Bedside clinical prediction tool for mortality in critically ill children

Abstract

Introduction

Mortality rates among critically ill pediatric patients remain a persistent challenge. It is imperative to identify patients at higher risk to effectively allocate appropriate resources. Our study aimed to develop a prediction score based on clinical parameters and hemogram to predict pediatric intensive care unit (PICU) mortality.

Methods

We conducted a retrospective study to develop a clinical prediction score using data from children aged 1 month to 18 years admitted for at least 24 hours to the PICU at Chiang Mai University between January 2018 and December 2022. PICU mortality was defined as death within 28 days of admission. The score was developed using multivariable logistic regression and assessed for calibration and discrimination.

Results

There were 29 deaths in 330 children (8.8%). Our model for predicting 28-day ICU mortality uses four key predictors: male gender, use of vasoactive drugs, red blood cell distribution width (RDW) ≥15.9%, and platelet distribution width (PDW), categorized as follows: <10% (0 points), 10–14.9% (2 points), and ≥15% (4 points). Scores range from 0 to 8, with a cutoff value of 5 to differentiate low-risk (<5) from high-risk (≥5) groups. The tool demonstrates excellent performance with an AuROC curve of 0.86 (95% CI: 0.80–0.91, p<0.001) showing excellent discrimination and calibration, 82.8% sensitivity, and 73.1% specificity, respectively.

Conclusions

The score, developed from clinical data and hemogram, demonstrated potential in predicting ICU mortality among critically ill children. However, further studies are necessary to externally validate the score before it can be confidentially implemented in clinical practices.

Introduction

Despite advances in medical technology, pediatric intensive care unit (PICU) mortality rates vary widely between developed and developing countries. The PICU mortality rates in developed countries typically range between 4.0% to 7.3%. However, these rates are significantly higher in developing countries, reaching as high as 13.0% to 25.0% [1,2]. This disparity reflects differences in healthcare infrastructure, resource availability, and socioeconomic status that impact patient outcomes.

To optimize the care of critically ill pediatric patients, it is crucial to prioritize “at risk” patients for appropriate resources, enabling timely intervention and optimal management. A reliable scoring system can be a vital tool for predicting PICU outcomes and guiding physicians in making decisions about resource allocation, particularly in settings with limited resources.

The current landscape of mortality prediction scoring systems in PICUs presents significant challenges. Widely used systems like Pediatric Risk of Mortality (PRISM) III/IV, Pediatric Index of Mortality (PIM)-2 and -3, and Pediatric Logistic Organ Dysfunction-2 (PELOD-2) [3,4] have limitations in their application [3]. This prediction scoring system relies on arterial blood gas parameters such as pH, partial pressure of arterial oxygen (PaO₂), partial pressure of arterial carbon dioxide (PaCO₂), and base excess for score calculation. However, obtaining arterial blood gas measurements can be challenging, particularly in young children and infants. Additionally, serum biomarkers such as serum ferritin, lactate, C-reactive protein (CRP), and microalbuminuria are commonly employed in assessing the necessity for admission and prognostication [5–8]. However, their utility is often restricted by cost and availability, making them feasible primarily in tertiary-care or university hospitals or in developed countries.

Complete blood count (CBC) is inexpensive and widely available laboratory test and measure characteristics of red blood cell (RBC), white blood cell (WBC) and platelets. Several studies showed that RBC and platelet parameters including red blood cell distribution width (RDW), mean platelet volume (MPV) and platelet distribution width (PDW) were associated with inflammatory diseases and could predict ICU mortality [9–15].

Therefore, the aim of our study was to develop a prediction score to determine ICU mortality in critically ill pediatric patients, using clinical and hemogram parameters. This approach is designed to enhance applicability, especially in primary-care hospitals, by facilitating prompt bedside assessment and decision-making.

Materials and methods

Study design and setting

This study was conducted retrospectively from critically pediatric patients who were admitted to two PICUs of the Faculty of Medicine, Chiang Mai University, Thailand, between January 2018 and December 2022. The study protocol was approved by the Research Ethics committee of the Faculty of Medicine, Chiang Mai University (approval no. 166/2566 on 10 May 2023) and was performed in accordance with the Declaration of Helsinki as a statement of ethical principles for medical research involving human subjects. Informed consent was waived due to minimal risk, and the data were analyzed anonymously. The data were accessed for research purpose during 1 June 2023 to 31 December 2023.

Eligibility criteria

Inclusion criteria.

All children aged 1 month to 18 years who admitted to our PICUs were included. These patients were required to have a CBC upon PICU admission, and their ICU stay had to be at least 24 hours. Subsequent readmissions were considered new episodes of admissions.

Exclusion criteria.

We excluded patients with the following conditions: children who had received a RBC transfusion within 21 days or a platelet transfusion within 7 days prior to the PICU admission; children diagnosed with beta-thalassemia major, beta-thalassemia/hemoglobin E (Hb E), and hemoglobin H (Hb H) disease, as these conditions can affect RDW; children diagnosed with congenital hemolytic anemia, such as red blood cell membrane defect, that impact RDW; and children diagnosed with congenital macrothrombocytopenia or myeloproliferative disorders, which can affect platelet count, MPV and PDW.

Data collection

We collected information on study participants, including patient characteristics such as age, sex, body weight, height, body mass index (BMI), reasons for ICU admission, pre-existing comorbidities, thalassemia trait, and iron deficiency anemia. ICU parameters were also recorded, including the use of mechanical ventilation (MV), duration of MV support, use of vasoactive drugs, use of continuous renal replacement therapy (CRRT), presence of multi-organ dysfunction (MOD), PIM-2 score, and ICU length of stay (ICU-LOS). Additionally, CBC results from within the first 24 hours of ICU admission were included in the analysis. The CBCs were assessed using the Sysmex XN-9000® analyzer (Sysmex Co., Kobe, Japan).

Candidate predictors

We identified 12 potential predictors for developing the ICU mortality prediction score based on prior knowledge, clinical expertise, and a comprehensive literature review. These predictors included age [16–18]; sex [18,19]; BMI [20]; comorbidities [21]; use of MV [22]; use of vasoactive drug [18,22,23]; presence of MOD [16,22]; RDW [9–12,24–28]; WBC [18,25]; platelet count [15,18,25,29]; MPV [15,30]; and PDW [15]. The prediction score for our cohort was calculated using data collected within the first 24 hours of ICU admission.

