Red cell distribution width–standard deviation to albumin ratio and mortality in acute pulmonary thromboembolism: a single-center retrospective cohort study

Abstract

Background

See More

Recent studies have shown a relation between red blood cell distribution width (RDW) / albumin (RAR) levels and worse outcomes in cases of pulmonary embolism (PE). Simplified pulmonary embolism severity index (sPESI) has been developed from more complex PESI score, predicting the risk of death in patients with acute PE (APE). This study aims to investigate whether RDW-Standard Deviation/Albumin (RDW-SD/Alb) can serve as a useful prognostic marker for APE and enhance the predictive capability of the sPESI.

Methods

This research is a single-center, retrospective analysis involving patients over 18 years. We enrolled 235 consecutive hospitalized patients with confirmed APE diagnoses. To evaluate the sensitivity and specificity of RDW-SD/Alb, sPESI, and the combined sPESI plus RDW-SD/Alb in predicting 6-month all-cause death, we used Receiver Operating Characteristic (ROC) curves. Additionally, we conducted Kaplan–Meier analysis to assess the impact of elevated RDW-SD/Alb levels (> 13.6) on patient survival time. We utilized multivariate Cox regression analysis to identify independent prognostic factors affecting patients survival.

Results

The mortality rate for RDW-SD/Alb > 13.6 group was significantly higher than that for the RDW-SD/Alb ≤ 13.6 group. Area under ROC (AUROC) of sPESI plus RDW-SD/Alb was statistically larger than AUROC of sPESI (p = 0.025). In the fully adjusted model, increased RDW-SD/Alb levels were consistently linked to all-cause mortality within six months of admission.

Conclusions

The predictive value of the sPESI for 6-month all-cause death improved when the RDW-SD/Alb > 13.6 parameter was included. RDW-SD/Alb, a novel inflammatory marker, was an independent prognostic factor for predicting 6-month all-cause mortality in patients with APE.

Introduction

Pulmonary embolism (PE) denotes obstruction of the pulmonary artery and can be induced by many origins, including tumors, fat, air, amniotic fluid, and septic emboli, though the extended majority are thromboembolic [1]. After ischemic heart disease and stroke, PE is the third foremost cause of cardiovascular mortality worldwide [2].

Red blood cell distribution width (RDW) is a parameter routinely reported in the complete blood count (CBC) test, indicating the variability in the size of red blood cells in peripheral blood, and it is traditionally utilized clinically for the differential diagnosis of anemias. Recent evidence confirms that anisocytosis is common in human disorders such as cancer, liver failure, diabetes, cardiovascular disease, venous thromboembolism, community-acquired pneumonia, chronic obstructive pulmonary disease, kidney failure, and other acute or chronic conditions. Furthermore, the value of RDW is now believed to be a decisive and independent risk factor for death in the general population [3]. Serum albumin is a protein produced by the liver that serves as a negative acute phase reactant, meaning its levels decrease in response to inflammation [4]. This protein plays several important roles, including acting as an anti-inflammatory agent, antioxidant, anticoagulant, and preventing platelet aggregation [5]. RDW/Albumin (RAR) is a novel inflammatory biomarker that has been linked to mortality in various diseases, including chronic obstructive pulmonary disease, myocardial infarction, diabetic ketoacidosis, acute respiratory distress syndrome, and PE [6, 7]. The simplified pulmonary embolism severity index (sPESI) has been developed from the more complex PESI score, which predicts the risk of death in patients with acute PE (APE) using only clinical indices [8]. Although nearly 20% of patients receiving treatment for PE die within 90 days, PE is not typically the direct cause of death; rather, it often occurs alongside other serious conditions such as cancer, sepsis, or illnesses that necessitate hospitalization [9]. Therefore, incorporating RAR into the sPESI score may enhance its prognostic value.

So far, only two studies have investigated the prognostic value of RAR in acute pulmonary embolism (APE) [6, 7]. Our goal is to determine whether RDW-Standard deviation/Albumin (RDW-SD/Alb) can serve as a valuable prognostic marker in APE and whether it can improve the prognostic power of the sPESI score.

Methods

This study is a single-center, retrospective analysis implicating patients 18 years of age or older. A total of 235 consecutive patients who were hospitalized with a confirmed diagnosis of APE and were consulted to the cardiology department from the emergency department between February 2023 and March 2024 were screened. These patients were followed until October 2024.

