Abstract
Background
The importance of the red cell distribution width (RDW)-to-platelet ratio (RPR) in clinical practice has been described in various diseases. However, the prognostic importance of the RPR in patients with acute respiratory failure (ARF) remains unclear. The aim of the present study was to investigate the association of the RPR with 30-day all-cause mortality among ARF patients in the intensive care unit (ICU).
Methods
Data were obtained from the Medical Information Mart for Intensive Care IV (MIMIC-IV) database (version 3.1). The study population was divided into different groups based on RPR quartiles (Q1–Q4). The outcome of interest was all-cause mortality within 30 days. Cox proportional regression analysis was performed to determine the relationship between the RPR and 30-day mortality, and Kaplan‒Meier (K‒M) analysis was conducted to obtain survival curves. Additionally, restricted cubic spline (RCS) analysis was utilized to determine whether there was a nonlinear correlation between the RPR (as a continuous variable) and 30-day mortality in ARF patients. Subgroup analysis was subsequently performed to determine whether the association between the RPR and 30-day mortality in ARF patients was stable across the subgroups.
Results
A total of 10,912 ARF patients were enrolled in this study. The 30-day mortality rate was 27.2%. Cox proportional regression analysis revealed that an elevated RPR was associated with an increased risk of 30-day mortality in ARF patients. Furthermore, K‒M survival analyses revealed that patients with higher RPR levels demonstrated significantly higher all-cause mortality rates within 30 days of admission than patients with lower RPRs. The RCS model illustrated the linear relationship between a higher RPR and increased risk of mortality in this specific patient population. The final subgroup analysis indicated no significant interaction effect of the RPR in most of the subgroups.
Conclusions
This study demonstrated that RPR levels were significantly associated with short-term prognosis in patients with ARF and that a high RPR could serve as an independent predictor of all-cause mortality at 30 days in patients with ARF.
Keywords: Acute respiratory failure, Red cell distribution width-to-platelet ratio, MIMIC-IV database, Intensive care unit, All-cause mortality
Introduction
Acute respiratory failure (ARF) is one of the most common and prevalent life-threatening clinical syndromes and is characterized by hypoxia with or without hypercapnia [1]. In this condition, the respiratory system fails to perform pulmonary ventilation and gas exchange functions, thus resulting in acid-base metabolic disorders and the failure of important organs, such as the heart and kidney [2]. The underlying causes of ARF vary and can be attributed to severe pneumonia, emphysema, lung injury, pulmonary edema, sepsis, congestive heart failure, and central nervous system depression [3]. ARF is a common reason for admission to the intensive care unit (ICU), and nearly 40% of ARF-related deaths occur during hospitalization [4]. An epidemiological report revealed that the United States government may spend up to $5 billion annually on the treatment of respiratory failure [5]. Due to the high treatment expenses and in-hospital mortality, it is necessary to discover an accurate predictor of ARF-related mortality, in order to implement effective management strategies.
Red blood cell distribution width (RDW) is a quantitative parameter that is typically used to measure the degree of volume variation in erythroid cells and to help in the diagnosis of anemia via hematology. The RDW has received considerable attention with respect to the clinical prediction of mortality risk in various diseases, such as coronavirus disease 2019, acute myocardial infarction, pulmonary hypertension, acute kidney injury, and acute pancreatitis [6–10]. Moreover, a recent study demonstrated that the RDW is a potential parameter for predicting adverse outcomes in patients with ARF [11]. The red blood cell distribution width-to-platelet ratio (RPR), which is a novel index that is used to assess the severity of inflammation, has been employed for prognostic prediction in various diseases, such as acute pancreatitis (AP) and fibrosis in chronic hepatitis [12, 13]. Nevertheless, to date, few studies have investigated the prognostic value of the RPR in patients with ARF. Hence, this study aimed to determine whether the RPR can be used to predict short-term mortality in patients with ARF.
Methods
Database
In this retrospective study, the data were obtained from a free and developed database known as the Medical Information Mart for Intensive Care IV (MIMIC-IV) database (version 3.1). The MIMIC-IV is a large, open clinical database containing information on patients who were admitted to the ICU at a major Boston hospital from 2008 to 2022. To gain access to the database, Author You completed the “Protecting Human Research Participants” exam (certification number: 61758815) and executed a data use agreement. All of the personal information of the patients was hidden.
Study population
Patients eligible for a diagnosis of ARF were initially screened according to the International Classification of Diseases 9 code (51851 and 51881) and 10 code (J960, J9600, J9601, and J9602). The exclusion criteria included patients with a length of stay in the ICU < 24 h; multiple admissions to the ICU (but only retaining the data from the first admission); no RDW or platelet value; and a diagnosis of leukemia, myelodysplastic syndrome, thrombocytosis or thrombocytopenia.
Data extraction
The following outcome data were extracted for the enrolled patients based on previous studies and clinical relevance: age, sex, race, vital signs [including temperature, heart rate (HR), respiratory rate (RR), mean systolic blood pressure (SBP), mean diastolic blood pressure (DBP), mean blood pressure (MBP), and oxyhemoglobin saturation (SpO2)], comorbidities [including hypertension, diabetes, heart failure (HF), myocardial infarction (MI), malignant neoplasm (MN), chronic kidney disease (CKD), and chronic obstructive pulmonary disease (COPD)], laboratory parameters [including the international normalized ratio (INR), partial thromboplastin time (PTT), red blood cell (RBC) count, white blood cell (WBC) count, and red blood cell distribution width (RDW), as well as hematocrit, hemoglobin, glucose, blood urea nitrogen (BUN), creatinine, chloride, sodium, potassium, calcium, and platelet levels], scores [including the Sequential Organ Failure Assessment (SOFA) score, the Systemic Inflammatory Response Syndrome (SIRS) score, and the Simplified Acute Physiology Score (SAPSII)], and clinical treatments (including vancomycin, norepinephrine, noninvasive therapies and invasive therapies). All of the data were extracted within 24 h of admission to the ICU. The 30-day mortality rate was considered to be the survival endpoint.