Study endpoint

The study endpoint was 28-day ICU mortality, defined as death occurring within 28 days of PICU admission.

Study size estimation

The study’s sample size was determined by examining factors predicting mortality in the PICU, based on a pilot study of 256 patients conducted at Chiang Mai University Hospital from June 2012 to June 2013. This pilot study found a mortality rate of 14.5%, with 219 survivors and 37 non-survivors. Sample size calculations were performed using two-sample proportion comparisons related to PICU mortality, specifically looking at thrombocytopenia (34.2% in survivors vs. 86.5% in non-survivors) and MV use (72.1% in survivors vs. 94.6% in non-survivors). With a significance level (alpha) of 0.05 (two-sided) and a desired power of 0.80, the estimated sample sizes needed were 64 based on thrombocytopenia and 218 based on MV support. To address potential data loss, we adjusted for 25% missing values, ensuring a minimum sample size of 290. We calculated the necessary sample size to develop a prediction model aimed at accurately estimating the population’s average outcome risk at a specific time point of interest. Based on an expected outcome incidence of 8%, the minimum required sample size for model development was determined to be 114, representing 10 events. To enhance precision, we plan to include all eligible patients within the study period [31].

Statistical analysis

Descriptive statistics were utilized to summarize the study population, including frequencies, percentages, and median and interquartile range (IQR), as appropriate. Univariable and multivariable logistic regression analyses, along with the calculation of the area under the receiver operating characteristic (AuROC) curve, were conducted using STATA Statistical software version 17 (StataCorp, College Station, Texas). Statistical significance was set at a p-value of less than 0.05.

Prediction score development

We tested for multicollinearity and included pre-selected predictors in the analytic model. Both univariable and multivariable logistic regression analyses were performed to explore the relationship between the independent predictors and 28-day ICU mortality. The initial model included all predictors, but backward elimination was applied to exclude those predictors with a p-value greater than 0.05. Given the 29 non-survivors in our cohort, a maximum of 4 variables was used to develop the prediction score. After model reduction, the point for each predictor was calculated by dividing each regression coefficient by the smallest regression coefficient, and rounding to the nearest integer. The final score was then obtained by summing these rounded points.

We divided the scores into two clinical risk categories for practical application: low-risk and high-risk. A score below 5 points indicates a lower risk of mortality, whereas a score of 5 or higher signifies a substantially increased risk. We also assessed and compared the diagnostic performance of the score and the PIM-2 score, examining sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and likelihood ratios through statistical analysis and graphical representation.

The predictive performance of the score was evaluated through calibration and discrimination, with calibration assessed using the Hosmer–Lemeshow goodness-of-fit test and a calibration plot, while discriminative ability was measured with the area under the AuROC. For internal validation, bootstrap resampling with 300 replicates was employed to estimate the model’s optimism.

Results

Study population

Four hundred and fifty-three pediatric patients were admitted to the ICU. Of these, 123 were excluded due to recent transfusions: 87 had received RBC transfusions within the preceding 21 days, 1 had received a platelet transfusion within the last 7 days, and 35 had received both. Consequently, 330 children (49.7% male) were included in the study. The median age of the participants was 4.33 (IQR: 2.91–8.67) years, and their median BMI was 14.60 (IQR: 12.99–16.56) kg/m². The most common reasons for ICU admission were post-operative care (40.3%), pneumonia with respiratory failure (32.4%), septic shock (17.6%), and heart failure (17.6%), respectively. Intubation was required for 248 children (75.2%), and 152 children (46.1%) required vasoactive drugs administration within the first 24 hours of ICU admission. The median length of stay in the PICU was significantly longer in the non-survival group compared to the survival group (8 days [IQR: 3–26 days] vs. 5 days [IQR: 3–11 days], p<0.001). The overall PICU mortality rate was found to be 8.8%. Detailed demographic characteristics of the study population are presented in Table 1.

[Figure omitted. See PDF.]

Continuous data are presented as mean ± SD. Otherwise (*) denotes a report as median and interquartile range (IQR). Other reasons for ICUs admission: severe asthma exacerbation, acute myocarditis, cardiogenic shock, arrhythmia, hypoxic spell, seizure, diabetes ketoacidosis, hypovolemic shock and CRRT or plasmapheresis. Other comorbidities: gastrointestinal disease, endocrine disease and nephrological disease.

The factors associated with mortality in critically ill children are demonstrated in Table 2 and 3. According to multivariable logistic regression analysis, the only factors significantly associated with mortality in critically ill children were male gender (mOR = 2.70; 95% CI 1.07–6.79, p= 0.034), use of vasoactive drugs (mOR = 3.69; 95% CI 1.40–9.72, p= 0.008), RDW≥15.9% (mOR = 4.08; 95% CI 1.43–11.61, p= 0.008), PDW 10–14.9% (mOR = 5.79; 95% CI 1.29–26.01, p= 0.022), and PDW ≥15% (mOR =34.72; 95% CI 6.28–191.98, p<0.001), respectively. Consequently, these parameters have led to the development of the predictive score for mortality in critically ill children, termed the “Early Screening Critically Ill Children (ESCIC) score”. The ESCIC score ranges from 0 to 8.

[Figure omitted. See PDF.]

[Figure omitted. See PDF.]

The discrimination capacity of the ESCIC score, as determined by AuROC curve, was 0.86 (95% CI: 0.80–0.91), while that of the PIM-2 was 0.64 (95% CI: 0.53–0.74). The ESCIC score significantly improved AuROC performance (p < 0.001), as illustrated in Fig 1A and 1B. Internal validation of the ESCIC score, conducted through a bootstrapping procedure with 300 replicates, revealed an apparent AuROC of 0.86 (range: 0.80–0.91) and a test AuROC of 0.85 (range: 0.78–0.91). The optimism of the AuROC was calculated to be 0.01.

[Figure omitted. See PDF.]

(A) Area under the receiver operating characteristics (AuROC) curve of Early Screening Critically Ill Children (ESCIC) score compared to PIM-2 score. (B) Model calibration plot of agreement between model predicted odds and observed proportion of children with mortality.

For practical application, the derived models are presented as the ESCIC, based on individual inputs. This application estimates the predicted probability of 28-day mortality in critically ill children. We categorized the scores into 2 clinical risk categories: low-risk and high-risk. An ESCIC score below 5 points indicates a low probability of mortality, while a score of 5 or higher signifies a substantially increased probability of mortality. The sensitivity and specificity levels of the ESCIC score were 82.8% (95% CI: 64.2–94.2) and 73.1% (95% CI: 67.7–78.0), respectively in Table 4.

