Association between the glucose-to-platelet ratio and short-term and long-term all-cause mortality in critically ill patients with pneumonia: a retrospective cohort study

Introduction

Pneumonia remains a major reason for intensive care unit (ICU) admission worldwide and continues to carry high short- and long-term mortality, even with advances in antimicrobial therapy and supportive care (1,2). In practice, ICU pneumonia is anything but uniform: patients present with varying degrees of systemic inflammation, metabolic derangement, and organ dysfunction, which translates into wide variability in outcomes (3,4). Although a number of clinical scores and laboratory markers are available, timely and dependable risk stratification early after ICU admission is still difficult (5,6). This has driven ongoing interest in simple, routinely available biomarkers that can complement existing prognostic tools.

Disturbances in glucose homeostasis are common in critical illness and consistently associate with worse outcomes across ICU populations (7-9). Stress hyperglycemia, reflecting acute neuroendocrine activation and inflammatory responses, has been linked to increased mortality regardless of pre-existing diabetes (10,11). Beyond single measurements, recent large-cohort studies—including analyses from Medical Information Mart for Intensive Care IV (MIMIC-IV)—suggest that composite metabolic indices such as the stress hyperglycemia ratio (SHR) and the triglyceride-glucose (TyG) index may add prognostic value on top of glucose alone (12-14). Together, these data support the idea that metabolism-informed composite markers can help refine early risk assessment in critically ill patients.

Platelets also provide clinically relevant signals in severe infection. In addition to their hemostatic function, platelets interact with leukocytes and endothelial cells and participate in inflammatory and immune responses (15). Accordingly, platelet abnormalities have been repeatedly associated with adverse outcomes in pneumonia: both thrombocytopenia and thrombocytosis on admission have been linked to higher short-term mortality in community-acquired and severe pneumonia cohorts (16,17). More recently, platelet-derived composite indices (e.g., ratios incorporating platelet size and count) have shown promise in outcome prediction, suggesting that integrating platelet-related information may be more informative than platelet count (PLT) alone in selected settings (18,19).

Against this background, the glucose-to-platelet ratio (GPR), calculated from blood glucose and PLT measured in routine care, may capture two clinically salient domains—metabolic stress and inflammation-related platelet dysregulation—within a single, easily obtained metric. However, its prognostic value in ICU patients with pneumonia has not been well defined. Using a large cohort from the MIMIC-IV database, we therefore examined the association between early GPR (within the first 24 hours after ICU admission) and both 30-day and 360-day all-cause mortality, and evaluated whether the relationship appeared linear across the observed range. We also explored consistency across clinically relevant subgroups to assess the robustness of the association. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1-2769/rc).

Methods

Study population

This retrospective cohort study was conducted using data from the MIMIC-IV (version 3.1), a large, publicly available critical care database maintained by the Massachusetts Institute of Technology (MIT) Laboratory for Computational Physiology. The database is available at https://physionet.org/content/mimiciv/3.1/. The database contains detailed clinical information from patients admitted to the ICUs of Beth Israel Deaconess Medical Center. Access to the database was granted after completion of the required training, and data extraction was performed by an authorized investigator.

Adult patients (aged ≥18 years) admitted to the ICU with a diagnosis of pneumonia were identified using the International Classification of Diseases, Ninth and Tenth Revision, Clinical Modification (ICD-9-CM and ICD-10-CM) codes. Because pneumonia subtype classification (e.g., community-acquired or hospital-acquired pneumonia) is not consistently available in structured fields within the MIMIC-IV database, all eligible pneumonia cases were included regardless of specific subtype. For patients with multiple ICU admissions, only the first ICU admission was included. Patients were excluded if the ICU length of stay was less than 24 hours or if PLT or blood glucose measurements within the first 24 hours after ICU admission were unavailable.

After applying the inclusion and exclusion criteria, 9,817 critically ill patients with pneumonia were included in the final analysis. Patients were subsequently stratified into tertiles according to GPR values measured within the first 24 hours after ICU admission (Figure 1). This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Figure 1 Flowchart of patient selection and study inclusion. ICU, intensive care unit.

