European Journal of Cardiovascular Medicine

Recognizing patients at risk of deterioration in a timely manner is essential for providing appropriate care and effectively allocating resources. In Western countries, several illness scoring systems, such as the Pediatric Logistic Organ Dysfunction (PELOD) score, have been widely used for decades to assess pediatric patient's conditions. Other commonly used scoring systems include the Pediatric Index of

Mortality (PIM) and the Pediatric Risk of Mortality (PRISM) scores. However, the major limitation of these scoring systems is their reliance on multiple physiological and biochemical parameters, making them both time-consuming and dependent on well-equipped laboratory facilities. (Tucker et al.¹, 2009; Akre et al.², 2010). To address these challenges, early clinical warning scores have been developed based on fundamental bedside clinical parameters. These scores assign points to key observations and provide an overall score to objectively assess a patient's condition, enabling timely intervention. Pediatric-specific scoring systems, such as the Children’s Early Warning Score and the Pediatric Early Warning Score (PEWS), were introduced using this approach. Despite numerous studies, no standardized PEWS scoring system exists across pediatric inpatient settings internationally. There is limited standardization in outcome comparisons, uncertainty in PEWS education and implementation, and a lack of consensus on the most effective system or features for pediatric patients. (Veronica Lambert., et al.³ (2017). Hence, the parameters which can be quickly assessed bedside in an emergency are used under the scope of this study. One such scoring system used for early detection of sepsis and patient deterioration in adults is the quick Sequential Organ Failure Assessment (qSOFA) score.

qSOFA is a simplified version of the full SOFA score and relies on three clinical parameters: altered mental status, systolic blood pressure ≤90 mmHg, and respiratory rate ≥22 breaths per minute. A qSOFA score of ≥2 suggests a higher risk of poor outcomes, including mortality. Compared to other scoring methods, qSOFA is advantageous due to its simplicity, reliance on bedside assessment, and lack of requirement for laboratory tests. However, its application in pediatric patients remains a topic of investigation, as it was initially designed for adult populations. To improve the applicability of qSOFA scoring system in all age paediatric patients, an age adapted qSOFA scoring system is also designed incorporating the effect of age on normal range of these biochemical markers.

While qSOFA's or age-adapted qSOFA simplicity allows for rapid bedside assessment, PEWS provides a more comprehensive evaluation by incorporating additional vital signs, including heart rate, CRT (capillary refill time) and oxygen saturation. This makes PEWS particularly useful in pediatric settings where early detection of multiple potential complications is crucial.

Despite the effectiveness of early warning scores like PEWS, qSOFA or age-adapted qSOFA limitations exist. These scores may not capture all nuances of a patient’s clinical trajectory and may sometimes yield false positives or negatives. Additionally, inter-observer variability can impact the accuracy of scoring, leading to inconsistent results.

Various scoring systems, such as the Pediatric Early Warning Score (PEWS), Quick Sequential Organ Failure Assessment (qSOFA), and others, have been developed to predict the severity of illness and the likelihood of adverse outcomes in pediatric patients. This research article aims to compare the effectiveness of these scoring methods in predicting the condition and outcomes of pediatric patients based on quick assessment of vital signs. This study incorporates findings from recent relevant studies, including meta-analyses, qualitative research, and observational studies, to provide a comprehensive analysis of the topic.

Age-Adapted Quick Sequential Organ Failure Assessment (qSOFA)

Overview

The Age-Adapted qSOFA is a modification of the original qSOFA score tailored for pediatric patients. It adjusts thresholds for respiratory rate, systolic blood pressure, and mental status based on age to improve the early identification of sepsis in children. This adaptation enhances the sensitivity and predictive accuracy for adverse outcomes.

Technique

A score of 1 is assigned for any of the following abnormalities based on age (Table 1):

Table 1

Limitations

While improving age-specific predictive capacity, age-adapted qSOFA lacks laboratory data integration, limiting diagnostic precision. Variables such as altered mental status and respiratory rate can lead to false positives. It also omits clinical factors like oxygen support needs, further constraining its utility. The lack of universal validation restricts its application across diverse settings. Weiss et al.⁴ (2017) found qSOFA less sensitive than PEWS in pediatric contexts. Jia et al.⁵ (2023) proposed adding glucose and lactate to improve qSOFA’s predictive accuracy, but broader validation is still needed.