[Figure omitted. See PDF.]

Discussion

In reality, mortality rates among critically ill children are influenced by various factors, such as case-mix, infrastructure, environmental hygiene, personnel expertise, facilities, and economic policies within the country. Given the advancements in pediatric care within the ICU, maintaining strict quality control is crucial to identify high-risk groups for mortality and optimize treatment adequacy, as well as to plan and efficiently utilize resources. Variations in mortality rates across PICUs often arise from differences in illness severity.

In this context, our study introduces an innovative prediction model termed the “ESCIC score,” designed for mortality screening in critically ill children. Our model integrates four readily available predictors—male gender, use of vasoactive drugs, RDW, and PDW—into a comprehensive risk assessment tool. The ESCIC score demonstrated robust discriminatory performance and excellent calibration. By incorporating these key variables, the ESCIC score provides a valuable tool for risk stratification and informed clinical decision-making within pediatric intensive care settings.

In our study, we observed a higher mortality rate among male patients in the PICUs. This finding aligns with existing literature that suggests gender differences in critical care outcomes [32–34]. However, the reasons behind increased mortality in male PICU patients are multifaceted. Biological factors, such as differences in immune responses between genders, may play a significant role. Additionally, behavioral and social factors could contribute to the observed disparities. It’s important to note that findings on gender differences in PICU mortality are not entirely consistent across studies. Given these complexities, further research is necessary to elucidate the underlying mechanisms contributing to gender disparities in PICU mortality. Understanding these factors is crucial for developing targeted interventions aimed at improving outcomes for all critically ill children.

In our study, we observed that three-quarters of non-surviving children were prescribed vasoactive drugs. Consistently, numerous studies have demonstrated that the use of vasoactive drugs strongly correlates with in-ICU death, indicating a higher ICU mortality rate among patients who received such medications [35,36]. The administration of vasoactive drugs in pediatric patients is associated with increased mortality [22,23]. This correlation may stem from the fact that children prescribed these medications often exhibit unstable hemodynamics, highlighting the severity of their condition. Vasoactive drugs play a crucial role in managing shock, with catecholamines being the most commonly utilized agents in the ICU [37,38].

Moreover, our study found the significant association between RDW and PICU mortality. Similarly, the clinical utility of RDW has also been affirmed in numerous studies involving critically ill patients [9–13,25–27]. RDW emerges as a straightforward and routinely reported metric, incurring no additional effort or costs, thereby presenting an advantageous prognostic factor. Elevated RDW levels may result from any pathological process disrupting RBC production, leading to the release of more immature RBCs into circulation. Furthermore, a high RDW has consistently been linked to increased inflammatory markers such as CRP levels, erythrocyte sedimentation rate (ESR), and interleukin-6 (IL-6) [39]. IL-6 and IL-1β have been shown to directly impair RBC survival in circulation, induce RBC membrane deformity, and suppress erythrocyte maturation [10,40,41]. Established research confirms that inflammatory cytokines disrupt RBC maturation within the bone marrow through various mechanisms, including inhibiting the production of or response to erythropoietin, causing impairments in iron metabolism and shortening RBC survival. Consequently, these inflammatory mediators can result in the release of newer and larger reticulocytes into peripheral circulation, thereby increasing RDW levels [39–44].

In addition, high PDW levels were observed within our cohort. Elevated PDW values, indicative of accelerated platelet turnover and reflecting a wide range of platelet sizes, demonstrated significant variation between sepsis survivors and non-survivors upon admission. Consistent with previous studies [15,45,46], these findings showed the association between high PDW and adverse outcomes. The increased PDW suggests platelet heterogeneity originating from bone marrow megakaryocytes, possibly due to platelet immaturity and swelling in circulation [45,47]. Moreover, a recent report has shown that IL-6, an acute-phase response inflammatory cytokine, regulates megakaryocytic maturation, platelet production, and platelet size [48]. These mechanisms may contribute to the observed elevation in PDW among critically ill patients.

Our study has several significant limitations. First, the retrospective design of the study limited the range of data that could be collected. Additionally, because the research was conducted at a single center, the findings may not be widely applicable. Second, with only 29 non-survivors in our cohort, we were constrained to use a maximum of 4 variables for developing the prediction score. Third, since neither thrombocytopenia nor mechanical ventilation parameters were associated with PICU mortality in our cohort, the initial sample size calculation may not have been adequate. To address this result, we recalculated our sample size of 330 to verify the power of statistical significance (0.8) using two-sample means comparisons of RDW between non-survivors (19.2 ± 3.9%) and survivors (16.6 ± 3.4%), with an overall PICU mortality rate of 8.8%. Our recalculation showed a power of 0.87, which is sufficient to support the reliability of our findings. While our sample size calculation ensures precise estimation of the average risk, it does not guarantee the prevention of overfitting or a small absolute difference between the adjusted and apparent R-squared values. Fourth, our study included patients with leukemia or solid tumors, as these patients often require intensive care due to septic shock or respiratory failure, reflecting their heightened vulnerability. We aimed to provide a comprehensive representation of the overall patient population and mortality patterns in our PICU. However, we acknowledge that the inclusion of these patients may introduce a potential bias, particularly concerning post-chemotherapy immunosuppression, which could impact clinical outcomes. Fifth, our study utilized 28-day PICU mortality as the primary outcome, which may differ from overall PICU mortality in certain aspects. To address this, we performed a multivariable analysis using a full model to identify factors associated with overall ICU mortality, as presented in Table 5, for further consideration. Lastly, we did not monitor the follow-up of clinical parameters, which would have enabled a more thorough assessment of the relationship between our model and patient outcomes. Despite these limitations, it remains crucial to identify at-risk patients in order to deliver high-quality care within resource constraints. Utilizing clinical parameters and inexpensive, widely available blood tests can aid in optimal management and resource allocation prioritization. Therefore, we advocate for the validation of our model in external datasets with a larger proportion of patients. Moreover, further prospective studies are warranted to validate our newly developed score and refine its clinical predictive accuracy.

[Figure omitted. See PDF.]

Conclusions

The ESCIC score, which integrates variables such as sex, vasoactive drug use, RDW, and PDW, demonstrates reliable and satisfactory predictive performance for mortality in critically ill children. This tool effectively identifies high-risk patients who may benefit from targeted interventions, thereby facilitating the delivery of high-quality care. However, external validation is necessary to confirm its generalizability.