Data collection

Data were extracted from the MIMIC-IV database using Structured Query Language (SQL) with PostgreSQL (version 13.7) and Navicat Premium (version 16). Extracted variables were grouped into four categories: (I) demographic characteristics, including age, sex, race, and weight; (II) comorbidities, including congestive heart failure, atrial fibrillation, cerebrovascular disease, chronic pulmonary disease, liver disease, malignant cancer, diabetes, and hypertension; (III) laboratory parameters, including red blood cell (RBC) count, white blood cell (WBC) count, hemoglobin, PLT, serum sodium, serum potassium, serum calcium, blood glucose, and other biochemical indices; (IV) severity-of-illness measures at ICU admission, including the Logistic Organ Dysfunction System (LODS), the Sequential Organ Failure Assessment (SOFA) score, and the Charlson Comorbidity Index (CCI).

Follow-up commenced at ICU admission and continued until death or the end of the predefined follow-up period. The GPR was calculated as blood glucose (mg/dL) divided by PLT (×10⁹/L). All laboratory variables and severity scores were obtained from records within the first 24 hours after ICU admission.

To reduce potential bias related to missing data, variables with more than 20% missing values were excluded from the analysis. For variables with less than 20% missing data, missing values were imputed using multiple imputation with a random forest approach, implemented with the “mice” package in R software.

Clinical outcome

The primary outcome was all-cause mortality within 30 days after ICU admission, and the secondary outcome was all-cause mortality within 360 days after ICU admission. The 360-day all-cause mortality was included as a secondary outcome to assess the long-term prognostic value of GPR beyond the acute phase of critical illness. Mortality information was obtained from the MIMIC-IV death records, which include both in-hospital and post-discharge deaths.

Statistical analysis

Continuous variables are presented as mean ± standard deviation or median with interquartile range (IQR), as appropriate, and categorical variables are summarized as counts and percentages. Distributional characteristics of continuous variables were assessed, and comparisons between groups were performed using Student’s t-test or one-way analysis of variance for normally distributed data, and the Mann-Whitney U test or Kruskal-Wallis test for non-normally distributed data.

Kaplan-Meier survival curves were constructed to compare 30-day and 360-day all-cause mortality across groups defined by tertiles of the GPR, with differences evaluated using the log-rank test. Cox proportional hazards regression models were used to examine the association between GPR and all-cause mortality, and hazard ratios (HRs) with 95% confidence intervals (CIs) were reported.

Three Cox models were prespecified. Model 1 was unadjusted. Model 2 adjusted for age, sex, race, and weight. Model 3 further adjusted for clinically relevant variables reflecting comorbidity burden, hemodynamic status, organ support, and laboratory indices, including chronic pulmonary disease, norepinephrine use, mechanical ventilation, diastolic blood pressure, oxygen saturation, WBC count, serum creatinine, international normalized ratio, and urine output, in addition to the variables included in Model 2. Covariates were selected a priori based on clinical relevance and existing literature rather than solely on univariate associations.

Restricted cubic spline (RCS) models with four knots were used to explore potential nonlinear associations between GPR and 30-day and 360-day all-cause mortality. GPR was analyzed both as a continuous variable and as a categorical variable based on tertiles, with the lowest tertile serving as the reference group. Linear trends across tertiles were assessed by modeling the median value of each tertile as a continuous variable.

Subgroup analyses were performed to evaluate the consistency of the association between GPR and the primary outcome across prespecified strata, including age, sex, chronic pulmonary disease, diabetes, hypertension, and norepinephrine use. Potential interactions were assessed using likelihood ratio tests. All tests were two-sided, and a P value <0.05 was considered statistically significant. Statistical analyses were conducted using R software (version 4.0.2) and SPSS software (version 22.0; IBM Corp., Armonk, NY, USA).

Results

A total of 9,817 critically ill patients with pneumonia were included in the analysis. The median age was 68.14 years (IQR, 56.75–78.88 years), and 5,485 patients (56.0%) were male. The median GPR was 0.71 (IQR, 0.48–1.10). Overall, all-cause mortality was 20% at 30 days and 39% at 360 days after ICU admission (Table 1).

