Platelet-to-White Blood Cell Ratio as a Predictor of ICU Mortality in Septic Shock and Heart Failure: A Real-World Validation Study Using the MIMIC-eICU and a Chinese Cohort

Abstract

Background

The platelet-to-white blood cell ratio (PWR) has been associated with adverse outcomes in critically ill patients. Nonetheless, its prognostic significance in individuals with concurrent septic shock and heart failure remains insufficiently validated externally. This study sought to examine this association utilizing a large public database and to validate the findings in an independent real-world cohort.

Methods

This retrospective cohort study included adult patients with septic shock and heart failure. The discovery cohort was derived from the MIMIC-IV and eICU databases. The validation cohort was retrospectively collected from The Central hospital of Enshi Tujia and Miao Autonomous Prefecture (August 2023–June 2025). The primary outcome was ICU mortality. Multivariable Cox regression, Kaplan-Meier analysis, and subgroup analyses were performed.

Results

The study incorporated 934 patients in the discovery cohort and 124 patients in the validation cohort. In the discovery cohort, restricted cubic spline analysis demonstrated a linear relationship between platelet-to-white blood cell ratio (PWR) and mortality (nonlinearity P=0.655), prompting a tertile-based classification (Q1: ≤9.95; Q2: >9.95–≤17.72; Q3: >17.72). Compared to Q1, Q3 was associated with significantly reduced ICU mortality in the fully adjusted model (HR 0.66, 95% CI 0.44–0.99). Applying the same cut-off (9.95) in the validation cohort, patients with a low PWR (≤9.95) exhibited significantly higher mortality compared to those with a high PWR (>9.95) (HR 3.11, 95% CI 1.46–6.66; log-rank P=0.002). Subgroup analyses substantiated the robustness of these findings.

Conclusion

A reduced platelet-to-white blood cell ratio serves as an independent predictor of elevated ICU mortality among patients experiencing septic shock and heart failure. This straightforward and cost-effective biomarker demonstrated a consistent association in an external real-world cohort, underscoring its potential utility for early risk stratification.

Introduction

Septic shock represents a dysregulated host response to infection, characterized by significant circulatory, cellular, and metabolic disturbances, and it remains a major contributor to mortality in intensive care units globally.¹ Heart failure frequently occurs in the context of septic shock, indicative of a complex interaction involving sepsis-induced myocardial injury, such as mitochondrial dysfunction, neuroendocrine activation, and microcirculatory malperfusion, all of which impair myocardial contractility and cardiac output.^2,3 The concurrence of these conditions leads to a considerably poorer prognosis compared to either condition in isolation, with research indicating notably higher mortality rates among septic shock patients who develop heart failure.⁴ Consequently, the identification of reliable prognostic markers in this high-risk cohort is crucial for early risk stratification and the implementation of targeted interventions.

The prognosis of ICU patients with septic shock and heart failure is primarily assessed using scoring systems like the Acute Physiology and Chronic Health Evaluation (APACHE) and Sequential Organ Failure Assessment (SOFA).⁵ However, these systems are limited by their reliance on numerous parameters, complex calculations, and limited real-time applicability, hindering their effectiveness in rapid clinical assessments.⁶ Thus, exploring simpler and more convenient methods to predict the prognosis of patients with septic shock and heart failure is essential. Recent studies indicate that hematological parameters, including platelet count,⁷ circulating neutrophil count,⁸ procalcitonin (PCT),⁹ and interleukin levels,¹⁰ may be valuable for assessing the prognosis of septic shock. Platelets and leukocytes are integral to the inflammatory and immune dysregulation processes that contribute to organ dysfunction in conditions such as septic shock and heart failure. Specifically, platelets influence leukocyte recruitment and activation, whereas unregulated leukocyte activation exacerbates tissue damage. The platelet-to-white blood cell ratio (PWR) serves as an indicator of the equilibrium between these two interconnected elements and has been associated with negative outcomes in various critical illness contexts.^11–13

In the context of septic shock and heart failure, PWR may influence the ICU mortality by modulating the inflammatory response and immune regulation. We propose that PWR may independently predict ICU mortality in patients with septic shock and heart failure. We aim to provide clinicians with a simple, cost-effective, and reliable tool to help identify patients at high risk of death, enabling timely intervention to improve patient outcomes.