Supporting information

S1 File. Minimaldataset_ESCIC.

https://doi.org/10.1371/journal.pone.0322050.s001

(XLS)

References

1. 1. Pollack MM, Ruttimann UE, Getson PR. Accurate prediction of the outcome of pediatric intensive care. A new quantitative method. N Engl J Med. 1987;316(3):134–9. pmid:3796684

* View Article

* PubMed/NCBI

* Google Scholar

2. 2. Punchak M, Hall K, Seni A, Buck WC, DeUgarte DA, Hartford E, et al. Epidemiology of Disease and Mortality From a PICU in Mozambique. Pediatr Crit Care Med. 2018;19(11):e603–10. pmid:30063654

* View Article

* PubMed/NCBI

* Google Scholar

3. 3. Shen Y, Jiang J. Meta-Analysis for the Prediction of Mortality Rates in a Pediatric Intensive Care Unit Using Different Scores: PRISM-III/IV, PIM-3, and PELOD-2. Front Pediatr. 2021;9:712276. pmid:34504815

* View Article

* PubMed/NCBI

* Google Scholar

4. 4. Toltzis P, Soto-Campos G, Shelton CR, Kuhn EM, Hahn R, Kanter RK, et al. Evidence-Based Pediatric Outcome Predictors to Guide the Allocation of Critical Care Resources in a Mass Casualty Event. Pediatr Crit Care Med. 2015;16(7):e207-16. pmid:26121100

* View Article

* PubMed/NCBI

* Google Scholar

5. 5. Tonial CT, Costa CAD, Andrades GRH, Crestani F, Bruno F, Piva JP, et al. Performance of prognostic markers in pediatric sepsis. J Pediatr (Rio J). 2021;97(3):287–94. pmid:32991837

* View Article

* PubMed/NCBI

* Google Scholar

6. 6. Aygun F. Procalcitonin Value Is an Early Prognostic Factor Related to Mortality in Admission to Pediatric Intensive Care Unit. Crit Care Res Pract. 2018;2018:9238947. pmid:30675399

* View Article

* PubMed/NCBI

* Google Scholar

7. 7. Nismath S, Rao SS, Baliga BS, Kulkarni V, Rao GM. Comparative validity of microalbuminuria versus clinical mortality scores to predict pediatric intensive care unit outcomes. Clin Exp Pediatr. 2020;63(1):20–4. pmid:31401824

* View Article

* PubMed/NCBI

* Google Scholar

8. 8. Siddiqui I, Jafri L, Abbas Q, Raheem A, Haque AU. Relationship of Serum Procalcitonin, C-reactive Protein, and Lactic Acid to Organ Failure and Outcome in Critically Ill Pediatric Population. Indian J Crit Care Med. 2018;22(2):91–5. pmid:29531448

* View Article

* PubMed/NCBI

* Google Scholar

9. 9. Sachdev A, Simalti A, Kumar A, Gupta N, Gupta D, Chugh P. Outcome Prediction Value of Red Cell Distribution Width in Critically-ill Children. Indian Pediatr. 2018;55(5):414–6. pmid:29845956

* View Article

* PubMed/NCBI

* Google Scholar

10. 10. Kim DH, Ha EJ, Park SJ, Jhang WK. Evaluation of the usefulness of red blood cell distribution width in critically ill pediatric patients. Medicine (Baltimore). 2020;99(36):e22075. pmid:32899077

* View Article

* PubMed/NCBI

* Google Scholar

11. 11. Cai N, Chen ZQ, Tao M, Fan WT, Liao W. Mean platelet volume and red blood cell distribution width is associated with prognosis in premature neonates with sepsis. Open Med (Wars). 2021;16(1):1175–81. pmid:34514164

* View Article

* PubMed/NCBI

* Google Scholar

12. 12. Dubey A, Kumar S, Acharya S, Wanjari AK, Bawankule S, Agrawal S, et al. Impact of Red Cell and Platelet Distribution Width in Patients of Medical Intensive Care Unit. J Lab Physicians. 2021;13(4):309–16. pmid:34975248

* View Article

* PubMed/NCBI

* Google Scholar

13. 13. Bulut O, Akcakaya A, Bulut N, Ovali F. Elevated Red Cell Distribution Width as a Useful Marker in Neonatal Sepsis. J Pediatr Hematol Oncol. 2021;43(5):180–5. pmid:33512870

* View Article

* PubMed/NCBI

* Google Scholar

14. 14. Liberski PS, Szewczyk M, Krzych ŁJ. Haemogram-Derived Indices for Screening and Prognostication in Critically Ill Septic Shock Patients: A Case-Control Study. Diagnostics (Basel). 2020;10(9):638. pmid:32867031

* View Article

* PubMed/NCBI

* Google Scholar

15. 15. Mangalesh S, Dudani S, Malik A. Platelet Indices and Their Kinetics Predict Mortality in Patients of Sepsis. Indian J Hematol Blood Transfus. 2021;37(4):600–8. pmid:33776267

* View Article

* PubMed/NCBI

* Google Scholar

16. 16. Albuali WH, Algamdi AA, Hasan EA, Al-Qahtani MH, Yousef AA, Al Ghamdi MA, et al. Use of a Mortality Prediction Model in Children on Mechanical Ventilation: A 5-Year Experience in a Tertiary University Hospital. J Multidiscip Healthc. 2020;13:1507–16. pmid:33204099

* View Article

* PubMed/NCBI

* Google Scholar

17. 17. Feinstein Y, Greenberg D, Ben-Shimol S, Mimran M, Dagan R, Givon-Lavi N. Characterization of children younger than 5 Years of age with severe community-acquired alveolar pneumonia (CAAP) requiring Pediatric Intensive Care Unit (PICU) admission. Pediatrics and neonatology. 2020;61(4):406-13. Epub 2020/05/11. pmid:32386941; PMCID: PMCPMC7194609

* View Article

* PubMed/NCBI

* Google Scholar

18. 18. Wang G, Liu J, Xu R, Fu Y, Liu X. Lactate/albumin ratio as a predictor of in-hospital mortality in critically ill children. BMC Pediatr. 2022;22(1):725. pmid:36539725

* View Article

* PubMed/NCBI

* Google Scholar

19. 19. Fagbamigbe AF, Morakinyo OM, Balogun FM. Sex inequality in under-five deaths and associated factors in low and middle-income countries: a Fairlie decomposition analysis. BMC Public Health. 2022;22(1):334. pmid:35172780