Baseline characteristics

Baseline characteristics of critically ill patients with pneumonia stratified by tertiles of GPR are summarized in Table 1. Patients were categorized according to GPR values measured within the first 24 hours after ICU admission (T1: 0.19–0.55; T2: 0.55–0.88; T3: 0.88–3.11). Median GPR values were 0.41 (IQR, 0.32–0.48) in T1, 0.71 (IQR, 0.62–0.81) in T2, and 1.38 (IQR, 1.10–1.98) in T3.

Patients in the highest GPR tertile were more frequently male compared with those in the lowest tertile. Race did not differ significantly across GPR tertiles. Higher GPR was also associated with a greater prevalence of diabetes, liver disease and atrial fibrillation. With respect to physiological and laboratory variables at ICU admission, patients with higher GPR exhibited lower diastolic blood pressure, higher blood glucose levels, lower PLTs, and higher serum creatinine levels. Clinical outcomes differed across GPR tertiles. Compared with patients in the lowest tertile, those in the highest tertile had longer hospital length of stay (median, 12.93 vs. 10.88 days; P<0.001) and ICU length of stay (median, 4.27 vs. 3.24 days; P<0.001). In addition, patients in the highest GPR tertile experienced higher rates of hospital mortality, ICU mortality, 30-day all-cause mortality, and 360-day all-cause mortality compared with those in the lowest tertile (23% vs. 14%, 16% vs. 10%, 25% vs. 18%, and 45% vs. 37%, respectively; all P<0.001).

Primary outcomes

Kaplan-Meier survival curves for 30-day and 360-day all-cause mortality stratified by tertiles of the GPR are shown in Figure 2. Survival probabilities differed significantly across GPR tertiles for both time horizons, with patients in the highest tertile consistently exhibiting the poorest survival (log-rank P<0.001 for both 30-day and 360-day mortality).

Figure 2 Kaplan-Meier survival curves for all-cause mortality stratified by tertiles of the GPR. Kaplan-Meier curves illustrate 30-day (A) and 360-day (B) all-cause mortality across GPR tertiles. Differences between groups were assessed using the log-rank test. GPR, glucose-to-platelet ratio.

Results of the Cox proportional hazards regression analyses are summarized in Table 2. Across prespecified models with increasing degrees of adjustment, higher GPR was consistently associated with an increased risk of all-cause mortality. When analyzed as a continuous variable, GPR remained significantly associated with both 30-day and 360-day mortality in unadjusted, partially adjusted, and fully adjusted models, indicating a robust and independent relationship.

Similar findings were observed when GPR was analyzed as a categorical variable. Compared with patients in the lowest tertile, those in the highest tertile had significantly higher risks of 30-day and 360-day all-cause mortality across all Cox models, with effect estimates attenuated but remaining statistically significant after full adjustment. In addition, a significant linear trend across increasing GPR tertiles was observed for both outcomes (P for trend <0.001), suggesting a stepwise increase in mortality risk with higher GPR levels (Table 2; Figure 3A,3B).

Figure 3 Restricted cubic spline analyses of the association between the GPR and all-cause mortality. Restricted cubic spline curves illustrate the association between baseline GPR and 30-day (A) and 360-day (B) all-cause mortality. Solid red lines represent adjusted hazard ratios (HRs), and shaded areas indicate 95% confidence intervals (CIs). The horizontal dashed line indicates an HR of 1.0. P values for overall association and nonlinearity are shown in each panel. CI, confidence interval; GPR, glucose-to-platelet ratio.

RCS analyses further supported these findings. A modest nonlinear association was observed between GPR and 30-day all-cause mortality (P for nonlinearity =0.047), whereas the association with 360-day all-cause mortality appeared approximately linear (P for nonlinearity =0.73).

Subgroup analysis

Subgroup analyses were performed to assess the robustness of the association between the GPR and all-cause mortality across predefined clinical strata, including age, sex, chronic pulmonary disease, diabetes, hypertension, and norepinephrine use.

As shown in Figure 4, higher GPR was consistently associated with an increased risk of 30-day all-cause mortality across all examined subgroups, with HRs of similar magnitude among strata defined by age (<65 vs. ≥65 years), sex, comorbid conditions, and norepinephrine administration. No statistically significant interactions were observed between GPR and any subgroup variable (all P for interaction >0.05), indicating that the association between GPR and short-term mortality was consistent across clinically relevant subpopulations.