Pediatric Early Warning Score (PEWS)

Overview

PEWS is a structured, easy-to-use tool developed to identify early signs of deterioration in pediatric patients. It evaluates physiological parameters such as heart rate, respiratory rate, CRT, systolic blood pressure, oxygen saturation, and oxygen requirements. Its simplicity allows use in various clinical settings and aids in standardized assessments.

Technique

PEWS scores are based on seven parameters: heart rate, systolic BP, respiratory rate, respiratory effort, oxygen therapy, oxygen saturation, and capillary refill time (CRT). Scores range from 0 to 26, with higher scores indicating greater severity.

Effectiveness

Kowalski et al.⁶ (2023) found PEWS effective in identifying patients needing ICU transfer. Parwaiz et al.⁷ (2023) linked higher PEWS scores in emergency rooms with increased need for escalated care within 48 hours.

Barriers & Facilitators

Reuland et al.⁸ (2023) identified challenges in low-resource settings including inadequate training, limited resources, and change resistance. Facilitators include leadership support and engaged staff.

Limitations

PEWS is not a standalone diagnostic tool; clinical judgment is essential. Variations across institutions reduce consistency.

Some parameters are subjective (e.g., respiratory effort), requiring experience to assess accurately.

Van Nassau et al.9 (2015) and Agulnik et al.10 (2017) validated PEWS for improving patient outcomes, including in oncology settings.

Akre et al.2 (2010) and Breslin et al.11 (2014) confirmed PEWS’ effectiveness in predicting ICU need. Chapman et al.12 (2010) highlighted scoring variability while Skaletzky et al.13 (2012) emphasized training needs for reliable use.

Table 2

Other Scoring Systems

Pediatric Risk of Mortality (PRISM)

PRISM uses a combination of physiological and lab parameters (e.g., heart rate, blood pressure, pH, glucose). It is highly effective in PICUs for mortality prediction with an AUROC of 0.92 (Pollack et al.14, 1996).

Pediatric Index of Mortality (PIM)

PIM evaluates mortality risk using parameters like systolic BP, pupillary reflexes, and mechanical ventilation. It has a high predictive value with an AUROC of 0.88 (Slater et al.15, 2003).

Pediatric Logistic Organ Dysfunction (PELOD)

PELOD assesses multi-organ dysfunction, including cardiovascular, respiratory, and neurological metrics. It offers strong mortality prediction with an AUROC of 0.90 (Leteurtre et al.16, 2003).

Comparison of Scoring Methods

Predictive Value

PEWS is best for general pediatric wards, excelling in early detection but limited by subjectivity. qSOFA is good for sepsis prediction but lacks pediatric sensitivity. PRISM, PIM, PELOD excel in PICUs for mortality prediction but are complex and resource-intensive.

Biochemical Marker Integration

PRISM, PIM, and PELOD include lab markers (e.g., glucose, pH), boosting predictive accuracy. PEWS and qSOFA lack these markers, reducing their precision but increasing usability in routine settings.

Study Design and Setting

This research was carried out between April 2023 and August 2024 in the PICU of pediatric unit of a hospital in Jaipur, Rajasthan. It was a forward-looking observational study designed to assess whether the Pediatric Early Warning Score (PEWS) or Age adapted qSOFA could effectively determine the appropriate level of medical care, analyse the outcome in form of stay in hospital, death/discharge and requirement of oxygen therapy or mechanical ventilation. The treatment of all the patients were independent of the study being carried out.

Study Population and Sampling

Children aged between 1 month and 18 years who visited the pediatric department were included. Study was carried out on a sample size of 279 patients. Ethical approval for the study was granted by the ethical committee of MGUMST.

Inclusion Criteria

All children of age 1 month to 18 years admitted in PICU

Exclusion Criteria

Patients who left without medical advice.

Patients with chronic illness like TB

Patients with congenital malformation Patients with malignant disorders 

Data Collection and Statistical Analysis

Data interpretation and statistical analysis were performed using Microsoft Excel (version 2008) and IBM SPSS version 24.0. Qualitative variables were represented as frequencies and percentages, while quantitative data were presented as mean and standard deviation. Statistical analysis and techniques used are elaborated below. 

Statistical analysis technique and methods

Specificity, sensitivity, ROC and correlation-based analysis for binary classification and correlation, p value analysis for continuous variables. 

