1. Background
Congenital heart disease (CHD) is a major cause of morbidity and mortality in children. Inflammatory serological markers may increase in CHD, as they do in many other chronic diseases, especially in advanced stages (1). An increasing number of studies, mostly performed on adult populations, have assessed the relationship between red cell distribution width (RDW) and disease prognosis. The RDW itself is affected by deficiencies in micronutrients such as iron, folic acid, and vitamin B12, and therefore can be used as a marker for evaluating nutritional status (2).
Some studies have shown racial differences in RDW ranges; however, the number of these studies is limited (3). In addition, they have not differentiated between RDW in blacks, whites, and Hispanics (4). Patients with chronic kidney disease show endothelial dysfunction in carotid intima-media thickness (CIMT) and flow-mediated dilatation (FMD) in the brachial artery with increased RDW, independent of anemia and inflammation (CRP) (5).
Recent studies have investigated the RDW Index for a wide range of acute and chronic diseases, including pulmonary hypertension, renal failure, acute myocardial infarction, heart failure, angina, stroke, and other cardiovascular diseases (2, 6-8). These studies have demonstrated a relationship between elevated RDW and increased postoperative complications and mortality (9). One-fifth of children undergoing cardiac surgery, especially those with cyanotic heart disease, may experience morbidities such as prolonged cardiopulmonary bypass (CPB) time and increased need for post-operative inotropic support (10).
2. Objectives
There are few studies concerning the relationship between distinctly pre-operative RDW and post-operative cardiac function in children with CHD. As no related study has been performed in our country to date, we decided to evaluate this relationship in children following cardiac surgery for CHD.
3. Methods
In this cross-sectional descriptive study, all pediatric (age < 18 years) CHD patients requiring surgical intervention admitted to Pediatric Cardiology Ward of Modarres Hospital from May 2019 to October 2020 were enrolled and evaluated. Patients with hereditary spherocytosis and other inherited anemias, hypothyroidism, liver cirrhosis, kidney failure, infectious diseases, and recent (within 6 months) red blood cell product transfusion were excluded. Ultimately, 76 children met the criteria for inclusion in the study. Written consent was obtained from their parents, and the research was conducted in accordance with the Declaration of Helsinki. The study was approved by the Ethics Committee of the School of Medicine. Additionally, Shahid Beheshti University of Medical Sciences financially supported our study. Our raw data are available in the form of SPSS, and data were analyzed using SPSS.
Demographic information for all patients was collected. Blood samples were drawn for complete blood count (CBC) checking and RDW measurement using the Sysmex KX-21N Hematology Analyzer (Sysmex Corporation, Japan). The reference range for RDW with this analyzer is 8 - 26. Patients’ demographic and laboratory data were recorded in the data sheet.
Bedside echocardiographic examination was performed by a pediatric cardiologist in the intensive care unit (ICU) within 24 hours after cardiac surgery using a Samsung HS70A ultrasound machine (Samsung Company, South Korea). Images were obtained using standard subcostal four-chamber, apical four-chamber, parasternal long-axis and short-axis, and suprasternal long-axis views, applying PE 2-4 and PA 3-8B phased array probes. All patients underwent comprehensive evaluation with two-dimensional, pulsed-wave Doppler, and tissue Doppler imaging (TDI) through the mitral and tricuspid valves and left ventricular outflow tract.
Left ventricular ejection fraction (LVEF) and tricuspid annular plane systolic excursion (TAPSE) were evaluated for left and right ventricular function evaluation. The Left Ventricular Myocardial Performance Index (LV MPI) was calculated based on the equation: "ICT+IRT/ET", where ICT stands for isovolumetric contraction time, IRT for isovolumetric relaxation time, and ET for ejection time, to evaluate global ventricular function. Peak velocity of blood at early diastole (E), late diastole (A), and aortic velocity time integral (VTI) were measured using pulsed-wave Doppler signals.