Patients were excluded if they had end-stage liver disease, a glomerular filtration rate (GFR) of less than 15 mL/min, a left ventricular ejection fraction of less than 30%, severe valvular heart disease, or congenital heart disease. Ultimately, we included 184 survivor patients and 51 non-survivor patients [male 46.2%, median age 66,0 (23.0) years; 49.0% male, median age 74,0 (18.0) years].

Blood samples were collected upon admission, and CBCs were performed using the Sysmex XT 4000i machine for each patient. The ratio of RDW-SD to albumin was calculated by dividing the RDW-SD (fL) by the albumin level (g/dL).

We gathered the clinical history, laboratory test results, transthoracic echocardiography (TTE) findings, and pulmonary computed tomographic angiography (CTA) results from the electronic medical reports in our hospital. All laboratory tests, TTE examinations, and CTA results for the patients were obtained within 24 h of their admission to the emergency department.

The sPESI score was calculated for each patient by the two different cardiologists. The score included variables such as age, history of cancer, chronic pulmonary disease, heart rate, systolic blood pressure (SBP), and oxygen saturation. 1 point is assigned for each of these situations; age is over 80 years, heart rate ≥ 110 beats/minute, SBP < 100 mmHg, or arterial oxygen saturation < 90%.

Active cancer refers to cancer that has been diagnosed or treated within the last six months.

Statistical Analysis

We examined all the data using SPSS version 22.0 software (SPSS Inc., Chicago, IL). Continuous data were delivered as either mean ± standard deviation for parametric data or median ± interquartile range for nonparametric data. Parametric continuous variables were assessed using the Independent Samples T-test, while nonparametric continuous variables were evaluated using the Mann–Whitney U test. We compared categorical data using Pearson's Chi-square test. Spearman's rank correlation test was utilized to assess the relationship between albumin and RDW-SD. Patients were additionally stratified according to the RDW-SD/Alb cut-off value of 13.6 to compare baseline characteristics and clinical outcomes.

The area under the receiver operating characteristic (ROC) curve was analyzed using the MedCalc program (MedCalc Software Ltd). The DeLong test compared the area under the ROC curves (AUROC). For patients with RDW-SD/Alb levels greater than 13.6, we assigned 1 point; otherwise, 0 points were counted for the sPESI score, resulting in a new score referred to as sPESI plus RDW-SD/Alb. We used ROC curves to determine the sensitivity and specificity of RDW-SD/Alb, sPESI, and the sPESI plus RDW-SD/Alb score in predicting 6-month all-cause death and compared the AUROC of both the sPESI and the sPESI plus RDW-SD/Alb score.

We performed Kaplan–Meier analysis to evaluate the impact of higher RDW-SD/Alb levels (> 13.6) on patient survival time. We selected variables with a P-value of less than 0.05 in the univariate Cox regression analysis for multivariate Cox regression analysis. Multivariate Cox regression analysis was utilized to specify the effect of independent prognostic factors on patient survival. Four different models were used: one unadjusted model and three adjusted models with increasing numbers of adjusted factors. The outcomes were evaluated at a 95% confidence interval, with significance considered at p < 0.05.

Results

We analyzed 235 consecutive patients diagnosed with APE, as revealed by pulmonary CTA, and separated them into the survivor group (n = 184, 46.2% male, mean age = 66.0 ± 23.0 years) and non-survivor group (n = 51, 49.0% male, mean age = 74.0 ± 18.0 years). Table 1 outlines the demographic and clinical data, while Table 2 provides the laboratory parameters for survivors and non-survivors. Hypertension (60.8% vs. 45.1%, p = 0.047), diabetes (35.3% vs. 21.7%, p = 0.047), and chronic obstructive pulmonary disease (27.5% vs. 14.7%, p = 0.033) were more prevalent in non-survivor group than survivor group. The neutrophil (7.83 ± 5.92 vs. 6.64 ± 4.85, p = 0.043), hs-troponin (48.7 ± 41.2 vs. 31.2 ± 44.3, p = 0.007), NT-Pro-Brain natriuretic peptide (NT-Pro-BNP) (2902 ± 2448 vs. 1802 ± 2609, p = 0.004), C-reactive protein (CRP) (68.5 ± 100.5 vs. 45.9 ± 63.5, p = 0.006) and RDW-SD/Alb (14.9 ± 6.8 vs. 12.5 ± 3.8, p = 0.006) were higher in non-survivor group than survivor group, while lymphocyte (0.98 ± 1.10 vs. 1.68 ± 1.23, p < 0.001) was lower. Spearman’s correlation analysis revealed a statistically significant inverse relationship between albumin and RDW-SD (r = −0.353, p < 0.001).