Figure 4 Forest plots of HRs for 30-day all-cause mortality across predefined subgroups in critically ill patients with pneumonia. CI, confidence interval; HR, hazard ratio.

Similar patterns were observed for 360-day all-cause mortality (Figure 5). Higher GPR remained significantly associated with increased long-term mortality across all predefined subgroups, and effect estimates were broadly comparable among different strata, with no evidence of effect modification by age, sex, comorbidities, or norepinephrine use (all P for interaction >0.05).

Figure 5 Forest plots of HRs for 360-day all-cause mortality across predefined subgroups in critically ill patients with pneumonia. CI, confidence interval; HR, hazard ratio.

Discussion

In this large retrospective cohort study based on the MIMIC-IV database, we found that the GPR measured at ICU admission was independently associated with both short-term and long-term all-cause mortality in critically ill patients with pneumonia. The association persisted after adjustment for demographic characteristics, comorbidities, indicators of disease severity, and laboratory variables, and was supported by consistent findings across multiple analytic approaches. RCS analyses suggested an approximately linear association between GPR and mortality risk overall, with a mild non-linear pattern observed for 30-day mortality, whereas the association with 360-day mortality appeared largely linear. In addition, subgroup analyses showed no meaningful effect modification for either short-term or long-term mortality, indicating that the observed association was robust across clinically relevant patient subgroups.

In recent years, increasing attention has been directed toward composite biomarkers that integrate metabolic stress and inflammatory or immune dysregulation, with the aim of improving prognostic assessment in critically ill patients. This shift reflects growing recognition that single laboratory parameters often fail to capture the complex and dynamic pathophysiology of severe infection, particularly in heterogeneous ICU populations. Within this context, glucose-related composite indices have been among the most extensively studied. Stress hyperglycemia has long been regarded as a hallmark of critical illness, reflecting acute neuroendocrine activation, insulin resistance, and systemic inflammation, and its association with adverse outcomes has been consistently reported across diverse ICU cohorts (7-9). Building on this concept, several large observational studies, including those based on the MIMIC-IV database, have demonstrated that glucose-centered composite indices such as the SHR and the TyG index are independently associated with mortality in sepsis and other critical conditions (10-14). Collectively, these findings support the notion that composite metabolic metrics may provide prognostic information beyond isolated glucose measurements.

Parallel to advances in metabolic risk stratification, a substantial body of literature has established the prognostic relevance of platelet-related parameters in infectious diseases. Beyond their traditional role in hemostasis, platelets actively participate in inflammatory signaling, immune modulation, and endothelial interactions during infection (15). In pneumonia and sepsis, platelet activation and consumption are closely linked to microvascular dysfunction and organ injury, a relationship further conceptualized within the framework of inflammation-coagulation crosstalk and immunothrombosis (20,21). Clinically, both thrombocytopenia and thrombocytosis at presentation have been associated with increased short-term mortality in patients with community-acquired and severe pneumonia (16,17). Moreover, platelet-derived composite indices, including ratios incorporating platelet volume or count, have been proposed as markers of disease severity and predictors of adverse outcomes (18,19). Together, these studies indicate that platelet abnormalities are not merely epiphenomena of severe infection, but reflect key pathophysiological processes relevant to prognosis.

Against this background, a small number of investigations have begun to explore composite indices that combine glucose with immune cell counts, extending the immune-metabolic integration concept beyond purely metabolic markers. Ratios such as the glucose-to-lymphocyte ratio have been examined in critically ill patients with sepsis, with reported associations between higher values and increased mortality risk. Importantly, however, these studies were typically conducted in heterogeneous sepsis cohorts, focused on alternative cellular components, and were not designed to address pneumonia-specific populations. Furthermore, the cellular element incorporated in these indices differs fundamentally from platelets, which occupy a unique position at the intersection of inflammation, coagulation, and innate immunity. Consequently, although prior work supports the general feasibility of combining glucose with immune-related parameters, direct evidence regarding the prognostic value of the GPR in infectious diseases—particularly in ICU patients with pneumonia—has remained limited.