Materials and Methods

Data Source

In this study, we employed two renowned publicly accessible critical care databases: Medical Information Mart for Intensive Care IV version 2.2 database (MIMIC-IV 2.2) and eICU Collaborative Research and Innovation Center database(eICU). The MIMIC-IV developed by MIT’s Laboratory for Computational Physiology, contains detailed clinical data of ICU patients from Beth Israel Deaconess Medical Center, spanning 2008 to 2019. This data includes physiological monitoring, laboratory test results, medication records, length of hospital stay, and demographic information. The eICU database, developed by the eICU Collaborative Research and Innovation Center, provides data from ICU patients across multiple hospitals, including detailed information on physiological parameters, laboratory results, and treatment protocols. To ensure patient privacy, all personal identifiable information has been de-identified, and data is anonymized with patient identifiers replaced by random numbers. Relevant data was extracted from these databases using the access credentials (ID57580527) granted for this study. To conduct external validation, we retrospectively gathered data from patients admitted to the Department of Pulmonary and Critical Care Medicine at The Central Hospital of Enshi Tujia and Miao Autonomous Prefecture between August 2023 and June 2025. This validation cohort comprised adult patients diagnosed with septic shock and heart failure, with clinical and laboratory data extracted from electronic medical records. All data usage strictly adhered to privacy protection and ethical guidelines. Ethical approval was granted by the Ethics Committee of Enshi Tujia and Miao Autonomous Prefecture Central Hospital.

Inclusion and Exclusion Criteria

The study included participants who met the following criteria: Age > 18 years; Only the initial ICU admission was considered for patients with multiple stays. Patients diagnosed with both “septic shock” and “heart failure” as per ICD-9 and ICD-10 classifications. The complete lists of diagnostic codes used for case identification in each database are provided in Table S1. Patients with complete records of platelet and white blood cell (WBC) data within 24 hours of ICU admission. The exclusion criteria were as follows: ICU stay of less than 24 hours; Missing data exceeding 30% of the total variables collected per patient, a threshold chosen to exclude records with substantial incompleteness that could compromise the validity of multivariable adjustment; Patients without platelet and white blood cell data within 24 hours of ICU admission, as shown in Figure 1.

Figure 1.

Research flowchart.

Abbreviations: MIMIC-IV, Medical Information Mart for Intensive Care IV; eICU, eICU Collaborative Research Database; ICU, intensive care unit.

Data Extraction

Data were extracted using pgAdmin 4 (v4.26). Clinical data within 24 hours of ICU admission were collected. The study extracted the following variables: (1) Vital signs, including heart rate, blood pressure, and respiratory rate; (2) Demographic information, such as gender, age, and weight; (3) Laboratory test parameters, comprising white blood cell count, platelet count, hemoglobin, aspartate aminotransferase (AST); (4) Comorbidities, including hypertension, Diabetes Type II, malignancy, chronic kidney disease, and liver failure. Additionally, the SOFA score, APACHE III score were recorded. Additionally, information regarding ICU stay duration and time of death was gathered.

Statistical Analysis

In this study, categorical variables were expressed as frequencies and percentages, and intergroup comparisons were performed using Pearson’s chi-square test or Fisher’s exact test. Continuous variables were summarized as mean ± standard deviation or interquartile range. Differences between variables were evaluated using the t-test or Mann–Whitney U-test, as suitable Restricted cubic splines (RCS) modeled the relationship between PWR and outcomes to explore potential nonlinear associations. Based on the RCS results, patients were divided into three groups according to PWR tertiles. To further investigate the relationship between PWR and the risk of ICU mortality, we used Cox proportional hazards models. Three models were constructed: an unadjusted model, an adjusted model 1, and an adjusted model 2. These models progressively incorporated different confounders to evaluate the independent role of PWR in predicting ICU mortality. The unadjusted model performed a univariate analysis based on PWR. Adjusted model 1 incorporated age, gender, and weight, while adjusted model 2 included additional clinical variables such as heart rate, systolic and diastolic blood pressure, oxygen saturation, temperature, SOFA, APSIII, AST/Platelet Ratio Index (APRI), Platelet to Hemoglobin Ratio (PHR), hypertension, diabetes Type II, malignant neoplasm, chronic kidney disease, and liver failure. Collinearity among covariates in the fully adjusted model was assessed using variance inflation factors (VIF); all VIF values were below 2, indicating no serious multicollinearity. Kaplan-Meier survival curves were employed to analyze survival differences between the low PWR and high PWR groups in the validation cohort. Subgroup analyses were performed to assess the prognostic value of PWR across different conditions, including age, gender, hypertension, Diabetes Type II, malignant neoplasm, and liver failure. All statistical analyses were conducted using R software. Statistical comparisons employed two-tailed tests, with significance determined at a p-value below 0.05.