Study outcomes
The endpoint of the study was all-cause mortality within 30 days of patient admission.
Missing data
All of the variables had fewer than 20 missing values; moreover, we used multiple interpolations to fill in missing values for the appropriate variables.
Statistical analysis
The study population was categorized into four groups based on the RPR quartile. First, continuous variables were compared using the Kruskal‒Wallis test or ANOVA F test, and the results are expressed as medians (interquartile ranges, or IQRs). Categorical variables were then compared via the chi-square test and are expressed as the number of cases and percentages. To examine the association between the RPR and outcome risk, we constructed progressively adjusted Cox proportional hazards models, and the results are expressed as hazard ratios (HRs) and 95% confidence intervals (CIs). Model 1 was unadjusted. Model 2 was adjusted for sex, age, and race. Model 3 further adjusted Model 2 for SBP, DBP, MBP, HR, RR, temperature, SpO2, WBC, hemoglobin, hematocrit, RBC, INR, PTT, creatinine, BUN, sodium, potassium, calcium, chloride, and glucose. Model 4 additionally adjusted Model 3 for hypertension, diabetes, HF, MI, MN, CKD, and COPD. Finally, Model 5 was built upon Model 4 by incorporating the SOFA score, SIRS score, SAPS, vancomycin, norepinephrine, noninvasive treatments, and invasive treatments. The restricted cubic spline (RCS) model was used to explore the nonlinear relationship between the RPR and the risk of death in patients with ARF. Survival analysis of clinical outcomes at different RPRs was performed via Kaplan‒Meier (K‒M) curves. Subgroup analyses were performed to compare the correlation between the 30-day RPR and mortality in patients with ARF. All of the statistical analyses were conducted using R software (version 4.4.1), and a two-sided P value of less than 0.05 was considered to be statistically significant.
Results
Baseline characteristics
The selection process of the study population is shown in Fig. 1. As shown in Table 1, we extracted 10,912 eligible ARF patients from the MIMIC-IV database. The data were categorized into four groups according to the quartiles of the RPR (Q1: RPR < 0.052; Q2: 0.052 ≤ RPR < 0.068; Q3: 0.068 ≤ RPR < 0.090; Q4: RPR ≥ 0.090). Baseline characteristics were subsequently categorized. 54% of the patients were male, and 62.1% were white. The in-hospital, 30-day, and intensive care unit mortality rates were 23.8%, 27.2%, and 17.2%, respectively. Additionally, patients in the high RPR group demonstrated more unstable weights for life and higher disease scores than those in the low RPR group. Patients with higher RPRs were more likely to have experienced COPD, chronic kidney disease, myocardial infarction, and hypertension; in contrast, they were less likely to have experienced hypertension and diabetes mellitus. Moreover, patients in the fourth quartile exhibited lower levels of white blood cells, red blood cells, hemoglobin, hematocrit, blood calcium, and blood glucose, as well as higher levels of creatinine, international normalized ratio (INR), PTT, and urea nitrogen. The fourth quartile was associated with the need for more invasive treatments, such as norepinephrine and vancomycin.
Fig. 1.
Study flowchart
Table 1.
Baseline characteristics according to RPR quartiles
| Characteristics | RPR | P value | ||||
|---|---|---|---|---|---|---|
| Q1 | Q2 | Q3 | Q4 | |||
| RPR ≤ 0.052 | 0.052 < RPR ≤ 0.068 | 0.068 < RPR ≤ 0.090 | RPR > 0.090 | Total | ||
| Participants, N | 2728 | 2730 | 2726 | 2728 | 10,912 | |
| Demographics | ||||||
| Age, years | 67.1 (55.6, 77.1) | 68.3 (56, 78.8) | 70.3 (59.4, 80.6) | 68.8 (57.6, 79.6) | 68.5 (57.1, 79.2) | < 0.001 |
| Gender, n (%) | < 0.001 | |||||
| Female | 1434 (52.6) | 1308 (47.9) | 1157 (42.4) | 1118 (41) | 5017 (46) | |
| Male | 1294 (47.4) | 1422 (52.1) | 1569 (57.6) | 1610 (59) | 5895 (54) | |
| Race, n (%) | 0.032 | |||||
| Black | 281 (10.3) | 322 (11.8) | 351 (12.9) | 335 (12.3) | 1289 (11.8) | |
| Other | 688 (25.2) | 741 (27.1) | 709 (26) | 707 (25.9) | 2845 (26.1) | |
| White | 1759 (64.5) | 1667 (61.1) | 1666 (61.1) | 1686 (61.8) | 6778 (62.1) | |
| Vital signs | ||||||
| Temperature, °C | 36.9 (36.7, 37.2) | 36.9 (36.7, 37.3) | 36.9 (36.7, 37.2) | 36.9 (36.6, 37.2) | 36.9 (36.6, 37.2) | < 0.001 |
| HR, beats/min | 89.2 (77.9, 101.3) | 84.8 (74.7, 97.2) | 83.4 (73.2, 95.6) | 86.4 (74.9, 99.5) | 85.9 (75, 98.6) | < 0.001 |