A threshold value is the cutoff point used in classification problems like specificity, sensitivity, and confusion matrix that decides whether a test result is classified as positive or negative.

Changing the threshold affects sensitivity (how many actual positives are detected) and specificity (how well negatives are correctly identified).

• Sensitivity (True Positive Rate): Measures the ability of a scoring method to correctly identify patients who died. Its mathematical interpretation is TP / (TP+FN)
• Specificity (True Negative Rate): Measures the ability of a scoring method to correctly identify patients who were discharged. Its mathematical interpretation is TN / (TN+FP)
• AUROC is useful to assess overall discriminative ability where different cutoff values can be tested. ROC Curve and AUC (Area Under Curve)

The ROC (Receiver Operating Characteristic) Curve represents the trade-off between sensitivity and specificity. It is plotted as a graph between TPR and FPR at each threshold and a very helpful tool to compare the predictive nature of different scoring technique.

In our case, The AUC (Area Under Curve) quantifies how well a scoring method can distinguish between death and discharge cases. A higher AUC value indicates better predictive performance. The ROC curve is a graphical representation of a scoring system’s diagnostic ability at various threshold settings. The Area Under the ROC Curve (AUROC) quantifies the overall discriminative ability of the system.

Figure 1- Distribution of patients - Gender based, Age based, Outcome based (Hospital stay duration) and Symptomatic distribution

This report aims to evaluate and compare 2 scoring methods—Age-Adapted qSOFA and PEWS—to determine their effectiveness in predicting patient outcomes (discharge vs. death), hospital stay duration, requirement of oxygen therapy and mechanical ventilation. The statistical techniques employed include sensitivity, specificity, AUC (Area Under Curve) and correlation analysis.

Depending upon the nature of outcome (continuous or binary), four different analyses were done based on most optimum technique.

Analysis 1- Outcome (Death vs. Discharge) prediction

The dataset was categorized into two outcome groups:

• Discharged (0) = 259 patients
• Died (1) = 20 patients

Using statistical analysis, the performance of each scoring method in predicting patient outcomes was evaluated. Since the outcome of each patient is binary in nature Sensitivity, Specificity, and AUC Analysis is done for comparison.  

Table 3 - Comparison of scoring methods on Threshold >=1

Table 4 - Comparison of scoring methods on Threshold >= 2

Sensitivity vs. Specificity Trade-off

At Threshold 1, both scores (PEWS, Age-adapted qSOFA) had higher sensitivity but lower specificity.

At Threshold 2, specificity improved significantly but its sensitivity dropped sharply (85% to 30% - age adapted qSOFA), meaning it missed many true deaths.

It can be concluded that for evaluating the condition of a patient by the method of age adapted qSOFA, the threshold value of 1 can be used as it offers the balance of specificity and sensitivity, but considering the lower value of specificity and AUROC, the expected discriminatory power will not be that good. Also, the range of score in PEWS scoring system can vary between 0 to 26, a threshold value of 2 will be quite low and may be biased towards sensitivity. Therefore, pews score is evaluated at each threshold in table 5.

Table 5 - Sensitivity and Specificity in PEWS score at each threshold

From this analysis, it is clear that PEWS ≥ 8 is the most optimum threshold at which balance of specificity and sensitivity is obtained. Considering the value of specificity and sensitivity at this threshold, i.e. more than 80%, PEWS score of 8 can be used in deciding the level of care/monitoring of patients.

Outcome analysis-

Since the number of death cases in the overall sample space is very less (only 20 cases), the analysis is done on the very skewed data. Due to this, to obtain an optimum value to balance specificity and sensitivity is difficult.

To compare the performance of both scores in effectiveness of their discriminating power, AUROC curve is plotted for both scores.

The ROC curve compares the diagnostic performance of the qSOFA and PEWS scores in predicting patient outcomes (death vs discharge). Both curves demonstrate how sensitivity (true positive rate) trades off with 1-specificity (false positive rate) across various thresholds. The Area Under the Curve (AUC) provides a single summary metric of each model's overall performance. In this analysis, qSOFA and PEWS both show reasonable discriminatory ability, with PEWS exhibiting a significantly higher AUC (0.88) than qSOFA (0.65), suggesting it may be a better tool at distinguishing between patients who are likely to die versus those who will be discharged.