Within this evolving literature, the present study should be viewed not as an isolated departure, but as a structured extension of existing immune-metabolic research. By leveraging a large, well-characterized ICU pneumonia cohort from the MIMIC-IV database, we provide a comprehensive evaluation of GPR in relation to both short-term and long-term all-cause mortality. Several features distinguish our work from prior studies and underpin its contribution. First, the study population is narrowly defined and clinically coherent, focusing specifically on critically ill patients with pneumonia rather than on mixed infection or sepsis cohorts. This distinction is nontrivial, as pneumonia represents a leading cause of ICU admission and death worldwide, with clinical decision-making highly dependent on early and reliable risk stratification (1,2). Second, our analysis simultaneously addresses 30-day and 360-day mortality, aligning with emerging evidence that the prognostic impact of severe infection extends well beyond the acute hospitalization period (22,23). Third, the association between GPR and mortality was examined using complementary statistical approaches, including multivariable Cox regression, RCS analyses to assess dose-response patterns, and prespecified subgroup analyses to evaluate consistency across clinically relevant strata.

From a novelty standpoint, it is important to emphasize that the contribution of this study does not rest on the introduction of an entirely unprecedented concept, but rather on the systematic integration of two established prognostic dimensions—metabolic stress and platelet-related inflammatory dysregulation—within a clinically meaningful population. While prior studies have validated glucose-based composite indices in sepsis (10-14) and platelet-related markers in pneumonia (16-19), evidence directly linking GPR to outcomes in ICU pneumonia has been lacking. Our findings therefore bridge this gap by demonstrating that GPR is independently associated with both early and late mortality, even after extensive adjustment for demographic factors, comorbidities, disease severity, and laboratory variables. Importantly, the approximately linear relationship observed between GPR and mortality risk further suggests that GPR conveys prognostic information across a broad range of values, rather than identifying only patients at extreme risk.

This positioning also addresses potential concerns regarding incremental novelty. Even if one accepts that composite biomarkers per se are not new, the value of GPR lies in its biological plausibility, analytical robustness, and clinical accessibility within the specific context of ICU pneumonia. By combining routinely measured glucose and platelet values, GPR reflects the convergence of metabolic derangement and immune-coagulation imbalance—two central features of severe pulmonary infection. In this respect, our study extends, rather than merely replicates, prior work on immune-metabolic biomarkers, providing evidence that this integrative approach retains prognostic relevance when applied to a large, disease-specific ICU cohort. Accordingly, the present findings support GPR as a simple, reproducible, and potentially useful adjunct to existing severity assessment tools in critically ill patients with pneumonia.

Although causal inference cannot be established in this observational setting, the association between elevated GPR and adverse outcomes is biologically plausible. GPR integrates two routinely measured parameters—blood glucose and PLT—that reflect key aspects of metabolic stress and inflammatory dysregulation in critical illness. Disturbances in glucose homeostasis are common in ICU patients and arise from neuroendocrine activation, insulin resistance, and systemic inflammation. Stress hyperglycemia has been linked to immune dysfunction, endothelial injury, oxidative stress, and microcirculatory impairment, all of which may contribute to organ failure and mortality in severe infections such as pneumonia. Persistent immune dysregulation has been increasingly recognized as a key mechanism contributing to adverse outcomes in critically ill patients (24). Inflammation and coagulation are closely interconnected processes in critical illness, and coagulation activation may further amplify immune responses and organ dysfunction (20,21). Coagulation abnormalities, including sepsis-induced coagulopathy, represent a clinically relevant pathway linking inflammation to organ dysfunction in critical illness (25,26).

Beyond their role in hemostasis, platelets provide additional prognostic information in critically ill patients. Thrombocytopenia may reflect platelet consumption, endothelial dysfunction, and systemic inflammatory activation, all of which are associated with worse clinical outcomes. Therefore, the GPR may capture the combined effects of metabolic disturbance and coagulation-inflammatory dysregulation, providing a more integrated indicator of disease severity.