The TDI also measured peak velocity of the mitral and tricuspid annulus in systole (S’) and peak early and late diastolic filling velocities at lateral and medial annular mitral and lateral annular tricuspid valves. The ratio of early to late diastolic velocity of mitral and tricuspid valves (E’/A’) and mitral peak velocity of early diastolic filling to early diastolic annular velocity (E/E’) were calculated in the four-chamber apical view. The latest recommendations from the American Association of Echocardiography (ASE) were used to measure echocardiographic indicators. Findings were recorded in a data sheet (11).
The CHD type, type of surgical procedure (palliative or corrective), total surgery duration, CPB pump duration, and length of ICU and hospital stay were recorded in the prepared data sheet.
3.1. Statistical Analysis
Data analysis was performed using SPSS 25 for Windows (SPSS Inc., Chicago, IL). The normal distribution of quantitative data was investigated using the Kolmogorov-Smirnov test. Continuous variables were described using mean and standard deviation (mean ± SD). Comparison of continuous variables was performed using the Mann-Whitney test. The correlation between RDW and various clinical and echocardiographic parameters, as well as hospital and ICU stay durations, was examined by Spearman correlation analysis (rho coefficient). All continuous variables were expressed as mean ± standard deviation, and P-values less than 0.05 were considered statistically significant. The ROC curve, with its vertical axis indicating the sensitivity of the RDW percent in patients who have expired and its horizontal axis (1 - RDW specificity), was applied. It must be at least 0.7 to be significant.
4. Results
In this study, 76 CHD patients, including 41 (53%) girls and 35 (47%) boys, were examined. The mean patient age was 35 ± 4 months (range 2 - 156 months). The mean RDW level was 14.6 ± 3 (range 10.7 - 27.8), with 14 ± 3 (range 10.7 - 27.8) for girls and 15 ± 3 (range 12.2 - 25.9) for boys, showing no significant difference according to sex (P = 0.646). Palliative surgery was performed in 21 (27%) patients, and complete corrective surgery in 55 (73%) patients. There was no significant relationship between preoperative angiographic data (cardiac performance indices, pressure indices, and pulmonary to systemic flow and resistance ratios) and RDW level. The total hospital stay was 19.09 ± 8.7 (range 8 - 52) days, and ICU stay was 8.07 ± 7.4 (range 2 - 36) days in general. The CPB time duration was 83.97 ± 42.56 (range 14 - 240) minutes, and aortic cross-clamp time duration was 55.37 ± 30.28 (range 12 - 139) minutes (Table 1).
Table 1.Pre-operative Angiographic and Surgery-Related Data
| Variables | Mean ± SD | P-Value | Spearman’s Rho with Significant P-Value |
|---|---|---|---|
| Angiographic data | |||
| QP/QS | 2.5 ± 0.67 | 0.061 | - |
| Rp/Rs | 0.1 ± 0.04 | 0.617 | - |
| RVSP (mmHg) | 83 ± 37 | 0.398 | - |
| LVSP (mmHg) | 112 ± 19 | 0.238 | - |
| Systolic PAP (mmHg) | 36 ± 17 | 0.220 | - |
| Diastolic PAP (mmHg) | 16 ± 10 | 0.093 | - |
| Mean PAP (mmHg) | 22 ± 12 | 0.113 | - |
| Surgical data | |||
| CPB time (min) | 83 ± 42 | 0.024 | R = 0.343 |
| Aortic cross clamp time (min) | 55 ± 30 | 0.138 | - |
| ICU stay (d) | 8 ± 7 | 0.007 | R = 0.32 |
| Hospital stay (d) | 19 ± 8 | 0.091 | - |
Abbreviations: Qp/Qs, pulmonary to systemic flow ratio; Rp/Rs, pulmonary to systemic resistance ratio; RVSP, right ventricle systolic pressure; LVSP, left ventricle systolic pressure; PAP, pulmonary artery pressure; CPB, cardiopulmonary bypass; ICU, intensive care unit.
Patients with higher RDW had longer CPB time (P = 0.01, R = 0.343) and ICU stay (P = 0.007, R = 0.32), but aortic cross-clamp time and total hospital stay were not correlated with RDW (P = 0.138 and 0.091, respectively). There was no significant difference between cyanotic and acyanotic patients according to CPB time (P = 0.125), ICU stay (P = 0.175), or total hospital stay (P = 0.362).