| RR, breaths/min | 20.5 (17.9, 23.7) | 20 (17.5, 23) | 19.9 (17.5, 22.6) | 19.9 (17.3, 22.8) | 20 (17.6, 23) | < 0.001 |
| SBP, mmHg | 115.6 (106.1, 127.6) | 116.3 (107.1, 129.2) | 115.9 (106.7, 127.6) | 112.7 (104, 123.9) | 115.2 (105.9, 127.1) | < 0.001 |
| DBP, mmHg | 63.2 (56.5, 71) | 63.1 (56.9, 70.5) | 62.4 (56, 70) | 61.2 (55, 68.3) | 62.5 (56.1, 69.9) | < 0.001 |
| MBP, mmHg | 77.8 (71.7, 85.8) | 78.2 (71.9, 85.7) | 77.7 (71.5, 85) | 75.7 (70, 82.8) | 77.3 (71.2, 84.9) | < 0.001 |
| SpO2, % | 96.9 (95.3, 98.4) | 97 (95.2, 98.5) | 96.9 (95.2, 98.5) | 97.1 (95.3, 98.6) | 97 (95.2, 98.5) | 0.361 |
| Comorbidities, n (%) | ||||||
| Hypertension | 522 (19.1) | 536 (19.6) | 483 (17.7) | 420 (15.4) | 1961 (18) | < 0.001 |
| CKD | 315 (11.5) | 398 (14.6) | 499 (18.3) | 522 (19.1) | 1734 (15.9) | < 0.001 |
| MI | 395 (14.5) | 450 (16.5) | 484 (17.8) | 462 (16.9) | 1791 (16.4) | 0.009 |
| MN | 789 (28.9) | 652 (23.9) | 603 (22.1) | 701 (25.7) | 2745 (25.2) | < 0.001 |
| Diabetes | 968 (35.5) | 1009 (37) | 1026 (37.6) | 917 (33.6) | 3920 (35.9) | 0.01 |
| HF | 903 (33.1) | 1012 (37.1) | 1164 (42.7) | 1112 (40.8) | 4191 (38.4) | < 0.001 |
| COPD | 355 (13) | 375 (13.7) | 428 (15.7) | 397 (14.6) | 1555 (14.3) | 0.031 |
| Laboratory events | ||||||
| WBC, 109/L | 14.4 (10.6, 19.3) | 12.2 (9.1, 16.3) | 10.8 (8.1, 14.5) | 9.3 (6.3, 13.4) | 11.7 (8.3, 16.1) | < 0.001 |
| RBC, 1012/L | 3.7 (3.2, 4.3) | 3.8 (3.3, 4.3) | 3.7 (3.2, 4.2) | 3.3 (2.8, 3.9) | 3.6 (3.1, 4.2) | < 0.001 |
| Creatinine, mg/dL | 1 (0.7, 1.6) | 1.1 (0.8, 1.7) | 1.1 (0.8, 1.9) | 1.4 (0.9, 2.4) | 1.1 (0.8, 1.9) | < 0.001 |
| Hemoglobin, g/dL | 11 (2.3) | 11.2 (2.3) | 11 (2.2) | 10.1 (2.2) | 10.8 (2.3) | < 0.001 |
| Hematocrit, % | 34 (6.6) | 34.5 (6.5) | 34 (6.6) | 31.2 (6.8) | 33.4 (6.8) | < 0.001 |
| INR | 1.2 (1.1, 1.5) | 1.2 (1.1, 1.4) | 1.2 (1.1, 1.5) | 1.4 (1.2, 1.8) | 1.3 (1.1, 1.6) | < 0.001 |
| PTT, sec | 30.2 (26.8, 35.5) | 29.8 (26.7, 35.4) | 30.4 (27.1, 36.8) | 32.7 (28.4, 41.2) | 30.7 (27.2, 37.1) | < 0.001 |
| BUN, mg/dL | 22 (15, 35) | 22 (15, 38) | 24 (16, 40) | 28 (18, 47) | 24 (15, 40) | < 0.001 |
| Sodium, mEq/L | 140 (137, 142) | 140 (137, 143) | 140 (138, 143) | 140 (137, 143) | 140 (137, 143) | < 0.001 |
| Potassium, mEq/L | 4.4 (4, 4.9) | 4.4 (4, 4.9) | 4.4 (4, 4.9) | 4.4 (4.1, 4.9) | 4.4 (4, 4.9) | 0.157 |
| Calcium, mEq/L | 8.7 (8.2, 9.1) | 8.6 (8.2, 9) | 8.5 (8.1, 9) | 8.4 (8, 9) | 8.6 (8.1, 9) | < 0.001 |
| Chloride, mEq/L | 104 (100, 108) | 105 (101, 109) | 105 (101, 109) | 105 (100, 110) | 105 (101, 109) | < 0.001 |
| Glucose, mg/dL | 157 (125, 208) | 154 (122, 208) | 146 (118, 198) | 147 (117, 194) | 151 (121, 202) | < 0.001 |
| Therapies, n (%) | ||||||
| Noninvasive | 179 (6.6) | 189 (6.9) | 216 (7.9) | 190 (7) | 774 (7.1) | 0.24 |
| Invasive | 1870 (68.5) | 1898 (69.5) | 1899 (69.7) | 1927 (70.6) | 7594 (69.6) | 0.419 |
| Norepinephrine | 886 (32.5) | 862 (31.6) | 817 (30) | 1033 (37.9) | 3598 (33) | < 0.001 |
| Vancomycin | 1978 (72.5) | 1855 (67.9) | 1846 (67.7) | 2006 (73.5) | 7685 (70.4) | < 0.001 |
| Scores | ||||||
| SOFA | 1 (0, 2) | 1 (0, 3) | 1 (0, 3) | 2 (0, 4) | 1 (0, 3) | < 0.001 |
| SIRS | 3 (2, 4) | 3 (2, 3) | 3 (2, 3) | 3 (2, 3) | 3 (2, 3) | < 0.001 |
| SAPSII | 39 (30, 49) | 37 (29, 47) | 39 (31, 48) | 42 (34, 52) | 39 (31, 49) | < 0.001 |
| Events | ||||||
| LOS ICU, days | 4 (2.1, 8.1) | 4 (2.1, 8.6) | 4 (2.1, 8.3) | 3.9 (2.2, 8.2) | 4 (2.1, 8.3) | 0.995 |
| LOS hospital, days | 10.9 (6.3, 18.9) | 10.7 (6.2, 18.7) | 10.3 (6, 17.8) | 11.1 (6, 20) | 10.8 (6.1, 18.8) | 0.02 |
| ICU death, n (%) | 474 (17.4) | 400 (14.7) | 414 (15.2) | 584 (21.4) | 1872 (17.2) | < 0.001 |
| Hospital death, n (%) | 653 (23.9) | 547 (20) | 575 (21.1) | 819 (30) | 2594 (23.8) | < 0.001 |
| 30-day death, n (%) | 745 (27.3) | 651 (23.8) | 676 (24.8) | 897 (32.9) | 2969 (27.2) | < 0.001 |
Abbreviations: RPR Red blood cell distribution width-to-platelet ratio, HR heart rate; RR, respiratory rate, SBP systolic blood pressure, DBP diastolic blood pressure, MBP mean blood pressure, SpO2 oxyhemoglobin saturation, HF heart failure, CKD chronic kidney disease, MI myocardial infarction, MN malignant neoplasm, COPD chronic obstructive pulmonary disease, WBC white blood cell, RBC red blood cell, INR international normalized ratio, PTT partial thromboplastin time, SOFA Sequential Organ Failure Assessment, SIRS Systemic Inflammatory Response Syndrome, SAPSII Simplified Acute Physiological Score II, ICU intensive care unit, LOS length of stay