Results

Baseline Demographic and Clinical Characteristics

Based on the inclusion criteria, 934 patients in the MIMIC_EICU discovery cohort were enrolled in the study, categorized into the survival group (N=707) and the mortality group (N=227) (Figure 1). The mean age of the participants was 73.18 ± 12.89 years. Among the participants, 558 were male and 376 were female. Baseline characteristics revealed no significant differences between the two groups regarding age, weight, heart rate, systolic and diastolic blood pressure, and oxygen saturation, with all p-values exceeding 0.05. Significant differences in SOFA and APSIII scores were noted between survival and mortality groups (p< 0.05), indicating a potential association with the prognosis of septic shock patients. Laboratory results indicated significant differences between the two groups in various indicators (p< 0.05), such as APRI and PWR (Table 1). The baseline characteristics of the Enshi validation cohort are presented in Table S2.

Table 1.

Baseline Characteristics of the MIMIC-EICU Discovery Cohort

Clinical Features, Laboratory Indicators	Survivors (N=707)	No-Survivors (N=227)	P-value
Age	71.89 (13.96)	72.74 (11.75)	0.408
Gender			0.599
Male	419 (59.26%)	139 (61.23%)
Female	288 (40.74%)	88 (38.77%)
Age			0.139
<60	123 (17.40%)	30 (13.22%)
≥60	584 (82.60%)	197 (86.78%)
Weight	81.45 (65.76–99.95)	79.05 (64.70–99.20)	0.986
Heart Rate	94.13 (22.07)	94.50 (22.54)	0.825
Systolic Blood Pressure	110.13 (24.02)	107.23 (22.42)	0.115
Diastolic Blood Pressure	64.47 (19.73)	64.84 (21.45)	0.812
Oxygen Saturation	97.00 (94.00–99.00)	98.00 (94.00–100.00)	0.323
Temperature	36.92 (0.84)	36.74 (0.84)	0.007
Sofa	9.31 (3.98)	10.64 (3.64)	<0.001
Apsiii	61.20 (22.73)	72.58 (24.68)	<0.001
APRI	0.23 (0.12–0.55)	0.35 (0.18–1.09)	0.040
PWR	13.70 (8.58–21.18)	12.35 (7.74–18.43)	0.052
PHR	17.63 (12.37–25.82)	17.76 (11.50–23.52)	0.101
Hypertension	128 (18.10%)	47 (20.70%)	0.382
Diabetes Type II	234 (33.10%)	88 (38.77%)	0.118
Malignant Neoplasm	82 (11.60%)	21 (9.25%)	0.326
Chronic Kidney Disease	268 (37.91%)	99 (43.61%)	0.126
Liver Failure	86 (12.16%)	56 (24.67%)	<0.001

Abbreviations: Sofa, Sequential Organ Failure Assessment; Apsiii, Acute Physiology Score III; APRI, AST/Platelet Ratio Index; PWR, Platelet to White Blood Cell Ratio; PHR, Platelet to Hemoglobin Ratio.

Evaluating the Linear Correlation Between PWR and ICU Mortality Using the RCS Curve

In the discovery cohort, a restricted cubic spline model was used to evaluate the association between PWR and ICU mortality in septic shock patients with heart failure. The model demonstrated a linear correlation (nonlinearity p=0.655) (Figure 2). Based on the RCS results, patients were divided into three groups: Q1 group (≤9.95), Q2 group (>9.95, ≤17.72), and Q3 group (>17.72).

Figure 2.