Six patients (7%) died in the operating room or during ICU stay, and the total morbidity rate was 15% (11 patients). Among corrective surgery patients, 3 (5%) died due to sepsis, hemodialysis complications, and tracheal perforation. The morbidity rate was 5% (3 patients) due to low cardiac output syndrome, prolonged ICU stay, sepsis, diaphragmatic paralysis, and complications of tracheostomy. Among palliative surgery cases, 3 (14%) died due to COVID-19, emergency shunt implantation complications, and complications of tracheostomy. The morbidity rate was 14% (3 patients), reported as prolonged ICU stay, postoperative seizures, and transient blindness.
The mean RDW was 14 ± 1 in deceased patients and 15 ± 3 in the survivors; the difference was not significant (P = 0.787). The area under the ROC curve did not predict mortality in patients (the area was 0.53). The mean RDW was 14 ± 2 in patients with morbidity. Mann-Whitney analysis showed no correlation between RDW and mortality (P = 0.78) or morbidity (P = 0.545). Fisher’s exact test analysis demonstrated no significant relationship between surgery type (palliative or corrective) and morbidity (P = 0.688) or mortality (P = 0.766). As Table 1 shows, pre-operative angiographic and surgery-related data.
Mean RDW was 14 ± 1 in patients with Down syndrome (12 patients) and 14 ± 3 in non-syndromic patients, with no significant difference (P = 0.191). The RDW was not significantly different according to hemoglobin level (P = 0.236) and different blood groups (P = 0.101). Mean RDW was 16 ± 4 in cyanotic and 14 ± 1 in acyanotic patients, with a significant difference (P = 0.009). The ICU stay time was shorter in patients with lower RDW (P = 0.007, R = 0.32), although this difference was not significant among cyanotic patients (P = 0.356). A relationship was observed between RDW and postoperative echocardiographic indices (Table 2).
Table 2.Relationship Between Red Cell Distribution Width and Post-operative Echocardiographic Indices
| Echocardiographic Indices | Mean ± SD | P-Value | Spearman’s Rho with Significant P-Value |
|---|---|---|---|
| LVEDD (cm) | 32 ± 10 | 0.926 | - |
| FS (%) | 30 ± 5 | 0.209 | - |
| LVEF (%) | 59 ± 7 | 0.701 | - |
| Stroke volume (mL/beat) | 35 ± 25 | 0.386 | - |
| Cardiac output (L/min) | 4.21 ± 3.03 | 0.359 | - |
| Cardiac Index (L/min/m2) | 2.69 ± 1.87 | 0.229 | - |
| Aortic VTI (cm) | 17 ± 6 | 0.035 | R = 0.271 |
| MV E | 74 ± 27 | 0.124 | - |
| MV A | 57 ± 28 | 0.469 | - |
| MV E/A | 1.52 ± 0.85 | 0.513 | - |
| Septal MV E’ | 14 ± 11 | 0.785 | - |
| Septal MV A’ | 14 ± 10 | 0.926 | - |
| Septal MV E’/A’ | 1.4 ± 1.5 | 0.484 | - |
| Septal MV E/E’ | 7.62 ± 7.19 | 0.846 | - |
| Septal MV S’ | 18 ± 20 | 0.469 | - |
| Lateral MV E’ | 14 ± 8 | 0.304 | - |
| Lateral MV A | 18 ± 16 | 0.131 | - |
| Lateral MV E’/ A’ | 1.4 ±1.02 | 0.929 | - |
| Lateral MV E/E’ | 7.43 ±11.43 | 0.161 | - |
| Lateral MV S’ | 13 ± 9 | 0.027 | R = 0.273 |
| MPI | 0.27 ± 0.12 | 0.423 | - |
| TAPSE (mm) | 8 ± 4 | 0.935 | - |
| TV E | 58 ± 24 | 0.699 | - |
| TV A | 49 ± 21 | 0.821 | - |
| TV E/A | 0.4 ± 0.5 | 0.100 | - |
| TV E’ | 16 ± 9 | 0.324 | - |
| TV A’ | 17 ± 12 | 0.171 | - |
| TV E’/ A’ | 1.3 ± 1.05 | 0.374 | - |
| TV E/E’ | 6.58 ±7.78 | 0.335 | - |
| TV S’ | 18 ± 19 | 0.070 | - |
Abbreviations: LVEF, left ventricular ejection fraction; VTI, velocity time integral; MV S’, mitral valve systolic velocity; TAPSE, tricuspid annular plane systolic excursion.