Association between the RPR quartile and 30-day mortality
According to the Kaplan‒Meier survival analysis curves, patients with ARF in group Q4 demonstrated a lower probability of survival than those in the other three quartile groups (log-rank P < 0.0001) (Fig. 2). Cox proportional risk analysis revealed that in the unadjusted model (Model 1), the risk of all-cause mortality in patients with ARF was significantly associated with the RPR (HR: 2.15 [95% CI: 1.76–2.62], P < 0.001). Furthermore, this association was significantly maintained in the fully adjusted model (Model 5) (HR: 1.73 [95% CI: 1.36–2.20], P < 0.001); moreover, for every 1-unit increase in the RPR, the risk of death in patients with ARF increased by a factor of 1.73. Additionally, the risk of death was significantly greater in the Q4 group than in the Q1 group; specifically, the risk in the Q4 group was 1.28 times greater than that in the Q1 group (Model 1 - HR: 1.26 [95% CI: 1.15–1.39], P = 0.001; Model 2 - HR: 1.23 [95% CI: 1.11–1.35], P < 0.001; Model 3 - HR: 1.26 [95% CI: 1.14–1.40], P < 0.001; Model 4 - HR: 1.33 [95% CI: 1.20–1.47], P < 0.001; Model 5 - HR: 1.28 [95% CI: 1.15–1.42], P < 0.001) (Table 2). To further explore the nonlinear association between the RPR and ARF all-cause mortality, we performed a restricted regression spline analysis by assuming three nodes for the 30-day all-cause mortality of the patients (Fig. 3). The RCS regression model demonstrated that there was a significant nonlinear association between the RPR and the 30-day all-cause mortality of the patients (p for nonlinearity < 0.001). The identification of a critical value (cutoff value = 0.071) suggested a threshold effect of the RPR in predicting the risk of 30-day mortality, which was observed to be clinically stratified and important (p < 0.001).
Fig. 2.
Kaplan‒Meier survival analysis curves
Table 2.
Cox proportional hazard ratios (HRs) for all-cause mortality based on the RPR
| Categories | Model 1 | Model 2 | Model 3 | Model 4 | Model 5 | |||||
|---|---|---|---|---|---|---|---|---|---|---|
| HR (95% CI) | P | HR (95% CI) | P | HR (95% CI) | P | HR (95% CI) | P | HR (95% CI) | P | |
| 30 days mortality | ||||||||||
| Continuous variable per 1 unit | 2.15 (1.76–2.62) | < 0.001 | 2.27 (1.88–2.73) | < 0.001 | 2.00 (1.64–2.44) | < 0.001 | 1.80 (1.48–2.20) | < 0.001 | 1.73 (1.36–2.20) | < 0.001 |
| Q1 | Reference | Reference | Reference | Reference | Reference | |||||
| Q2 | 0.86 (0.77–0.95) | 0.005 | 0.84 (0.76–0.94) | 0.002 | 0.94 (0.84–1.04) | 0.248 | 0.98 (0.88–1.09) | 0.748 | 1.01 (0.90–1.12) | 0.910 |
| Q3 | 0.90 (0.81-1.00) | 0.051 | 0.86 (0.77–0.95) | 0.004 | 0.97 (0.87–1.08) | 0.549 | 1.04 (0.93–1.16) | 0.486 | 1.06 (0.95–1.18) | 0.278 |
| Q4 | 1.26 (1.15–1.39) | < 0.001 | 1.23 (1.11–1.35) | < 0.001 | 1.26 (0.76–0.94) | < 0.001 | 1.33 (1.15–1.42) | < 0.001 | 1.28 (1.15–1.42) | < 0.001 |
Model 1 was unadjusted
Model 2 was adjusted for sex, age and race
Model 3 was adjusted for sex, age, race, SBP, DBP, MBP, HR, RR, temperature, SpO2, WBC, hemoglobin, hematocrit, RBC, INR, PTT, creatinine, BUN, sodium, potassium, calcium, chloride, and glucose
Model 4 was adjusted for sex, age, race, SBP, DBP, MBP, HR, RR, temperature, SpO2, WBC, hemoglobin, hematocrit, RBC, INR, PTT, creatinine, BUN, sodium, potassium, calcium, chloride, glucose, hypertension, diabetes, HF, MI, MN, CKD, and COPD
Model 5 was adjusted for sex, age, race, SBP, DBP, MBP, HR, RR, temperature, SpO2, WBC, hemoglobin, hematocrit, RBC, INR, PTT, creatinine, urea nitrogen, sodium, potassium, calcium, chloride, glucose, hypertension, diabetes, HF, MI, MN, CKD, COPD, SOFA, SIRS, SAPS, vancomycin, norepinephrine, noninvasive treatment, and invasive treatment
Abbreviations: HR hazard ratio, CI confidence interval, HR heart rate, RR respiratory rate, SBP systolic blood pressure, DBP diastolic blood pressure, MBP mean blood pressure, SpO2 oxyhemoglobin saturation, HF heart failure, CKD chronic kidney disease, MI myocardial infarction, MN malignant neoplasm, COPD chronic obstructive pulmonary disease, WBC white blood cell, RBC red blood cell, INR international normalized ratio, PTT partial thromboplastin time, SOFA Sequential Organ Failure Assessment, SIRS Systemic Inflammatory Response Syndrome, SAPSII Simplified Acute Physiological ScoreII
Fig. 3.