Restricted cubic spline (RCS) analysis of the association between PWR and ICU mortality in the discovery cohort. The solid line represents the estimated hazard ratio, and the shaded area indicates the 95% confidence interval. The non-linearity P value was 0.655, supporting a linear relationship.

Abbreviations: RCS, Restricted cubic spline; PWR, platelet-to-white blood cell ratio; ICU, intensive care unit.

Univariate Cox Regression Analysis to Predict ICU Mortality Risk in Patients

In the discovery cohort, univariate analysis identified several predictors of ICU mortality, including age ≥ 60 years (HR 1.49, 95% CI 1.02–2.20), weight (HR 0.99, 95% CI 0.99–1.00), SOFA score (HR 1.07, 95% CI 1.03–1.11), APSIII score (HR 1.01, 95% CI 1.01–1.02), and PWR (HR 0.98, 95% CI 0.97–1.00)(Table 2).

Table 2.

Univariate Cox Regression Analysis for ICU Mortality in the MIMIC-EICU Discovery Cohort

Characteristics	Statistics	HR (95% CI)	P value
Age	72.10 ± 13.45	1.02 (1.01, 1.03)	0.001
Gender
Male	558 (59.74%)	1
Female	376 (40.26%)	1.17 (0.89, 1.53)	0.255
Age
<60	153 (16.38%)	1
≥60	781 (83.62%)	1.49 (1.02, 2.20)	0.042
Weight	86.48 ± 31.04	0.99 (0.99, 1.00)	0.024
Heart Rate	94.22 ± 22.18	1.00 (0.99, 1.01)	0.827
Systolic Blood Pressure	109.43 ± 23.67	1.00 (0.99, 1.00)	0.175
Diastolic Blood Pressure	64.56 ± 20.15	1.00 (1.00, 1.01)	0.377
Oxygen Saturation	95.94 ± 5.00	1.03 (1.00, 1.07)	0.048
Temperature	36.88 ± 0.84	0.77 (0.66, 0.91)	0.002
Sofa	9.63 ± 3.94	1.07 (1.03, 1.11)	0.000
Apsiii	64.00 ± 23.72	1.01 (1.01, 1.02)	<0.001
APRI	0.83 ± 1.67	1.04 (0.98, 1.11)	0.204
PWR	15.89 ± 10.90	0.98 (0.97, 1.00)	0.019
PHR	20.07 ± 11.06	0.99 (0.98, 1.01)	0.388
Hypertension	175 (18.74%)	0.97 (0.66, 1.41)	0.867
Diabetes Type II	322 (34.48%)	0.95 (0.71, 1.26)	0.712
Malignant Neoplasm	103 (11.03%)	1.11 (0.70, 1.74)	0.663
Chronic Kidney Disease	367 (39.29%)	1.11 (0.84, 1.46)	0.456
Liver Failure	142 (15.20%)	1.39 (1.03, 1.89)	0.033

Abbreviations: HR, Hazard Ratio; 95% CI,95% Confidence Interval; Sofa, Sequential Organ Failure Assessment; Apsiii, Acute Physiology Score III; APRI, AST/Platelet Ratio Index; PWR, Platelet to White Blood Cell Ratio; PHR, Platelet to Hemoglobin Ratio.

Multivariate Cox Regression Model for Different PWR Levels

In the MIMIC_EICU discovery cohort, multivariate Cox regression analysis indicated that elevated PWR levels were associated with a decreased risk of mortality in the ICU. In the unadjusted model, the third tertile (Q3) group (PWR > 17.72) exhibited a statistically significant 30% reduction in mortality risk compared to the first tertile (Q1) group (HR 0.70, 95% CI 0.51–0.98, P=0.035). This association persisted after adjusting for age, gender, and weight in Model 1 (HR 0.70, 95% CI 0.51–0.98, P = 0.037). In the fully adjusted Model 2, which controlled for all potential confounders, the Q3 group demonstrated a 34% reduction in ICU mortality risk (HR 0.66, 95% CI 0.44–0.99, P = 0.044), while the second tertile (Q2) group exhibited a non-significant 10% reduction in risk (HR 0.90, 95% CI 0.63–1.28, P = 0.552).

