Value of nitroglycerin test in the diagnosis of heart failure in emergency department patients with undifferentiated dyspnea

Abstract Background Rapid diagnosis of heart failure (HF) in acutely dyspneic patients can be challenging for emergency department (ED) physicians. Hypothesis Cardiac output (CO) change with sublingual nitroglycerin (NTG) could be helpful in the diagnosis of HF in patients with acute undifferentiated dyspnea. Materials and Methods A prospective study of patients >18 years admitted to the ED for acute dyspnea. Using thoracic bioimpedance, we measured CO change at baseline and after sublingual administration of 0.6 mg of NTG. HF was defined on the basis of clinical examination, pro‐brain natriuretic peptide levels, and echocardiographic findings. Diagnostic performance of delta CO was calculated by sensitivity, specificity, likelihood ratio and receiver operating characteristic (ROC) curve. Results This study included 184 patients with mean age of 64 years. Baseline CO was comparable between the HF group and the non‐HF group. At its best cutoff (29%), delta CO showed good accuracy in the diagnosis of HF with a sensitivity, specificity, positive and negative likelihood ratios of 80%, 44%, 57%, and 66% respectively. Area under ROC curve was 0.701 [95% CI 0.636–0.760]. The decrease of CO with sublingual NTG was significantly higher in patients with HFpEF compared with those with HFrEF. Multivariate analysis, showed that delta CO was an independent factor associated with HF diagnosis [OR 0.19 (95% CI 0.11–0.29); p < .001]. Conclusions Our study showed that CO change with sublingual nitroglycerin is a simple tool that may be helpful for the diagnosis of HF in ED patients with undifferentiated dyspnea.

Conclusions: Our study showed that CO change with sublingual nitroglycerin is a simple tool that may be helpful for the diagnosis of HF in ED patients with undifferentiated dyspnea.

| INTRODUCTION
Heart failure (HF) is among the most costly chronic diseases with a major impact on the patients' quality of life. Its evolution is characterized by recurrent exacerbations which represent a frequent reason for admission to the emergency department (ED). 1 Dyspnea is the main symptom of HF exacerbations, but differentiation of dyspnea related to HF from other causes can be challenging despite improved diagnostic tools ranging from chest X-ray to more sophisticated and less accessible techniques such as cardiac ultrasound or pulmonary arterial catheterization. 2 Early diagnosis is nevertheless a prerequisite for appropriate management. Delays in establishing the correct diagnosis and appropriate treatment may influence subsequent length of hospital stay, morbidity, and mortality. 3,4 Numerous studies have shown that the measurement of cardiac output (CO) or its surrogate after preload-modifying manoeuvers could be a reliable and accurate method to investigate hemodynamic status. Depending on cardiac function, a given change in preload could lead to either a significant or a negligible increase in stroke volume. [5][6][7][8] In a normal heart, the more the left ventricle is filled, the better it contracts and the higher is the volume of systolic ejection (steep slope of the Frank-Starling curve).
Preload dependence of CO is less detectable in the failing heart with regard to the flat cardiac function curve. [9][10][11] Postural maneuvers that change cardiac venous return have been proposed as a useful means for the diagnosis of HF in dyspneic patients, but this has rarely been described with pharmacological interventions. [12][13][14] The aim of our study was to evaluate the utility of sublingual nitroglycerin administration (NTG test) and its effects on non-invasive cardiac output measurement in the diagnosis of acute HF in ED patients presenting with acute dyspnea.