Figure 2-

Analysis 2 - Hospital stay duration prediction

Since the duration of stay in hospital is a linear and continuous variable so classification-based technique/statistical tools as used in previous discussion may not be directly used. Therefore, the stay duration is classified in 2 categories i.e. Stay> 7 days and Stay <7 days for analysis purpose. The analysis trying to state any relation between scoring systems and categorized stay duration is listed below.

Both Age adapted qSOFA and PEWS have some ability to predict whether a stay will be long (≥7 days). PEWS performs slightly better (AUROC = 0.65) compared to pqsofa (AUROC = 0.60).

In general, an AUROC between 0.6–0.7 suggests poor to fair discrimination — but PEWS is relatively better.

From this chart, it is clear that these techniques are not useful in predicting stay more or less than 7 days. It is also expected in the analysis as the stay slightly more or less than a particular duration is not significantly different eg. The condition of patient with 6 days stay or 8 days stay may not be significantly different but in this analysis purpose by converting it into classification problem we have made these durations as completely opposite.

Figure 3-

Analysis 3 - Hospital stay duration prediction without classification technique

As the above analysis which was done by converting stay duration into binary classification (stay< 7 days or more) has failed to provide any useful result, we applied Spearman's correlation to assess the relationship between each scoring method and hospital stay duration. This correlation technique is effective in continuous variables.

Table 5 - Correlation Results

Spearman's correlation coefficient (ρ) measures the strength and direction of a monotonic relationship between two variables. A coefficient between ±0.20 to ±0.39 indicates a weak association, while ±0.40 to ±0.59 suggests a moderate one. Values closer to ±1.00 show very strong relationships, and 0 means no monotonic relationship.

From the above table, it is clear that since both correlation coefficients are positive value, it indicates positive correlation of scores with stay duration.

Since the value is near to 0.2 and 0.3, it indicates very weak correlation between scores and stay duration while PEWS comes out as a slightly better predictor of hospital stay duration.

Scatter plots with trend lines are also shown in figure-4 to visualise the correlation

Figure 4-

The scatter plots show that both Age adapted qSOFA and PEWS scores have a slight positive trend with hospital stay duration. PEWS shows a clearer upward pattern compared to PQSOFA, suggesting a better predictive ability. Trend line confirm that higher scores tend to relate to longer hospital stays, especially for PEWS.

The dots representing the patients data is also scattered signifying some aberration but majority of the cases are in line with trend line.

Analysis 4: Comparison of scores in predicting relation/effect on requirement of oxygen therapy

The requirement of oxygen therapy requirement is derived from 3 variables from master chart:

"FiO2", "Flow" , "Requiring invasive ventilation" 1.

To process the data for analysis on the relation between the scoring system and requirement of oxygen support the outcome variable i.e. requirement of oxygen support is converted into binary classification:

The dataset has 174 patients who did not require oxygen therapy (0) and 105 patients required oxygen therapy (1).

Figure 5 -

Among the tested thresholds, PEWS ≥ 6 provides the best balance, achieving 78% sensitivity, 85% specificity, and the AUROC of 0.89, making it an effective predictor.

Age-adapted qSOFA also performs reasonably well at Threshold ≥ 1, with 73% sensitivity, 48% specificity, and an AUROC of 0.61, making it a viable secondary option.

Table 6 – Specificity, Sensitivity and AUROC

Figure 6-

Analysis 5: Comparison of scores in predicting relation/effect on requirement of invasive ventilation

No requirement for invasive ventilation (0)

Requirement for invasive ventilation (1) (including all cases requiring ventilation, regardless of duration)

Similar analysis of Sensitivity and specificity is done at each threshold to determine the balance between correctly identifying patients in need and minimizing false positives. AUROC was used to measure the overall predictive capability of each scoring system.

Table 7 – Specificity, Sensitivity and AUROC

Age-Adapted qSOFA

At Threshold ≥1, age adapted qSOFA demonstrated 90% sensitivity but low specificity (42%), leading to a high rate of false positives. At Threshold ≥3, sensitivity sharply declined, approaching 5% but specificity improves (98%). This shows with such low sensitivity it will miss many true positive cases, indicating that qSOFA becomes ineffective in capturing cases requiring invasive ventilation. The AUROC for qSOFA remained 0.60, showing its limited predictive power for this outcome.

PEWS

At Threshold ≥7, an optimal balance was achieved with 85% sensitivity and 75% specificity. PEWS outperformed qSOFA with an AUROC of 0.88, indicating a stronger predictive ability.