In the Enshi validation cohort, a comparable protective pattern was identified. Utilizing identical tertile cut-off values, the Q2 group exhibited a significantly reduced risk of ICU mortality compared to the Q1 group in the fully adjusted model (HR 0.08, 95% CI 0.01–0.84, P=0.035). Although the Q3 group showed a trend towards a decreased mortality risk, this did not achieve statistical significance (HR 0.40, 95% CI 0.09–1.76, P=0.225), likely due to the smaller sample size of the validation cohort. Comprehensive results for both cohorts are detailed in Table 3.

Table 3.

Association Between PWR Levels and Mortality in ICU Patients in Different Adjusted Multivariate Cox Regression Models

PWR	MIMIC_EICU cohort						Enshi cohort
	Non-adjusted (N=934)		Model 1(N=928)		Model 2(N=827)		Non-adjusted (N=124)		Model 1(N=124)		Model 2(N=124)
	HR (95% CI)	P-value	HR (95% CI)	P-value	HR (95% CI)	P-value	HR (95% CI)	P-value	HR (95% CI)	P-value	HR (95% CI)	P-value
Q1 group (≤9.95)	1		1		1		1		1		1
Q2 group (>9.95,≤17.72)	0.90 (0.66, 1.23)	0.523	0.85 (0.63, 1.16)	0.319	0.90 (0.63, 1.28)	0.552	0.07 (0.01, 0.55)	0.011	0.08 (0.01, 0.60)	0.014	0.08 (0.01, 0.84)	0.035
Q3 group (>17.72)	0.70 (0.51, 0.98)	0.035	0.70 (0.51, 0.98)	0.037	0.66 (0.45, 0.99)	0.044	0.46 (0.21, 1.01)	0.053	0.48 (0.22, 1.07)	0.074	0.40 (0.09, 1.76)	0.225

Notes: Model 1 adjust for: Age, Gender, Weight. Model 2 adjust for: Age, Gender, Weight, Heart Rate, Systolic Blood Pressure, Diastolic Blood Pressure, Oxygen Saturation, Temperature, Sofa, Apsiii, APRI, PHR, Hypertension, Diabetes Type II, Malignant Neoplasm, Chronic Kidney Disease, Liver Failure.

Abbreviations: HR, Hazard Ratio; 95% CI, 95% Confidence Interval; ICU, Intensive Care Unit; Sofa, Sequential Organ Failure Assessment; Apsiii, Acute Physiology Score III; APRI, AST/Platelet Ratio Index; PWR, Platelet to White Blood Cell Ratio; PHR, Platelet to Hemoglobin Ratio.

Kaplan-Meier Survival Analysis in the Validation Cohort

A Kaplan-Meier survival analysis was conducted on the Enshi validation cohort to evaluate ICU outcomes between patients with low PWR (≤9.95) and high PWR (>9.95). As illustrated in Figure 3, the median ICU survival time for patients in the low PWR group was significantly shorter at 16.0 days (95% CI 15.0–not applicable [NA]) compared to 47.0 days (95% CI NA–NA) for those in the high PWR group. Further, Cox regression analysis indicated that patients with low PWR had a 3.11-fold increased risk of ICU mortality compared to those with high PWR (HR 3.11, 95% CI 1.46–6.66). The Log rank test corroborated a statistically significant difference between the two groups (P=0.002).

Figure 3.

Kaplan-Meier survival curves comparing ICU mortality between patients with low PWR (≤9.95) and high PWR (>9.95) in the Enshi validation cohort. The Log rank test showed a statistically significant difference between the two groups (P = 0.002).

Abbreviations: PWR, platelet-to-white blood cell ratio; ICU, intensive care unit; HR, hazard ratio; CI, confidence interval.

Subgroup Analysis in the Validation Cohort

Subgroup analyses were conducted within the Enshi validation cohort to assess the relationship between low PWR (≤9.95) and ICU mortality across various clinical subgroups, using high PWR (>9.95) as the reference category (in Figure 4). Overall, patients exhibiting low PWR demonstrated a significantly elevated risk of ICU mortality (HR 3.11, 95% CI 1.46–6.66, P=0.003). This increased risk associated with low PWR was consistently observed across most subgroups; however, statistical significance was not achieved in certain strata, likely due to limited sample sizes. Importantly, no significant interactions were identified for any of the stratified variables (all P for interaction >0.05), suggesting that the prognostic value of PWR remains consistent and is not influenced by these clinical characteristics.