Nonlinear relationship analysis between the RPR and all-cause mortality due to ARF
Subgroup analyses
Additionally, to confirm the relationship between the RPR and 30-day mortality, subgroup analyses were performed according to age, sex, race, hypertension, diabetes, heart failure, myocardial infarction, malignant neoplasm, chronic kidney disease and chronic obstructive pulmonary disease (Fig. 4). The results revealed that the RPR was significantly associated with an increased risk of death within 30 days in specific subgroups of ARF patients. These subgroups included patients younger than 65 years of age (HR: 3.21 [95% CI: 2.38–4.34]), patients of other races (HR: 3.82 [95% CI: 2.56–5.69]), and patients with no malignancy (HR: 2.82 [95% CI: 2.21–3.58]) (P for all interactions < 0.05). Most of the other subgroups were not significantly different from the RPR cross-stratification (P for interaction > 0.05). Therefore, the subgroup analyses suggested that our results were relatively stable and less affected by confounders.
Fig. 4.
Subgroup analyses
Discussion
In this study, the RPR was demonstrated to be an independent predictor of 30-day all-cause mortality in patients with ARF. After confounders such as age, sex, race, laboratory test indices, comorbidities, disease severity scores (such as APACHE II and SOFA scores), and treatment regimens were corrected, ARF patients with high RPR values demonstrated significantly greater 30-day (HR: 1.28 [95% CI 1.15–1.42]; p < 0.001) and all-cause mortality rates. Additionally, RCS analysis revealed a significant nonlinear relationship between the RPR and 30-day all-cause mortality in ARF patients, with a steeply increasing trend in the mortality risk curve being observed when the RPR > 0.071, thus suggesting that the RPR may have a clear clinical cutoff value. In a previous study involving acute respiratory distress syndrome patients, Qin et al. [14] reported that the RPR was independently associated with 28-day all-cause mortality after correction for correlates in a multivariate model (HR: 2.74 [95% CI: 1.46–5.15]; p = 0.002). Moreover, Chen et al. [15] reported that the RPR was significantly associated with 28-day all-cause mortality (OR: 1.36 [95% CI: 1.01–1.84]; p = 0.042). These studies are consistent with the results of the present study. In the present study, the results revealed that the mortality rate of patients in the fourth quartile group of the RPR was significantly greater than that of the other three groups (p < 0.001); additionally, the all-cause mortality rate of patients with higher RPRs was significantly greater than that of patients with lower RPRs in the other three groups within 30 days after admission (log-rank test, p < 0.0001). The abovementioned results suggest the robustness of the RPR as an independent predictor of prognosis in patients with ARF. The RPR not only demonstrates good prognostic efficacy in patients with acute respiratory failure but also exhibits important clinical predictive value in other respiratory diseases.
The RPR is a simple biological index that is defined as the RDW divided by the platelet count. As a potential biomarker of the inflammatory response, the RPR has previously been regarded as a prognostic indicator in inflammation-related diseases. Specifically, Chen et al. described the RPR as an independent factor corresponding to hepatic fibrosis and cirrhosis in chronic hepatitis B patients and as a promising marker reflecting inflammation severity [12]. Similarly, Wang et al. demonstrated that the RPR was significantly associated with the histological severity of primary biliary cirrhosis (PBC) [16]. Moreover, Takeuchi et al. reported that an elevated RPR could be used as an auxiliary assessment tool to predict poor prognosis in patients with breast carcinoma [17]. Other studies have revealed that the RPR is a valuable predictive marker of mortality in acute pancreatitis, sepsis and acute myocardial infarction (AMI) patients [13, 17–20]. Furthermore, previous research has demonstrated that the RPR is more effective than the RDW or platelet count alone for predicting the prognosis of acute kidney injury, which is possibly due to the fact that the RPR comprehensively reflects the inflammatory load and metabolic state of patients [21]. To the best of our knowledge, no such association has yet been reported in patients with ARF. In this large public retrospective cohort analysis of MIMIC-IV datasets, we revealed for the first time that an elevated RPR is independently related to a significant increase in the risk of 30-day mortality in patients with ARF. Notably, the RCS analysis revealed a nonlinear correlation between the RPR and the risk of 30-day mortality in ARF patients. Subsequent subgroup analysis indicated no obvious interactions. Our findings are similar to those of observational studies of other diseases and have important clinical implications for the appropriate management of ARF patients.
The utility of the RPR as a prognostic predictor of ARF is associated with the biological functions of platelets and the RDW. The RDW is an easily obtainable parameter of the complete blood count test and routinely reflects the systemic inflammatory state and oxidative stress characteristics of patients (to a certain extent). Previous studies have demonstrated that the RDW is an independent predictor of both short-term and long-term mortality in ARF patients. For example, Zhang et al. reported that a high RDW at ICU admission was associated with an increased risk of 3-year mortality in patients with ARF [11]. Additionally, Shi et al. reported that an elevated RDW was associated with a higher 30-day mortality rate in ARF patients [22]. Several factors may partially explain the mechanism by which the RDW influences the prognosis of ARF in patients. Abnormal inflammatory reactions and oxidative stress inhibit erythrocyte maturation, induce increased production of new reticulocytes, reduce the survival rate of erythrocytes and cause the release of large premature red blood cells into the peripheral circulation, thereby resulting in an increase in the RDW [23, 24]. Additionally, systemic inflammation contributes to RDW elevation by changing the membrane glycoproteins and ion channels of RBCs, with consequent morphological alterations being elicited [25, 26]. Moreover, the RDW has been associated with elevated levels of several inflammatory markers, such as CRP and interleukins [27]. The platelet count represents another widely used complete blood count parameter and has historically been used as an important indicator of both the inflammatory process and hemostasis [28]. The relationship between the platelet count and poor prognosis of ARF in patients has attracted increasing attention. Both Xiao et al. and Zhou et al. emphasized a low platelet count as being a risk factor for mortality in ARF patients when the platelet count is detected within a certain range [29, 30]. Additionally, Lopez-Delgado et al. demonstrated that ARF patients with H1N1 influenza complicated with thrombocytopenia exhibited an increased risk of in-hospital mortality in a prospective observational study [31]. In general, numerous platelets are damaged by various toxins and inflammatory factors that are produced due to severe infection and/or acute inflammation, thus leading to dysfunction of the bone marrow microenvironment, insufficient production of thrombopoietin, suppression of stem cell/progenitor cell proliferation, and a decrease in the number of platelets [32–36].