Figure 7-

Numerous studies have evaluated early warning scores for identifying clinical deterioration in pediatric intensive care units (PICUs). Among these, the Pediatric Early Warning Score (PEWS) has consistently demonstrated superior performance. Parshuram et al.17, (2011) first validated its effectiveness in identifying critically ill children requiring urgent intervention. Subsequent studies, including those by Lambert et al.18, (2017), Duncan et al. 19, (2009), and Reuland et al. 8, (2023),  further confirmed its utility in predicting clinical deterioration, reducing mortality, and guiding early intervention, particularly in resource-limited settings. In contrast, the age-adapted quick Sequential Organ Failure Assessment (qSOFA), initially developed for adult sepsis detection, has shown limited applicability in pediatric populations. Although efforts have been made to tailor qSOFA for children, its predictive accuracy remains suboptimal compared to PEWS. Our findings corroborate this, establishing PEWS as a more reliable tool for pediatric risk stratification.

Predictive Performance

In mortality prediction, PEWS significantly outperformed age-adapted qSOFA, with an Area Under the Receiver Operating Characteristic (AUROC) of 0.88 versus 0.65. At a threshold ≥8, PEWS achieved optimal sensitivity (80%) and specificity (81%). qSOFA, despite reasonable sensitivity at a threshold of 1 (90%), suffered from poor specificity (42%), limiting its standalone clinical value.

In terms of respiratory support needs, PEWS again outperformed qSOFA. For oxygen therapy prediction, PEWS achieved an AUROC of 0.89, with an optimal threshold of 6 yielding 78% sensitivity and 85% specificity. For invasive ventilation prediction, PEWS at a threshold ≥7 yielded an AUROC of 0.88, with high sensitivity (85%) and specificity (75%). In comparison, qSOFA’s AUROC for predicting ventilation need was only 0.60, with low specificity (42%) despite high sensitivity. These findings align with previous studies. Seiger et al. 20, (2013) and Schlapbach et al. 21, (2017) also reported the superior predictive capacity of PEWS for respiratory distress and mechanical ventilation requirements.

Comparative Evidence

Despite its effectiveness, PEWS is not without limitations. Kowalski et al. 6 (2023) observed that nearly half (47%) of PEWS assessments underestimated illness severity prior to PICU transfers. Conversely, a prospective study by Parwaiz et al.7 (2023) demonstrated that a PEWS >6 at emergency admission accurately predicted PICU need, with specificity of 98.42%, supporting its application in early triage, especially in low-resource settings.

Age-adapted qSOFA has shown moderate predictive power in emergency settings. Eun et al.22 (2023) reported an AUROC of 0.733 for mortality prediction, with improved accuracy when the AVPU scale was used instead of the Glasgow Coma Scale. However, moderate sensitivity limited its clinical utility as a standalone tool.

Implementation Challenges and Innovation

Reuland et al. 8 (2023) emphasized implementation challenges for PEWS in resource-constrained settings, including staff shortages, lack of equipment, and inefficient workflows. Their study at the Philippine Children’s Medical Center highlighted the need for system-level adaptations and context-specific tools.

Emerging technologies such as artificial intelligence (AI) and machine learning (ML) present viable solutions. Kim et al.23 (2019) introduced the PROMPT deep learning model, which predicted ICU mortality up to 60 hours in advance with AUROC as high as 0.97, surpassing conventional scores like PIM 3. Similarly, Lee et al. 24 (2021) developed a Random Forest model across four tertiary hospitals, achieving AUROC of 0.942—significantly higher than PIM 3 (0.892). These models identified key physiological and laboratory variables, showing the potential of dynamic, real-time systems in pediatric critical care.

Schlapbach et al.21 (2017) also compared multiple scoring systems in children with suspected infections, finding that SOFA (AUROC 0.829) and PELOD-2 (AUROC 0.816) outperformed both qSOFA (0.739) and SIRS (0.710). This supports the superiority of organ dysfunction-based tools and their alignment with Sepsis-3 definitions.

Clinical Implications

The present findings affirm PEWS as a robust tool for early identification of deterioration in pediatric patients. The thresholds identified (≥8 for mortality, ≥6 for oxygen therapy) provide practical guidance for clinical decision-making. PEWS' predictive reliability is particularly valuable in resource-limited environments where early risk identification is crucial for prioritizing care.