Figure 4.

Subgroup analysis evaluating the association between low platelet-to-white blood cell ratio (PWR; ≤9.95) and ICU mortality in the validation cohort, with high PWR (>9.95) as the reference. Hazard ratios (HRs) with 95% confidence intervals (CIs) are shown for each subgroup. No significant interactions were identified for any stratified variable (all P for interaction >0.05).

Abbreviations: PWR, platelet-to-white blood cell ratio; ICU, intensive care unit; HR, hazard ratio; CI, confidence interval.

Discussion

This study conducted an analysis of data from the MIMIC-EICU discovery cohort and subsequently validated the findings in an independent Enshi cohort, demonstrating that the PWR serves as an independent predictor of ICU mortality in patients experiencing septic shock with concurrent heart failure. The results indicate a significant association between low PWR levels and increased ICU mortality, a relationship that was corroborated in the validation cohort through multivariate regression and subgroup analyses.

In the context of septic shock, platelets and leukocytes play crucial roles in immune response and inflammation. Specifically, platelets are not only key components in coagulation but also involved in regulating the complex processes of immune response.¹⁴ Studies show that platelets enhance the body’s immune defense in septic shock by interacting with leukocytes, promoting leukocyte recruitment and activation, and thus taking on important roles during the early stages of infection. This interaction has a significant impact on the severity of the disease and patient prognosis.¹⁵

Leukocytes are crucial in inflammation, with neutrophils as the predominant white blood cells, essential for fighting bacterial infections and contributing to the inflammatory process. In septic shock, the overactivation and misregulation of leukocytes frequently result in tissue damage and multiple organ dysfunction, significantly contributing to the high mortality rate among patients.^16,17 In the pathophysiology of heart failure, inflammation also plays a critical role. Cardiac injury and stress responses trigger both local and systemic inflammation, and the release of inflammatory mediators promotes myocardial cell apoptosis and fibrosis¹⁸. The levels of inflammatory markers in heart failure patients are significantly elevated, reflecting the body’s response to cardiac injury, particularly the quantity and activity of leukocytes, which are closely related to the severity of heart failure.¹⁹

Getawa noted that common blood abnormalities in heart failure patients include neutrophilia, which may be linked to cell activation during the inflammatory process.²⁰ Additionally, platelet function is also affected in heart failure, with its synthesis, activation, and interaction with leukocytes playing a critical role in the inflammatory response.¹⁶ These changes not only reflect the pathophysiological characteristics of heart failure but may also influence patient prognosis. Similar to the findings of this study, platelets and leukocytes play important regulatory roles in both septic shock and heart failure, and their interactions not only impact the inflammatory response but may also have profound effects on patient prognosis.

Recent research suggests that the PWR holds promise as a prognostic biomarker for a variety of diseases. An elevated PWR has been independently correlated with enhanced survival outcomes in acute myeloid leukemia,¹¹ a reduced risk of aneurysm rupture,²¹ decreased 90-day mortality in cases of acute ischemic stroke,¹² fewer complications associated with pyogenic liver abscess,¹³ and lower short-term mortality in patients with hepatitis virus-related decompensated cirrhosis.²² These findings imply that PWR may serve as an indicator of underlying inflammatory and immune status across diverse clinical settings, thereby supporting its potential role as a universal prognostic marker. The correlation between PWR and ICU mortality has been confirmed by several studies. Hungthienthong identified a significant link between PWR and in-hospital mortality in a retrospective cohort study of critically ill COVID-19 patients. Non-survivors had a significantly lower PWR compared to survivors (15.8 vs. 29.0, p < 0.001), suggesting PWR’s potential in predicting outcomes for critically ill patients.²³ Similarly, Schupp et alin their study of heart failure patients identified several novel inflammatory markers, including neutrophil and platelet-related markers, that were correlated with cardiac mortality. Their variations also reflected the severity of the condition and prognosis.²⁴ Getawa highlighted the association between platelet and white blood cell abnormalities and clinical outcomes in heart failure patients, suggesting their prognostic potential.²⁰ These findings indicate that PWR is crucial for monitoring septic shock patients and may independently predict ICU mortality. This study establishes PWR as an independent prognostic indicator for ICU mortality in septic shock patients with heart failure, offering a valuable tool for clinical prognosis evaluation. This finding is clinically significant as it may facilitate early identification of high-risk patients and inform treatment strategies, thereby enhancing outcomes for patients with septic shock and heart failure.