From a clinical standpoint, GPR has several practical advantages. It is derived from routinely available laboratory tests, can be calculated at ICU admission without additional cost, and does not require specialized assays. The approximately linear association between GPR and mortality suggests that prognostic information is conveyed across a broad range of values rather than being limited to extreme thresholds. Moreover, the consistency of the association across subgroups supports the potential applicability of GPR in heterogeneous ICU populations with pneumonia. While GPR is not intended to replace established severity scoring systems, it may serve as a complementary marker that incorporates metabolic and inflammatory dimensions not fully captured by traditional scores.

Several limitations of this study should be acknowledged. First, as a retrospective observational analysis based on an administrative database, our findings demonstrate association but cannot establish causality. Despite multivariable adjustment for a wide range of clinically relevant confounders, residual confounding due to unmeasured or imperfectly measured factors (e.g., detailed antimicrobial regimens, corticosteroid use, fluid management strategies, and inflammatory biomarkers like C-reactive protein or procalcitonin) may persist. Second, the GPR was calculated using the first recorded values within 24 hours of ICU admission. This single-time-point measurement does not capture the dynamic changes in glucose and PLTs during the ICU stay, which may also carry prognostic information. Third, the study population was derived from the MIMIC-IV database, which, despite being widely used and validated, originates from a single academic medical center in the United States. This may limit the generalizability of our findings to other healthcare settings, regions, or patient populations with different demographic and clinical characteristics. Fourth, certain granular clinical data were not available in the database, including the specific etiology of pneumonia (bacterial vs. viral vs. fungal), detailed severity scores specific to pneumonia (e.g., CURB-65, PSI), and long-term functional outcomes. In addition, detailed classification of pneumonia subtypes (e.g., community-acquired vs. hospital-acquired pneumonia) was not consistently available in the database, which may have introduced heterogeneity into the study population. These factors could influence mortality and are potential unmeasured confounders.

Conclusions

Elevated GPR measured early after ICU admission is independently associated with increased short- and long-term mortality in critically ill patients with pneumonia. GPR may be a useful and readily available marker for early risk stratification, although prospective validation is warranted.

Acknowledgments

The authors would like to express their sincere gratitude to their teachers, particularly Professor Linyan Zhao, for guidance and support throughout this study. They also thank their families and friends for their encouragement and assistance. The authors thank Yinghua Yan for her support and encouragement during the study.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1-2769/rc