* View Article

* PubMed/NCBI

* Google Scholar

20. 20. Ventura JC, Hauschild DB, Barbosa E, Bresolin NL, Kawai K, Mehta NM, et al. Undernutrition at PICU Admission Is Predictor of 60-Day Mortality and PICU Length of Stay in Critically Ill Children. J Acad Nutr Diet. 2020;120(2):219–29. pmid:31522971

* View Article

* PubMed/NCBI

* Google Scholar

21. 21. Dendir G, Awoke N, Alemu A, Sintayhu A, Eanga S, Teshome M, et al. Factors Associated with the Outcome of a Pediatric Patients Admitted to Intensive Care Unit in Resource-Limited Setup: Cross-Sectional Study. Pediatric Health Med Ther. 2023;14:71–9. pmid:36890923

* View Article

* PubMed/NCBI

* Google Scholar

22. 22. Costa GA, Delgado AF, Ferraro A, Okay TS. Application of the pediatric risk of mortality (PRISM) score and determination of mortality risk factors in a tertiary pediatric intensive care unit. Clinics (Sao Paulo). 2010;65(11):1087–92. pmid:21243277

* View Article

* PubMed/NCBI

* Google Scholar

23. 23. Shah P, Petersen TL, Zhang L, Yan K, Thompson NE. Using Aggregate Vasoactive-Inotrope Scores to Predict Clinical Outcomes in Pediatric Sepsis. Front Pediatr. 2022;10:778378. pmid:35311061

* View Article

* PubMed/NCBI

* Google Scholar

24. 24. Li G, Jia P, Zhao J, Wu X, Duan Y, Liu D, et al. Usefulness of RBC distribution width and C-reactive protein to predict mortality in pediatric non-cardiac critical illness. Am J Emerg Med. 2019;37(12):2143–50. pmid:30772131

* View Article

* PubMed/NCBI

* Google Scholar

25. 25. Kim YC, Song JE, Kim EJ, Choi H, Jeong WY, Jung IY, et al. A Simple Scoring System Using the Red Blood Cell Distribution Width, Delta Neutrophil Index, and Platelet Count to Predict Mortality in Patients With Severe Sepsis and Septic Shock. J Intensive Care Med. 2019;34(2):133–9. pmid:30021478

* View Article

* PubMed/NCBI

* Google Scholar

26. 26. Ramby AL, Goodman DM, Wald EL, Weiss SL. Red Blood Cell Distribution Width as a Pragmatic Marker for Outcome in Pediatric Critical Illness. PLoS One. 2015;10(6):e0129258. pmid:26057629

* View Article

* PubMed/NCBI

* Google Scholar

27. 27. Wang H, Huang J, Liao W, Xu J, He Z, Liu Y, et al. Prognostic Value of the Red Cell Distribution Width in Patients with Sepsis-Induced Acute Respiratory Distress Syndrome: A Retrospective Cohort Study. Dis Markers. 2021;2021:5543822. pmid:34122669

* View Article

* PubMed/NCBI

* Google Scholar

28. 28. Loveday S, Sinclair L, Badrick T. Does the addition of RDW improve current ICU scoring systems?. Clin Biochem. 2015;48(9):569–74. pmid:25869493

* View Article

* PubMed/NCBI

* Google Scholar

29. 29. Agrawal S, Sachdev A, Gupta D, Chugh K. Platelet counts and outcome in the pediatric intensive care unit. Indian J Crit Care Med. 2008;12(3):102–8. pmid:19742257

* View Article

* PubMed/NCBI

* Google Scholar

30. 30. Zheng Y, Luo Z, Cao Z. Mean platelet volume as a predictive biomarker for in-hospital mortality in patients receiving invasive mechanical ventilation. BMC Pulm Med. 2022;22(1):353. pmid:36115956

* View Article

* PubMed/NCBI

* Google Scholar

31. 31. Riley RD, Ensor J, Snell KIE, Harrell FE Jr, Martin GP, Reitsma JB, et al. Calculating the sample size required for developing a clinical prediction model. BMJ. 2020;368:m441. pmid:32188600

* View Article

* PubMed/NCBI

* Google Scholar

32. 32. Tyagi P, Tullu MS, Agrawal M. Comparison of Pediatric Risk of Mortality III, Pediatric Index of Mortality 2, and Pediatric Index of Mortality 3 in Predicting Mortality in a Pediatric Intensive Care Unit. J Pediatr Intensive Care. 2018;7(4):201–6. pmid:31073495

* View Article

* PubMed/NCBI

* Google Scholar

33. 33. Gissler M, Järvelin MR, Louhiala P, Hemminki E. Boys have more health problems in childhood than girls: follow-up of the 1987 Finnish birth cohort. Acta Paediatr. 1999;88(3):310–4. pmid:10229043

* View Article

* PubMed/NCBI

* Google Scholar

34. 34. de-Miguel-Díez J, López-de-Andrés A, Hernández-Barrera V, de-Miguel-Yanes JM, Carabantes-Alarcón D, Ji Z, et al. Sex-differences in incidence of hospitalizations and in hospital mortality of community-acquired pneumonia among children in Spain: a population-based study. Eur J Pediatr. 2022;181(7):2705–13. pmid:35469030

* View Article

* PubMed/NCBI

* Google Scholar

35. 35. Motiejunaite J, Deniau B, Blet A, Gayat E, Mebazaa A. Inotropes and vasopressors are associated with increased short-term mortality but not long-term survival in critically ill patients. Anaesth Crit Care Pain Med. 2022;41(1):101012. pmid:34952218

* View Article

* PubMed/NCBI

* Google Scholar

36. 36. Pölkki A, Pekkarinen PT, Lahtinen P, Koponen T, Reinikainen M. Vasoactive Inotropic Score compared to the sequential organ failure assessment cardiovascular score in intensive care. Acta Anaesthesiol Scand. 2023;67(9):1219–28. pmid:37278095

* View Article

* PubMed/NCBI

* Google Scholar

37. 37. Annane D, Ouanes-Besbes L, de Backer D, DU B, Gordon AC, Hernández G, et al. A global perspective on vasoactive agents in shock. Intensive Care Med. 2018;44(6):833–46. pmid:29868972