In the corrective surgery group, children showed a significant positive correlation between RDW and ICU stay (P = 0.01) and hospital stay (P = 0.042) periods; meanwhile, palliative surgery patients showed no significant relationship between RDW and ICU stay (P = 0.509) or hospital stay (P = 0.535). According to post-operative echocardiographic evaluation, RDW was correlated with some postoperative left ventricular systolic indices, including aortic VTI (P = 0.035, R = 0.271) and lateral mitral valve systolic velocity (MV S’; P = 0.027, R = 0.273). Other parameters were not affected postoperatively (Table 2).
5. Discussion
The RDW indicates variation in the size of circulating red blood cells as a laboratory parameter. It has traditionally been employed for the investigation of causes of anemia and conditions leading to red blood cell destruction. However, it has recently been applied in a wider spectrum of clinical settings, such as the prediction of poor prognosis and mortality in cardiac disorders (12).
In the present study, we investigated the relationship between preoperative RDW levels and postoperative echocardiographic findings. We noticed that RDW showed significant alteration in functional indices such as aortic VTI and MV systolic velocity in the early period after cardiac surgery. Additionally, patients with higher preoperative RDW demonstrated a longer duration of surgery and ICU stay compared to those with normal RDW levels.
Polat et al. studied 107 pediatric CHD patients and found a significant correlation between RDW and the length of ICU and hospital stay, a finding that is consistent with our study. However, they also noticed that the RDW range was significantly higher in deceased patients and suggested RDW as a predictive parameter for morbidity and mortality in the pre- and post-operative periods of CHD surgery, which contrasts with our findings. This discrepancy may be due to variations in causes and conditions resulting in mortality, including COVID-19, in our study (13).
Massin et al., in their study on 688 CHD children who underwent cardiac surgery, showed a strong positive correlation between pre-operative high RDW and adverse outcomes. The mortality rate was five times higher among patients with RDW of 16% or more. Patients with higher RDW levels had longer durations of ICU stay (especially acyanotic children younger than 6 months) and hospital stay (especially acyanotic children with normal hemoglobin levels). They suggested RDW as an inflammatory marker predicting post-operative prognosis in CHD patients. In our study, the mean RDW range was significantly higher in cyanotic patients, and the duration of ICU stay was shorter with lower RDW, but it was not associated with cyanosis. Other factors, such as infections, may have influenced our study. In Massin’s study, the risk of postoperative mortality for RDW ≥ 16% was five times higher, while in our study, RDW ≥ 14% did not show increased mortality. As the mean age of studied patients in both surveys was similar, this difference in mortality could probably be due to differences in race and nutritional status (14).
Oh et al. studied 100 adult patients with acute heart failure. Their study revealed a significant positive correlation between RDW and MV E and E/E’. Their study group, which was composed of patients mostly suffering from ischemic heart disease with RDW ≥ 13.45% and NT-pro BNP ≥ 2456 pg/mL, showed E/E’ ≥ 15, suggesting increased LV filling pressure and diastolic dysfunction. They proposed RDW as a simple marker demonstrating hemodynamic status in these patients. In our study, however, there was no significant relationship between these parameters (15). Our study showed a positive significant relationship between RDW and aortic VTI and mitral lateral S’, indicating ventricular systolic function. As we performed echocardiography in the early postoperative period, this might be the reason why our findings differ from theirs.
Mawlana et al., in their study on 31 children with heart failure, demonstrated that an RDW level higher than 16.1% was related to abnormal LV functional indices such as LVFS, MV A, and MV E/A. They suggested using RDW as a simple and inexpensive marker of LV function (16). In our study, there was a significant relationship between RDW and two indicators of left ventricular systolic function, but there was no relationship between RDW and ventricular diastolic function. This could be related to different sample sizes.