Although qSOFA offers ease of use, its low specificity and limited AUROC values restrict its use as a standalone tool. These results suggest the need to refine qSOFA or integrate it with other clinical parameters for pediatric applications.

Limitations

This study’s findings must be interpreted in light of certain limitations:

• Single-center design: Limits generalizability to broader populations and settings.
• Small sample size: Especially in critical outcome groups (e.g., mortality), reducing statistical power.
• Observer bias: Scores were recorded by a single observer, possibly introducing subjectivity.
• Single time-point measurement: Dynamic changes in patient status were not captured. AI-based studies [14, 30] using continuous monitoring have shown greater predictive precision and could be more effective in future clinical use.

Future Scope

The integration of AI and Machine Learning (ML) holds promise for enhancing pediatric scoring systems. These technologies can leverage real-time data from electronic health records (EHRs) to generate more accurate, automated predictions. Kim et al.23 (2019) developed a deep learning model for real-time mortality prediction in PICUs, showing strong performance. Lee et al.24 (2021) created an ML model that outperformed traditional systems in early ICU admissions. AI can help overcome limitations of manual scoring by standardizing interpretation, improving early warning scores' reliability, especially in resource-constrained environments.

The comparison between PEWS and Age-Adapted qSOFA for predicting pediatric patient outcomes highlights the superior performance of PEWS in several critical domains. For mortality prediction, PEWS achieved a high AUC of 0.88, with a threshold score of ≥8 offering the best sensitivity-specificity balance. Age-Adapted qSOFA, by contrast, showed lower predictive accuracy and was less effective as a standalone tool. Similarly, for oxygen therapy requirements, PEWS ≥6 yielded an excellent AUROC of 0.89, outperforming Age-Adapted qSOFA (AUROC 0.61), largely due to PEWS including multiple respiratory parameters directly related to oxygen need. In predicting the requirement for invasive ventilation, PEWS again proved superior (AUROC 0.88 vs. 0.60), reinforcing its value as a more reliable tool for pediatric critical care triage and intervention planning.

When analyzing hospital stay duration, initial binary classification (stay >7 vs. <7 days) revealed limited predictive utility for both scores, with PEWS (AUROC 0.65) only slightly outperforming Age-Adapted qSOFA (AUROC 0.60). Recognizing the limitations of binary classification for a continuous variable, the study shifted to Spearman’s correlation, revealing weak but positive associations—PEWS (ρ = 0.284) and Age-Adapted qSOFA (ρ = 0.221)—with PEWS again marginally better. Scatter plots supported these trends, showing a mild correlation between higher scores and longer hospital stays. While neither score showed strong predictive power for stay duration, PEWS consistently outperformed Age-Adapted qSOFA across all clinical outcome domains.