Table 1 Baseline Patient Characteristics of Participants, Survivors and Non-Survivors

Table 2 Laboratory Parameters of Survivors and Non-Survivors

The cut-off value of RDW-SD/Alb for 6-month all-cause death was 13.6. RDW-SD/Alb higher than 13.6 predicted the 6-month all-cause death with a sensitivity of 70.6% and specificity of 67.0% (p < 0.001). Patients with RDW-SD/Alb > 13.6 (n = 100) had significantly higher sPESI scores (2.0 ± 2.0 vs. 1.0 ± 2.0, p < 0.001) and higher 6-month all-cause death (36/100; 36.0% vs. 15/135; 11.1%, p < 0.001), compared to those with RDW-SD/Alb ≤ 13.6 (n = 135). The prevalence of active cancer was also higher in the RDW-SD/Alb > 13.6 group (24/100; 24.0% vs. 19/135; 14.1%), although this difference did not reach statistical significance (p = 0.052).

AUROC values for the variables analysed by ROC curve analysis in terms of predicting 6-month all-cause death in APE were as follows: RDW-SD/Alb: AUC: 0.727 (95% CI: 0.666–0.784, p < 0.001), sPESI: AUC: 0.708 (95% CI: 0.645–0.765, p < 0.001), sPESI plus RDW-SD/Alb score: AUC: 0.740 (95% CI: 0.679–0.795, p < 0.001). After the pairwise comparison by the Delong test, AUROC of sPESI plus RDW-SD/Alb score was statistically larger than the AUROC of sPESI (p = 0.025) (Fig. 1).

Fig. 1

ROC Curves of sPESI, RDW-SD/Alb, sPESI + RDW-SD/Alb for predicting 6-month all-cause death

We generated Kaplan–Meier survival curves for patients (Fig. 2). The mortality rate in the RDW-SD/Alb > 13.6 group was significantly greater compared to the RDW-SD/Alb ≤ 13.6 group (Log-rank, p < 0.001). In the fully adjusted model, RDW-SD/Alb was definitively associated with all-cause mortality within a six-month period following admission. We have installed three adjustment models (Table 3). In Model 1, adjusted for age, diabetes, active cancer and Model 2, adjusted for age, diabetes, active cancer, heart rate, aspartate aminotransferase (AST), we observed a statistically remarkable increase in the risk of all-cause mortality within 6 months after hospital admission for patients with elevated RDW-SD/Alb levels. Additionally, in the multivariate Cox regression analysis, the RDW-SD/Albumin was identified as an independent predictor of 6-month mortality. Each 1-unit increase in the RDW-SD/Alb was associated with a 3.364-fold increase in the risk of death (HR: 3.364, 95% CI: 1.944–5.821, p < 0.001). When modeled on a logarithmic scale (log₂), each doubling of the RDW-SD/Alb ratio corresponded to a 2.32-fold increase in mortality risk (HR: 2.32, 95% CI: 1.59–3.39, p < 0.001) in the Model 3, adjusted for age, diabetes, active cancer heart rate, AST, troponin, CRP, and the sPESI score. Other independent prognostic factors on survival in Model 3 were age, AST, and diabetes (HR 1.037, 95% CI 1.010, 1.064, p = 0.007; HR 1.016, 95% CI 1.004, 1.029, p = 0.008 and 1.965, 95% CI 1.059, 3.648, p = 0.032, respectively). Furthermore, there was a statistically significant difference between Model 1 and Model 2, but no significant difference between Model 2 and Model 3 (p = 0.001; p = 0.473, respectively).