* View Article

* PubMed/NCBI

* Google Scholar

38. 38. Ranjit S, Kissoon N, Argent A, Inwald D, Ventura AMC, Jaborinsky R, et al. Haemodynamic support for paediatric septic shock: a global perspective. Lancet Child Adolesc Health. 2023;7(8):588–98. pmid:37354910

* View Article

* PubMed/NCBI

* Google Scholar

39. 39. Jo YH, Kim K, Lee JH, Kang C, Kim T, Park H-M, et al. Red cell distribution width is a prognostic factor in severe sepsis and septic shock. Am J Emerg Med. 2013;31(3):545–8. pmid:23380094

* View Article

* PubMed/NCBI

* Google Scholar

40. 40. Förhécz Z, Gombos T, Borgulya G, Pozsonyi Z, Prohászka Z, Jánoskuti L. Red cell distribution width in heart failure: prediction of clinical events and relationship with markers of ineffective erythropoiesis, inflammation, renal function, and nutritional state. Am Heart J. 2009;158(4):659–66. pmid:19781428

* View Article

* PubMed/NCBI

* Google Scholar

41. 41. Vreugdenhil G, Löwenberg B, van Eijk HG, Swaak AJ. Anaemia of chronic disease in rheumatoid arthritis. Raised serum interleukin-6 (IL-6) levels and effects of IL-6 and anti-IL-6 on in vitro erythropoiesis. Rheumatol Int. 1990;10(3):127–30. pmid:2392639

* View Article

* PubMed/NCBI

* Google Scholar

42. 42. Sadaka F, O’Brien J, Prakash S. Red cell distribution width and outcome in patients with septic shock. J Intensive Care Med. 2013;28(5):307–13. pmid:22809690

* View Article

* PubMed/NCBI

* Google Scholar

43. 43. Perlstein TS, Weuve J, Pfeffer MA, Beckman JA. Red blood cell distribution width and mortality risk in a community-based prospective cohort. Arch Intern Med. 2009;169(6):588–94. pmid:19307522

* View Article

* PubMed/NCBI

* Google Scholar

44. 44. Fujita B, Strodthoff D, Fritzenwanger M, Pfeil A, Ferrari M, Goebel B, et al. Altered red blood cell distribution width in overweight adolescents and its association with markers of inflammation. Pediatr Obes. 2013;8(5):385–91. pmid:23239558

* View Article

* PubMed/NCBI

* Google Scholar

45. 45. Gao Y, Li Y, Yu X, Guo S, Ji X, Sun T, et al. The impact of various platelet indices as prognostic markers of septic shock. PLoS One. 2014;9(8):e103761. pmid:25118886

* View Article

* PubMed/NCBI

* Google Scholar

46. 46. Zhang S, Cui Y-L, Diao M-Y, Chen D-C, Lin Z-F. Use of Platelet Indices for Determining Illness Severity and Predicting Prognosis in Critically Ill Patients. Chin Med J (Engl). 2015;128(15):2012–8. pmid:26228211

* View Article

* PubMed/NCBI

* Google Scholar

47. 47. Wattel F, Durocher A, Crasquin O, Dewailly P, Romon M. Disorders of hemostasis during intake of lipids. Ann Anesthesiol Fr. 1977;18(11):916–20. pmid:24382

* View Article

* PubMed/NCBI

* Google Scholar

48. 48. Bester J, Pretorius E. Effects of IL-1β, IL-6 and IL-8 on erythrocytes, platelets and clot viscoelasticity. Sci Rep. 2016;6:32188. pmid:27561337

* View Article

* PubMed/NCBI

* Google Scholar

Citation: Sunkonkit K, Chai-adisaksopha C, Natesirinilkul R, Phinyo P, Trongtrakul K (2025) Bedside clinical prediction tool for mortality in critically ill children. PLoS ONE 20(4): e0322050. https://doi.org/10.1371/journal.pone.0322050

About the Authors:

Kanokkarn Sunkonkit

Roles: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Validation, Visualization, Writing – original draft, Writing – review & editing

Affiliations: Department of Pediatrics, Division of Pulmonary and Sleep Medicine, Faculty of Medicine, Chiang Mai University, Chiang Mai, Thailand, Center for Clinical Epidemiology and Clinical Statistics, Faculty of Medicine, Chiang Mai University, Chiang Mai, Thailand, Department of Biomedical Informatics and Clinical Epidemiology (BioCE), Faculty of Medicine, Chiang Mai University, Chiang Mai, Thailand

ORICD: https://orcid.org/0000-0003-2725-4636

Chatree Chai-adisaksopha

Roles: Conceptualization, Supervision, Validation, Visualization, Writing – review & editing

Affiliations: Center for Clinical Epidemiology and Clinical Statistics, Faculty of Medicine, Chiang Mai University, Chiang Mai, Thailand, Department of Biomedical Informatics and Clinical Epidemiology (BioCE), Faculty of Medicine, Chiang Mai University, Chiang Mai, Thailand, Department of Internal Medicine, Division of Hematology, Faculty of Medicine, Chiang Mai, Thailand

ORICD: https://orcid.org/0000-0001-7209-9722

Rungrote Natesirinilkul

Roles: Conceptualization, Supervision, Validation, Visualization, Writing – review & editing

Affiliation: Department of Pediatrics, Division of Hematology and Oncology, Faculty of Medicine, Chiang Mai University, Chiang Mai, Thailand

Phichayut Phinyo

Roles: Conceptualization, Formal analysis, Methodology, Software, Supervision, Validation, Visualization, Writing – review & editing

Affiliations: Center for Clinical Epidemiology and Clinical Statistics, Faculty of Medicine, Chiang Mai University, Chiang Mai, Thailand, Department of Biomedical Informatics and Clinical Epidemiology (BioCE), Faculty of Medicine, Chiang Mai University, Chiang Mai, Thailand

ORICD: https://orcid.org/0000-0002-8543-6254

Konlawij Trongtrakul

Roles: Conceptualization, Formal analysis, Methodology, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing

E-mail: [email protected]

Affiliations: Center for Clinical Epidemiology and Clinical Statistics, Faculty of Medicine, Chiang Mai University, Chiang Mai, Thailand, Department of Biomedical Informatics and Clinical Epidemiology (BioCE), Faculty of Medicine, Chiang Mai University, Chiang Mai, Thailand, Department of Internal Medicine, Division of Pulmonary, Critical Care, and Allergy, Faculty of Medicine, Chiang Mai University, Chiang Mai, Thailand