| Participants and methods
Patients' demographic characteristics as well as clinical examination data including age, sex, height, weight, history, current drug treatment, blood pressure, heart rate, respiratory rate, SpO2, and temperature were collected. Standard biological examinations and blood gases as well as cardiac enzymes and the pro-brain natriuretic peptide (pro-BNP) were measured at admission. The definitive diagnosis of HF was made by two clinical experts on the basis of clinical data, echocardiographic findings and pro-BNP level. CO measurement was performed by transthoracic impedance using the Biopac system (Biopac Student Lab software version 3.7.2). Artifacts in the bioimpedance signal were detected and excluded from the study. In practice, the device was connected using four electrodes placed at the base of the neck (posterior) and at the base of the thorax (posterior). At each level, two electrodes were placed 5 cm apart and the 2 levels were separated by a distance of 28 cm. The electrocardiographic recording was taken simultaneously with 2 other electrodes placed at the level of the right upper limb and lower left limb. The parameter measured by the transthoracic impedance was stroke volume (SV). It was calculated from the formula: SV = VTEP Â VET Â ([dZ/dtmax]/ BCI) where VTEP is the volume of participating electrical tissue, VET is the left ventricular ejection time, dZ/dtmax is the rate of impedance change during systole, and BCI is the baseline chest impedance. Cardiac output was calculated instantaneously from the systolic ejection volume (SEV) and the cardiac frequency. The device was set up in such a way that data acquisition was averaged over 10 cardiac cycles. The average of 3 measurements of the SEV was retained provided that the values did not differ more than 10% from each other. All the measurements were made by the same investigators who were not aware of the clinical and biological details of the patient. The treating physicians were blinded to the results of the thoracic bioimpedance. Each patient was initially placed in a semi-sitting position at 30 for 5 min and then CO is measured (baseline CO). NTG (0.6 mg) was then given to the patient sublingually and CO measurement was repeated. CO was calculated by averaging three measurements at 1 min intervals at baseline and after NTG administration. Delta CO was defined as the percent of change of baseline CO after NTG test.

| Statistical analysis
Variables were expressed as mean ± standard deviation, or median and interquartile range as appropriate. The patients were divided into two groups: a group of patients identified as having HF (HF group) and a group of patients where the diagnosis of HF was excluded (non HF group). Comparisons between both groups were made among continuous variables using Student's t test for independent samples.
Chi-square was used for discrete variables. Receiver operating characteristic (ROC) curves for predicting HF were constructed and the area under the curve (AUC) was measured for delta CO. Sensitivity, specificity, positive and negative predictive values, and likelihood ratios of positive and negative results were calculated using the optimal cutoff value of delta CO. Multivariate logistic regression analysis was used to identify the independent predictors of HF in patients with acute dyspnea. Age, sex, and clinical variables that were statistically significant after univariate analysis were included in this analysis. Each variable of interest was examined with univariate analysis, and variables that were significant at the 0.20 level were included in the logistic regression. We estimated a sample size of 110 subjects with acute dyspnea, assuming an anticipated delta CO sensitivity of at least 90% with a power (β) = 80%, and a probability of type-1 error (α) = 0.05. A pvalue of .05 was considered statistically significant. Calculations were performed with SPSS version 20 software package for Windows (SPSS Inc, Chicago, IL).

| RESULTS
Two hundred and eighty-four patients were included Table 1 summarizes the demographic and clinical characteristics of the overall population and the two study groups. Patients in the HF group (n = 143) were significantly older, with significantly higher rates of diabetes, hypertension, heart failure and coronary disease. Male predominance was observed in both groups. Mean left ventricular ejection fraction was 48 ± 14% and 63 ± 12% in the HF and the non-HF groups respectively. In the HF group, 56 (39%) had reduced LVEF (HFrEF), and 86 (61%) had preserved LVEF (HFpEF). Baseline CO was comparable between the two groups (6.08 and 6.04 L/min respectively for the HF group and the non-HF group. Cardiac output decreased significantly in both groups, but this decrease was greater in the non-HF group ( Figure 1). In patients with HF, baseline CO was not significantly different between the HFrEF and HFpEF groups (6.0 ± 1.3 versus 5.9 ± 1.1, respectively). Cardiac output decreased significantly in both sub-groups but the decrease was significantly higher in the HFpEF group compared with the HFrEF group (À45 ± 1.5 versus

| CONCLUSION
In this study, we showed that NTG test coupled to a fast and noninvasive measurement of CO may be a significant aid to the conventional diagnostic approach of HF. However, further studies are needed to validate our findings.

CONFLICT OF INTEREST
The authors declare no potential conflict of interest.

DATA AVAILABILITY STATEMENT
Research data are not shared.