the values of the patients at the initial examination were used in the analysis. Changes during follow-up may have affected the patients’ prognosis. Only the ACC/AHA stage C and stage D patients were included in the study. Although the sample size is acceptable, further studies need to represent the general population. Since the GWTG-HF score was not developed for a specific HF population, we did not evaluate the effect of specific devices (cardioverter defibrillator and cardiac resynchronization therapy), and no comments could be made on this issue.

Conclusion

Both the GWTG-HF and GWTG-HF+L scores have acceptable discriminatory power in patients with acute symptomatic HF. The GWTG-HF score, SI, MSI, and Age-SI can be used together with lactate to predict mortality in patients with acute HF. The presence of CVD, SBP, sodium, BUN, lactate, and ICU admission are independent predictors of in-hospital mortality among patients with acute HF.

Footnotes

Ethics Committee Approval: The study protocol was approved by the Ethics Committee of Bakırköy Dr. Sadi Konuk Training and Research Hospital (approval ID: 2021/567; date: January 3, 2022). This study was performed in line with the Declaration of Helsinki.

Informed Consent: The informed consent could not be obtained because of the retrospective nature of the study.

Peer-review: Externally peer-reviewed.

Author Contributions: Study Concept and Design – B.İ., G.B.B., A.B., H.D., A.K.; Acquisition of Data – B.İ., G.B.B.; Analysis and Interpretation of Data – B.İ., G.B.B.; Drafting of Manuscript – B.İ., G.B.B., H.D.; Critical Revision of Manuscript – B.İ., G.B.B., A.B., H.D., A.K.

Declaration of Interests: The authors have no conflicts of interest to declare.

References

1. Ambrosy AP, Fonarow GC, Butler J. The global health and economic burden of hospitalizations for heart failure: lessons learned from hospitalized heart failure registries. J Am Coll Cardiol. 2014;63(12):1123-1133. https://doi.org/10.1016/j.jacc.2013.11.053
2. Virani SS, Alonso A, Aparicio HJ. Heart disease and stroke statistics-2021 update: a report from the American Heart Association. Circulation. 2021;143(8):e254-e743. https://doi.org/10.1161/CIR.0000000000000950
3. Shiraishi Y, Kohsaka S, Abe T. Validation of the Get with the Guideline-Heart Failure risk score in Japanese patients and the potential improvement of its discrimination ability by the inclusion of B-type natriuretic peptide level. Am Heart J. 2016;171(1):33-39. https://doi.org/10.1016/j.ahj.2015.10.008
4. Chang PP, Chambless LE, Shahar E. Incidence and survival of hospitalized acute decompensated heart failure in four US communities (from the Atherosclerosis Risk in Communities Study). Am J Cardiol. 2014;113(3):504-510. https://doi.org/10.1016/j.amjcard.2013.10.032
5. Vivo RP, Krim SR, Liang L. Short- and long-term rehospitalization and mortality for heart failure in 4 racial/ethnic populations. J Am Heart Assoc. 2014;3(5):e001134-. https://doi.org/10.1161/JAHA.114.001134
6. Heidenreich PA, Albert NM, Allen LA. Forecasting the impact of heart failure in the United States: a policy statement from the American Heart Association. Circ Heart Fail. 2013;6(3):606-619. https://doi.org/10.1161/HHF.0b013e318291329a
7. McDonagh TA, Metra M, Adamo M. 2021 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure. Eur Heart J. 2021;42(36):3599-3726. https://doi.org/10.1093/eurheartj/ehab368
8. Bello NA, Claggett B, Desai AS. Influence of previous heart failure hospitalization on cardiovascular events in patients with reduced and preserved ejection fraction. Circ Heart Fail. 2014;7(4):590-595. https://doi.org/10.1161/CIRCHEARTFAILURE.113.001281
9. Miró Ò, Peacock FW, McMurray JJ. European Society of Cardiology - Acute Cardiovascular Care Association position paper on safe discharge of acute heart failure patients from the emergency department. Eur Heart J Acute Cardiovasc Care. 2017;6(4):311-320. https://doi.org/10.1177/2048872616633853
10. Cetinkaya HB, Gunes H. Use of shock index and lactate to predict mortality in acute heart failure patients in emergency department. J Coll Physicians Surg Pak. 2021;31(3):262-266. https://doi.org/10.29271/jcpsp.2021.03.262
11. Lyle M, Wan SH, Murphree D. Predictive value of the Get With The Guidelines Heart Failure risk score in unselected cardiac intensive care unit patients. J Am Heart Assoc. 2020;9(3):e012439-. https://doi.org/10.1161/JAHA.119.012439
12. El-Menyar A, Sulaiman K, Almahmeed W. Shock index in patients presenting with acute heart failure: a multicenter multinational observational study. Angiology. 2019;70(10):938-946. https://doi.org/10.1177/0003319719857560
13. Peterson PN, Rumsfeld JS, Liang L. A validated risk score for in-hospital mortality in patients with heart failure from the American Heart Association get with the guidelines program. Circ Cardiovasc Qual Outcomes. 2010;3(1):25-32. https://doi.org/10.1161/CIRCOUTCOMES.109.854877
14. Heidenreich PA, Bozkurt B, Aguilar D. 2022 AHA/ACC/HFSA guideline for the management of heart failure: a report of the American College of Cardiology/American Heart Association Joint Committee on clinical practice guidelines. J Am Coll Cardiol. 2022;79(17):e263-e421. https://doi.org/10.1016/j.jacc.2021.12.012
15. Ho KK, Anderson KM, Kannel WB, Grossman W, Levy D. Survival after the onset of congestive heart failure in Framingham heart study subjects. Circulation. 1993;88(1):107-115. https://doi.org/10.1161/01.cir.88.1.107
16. McKee PA, Castelli WP, McNamara PM, Kannel WB. The natural history of congestive heart failure: the Framingham study. N Engl J Med. 1971;285(26):1441-1446. https://doi.org/10.1056/NEJM197112232852601
17. Adams KF, Fonarow GC, Emerman CL. Characteristics and outcomes of patients hospitalized for heart failure in the United States: rationale, design, and preliminary observations from the first 100,000 cases in the Acute Decompensated Heart Failure National Registry (ADHERE). Am Heart J. 2005;149(2):209-216. https://doi.org/10.1016/j.ahj.2004.08.005
18. Lagu T, Pekow PS, Shieh MS. Validation and comparison of seven mortality prediction models for hospitalized patients with acute decompensated heart failure. Circ Heart Fail. 2016;9(8):-. https://doi.org/10.1161/CIRCHEARTFAILURE.115.002912
19. Elias A, Agbarieh R, Saliba W. SOFA score and short-term mortality in acute decompensated heart failure. Sci Rep. 2020;10(1):20802-. https://doi.org/10.1038/s41598-020-77967-2
20. Uyar H, Yesil E, Karadeniz M. The effect of high lactate level on mortality in acute heart failure patients with reduced ejection fraction without cardiogenic shock. Cardiovasc Toxicol. 2020;20(4):361-369. https://doi.org/10.1007/s12012-020-09563-9
21. Fonarow GC, Adams KF, Abraham WT, Yancy CW, Boscardin WJ. Risk stratification for in-hospital mortality in acutely decompensated heart failure: classification and regression tree analysis. JAMA. 2005;293(5):572-580. https://doi.org/10.1001/jama.293.5.572