ORICD: https://orcid.org/0000-0002-2544-7322

References

1. Pollack MM, Ruttimann UE, Getson PR. Accurate prediction of the outcome of pediatric intensive care. A new quantitative method. N Engl J Med. 1987;316(3):134–9. pmid:3796684

2. Punchak M, Hall K, Seni A, Buck WC, DeUgarte DA, Hartford E, et al. Epidemiology of Disease and Mortality From a PICU in Mozambique. Pediatr Crit Care Med. 2018;19(11):e603–10. pmid:30063654

3. Shen Y, Jiang J. Meta-Analysis for the Prediction of Mortality Rates in a Pediatric Intensive Care Unit Using Different Scores: PRISM-III/IV, PIM-3, and PELOD-2. Front Pediatr. 2021;9:712276. pmid:34504815

4. Toltzis P, Soto-Campos G, Shelton CR, Kuhn EM, Hahn R, Kanter RK, et al. Evidence-Based Pediatric Outcome Predictors to Guide the Allocation of Critical Care Resources in a Mass Casualty Event. Pediatr Crit Care Med. 2015;16(7):e207-16. pmid:26121100

5. Tonial CT, Costa CAD, Andrades GRH, Crestani F, Bruno F, Piva JP, et al. Performance of prognostic markers in pediatric sepsis. J Pediatr (Rio J). 2021;97(3):287–94. pmid:32991837

6. Aygun F. Procalcitonin Value Is an Early Prognostic Factor Related to Mortality in Admission to Pediatric Intensive Care Unit. Crit Care Res Pract. 2018;2018:9238947. pmid:30675399

7. Nismath S, Rao SS, Baliga BS, Kulkarni V, Rao GM. Comparative validity of microalbuminuria versus clinical mortality scores to predict pediatric intensive care unit outcomes. Clin Exp Pediatr. 2020;63(1):20–4. pmid:31401824

8. Siddiqui I, Jafri L, Abbas Q, Raheem A, Haque AU. Relationship of Serum Procalcitonin, C-reactive Protein, and Lactic Acid to Organ Failure and Outcome in Critically Ill Pediatric Population. Indian J Crit Care Med. 2018;22(2):91–5. pmid:29531448

9. Sachdev A, Simalti A, Kumar A, Gupta N, Gupta D, Chugh P. Outcome Prediction Value of Red Cell Distribution Width in Critically-ill Children. Indian Pediatr. 2018;55(5):414–6. pmid:29845956

10. Kim DH, Ha EJ, Park SJ, Jhang WK. Evaluation of the usefulness of red blood cell distribution width in critically ill pediatric patients. Medicine (Baltimore). 2020;99(36):e22075. pmid:32899077

11. Cai N, Chen ZQ, Tao M, Fan WT, Liao W. Mean platelet volume and red blood cell distribution width is associated with prognosis in premature neonates with sepsis. Open Med (Wars). 2021;16(1):1175–81. pmid:34514164

12. Dubey A, Kumar S, Acharya S, Wanjari AK, Bawankule S, Agrawal S, et al. Impact of Red Cell and Platelet Distribution Width in Patients of Medical Intensive Care Unit. J Lab Physicians. 2021;13(4):309–16. pmid:34975248

13. Bulut O, Akcakaya A, Bulut N, Ovali F. Elevated Red Cell Distribution Width as a Useful Marker in Neonatal Sepsis. J Pediatr Hematol Oncol. 2021;43(5):180–5. pmid:33512870

14. Liberski PS, Szewczyk M, Krzych ŁJ. Haemogram-Derived Indices for Screening and Prognostication in Critically Ill Septic Shock Patients: A Case-Control Study. Diagnostics (Basel). 2020;10(9):638. pmid:32867031

15. Mangalesh S, Dudani S, Malik A. Platelet Indices and Their Kinetics Predict Mortality in Patients of Sepsis. Indian J Hematol Blood Transfus. 2021;37(4):600–8. pmid:33776267

16. Albuali WH, Algamdi AA, Hasan EA, Al-Qahtani MH, Yousef AA, Al Ghamdi MA, et al. Use of a Mortality Prediction Model in Children on Mechanical Ventilation: A 5-Year Experience in a Tertiary University Hospital. J Multidiscip Healthc. 2020;13:1507–16. pmid:33204099

17. Feinstein Y, Greenberg D, Ben-Shimol S, Mimran M, Dagan R, Givon-Lavi N. Characterization of children younger than 5 Years of age with severe community-acquired alveolar pneumonia (CAAP) requiring Pediatric Intensive Care Unit (PICU) admission. Pediatrics and neonatology. 2020;61(4):406-13. Epub 2020/05/11. pmid:32386941; PMCID: PMCPMC7194609

18. Wang G, Liu J, Xu R, Fu Y, Liu X. Lactate/albumin ratio as a predictor of in-hospital mortality in critically ill children. BMC Pediatr. 2022;22(1):725. pmid:36539725

19. Fagbamigbe AF, Morakinyo OM, Balogun FM. Sex inequality in under-five deaths and associated factors in low and middle-income countries: a Fairlie decomposition analysis. BMC Public Health. 2022;22(1):334. pmid:35172780

20. Ventura JC, Hauschild DB, Barbosa E, Bresolin NL, Kawai K, Mehta NM, et al. Undernutrition at PICU Admission Is Predictor of 60-Day Mortality and PICU Length of Stay in Critically Ill Children. J Acad Nutr Diet. 2020;120(2):219–29. pmid:31522971

21. Dendir G, Awoke N, Alemu A, Sintayhu A, Eanga S, Teshome M, et al. Factors Associated with the Outcome of a Pediatric Patients Admitted to Intensive Care Unit in Resource-Limited Setup: Cross-Sectional Study. Pediatric Health Med Ther. 2023;14:71–9. pmid:36890923

22. Costa GA, Delgado AF, Ferraro A, Okay TS. Application of the pediatric risk of mortality (PRISM) score and determination of mortality risk factors in a tertiary pediatric intensive care unit. Clinics (Sao Paulo). 2010;65(11):1087–92. pmid:21243277

23. Shah P, Petersen TL, Zhang L, Yan K, Thompson NE. Using Aggregate Vasoactive-Inotrope Scores to Predict Clinical Outcomes in Pediatric Sepsis. Front Pediatr. 2022;10:778378. pmid:35311061