Fig. 2

Kaplan–Meier analysis of the impact of RDW-SD/Alb > 13.6 cut-off value on survival

Table 3 Hazard ratios (HRs) for 6-month mortality based on RDW-SD/Alb in acute pulmonary embolism

Discussion

The results of our study identified three main findings: First, the RDW-SD/Alb ratio was higher in the non-survivor group than in the survivor group among patients with APE. Second, the cut-off value RDW-SD/Alb for the predicting 6-month all-cause death was determined to be 13.6. Third, the predictive value of the sPESI score for 6-month all-cause death improved when the RDW-SD/Alb > 13.6 parameter was included. Additionally, each doubling of the RDW-SD/Alb ratio corresponded to a 2.32-fold increase in 6-month mortality. Finally, RDW-SD/Alb level, along with age, diabetes and AST, were independent predictors of 6-month mortality in patients with APE.

The relationship between RDW and venous thromboembolism is complex and is still under investigation. It is well established that various nutritional deficiencies can occur in patients who experience chronic immobilization and declining renal function, both of which are common in individuals with deep vein thrombosis (DVT) and PE. Additionally, hypoxia caused by the obstruction of pulmonary arteries may lead to the hyperactivation of adrenergic and neurohormonal pathways, triggering the release of pro-inflammatory cytokines. Salvano et al. observed that RDW may be elevated due to the intricate interactions among these underlying conditions in patients with APE. [3] Albumin is a medium-sized protein synthesized by the liver and plays a key role in supporting colloid osmotic pressure in the body. Like RDW, albumin is important for the inflammatory response and oxidative stress [10].Numerous studies have indicated that low levels of albumin are related to poor prognosis in patients with APE [6]. In this study, the combination of RDW-SD and albumin, referred to as RDW-SD/Alb, was linked to all-cause mortality in patients with APE, even after adjusting for various other factors.

We found a moderate inverse correlation between RDW-SD and albumin levels (r = −0.353, p < 0.001). This finding suggests a potential link between changes in red blood cell characteristics and the inflammatory or nutritional status of patients with APE. This association may indicate a possible pathophysiological connection and supports the idea that the RDW-SD/Alb could be a useful prognostic marker. However, more research, including larger and more comprehensive studies, is needed to better understand its clinical significance.

APE is classified into three categories based on imaging findings and patient characteristics: high-risk, intermediate-risk, and low-risk. High-risk PE is defined by sustained hypotension, indicated by a SBP of less than 90 mm Hg for at least 15 min or a vasopressor requirement with no other plausible explanation for the shock. Intermediate-risk PE includes patients who exhibit signs of right ventricular (RV) dysfunction through imaging or biomarker evidence (such as myocardial necrosis or chamber dilation) but do not show sustained hypotension. A subcategory of intermediate-risk, intermediate-high-risk PE involves patients with both imaging and biomarker evidence of RV dysfunction. Low-risk PE does not fulfill the criteria for high-risk or intermediate-risk PE [8]. Severe PE patients with SBP greater than 90 mm Hg are categorized according to their risk of hemodynamic collapse and death. Individual risk factor-based clinical prediction scores can be used to identify patient subgroups at high risk for hemodynamic instability. Despite the large number of published clinical prediction scores, only a few numbers have been prospectively validated and shown to be trustworthy [11]. These include the Bova score, the PESI, and sPESI scores. The Bova score combines high troponin levels and RV dysfunction with hemodynamic instability criteria, whereas the PESI and sPESI scores combine markers of cardiopulmonary impairment with age and comorbidities [12, 13]. The European Society of Cardiology guidelines recommend using PESI, an algorithm that calculates various clinical and hemodynamic indicators, to predict 30-day mortality. Due to the complexity of the original PESI, which consists of 11 differently weighted variables, a simplified version called sPESI has been developed and validated. A sPESI score of 0 identifies a low-risk group, with an approximately 1.0% mortality risk within 30 days. In contrast, a sPESI score of 1 or higher identifies a high-risk group, with a 30-day mortality risk increasing to 10.9% [14]. Additionally, markers of RV dysfunction, including ECG, TTE, BNP, and troponin levels, are commonly used for prognostic evaluation in patients with APE. According to retrospective investigations, scoring systems that use biomarkers like the BOVA score are better than the sPESI score at predicting the likelihood of unfavourable events in the early stages [15,16,17,18]. Although the sPESI score is a powerful tool for excluding adverse events in patients with a score of 0, its predictive power for mortality in patients with a score of ≥ 1 is still in doubt. Trimaille et al. examined whether the sPESI score's capacity to predict death was enhanced by including renal dysfunction. [19] The findings demonstrated that a group with extremely high mortality was found when renal failure was included as a biomarker to the sPESI score. This implies that more biomarkers are required to help the sPESI score more accurately predict the likelihood of unfavorable outcomes. The sPESI score was initially developed and validated to predict 30-day mortality in patients with PE. In our study, we did not aim to enhance or recalibrate the predictive ability of the sPESI score. Instead, we utilized it as a baseline tool and included an additional analysis to assess its association with 6-month mortality. Our objective was to investigate whether the sPESI score could provide prognostic insights beyond the originally validated 30-day timeframe. Thus, the study does not change the original purpose of the score but rather examines its extended utility for long-term risk assessment. In the present study, although the sPESI score was higher in the non-survivor group, both groups had high-risk sPESI scores. In light of these findings, RDW-SD/Alb might be thought of as a predictor of all-cause mortality among patients with high-risk features. Furthermore, we observed that the predictive value of the sPESI score for six-month all-cause death improved when RDW-SD/Alb > 13.6 was included as a parameter. Therefore, in addition to the sPESI score, RDW-SD/Alb can be utilized as a biomarker to predict early mortality in individuals with APE more precisely.