Significant associations were observed in most of the subgroups, thus indicating the stability of the results. Additionally, the prognostic impact of the RPR was much greater in the subgroups of individuals younger than the age of 65 years, individuals of other races, and individuals with no malignancies, thus suggesting potential interactions between the RPR and these risk factors. Based on the findings of this study, it is recommended that specific RPR thresholds involving age, other races, and a lack of malignancy be integrated into risk stratification models of ARF; moreover, multidimensional intervention protocols including immunomodulatory therapy should be initiated in these specific groups of patients when their RPR levels exceed the 0.071 threshold. To the best of our knowledge, this is the first study to explore these interactions. However, the underlying mechanisms of these interactions remain unclear. Therefore, future studies should elucidate the molecular regulatory network of RPR abnormalities via multiomics techniques and conduct prospective cohort studies to validate the clinical value of targeted intervention strategies.
In this study, we observed that the RPR was significantly associated with mortality (HR = 2.15–2.27) in the unadjusted model (Model 1) and in the base-adjusted model (Model 2, adjusted for age, sex, and race). However, in the fully adjusted model (Model 5), which was further corrected for blood pressure parameters (including systolic, diastolic, and mean arterial pressure parameters), laboratory indices, and comorbidities, the effect value was attenuated (HR = 1.73). This attenuation may partly be due to the presence of multicollinearity with respect to the variables included in the model, such as the mean arterial pressure (MBP), which is essentially a derivative of the systolic (SBP) and diastolic (DBP) pressure parameters, as well as the scoring system, which includes vital signs and test results (in which the high degree of correlation may have diluted the independent effect of the RPR). Although we controlled for confounders by a stepwise adjustment strategy, the inherent correlation of such physiological parameters may still affect model stability.
Previous studies have demonstrated that the systemic immunoinflammatory index, platelet-lymphocyte ratio, and neutrophil-lymphocyte ratio are predictive of mortality in pneumosepsis patients in the intensive care unit [37]. Indices related to hemodynamic deterioration, such as the shock index and its derivatives, demonstrate an impact on the prognosis of ARF [38]. The combination of these indices with the RPR may enhance clinical decision-making.
The RPR is a conveniently accessible dual marker of inflammation and coagulation, and this study demonstrated its independent predictive value in the short- and medium-term prognostic assessment of ARF patients. Its nonlinear risk relationship and robust characteristics among different subgroups provide a theoretical basis for the establishment of RPR-based risk stratification models in clinical practice. Future studies could further explore the biological mechanisms of the RPR in respiratory diseases and provide guidance for individualized treatment decisions.
Limitations
This study explored the association between the RPR and the risk of death in ARF patients admitted to the ICU. Therefore, the novelty of this study is reflected in the establishment of a quantitative relationship between novel inflammatory indicators and the prognosis in critically ill patients. However, this study demonstrates several shortcomings. First, the parameter collection implemented in this study was limited to the baseline data collected at the time of the patient’s first admission to the ICU, and patients with multiple hospitalizations were not included; moreover, parameters that may exhibit dynamic changes during treatment were not incorporated, and this single time-point measurement may have weakened the strength of the observed temporal association between the index and prognosis. Second, the retrospective study design and single-center data source may have led to selection bias, and the limited sample size may have reduced the statistical power for rare clinical outcomes. Finally, although major confounders were corrected for via multivariate modeling, there may still be unmeasured residual confounders. Based on these limitations, a future multicenter, prospective cohort study with a large sample size should be conducted to obtain RPR trajectory data at multiple time points via a standardized dynamic monitoring protocol and to construct a composite predictive model in conjunction with other inflammation and organ function metrics. Such a study would further validate the external validity of the findings of the present study and the potential clinical translation of the RPR.
Conclusion
The RPR is significantly associated with the short-term prognosis of ARF in patients, and high RPRs can be used as an independent predictor of all-cause mortality at 30 days among ARF patients. The RPR can be used for risk stratification among ICU patients and serves as an objective assessment tool for the dynamic monitoring of disease progression. These findings need to be validated in future prospective studies.
Acknowledgements
I would like to express my sincere gratitude to the AJE (Academic Journal Experts) team for their professional editing services, which have significantly enhanced the clarity and readability of this manuscript.
Abbreviations
- RDW
Red cell distribution width
- RPR
Red cell distribution width-to-platelet ratio
- ARF
Acute respiratory failure
- ICU
Intensive care unit
- K‒M
Kaplan‒Meier
- RCS
Restricted cubic spline
- AP
Acute pancreatitis
- HR
Heart rate
- RR
Respiratory rate
- SBP
Systolic blood pressure
- DBP
Diastolic blood pressure
- MBP
Mean blood pressure
- SpO2
Oxyhemoglobin saturation
- HF
Heart failure
- MI
Myocardial infarction
- MN
Malignant neoplasm
- CKD
Chronic kidney disease
- COPD
Chronic obstructive pulmonary disease
- INR
International normalized ratio
- PTT
Partial thromboplastin time
- BUN
blood urea nitrogen
- RBC
Red blood cell
- WBC
White blood cell
- SOFA
Sequential Organ Failure Assessment
- SIRS
Systemic Inflammatory Response Syndrome
- SAPSII
Simplified Acute Physiology Score
- HR
Hazard ratio
- CI
Confidence interval
- PBC
Primary biliary cirrhosis
- AMI
Acute myocardial infarction
Authors’ contributions
YY completed the data analysis and original draft. KR and HC completed the data collection and data analysis. MW and SC completed the validation and review. LC provided funding, supervision, writing, reviewing and editing.