Celik et al. evaluated 71 adult patients with diastolic dysfunction, in whom RDW, NT-proBNP, and CRP levels were significantly increased. They suggested RDW ≥ 13.6% and NT-proBNP ≥ 125 pg/mL as factors indicating diastolic dysfunction, possibly due to neurohormonal, renal, or filling pressure variables (17). Although diastolic dysfunction was observed in some of our patients, it was not related to RDW level, and this difference may be due to the younger age range in our study (17). In his study, van Kimmenade et al. investigated 205 adult patients with acute heart failure. They demonstrated a reverse relationship between RDW level and 1-year survival rate, and otherwise no relationship between RDW and nutritional status, history of red blood transfusions, and inflammatory variables. They recommended considering RDW as a prognostic factor in patients with acute heart failure (like NT-pro BNP). In the present study, the patients were younger in age, and RDW level was not correlated with mortality (18).
Allen et al. in the Study of Anemia in a Heart Failure Population (STAMINA-HFP) registry introduced RDW as a prognostic factor for increased morbidity and considered that RDW rise in adult patients with chronic heart failure may be due to inflammation and iron metabolism impairment. Different results in our study might be due to the exclusion of anemic patients in our survey (19).
Alshawabkeh et al. studied 696 adult patients with CHD and showed worsening of HF and NYHA functional class, increased mortality, arrhythmia, and cardiac hospitalization with increased RDW levels (more than 15%), especially in patients with Eisenmenger syndrome or complex cyanotic heart disease. Our patients did not show any change in morbidity or mortality according to RDW variation, perhaps due to different sample size and age range; however, operation time and ICU stay were longer among them (20).
Kumar et al. in their study on 94 children with Tetralogy of Fallot (TF), demonstrated a reverse relationship between RDW levels and recovery time after TF repair (21). They showed increased ICU and hospital stay, duration of assisted ventilation implementation, and surgical site infection with RDW more than 17.8%. They suggested non-infectious mechanisms (such as chronic inflammation) as a probable reason for delay in wound healing and prolonged hospital stay, as microbial infection was not documented in the majority of their patients. They also suggested abnormal erythropoiesis and an increase in free iron release due to CPB implementation, especially in cyanotic patients, as the cause of prolonged recovery after surgery. In our study, 55 (72%) children underwent CPB, in whom RDW level was significantly related to the ICU and hospital stay duration. It seems that inflammation and oxidative stress both play important roles in elongating recovery time. However, a fundamental yet unanswered question remains: Whether the relationship between RDW and cardiovascular disease is a cause or effect, and if anisocytosis is the result of common metabolic and nutritional disorders or inflammatory cytokines, oxidative stress, and malnutrition in cardiovascular patients. As most CHD patients are already in an inflammatory state, another hypothesis is that elevated RDW may be a comorbidity rather than merely a causative factor in the pathogenesis of cardiovascular disease. Certainly, the clinical role of RDW in the prognosis of patients with cardiac disease is undeniable (21).
5.1. Conclusions
Various studies have shown RDW as an indicator of prognosis for cardiovascular disease. The RDW can be an important preoperative parameter in children with CHD to predict the patient’s duration of operation and ICU stay. It also has a correlation with cardiac function in the early postoperative period. It seems desirable to consider a longer time for ICU hospitalization in children with higher RDW.
5.2. Limitations
This study faced several potential limitations. Due to the available facilities, sample volume, and the cross-sectional nature of the study, a cause-and-effect relationship could not be concluded. Confounding factors might have affected the results, regardless of the adjusted analysis. We did not evaluate the nutritional and cytokine proinflammatory status of the patients in our study, which might have influenced the RDW. Other limitations include the small sample size, single-center design, and lack of inflammatory markers (e.g., CRP, cytokines).
5.3. Suggestions
We suggest the simultaneous measurement of levels of other inflammatory factors such as TNF, INF, and HS-CRP in upcoming studies. We also recommend conducting prospective studies with larger patient populations and longer follow-up periods across different Iranian ethnic groups to evaluate the effects of RDW on the long-term prognosis of children with cardiac disease. Finally, we suggest investigating the effect of RDW on delayed outcomes and long-term prognosis in CHD children in future studies.