It is important to note that this study still has some limitations. Firstly, while the MIMIC and eICU databases offer comprehensive clinical data, they predominantly originate from medical institutions within the United States, potentially introducing regional and population biases. To mitigate this limitation, we incorporated an external validation cohort from a Chinese tertiary hospital. However, the relatively small sample size of this validation cohort may constrain the statistical power and generalizability of our findings. Secondly, although the potential value of PWR as a biomarker in sepsis has been explored in several studies, its specific mechanisms in patients with heart failure complicated by septic shock remain unclear. The interaction between platelets and white blood cells may be regulated by various physiological and pathological factors, making the PWR alone insufficient to comprehensively reflect the clinical status of patients. Finally, PWR was measured at a single time point within the first 24 hours of ICU admission. Dynamic changes in PWR during the ICU stay were not explored, and their prognostic value warrants investigation in future studies. Moreover, combining other biomarkers and clinical scoring systems may provide a more accurate prognostic tool for patients with septic shock complicated by heart failure, enabling more effective individualized treatment strategies.

Conclusion

This study illustrates that a reduced platelet-to-white blood cell ratio is independently correlated with heightened ICU mortality among patients experiencing septic shock and heart failure. As an accessible and cost-effective hematological marker, the PWR holds potential as an adjunctive tool for early risk stratification in this high-risk cohort. However, further validation through larger multicenter studies is necessary before recommending its routine clinical implementation.

Funding Statement

This work was supported by grants from the Natural Science Foundation of Enshi, China(No.D20230070), and the Natural Science Foundation of Hubei Province, China (No.2025AFB807).

Reporting Guidelines

This study is reported following the RECORD (REporting of studies Conducted using Observational Routinely-collected health Data) guidelines,²⁵ to ensure comprehensive and transparent reporting of this observational study using routinely collected health data.

Abbreviations

PWR, Platelet to White Blood Cell Ratio; ICU, Intensive Care Unit; APACHE, Acute Physiology and Chronic Health Evaluation; Sofa, Sequential Organ Failure Assessment; Apsiii, Acute Physiology Score III; MIMIC-IV 2.2, Medical Information Mart for Intensive Care IV version 2.2 database; WBC, White blood cell; RCS, Restricted cubic splines; K-M curves, Kaplan-Meier survival curves; AST, Aspartate Aminotransferase; APRI, AST/Platelet Ratio Index; PHR, Platelet to Hemoglobin Ratio; AML, Acute myeloid leukemia; HR, Hazard Ratio; 95% CI, 95% Confidence Interval; NA, not applicable.

Data Sharing Statement

This study utilized publicly accessible datasets, which are available at the following link: https://physionet.org/about/database/. The Enshi validation cohort data used in this study are available from the corresponding author Peijun Liu upon reasonable request.

Approval for Ethics and Consent to Participate

This retrospective study complied with the Declaration of Helsinki and was approved by the Ethics Committee of Enshi Autonomous Prefecture Central Hospital (Approval No. 2024-064-01), which granted a waiver of informed consent.

Consent for Publication

All authors have given their consent for publication.

Author Contributions

QFL: Conceptualization, Investigation, Methodology, Writing-Original Draft, Writing-Review & Editing. WHT: Data Curation, Validation, Visualization, Writing -Original Draft. WT: Investigation, Methodology, Writing-Original Draft. MM: Resources, Software, Writing-Original Draft, Writing-Review & Editing. XT: Resources, Software, Writing-Review & Editing. MHZ: Conceptualization, Project Administration, Supervision, Writing-Review & Editing. PJL: Conceptualization, Project Administration, Visualization, Supervision, Funding, Writing-Review & Editing. All authors reviewed and approved the final manuscript. All authors gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors declare that there are no conflicts of interest that could have influenced the results or interpretations of this study.