Funding
The authors declare that financial support was received for the research, authorship, and/or publication of this article. This study was supported by funding from Science and Technology Projects in Guangzhou (No. 2025A03J4224).
Data availability
The data utilized in this study are publicly accessible through the MIMIC-IV database version 3.1 The data utilized in this study are publicly accessible through the MIMIC-IV database version 3.1 (https://physionet.org/content/mimiciv/3.1/).
Declarations
Ethics approval and consent to participate
The studies involving humans were approved by the Institutional Review Boards of Beth Israel Deaconess Medical Center and the Massachusetts Institute of Technology. The studies were conducted in accordance with local legislation and institutional requirements. The ethics committee/institutional review board waived the requirement of written informed consent for participation from the participants or the participants’ legal guardians/next of kin due to the use of anonymous clinical data.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Yong You, Kengliang Rao, Hongjia Chen and Min Wang contributed equally to this work.
References
- 1.van Gemert JP, van den Berk IAH, Nossent EJ, et al. Cyclophosphamide for interstitial lung disease-associated acute respiratory failure: mortality, clinical response and radiological characteristics. BMC Pulm Med. 2021;21(1):249. 10.1186/s12890-021-01615-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Azoulay E, Mokart D, Kouatchet A, et al. Acute respiratory failure in immunocompromised adults. Lancet Respir Med. 2019;7(2):173–86. 10.1016/S2213-2600(18)30345-X. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Ferreyro BL, Munshi L. Causes of acute respiratory failure in the immunocompromised host. Curr Opin Crit Care. 2019;25(1):21–8. 10.1097/MCC.0000000000000569. [DOI] [PubMed] [Google Scholar]
- 4.Liu K, Ma XY, Xiao H, et al. Association between the ROX index and mortality in patients with acute hypoxemic respiratory failure: a retrospective cohort study. Respir Res. 2024;25(1):143. 10.1186/s12931-024-02771-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Stefan MS, Shieh MS, Pekow PS, et al. Epidemiology and outcomes of acute respiratory failure in the United States, 2001 to 2009: a national survey. J Hosp Med. 2013;8(2):76–82. 10.1002/jhm.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Huang S, Zhou Q, Guo N, et al. Association between red blood cell distribution width and in-hospital mortality in acute myocardial infarction. Medicine. 2021;100(15):e25404. 10.1097/MD.0000000000025404. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Yang J, Liu C, Li L, et al. Red blood cell distribution width predicts pulmonary hypertension secondary to chronic obstructive pulmonary disease. Can Respir J. 2019;2019:3853454. 10.1155/2019/3853454. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Lorente L, Martín MM, Argueso M, et al. Association between red blood cell distribution width and mortality of COVID-19 patients. Anaesth Crit Care Pain Med. 2021;40(1):100777. 10.1016/j.accpm.2020.10.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Zhang FX, Li ZL, Zhang ZD, Ma XC. Prognostic value of red blood cell distribution width for severe acute pancreatitis. World J Gastroenterol. 2019;25(32):4739–48. 10.3748/wjg.v25.i32.4739. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Jia L, Cui S, Yang J, et al. Red blood cell distribution width predicts long-term mortality in critically ill patients with acute kidney injury: a retrospective database study. Sci Rep. 2020;10(1):4563. 10.1038/s41598-020-61516-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Zhang W, Wang Y, Wang J, Wang S, Chen B, Ye Y, Zhang J, Ying L, Chen Y. Association between red blood cell distribution width and long-term mortality in acute respiratory failure patients. Sci RepPLoS One. 20202013;10(1):21185. 10.1038/s41598-020-78321-2. [DOI] [PMC free article] [PubMed]
- 12.Chen B, Ye B, Zhang J, Ying L, Chen Y. RDW to platelet ratio: a novel noninvasive index for predicting hepatic fibrosis and cirrhosis in chronic hepatitis B. PLoS One. 2013;8(7):e68780. 10.1371/journal.pone.0068780. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Cetinkaya E, Senol K, Saylam B, Tez M. Red cell distribution width to platelet ratio: new and promising prognostic marker in acute pancreatitis. World J Gastroenterol. 2014;20(39):14450–4. 10.3748/wjg.v20.i39.14450. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Qin T, Liao L, Pinhu L. Correlation between red blood cell distribution width-to-platelet ratio, and mortality in patients with acute respiratory distress syndrome. A retrospective cohort study. Shock. 2022;58(6):498–506. 10.1097/SHK.0000000000002016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Chen S, Shi Y, Hu B, Huang J. A prediction model for in-hospital mortality of acute exacerbations of chronic obstructive pulmonary disease patients based on red cell distribution width-to-platelet ratio. Int J Chron Obstruct Pulmon Dis. 2023;18:2079–91. 10.2147/COPD.S418162. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Wang H, Xu H, Wang X, et al. Red blood cell distribution width to platelet ratio is related to histologic severity of primary biliary cirrhosis. Medicine. 2016;95(11):e3114. 10.1097/MD.0000000000003114. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Takeuchi H, Abe M, Takumi Y, et al. Elevated red cell distribution width to platelet count ratio predicts poor prognosis in patients with breast cancer. Sci Rep. 2019;9(1):3033. 10.1038/s41598-019-40024-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Celık T, Balta S, Demır M, et al. Predictive value of admission red cell distribution width-platelet ratio for no-reflow phenomenon in acute ST segment elevation myocardial infarction undergoing primary percutaneous coronary intervention. Cardiol J. 2016;23(1):84–92. 10.5603/CJ.a2015.0070. [DOI] [PubMed] [Google Scholar]