24. Li G, Jia P, Zhao J, Wu X, Duan Y, Liu D, et al. Usefulness of RBC distribution width and C-reactive protein to predict mortality in pediatric non-cardiac critical illness. Am J Emerg Med. 2019;37(12):2143–50. pmid:30772131

25. Kim YC, Song JE, Kim EJ, Choi H, Jeong WY, Jung IY, et al. A Simple Scoring System Using the Red Blood Cell Distribution Width, Delta Neutrophil Index, and Platelet Count to Predict Mortality in Patients With Severe Sepsis and Septic Shock. J Intensive Care Med. 2019;34(2):133–9. pmid:30021478

26. Ramby AL, Goodman DM, Wald EL, Weiss SL. Red Blood Cell Distribution Width as a Pragmatic Marker for Outcome in Pediatric Critical Illness. PLoS One. 2015;10(6):e0129258. pmid:26057629

27. Wang H, Huang J, Liao W, Xu J, He Z, Liu Y, et al. Prognostic Value of the Red Cell Distribution Width in Patients with Sepsis-Induced Acute Respiratory Distress Syndrome: A Retrospective Cohort Study. Dis Markers. 2021;2021:5543822. pmid:34122669

28. Loveday S, Sinclair L, Badrick T. Does the addition of RDW improve current ICU scoring systems?. Clin Biochem. 2015;48(9):569–74. pmid:25869493

29. Agrawal S, Sachdev A, Gupta D, Chugh K. Platelet counts and outcome in the pediatric intensive care unit. Indian J Crit Care Med. 2008;12(3):102–8. pmid:19742257

30. Zheng Y, Luo Z, Cao Z. Mean platelet volume as a predictive biomarker for in-hospital mortality in patients receiving invasive mechanical ventilation. BMC Pulm Med. 2022;22(1):353. pmid:36115956

31. Riley RD, Ensor J, Snell KIE, Harrell FE Jr, Martin GP, Reitsma JB, et al. Calculating the sample size required for developing a clinical prediction model. BMJ. 2020;368:m441. pmid:32188600

32. Tyagi P, Tullu MS, Agrawal M. Comparison of Pediatric Risk of Mortality III, Pediatric Index of Mortality 2, and Pediatric Index of Mortality 3 in Predicting Mortality in a Pediatric Intensive Care Unit. J Pediatr Intensive Care. 2018;7(4):201–6. pmid:31073495

33. Gissler M, Järvelin MR, Louhiala P, Hemminki E. Boys have more health problems in childhood than girls: follow-up of the 1987 Finnish birth cohort. Acta Paediatr. 1999;88(3):310–4. pmid:10229043

34. de-Miguel-Díez J, López-de-Andrés A, Hernández-Barrera V, de-Miguel-Yanes JM, Carabantes-Alarcón D, Ji Z, et al. Sex-differences in incidence of hospitalizations and in hospital mortality of community-acquired pneumonia among children in Spain: a population-based study. Eur J Pediatr. 2022;181(7):2705–13. pmid:35469030

35. Motiejunaite J, Deniau B, Blet A, Gayat E, Mebazaa A. Inotropes and vasopressors are associated with increased short-term mortality but not long-term survival in critically ill patients. Anaesth Crit Care Pain Med. 2022;41(1):101012. pmid:34952218

36. Pölkki A, Pekkarinen PT, Lahtinen P, Koponen T, Reinikainen M. Vasoactive Inotropic Score compared to the sequential organ failure assessment cardiovascular score in intensive care. Acta Anaesthesiol Scand. 2023;67(9):1219–28. pmid:37278095

37. Annane D, Ouanes-Besbes L, de Backer D, DU B, Gordon AC, Hernández G, et al. A global perspective on vasoactive agents in shock. Intensive Care Med. 2018;44(6):833–46. pmid:29868972

38. Ranjit S, Kissoon N, Argent A, Inwald D, Ventura AMC, Jaborinsky R, et al. Haemodynamic support for paediatric septic shock: a global perspective. Lancet Child Adolesc Health. 2023;7(8):588–98. pmid:37354910

39. Jo YH, Kim K, Lee JH, Kang C, Kim T, Park H-M, et al. Red cell distribution width is a prognostic factor in severe sepsis and septic shock. Am J Emerg Med. 2013;31(3):545–8. pmid:23380094

40. Förhécz Z, Gombos T, Borgulya G, Pozsonyi Z, Prohászka Z, Jánoskuti L. Red cell distribution width in heart failure: prediction of clinical events and relationship with markers of ineffective erythropoiesis, inflammation, renal function, and nutritional state. Am Heart J. 2009;158(4):659–66. pmid:19781428

41. Vreugdenhil G, Löwenberg B, van Eijk HG, Swaak AJ. Anaemia of chronic disease in rheumatoid arthritis. Raised serum interleukin-6 (IL-6) levels and effects of IL-6 and anti-IL-6 on in vitro erythropoiesis. Rheumatol Int. 1990;10(3):127–30. pmid:2392639

42. Sadaka F, O’Brien J, Prakash S. Red cell distribution width and outcome in patients with septic shock. J Intensive Care Med. 2013;28(5):307–13. pmid:22809690

43. Perlstein TS, Weuve J, Pfeffer MA, Beckman JA. Red blood cell distribution width and mortality risk in a community-based prospective cohort. Arch Intern Med. 2009;169(6):588–94. pmid:19307522

44. Fujita B, Strodthoff D, Fritzenwanger M, Pfeil A, Ferrari M, Goebel B, et al. Altered red blood cell distribution width in overweight adolescents and its association with markers of inflammation. Pediatr Obes. 2013;8(5):385–91. pmid:23239558

45. Gao Y, Li Y, Yu X, Guo S, Ji X, Sun T, et al. The impact of various platelet indices as prognostic markers of septic shock. PLoS One. 2014;9(8):e103761. pmid:25118886

46. Zhang S, Cui Y-L, Diao M-Y, Chen D-C, Lin Z-F. Use of Platelet Indices for Determining Illness Severity and Predicting Prognosis in Critically Ill Patients. Chin Med J (Engl). 2015;128(15):2012–8. pmid:26228211

47. Wattel F, Durocher A, Crasquin O, Dewailly P, Romon M. Disorders of hemostasis during intake of lipids. Ann Anesthesiol Fr. 1977;18(11):916–20. pmid:24382

48. Bester J, Pretorius E. Effects of IL-1β, IL-6 and IL-8 on erythrocytes, platelets and clot viscoelasticity. Sci Rep. 2016;6:32188. pmid:27561337