Recently, Ding et al. examined the relationship between RAR levels and the risk of all-cause mortality in patients with APE admitted to the intensive care unit (ICU). The researchers categorized the patients into three groups according to their RAR levels: Low RAR (2.68–4.71), Middle RAR (4.71–6.09), and High RAR (6.09–15.45). The study found a significant association between the Middle and High RAR groups and an increased risk of all-cause mortality in patients with PE. [6] Furthermore, our study indicated that higher RDW-SD/Alb (> 13.6) was associated with 6-month all-cause mortality, not only in ICU patients diagnosed with APE but also in all hospitalized patients with this condition.

The RDW-SD/Alb ratio demonstrated significant prognostic value in our cohort; however, the cut-off value of 13.6 was determined through ROC analysis and may not be applicable universally to all populations. Therefore, it should be interpreted with caution and validated in future studies involving independent groups.

Eraslan et al. evaluated the impact of RAR on mortality in patients with PE. [7] Their analysis identified a cut-off value of 5.294 using ROC analysis. Patients with an RAR of 5.294 or higher had a significantly shorter mean survival time compared to those with an RAR below this threshold. In our study, we established that the RDW-SD/Alb cut-off value predicting all-cause death at six months was 13.6. It is important to note that Eraslan et al. calculated the RAR using the RDW-CV (%) unit, whereas we utilized the RDW-SD (fL) unit. We believe the differing cut-off values between the two studies may be attributed to this distinction. Recent studies indicated that RDW-SD may offer additional diagnostic value, as it is a direct measurement that is not influenced by mean corpuscular volume and more accurately reflects variations in red cell size. [20, 21] There is insufficient data in the literature about whether RDW-SD or RDW-CV should be used for calculating RAR.

In this study, we observed a significantly higher 6-month mortality rate among APE patients with elevated RDW-SD/Alb levels. Since hypoalbuminemia is commonly found in cancer patients and malignancy is a well-established risk factor for both mortality and venous thromboembolism (VTE), it may be hypothesized that cancer primarily drives this association [22]. However, the RDW-SD/Alb ratio remained a significant and independent predictor of 6-month mortality, even after adjusting for active cancer in multivariable Cox regression analyses. These findings suggest that the RDW-SD/Alb ratio may act as a reliable prognostic biomarker in patients with APE, regardless of the presence of underlying cancer.

Our study has several limitations. First, it was a single-center, retrospective, observational study, which inherently has limitations due to its design. Second, data on the study population were obtained from our hospital's electronic medical records, which may introduce a selective bias. Third, RDW-SD/Alb is a recently identified inflammatory marker, and a prospective study is necessary to more accurately determine its predictive cut-off value in patients with APE.

Conclusion

Our study has identified RDW-SD/Alb as a new independent inflammatory marker that can predict 6-month all-cause mortality in patients with APE. This marker is low-cost and easy to use, providing valuable early information. As a novel parameter, RDW-SD/Alb may improve the management of APE patients by aiding in the prediction of 6-month mortality risk. It may be advisable to closely monitor APE patients with an RDW/Alb > 13.6.