1. Tucker KM, Brewer TL, Baker RB, Demeritt B, Vossmeyer MT. Prospective evaluation of a pediatric inpatient early warning scoring system. J Spec Pediatr Nurs. 2009;14(2):79–85.
2. Akre M, Finkelstein M, Erickson M, Liu M, Vanderbilt L, Billman G. Sensitivity of the pediatric early warning score to identify patient deterioration. Pediatrics. 2010;125(4):e763–9.
3. Lambert V, Matthews A, MacDonell R, Fitzsimons J. Pediatric early warning systems for detecting and responding to clinical deterioration in children: a systematic review. BMJ Open. 2017 Mar 13;7(3):e014497. doi: 10.1136/bmjopen‑2016‑014497. PMID: 28289051; PMCID: PMC5353324.
4. Weiss SL, Fitzgerald JC, Maffei FA, Kane JM, Rodriguez‑Nunez A, Hsing DD, et al. Quick Sepsis‑related Organ Failure Assessment, Systemic Inflammatory Response Syndrome, and Early Warning Scores for Detecting Clinical Deterioration in Infected Patients Outside the Intensive Care Unit. Am J Respir Crit Care Med. 2017;195(7):906–11.
5. Jia W, Zhang X, Sun R, Li P, Wang D, Gu X, Song C. Value of modified qSOFA, glucose, and lactate in predicting prognosis in children with sepsis in the PICU. Pediatr Res. 2023;93(2):456–62.
6. Kowalski RL, Lee L, Spaeder MC, Moorman JR, Keim‑Malpass J. Accuracy and monitoring of Pediatric Early Warning Score (PEWS) scores prior to emergent pediatric intensive care unit (ICU) transfer: retrospective analysis. Pediatr Crit Care Med. 2023;24(5):e234–41.
7. Parwaiz A, Agrawal N, Gupta A, Simalti A, Kedarnath M. Utility of pediatric early warning score at emergency room in predicting the level of care required for next 48 h: a single‑center, prospective, observational study. Indian J Pediatr. 2023;90(4):345–50.
8. Reuland C, Shi G, Deatras M, Ang M, Evangelista PPG, Shilkofski N. A qualitative study of barriers and facilitators to pediatric early warning score (PEWS) implementation in a resource‑limited setting. BMC Pediatr. 2023;23(1):45.
9. Van Nassau SC, van Beek RH, Driessen GJ, van de Pol AC. Validation of the Pediatric Early Warning Score: a multicenter study. Intensive Care Med. 2015;41(7):1208–16.
10. Agulnik A, Mora Robles LN, Forbes PW, Soberanis Vasquez DJ, Mack R, Antillon‑Klussmann F, Rodriguez‑Galindo C. Improved outcomes after successful implementation of a pediatric early warning system (PEWS) in a resource‑limited pediatric oncology hospital. Cancer. 2017;123(15):2965–74.
11. Breslin K, Marx J, Hoffman H, McBeth R, Pavuluri P. Pediatric early warning score at time of emergency department disposition is associated with level of care. Pediatr Emerg Care. 2014;30(2):97–103.
12. Chapman SM, Grocott MP, Franck LS. Systematic review of pediatric early warning systems (PEWS) for hospitalized children. J Pediatr Nurs. 2010;25(6):573–82.
13. Skaletzky SM, Raszynski A, Totapally BR. Validation of a modified pediatric early warning system score: a retrospective case‑control study. Clin Pediatr (Phila). 2012;51(5):431–5.
14. Pollack MM, Ruttimann UE, Getson PR. Pediatric risk of mortality (PRISM) score. Crit Care Med. 1996;14(11):1118–26.
15. Slater A, Shann F, Pearson G. PIM2: a revised version of the Pediatric Index of Mortality. Intensive Care Med. 2003;29(2):278–85.
16. Leteurtre S, Martinot A, Duhamel A, Gauvin F, Grandbastien B, Van Thanh T, Proulx F. Development of a pediatric multiple organ dysfunction score: use of two strategies. [Journal missing]. 2003.
17. Parshuram CS, Bayliss A, Reimer J, Middaugh K, Blanchard N. Implementing the Bedside Pediatric Early Warning System in a community hospital: a prospective observational study. Paediatr Child Health. 2011;16(3):e18–22.
18. Lambert V, Matthews A, MacDonell R, Fitzsimons J. Pediatric early warning systems for detecting and responding to clinical deterioration in children: a systematic review. BMJ Open. 2017;7(3):e014497.
19. Duncan HP, Frew E. The Pediatric Early Warning Score: a severity of illness score to predict urgent medical need in hospitalized children. J Crit Care. 2009;24(2):222–9.
20. Seiger N, Maconochie I, Oostenbrink R, Moll HA. Validity of different pediatric early warning scores in the emergency department. Pediatrics. 2013;132(4):e841–50.
21. Schlapbach LJ, Straney L, Bellomo R, MacLaren G, Pilcher D. Prognostic accuracy of age‑adapted SOFA, SIRS, PELOD‑2, and qSOFA for in‑hospital mortality among children with suspected infection admitted to the intensive care unit. [Journal missing]. 2017.
22. Eun S, Kim H, Kim HY, Lee M, Bae GE, Kim H, et al. Age‑adjusted quick Sequential Organ Failure Assessment score for predicting mortality and disease severity in children with infection: a systematic review and meta‑analysis. Crit Care Med. 2023;51(6):789–98.
23. Kim XY, Kim S, Cho J, Kim YS, Soi IS, Song Y, Cho H, Pa M, Jung H, Kim YH, Kim KW, John MH. A deep learning model for real‑time mortality prediction in critically ill children. Crit Care. 2019;23(1):45.
24. Lee B, Kim K, Hwang H, Kim YS, Chung EH, Yoon JS, Cho HJ, Park JD. Development of a machine learning model for predicting pediatric mortality in the early stages of intensive care unit admission. Pediatr Crit Care Med. 2021;22(5):456–62.