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1-2769/prf

Funding: This study was supported by the “1+X” Clinical Technology Capacity Enhancement Program of the Second Hospital of Dalian Medical University (No. 2022LCJSGC21).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1-2769/coif). All authors report that this study was supported by the “1+X” Clinical Technology Capacity Enhancement Program of the Second Hospital of Dalian Medical University (No. 2022LCJSGC21). The authors have no other conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. Data were obtained from the Medical Information Mart for Intensive Care IV (MIMIC-IV) database, which has received ethical approval from the institutional review boards of the Massachusetts Institute of Technology and Beth Israel Deaconess Medical Center. Because the database contains de-identified and publicly available patient information, the requirement for informed consent was waived.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Torres A, Cilloniz C, Niederman MS, et al. Pneumonia. Nat Rev Dis Primers 2021;7:25. [Crossref] [PubMed]
2. Mongardon N, Max A, Bouglé A, et al. Epidemiology and outcome of severe pneumococcal pneumonia admitted to intensive care unit: a multicenter study. Crit Care 2012;16:R155. [Crossref] [PubMed]
3. Angus DC, van der Poll T. Severe sepsis and septic shock. N Engl J Med 2013;369:2063. [Crossref] [PubMed]
4. Vincent JL, Singer M. Critical care: advances and future perspectives. Lancet 2010;376:1354-61. [Crossref] [PubMed]
5. Chalmers JD, Singanayagam A, Akram AR, et al. Severity assessment tools for predicting mortality in hospitalised patients with community-acquired pneumonia. Systematic review and meta-analysis. Thorax 2010;65:878-83.
6. Wang X, Feng Z, Wang L, et al. Risk prediction models for mortality in patients with severe pneumonia: a systematic review and meta-analysis. Front Med (Lausanne) 2025;12:1564545. [Crossref] [PubMed]
7. Marik PE, Bellomo R. Stress hyperglycemia: an essential survival response! Crit Care 2013;17:305. [Crossref] [PubMed]
8. van den Berghe G, Wouters P, Weekers F, et al. Intensive insulin therapy in critically ill patients. N Engl J Med 2001;345:1359-67. [Crossref] [PubMed]
9. NICE-SUGAR Study Investigators. Intensive versus conventional glucose control in critically ill patients. N Engl J Med 2009;360:1283-97. [Crossref] [PubMed]
10. Bai H, Niu D, Li B, et al. Elevated stress hyperglycemia ratio associated with higher hospital mortality in patients with respiratory failure. Sci Rep 2025;15:27972. [Crossref] [PubMed]
11. Krinsley JS. Association between hyperglycemia and increased hospital mortality in a heterogeneous population of critically ill patients. Mayo Clin Proc 2003;78:1471-8. [Crossref] [PubMed]
12. Zheng R, Qian S, Shi Y, et al. Association between triglyceride-glucose index and in-hospital mortality in critically ill patients with sepsis: analysis of the MIMIC-IV database. Cardiovasc Diabetol 2023;22:307. [Crossref] [PubMed]
13. Liao J, Lu D, Wang M, et al. Positive association between stress hyperglycemia ratio and ICU mortality in patients with pulmonary embolism: A retrospective study. PLoS One 2025;20:e0320644. [Crossref] [PubMed]
14. Chen X, Yang Z, Shi R, et al. Stress hyperglycemia ratio association with all-cause mortality in critically ill patients with coronary heart disease: an analysis of the MIMIC-IV database. Sci Rep 2024;14:29110. [Crossref] [PubMed]
15. Semple JW, Italiano JE Jr, Freedman J. Platelets and the immune continuum. Nat Rev Immunol 2011;11:264-74. [Crossref] [PubMed]
16. Mirsaeidi M, Peyrani P, Aliberti S, et al. Thrombocytopenia and thrombocytosis at time of hospitalization predict mortality in patients with community-acquired pneumonia. Chest 2010;137:416-20. [Crossref] [PubMed]
17. Viasus D, Garcia-Vidal C, Simonetti A, et al. Prognostic value of serum albumin levels in hospitalized adults with community-acquired pneumonia. J Infect 2013;66:415-23. [Crossref] [PubMed]
18. Cho J, Lee S, Uh Y, et al. Usefulness of mean platelet volume to platelet count ratio for predicting the risk of mortality in community-acquired pneumonia. Arch Med Sci 2020;16:1327-35. [Crossref] [PubMed]
19. Cui XJ, Xie B, Zhu KW, et al. Prognostic value of the platelet, neutrophil, monocyte, basophil, and eosinophil to lymphocyte ratios in patients with severe community-acquired pneumonia (SCAP). Sci Rep 2024;14:30406. [Crossref] [PubMed]
20. Levi M, van der Poll T. Inflammation and coagulation. Crit Care Med 2010;38:S26-34. [Crossref] [PubMed]
21. Engelmann B, Massberg S. Thrombosis as an intravascular effector of innate immunity. Nat Rev Immunol 2013;13:34-45. [Crossref] [PubMed]
22. Prescott HC, Angus DC. Enhancing Recovery From Sepsis: A Review. JAMA 2018;319:62-75. [Crossref] [PubMed]
23. Yende S, D'Angelo G, Kellum JA, et al. Inflammatory markers at hospital discharge predict subsequent mortality after pneumonia and sepsis. Am J Respir Crit Care Med 2008;177:1242-7. [Crossref] [PubMed]
24. Delano MJ, Ward PA. Sepsis-induced immune dysfunction: can immune therapies reduce mortality? J Clin Invest 2016;126:23-31. [Crossref] [PubMed]
25. Gando S, Levi M, Toh CH. Disseminated intravascular coagulation. Nat Rev Dis Primers 2016;2:16037. [Crossref] [PubMed]
26. Iba T, Levy JH, Warkentin TE, et al. Diagnosis and management of sepsis-induced coagulopathy and disseminated intravascular coagulation. J Thromb Haemost 2019;17:1989-94. [Crossref] [PubMed]