- 19.Ge S, Lin S, Zhang L, Zeng M. The association of red blood cell distribution width to platelet count ratio and 28-day mortality of patients with sepsis: a retrospective cohort study. Ther Clin Risk Manag. 2020;16:999–1006. 10.2147/TCRM.S268523. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Zhou Y, Zhong L, Chen W, et al. Enhanced red blood cell distribution width to platelet ratio is a predictor of mortality in patients with sepsis: a propensity score matching analysis based on the MIMIC-IV database. BMJ Open. 2022;12(9):e062245. 10.1136/bmjopen-2022-062245. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Wu J, Huang L, He H, et al. Red cell distribution width to platelet ratio is associated with increasing In-Hospital mortality in critically ill patients with acute kidney injury. Dis Markers. 2022;2022:4802702. 10.1155/2022/4802702. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Shi Y, Shi L, Chen F, et al. Association of red cell distribution width and its changes with the 30-day mortality in patients with acute respiratory failure: an analysis of MIMIC-IV database. PLoS ONE. 2023;18(11):e0293986. 10.1371/journal.pone.0293986. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Ku NS, Kim HW, Oh HJ, et al. Red blood cell distribution width is an independent predictor of mortality in patients with gram-negative bacteremia. Shock. 2012;38(2):123–7. 10.1097/SHK.0b013e31825e2a85. [DOI] [PubMed] [Google Scholar]
- 24.Ghaffari S. Oxidative stress in the regulation of normal and neoplastic hematopoiesis. Antioxid Redox Signal. 2008;10(11):1923–40. 10.1089/ars.2008.2142. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Song CS, Park DI, Yoon MY, et al. Association between red cell distribution width and disease activity in patients with inflammatory bowel disease. Dig Dis Sci. 2012;57(4):1033–8. 10.1007/s10620-011-1978-2. [DOI] [PubMed] [Google Scholar]
- 26.Şenol K, Saylam B, Kocaay F, Tez M. Red cell distribution width as a predictor of mortality in acute pancreatitis. Am J Emerg Med. 2013;31(4):687–9. 10.1016/j.ajem.2012.12.015. [DOI] [PubMed] [Google Scholar]
- 27.Lippi G, Targher G, Montagnana M, et al. Relation between red blood cell distribution width and inflammatory biomarkers in a large cohort of unselected outpatients. Arch Pathol Lab Med. 2009;133(4):628–32. 10.5858/133.4.628. [DOI] [PubMed] [Google Scholar]
- 28.Thomas MR, Storey RF. The role of platelets in inflammation. Thromb Haemost. 2015;114(3):449–58. 10.1160/TH14-12-1067. [DOI] [PubMed] [Google Scholar]
- 29.Xiao C, Qin Z, Xiao J, et al. Association between basal platelet count and all-cause mortality in critically ill patients with acute respiratory failure: a secondary analysis from the eICU collaborative research database. Am J Transl Res. 2022;14(3):1685–94. [PMC free article] [PubMed] [Google Scholar]
- 30.Zhou P, Guo QQ, Wang FX, Zhou L, et al. Nonlinear relationship between platelet count and 30-day in-hospital mortality in ICU acute respiratory failure patients: a multicenter retrospective cohort study. Eur J Med Res. 2024;29(1):312. 10.1186/s40001-024-01909-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Lopez-Delgado JC, Rovira A, Esteve F, et al. Thrombocytopenia as a mortality risk factor in acute respiratory failure in H1N1 influenza. Swiss Med Wkly. 2013;143:w13788. 10.4414/smw.2013.13788. [DOI] [PubMed] [Google Scholar]
- 32.Wei Y, Tejera P, Wang Z, et al. A missense genetic variant in LRRC16A/CARMIL1 improves acute respiratory distress syndrome survival by attenuating platelet count decline. Am J Respir Crit Care Med. 2017;195(10):1353–61. 10.1164/rccm.201605-0946OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Svensson L, Baumgarten M, Mörgelin M, et al. Platelet activation by Streptococcus pyogenes leads to entrapment in platelet aggregates, from which bacteria subsequently escape. Infect Immun. 2014;82:4307–14. 10.1128/IAI.02020-14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Xu XR, Zhang D, Oswald BE, et al. Platelets are versatile cells: new discoveries in hemostasis, thrombosis, immune responses, tumor metastasis and beyond. Crit Rev Clin Lab Sci. 2016;53:409–30. 10.1080/10408363.2016.1200008. [DOI] [PubMed] [Google Scholar]
- 35.Zhu X, Wang Y, Jiang Q, et al. All-trans retinoic acid protects mesenchymal stem cells from immune thrombocytopenia by regulating the complement-interleukin-1β loop. Haematologica. 2019;104:1661. 10.3324/haematol.2018.204446. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Koyama K, Katayama S, Muronoi T, et al. Time course of immature platelet count and its relation to thrombocytopenia and mortality in patients with sepsis. PLoS One. 2018;13:e0192064. 10.1371/journal.pone.0192064. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Wang Q, Zhu D. The prognostic value of systemic immune-inflammation index (SII) in patients after radical operation for carcinoma of stomach in gastric cancer. J Gastrointest Oncol. 2019;10(5):965–78. 10.21037/jgo.2019.05.03. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Ulu K. Prognostic value of shock index, modified shock index, and, age shock index in, patients with sepsis in the intensive care unit[J]. Journal of Health Sciences of Kocaeli University / Kocaeli Üniversitesi Sağlık Bilimleri Dergisi. 2024;10(3). 10.30934/kusbed.1518819.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data utilized in this study are publicly accessible through the MIMIC-IV database version 3.1 The data utilized in this study are publicly accessible through the MIMIC-IV database version 3.1 (https://physionet.org/content/mimiciv/3.1/).