Combining Cribari matrix and Need For Trauma Intervention (NFTI) to accurately assess undertriage in pediatric trauma

Prehospital and Emergency Department Vital Sign Abnormalities Among Injured Children

OBJECTIVES

Vital signs are a critical component in the assessment of the injured child. We compared vital sign abnormalities among injured children in the prehospital setting to those in the emergency department (ED) and evaluated the predictive value of each for the presence of major trauma.

METHODS

We performed a multi-agency and multicenter retrospective study of injured children within a county-based emergency medical services (EMS) system between 2010 and 2021, including injured children (<18 years) transported to the hospital. We compared prehospital vital signs for heart rate (HR), respiratory rate (RR), and systolic blood pressure (SBP) in the prehospital and ED setting. Using the Standard Triage Assessment Tool to define major trauma, we constructed multivariable models to evaluate the association of prehospital and ED vital sign abnormalities for major trauma.

RESULTS

We included 21,298 encounters (median age 13 years, IQR 6-16), with major trauma was present in 3,606 (16.9%). In the prehospital setting, abnormal vital signs were reported in 25.7% for HR, 14.6% for RR, and 24.3% for SBP. ED measurements recorded a higher proportion of abnormal HR (28.2%) and RR (21.3%), and slightly lower proportion with abnormal SBP (21.8%). Cohen's Kappa was fair for HR (0.27) and SBP (0.20), but slight for RR (0.09). Prehospital vital signs most strongly associated with major trauma included tachypnea (odds ratio [OR] 2.7, 95% confidence interval (CI 2.4-3.1) and bradypnea (OR 1.7, 95% CI 1.4-1.9). ED vital signs most strongly associated with major trauma included hypotension (OR 2.4, 95% CI 2.1-2.7) and tachypnea (OR 1.8, 95% CI 1.6-2.0). Prehospital and ED vital signs demonstrated similar performance in predicting major trauma (area under the receiver operator characteristic curve (AUROC 0.63); 95% CI 0.61-0.64 for prehospital; 0.63; 95% CI, 0.61-0.64 for ED). When combining prehospital and ED vital signs into a single model, predictive power increased (AUROC 0.66, 95% CI 0.65-0.67).

CONCLUSIONS

We identified poor correlation between prehospital and ED vital signs. In both settings, vital sign abnormalities were associated with major trauma. The combined use of prehospital and ED vital signs improved predictive value for major trauma, suggesting potential for future integration into trauma triage tools.

An Activation Failure: Factors Associated With Undertriage of Pediatric Major Trauma Victims

INTRODUCTION

Undertriage of children contributes to poorer clinical outcomes. The objective of this study was to determine factors associated with undertriage of pediatric major trauma victims.

METHODS

We performed a retrospective cross-sectional study of children (aged < 16 ys) using the 2021 American College of Surgeons National Trauma Data Bank. We identified children who met the definition of major trauma defined by the Standard Triage Assessment Tool. We performed multivariable logistic regression to determine factors associated with undertriage, defined as encounters which met criteria, but did not receive highest-level activation.

RESULTS

Of 97,812 included children, 5.3% met major trauma criteria. Undertriage occurred in 34.4% of encounters with major trauma. Factors associated with undertriage included fall and striking mechanisms, missing blood pressure, private vehicle arrival, and incoming interfacility transfers. Hypotension, decreased level of consciousness, prehospital and in-hospital intubation, tachycardia, hypothermia, penetrating mechanism, presentation to a pediatric level 2 or adult level 1 trauma center relative to pediatric level 1 center, and arrival by flight were associated with lower odds of undertriage.

CONCLUSIONS

Many children with major trauma were undertriaged, particularly those presenting with lower-risk histories, such as private vehicle arrivals and fall mechanisms. Future work should seek to develop risk-stratification systems that can better identify children with major trauma, with an emphasis on those with blunt traumatic mechanisms.

SIPA Poorly Predicts Outcomes in Young Pediatric Trauma Patients

INTRODUCTION

The shock index pediatric adjusted score (SIPA) predicts the need for blood transfusion (BT), hemorrhage control interventions (HCI), morbidity/mortality among older pediatric trauma patients but is less predictive in younger patients. We hypothesize that SIPA will be predictive among older pediatric patients for BT, HCI, mortality, and need for trauma intervention (NFTI), however we aim to further delineate the gap in utilizing SIPA in younger patients.

METHODS

Using the ACS NTDB for 2017-2021 we evaluated patients 1-14 years old who were transported by EMS from the scene for definitive care. Patients were divided into three age groups: 1-4, 5-9, and 10-14 years. Recursive partitioning was used to identify separate SIPA cut-points predictive of BT, HCI, NFTI, and morbidity/mortality. Cut-points from the partitions were evaluated using Area-under-curve (AUC) statistics and response probabilities were obtained from corresponding Leaf Reports.

RESULTS

Four SIPA cut-points from the recursive partitioning were selected for each age group. SIPA was more predictive of the need for HCI. BT showed similar results consistent with previous literature. SIPA alone showed poor discrimination in relation to NFTI and mortality, and again predictive value was slightly higher in older children.

CONCLUSION

While SIPA alone showed discrimination of specific outcomes of BT and HCI, it was poorly predictive of both the NFTI and mortality in children. The youngest pediatric patients continue to be elusive. Utilizing SIPA in combination with additional scores may be necessary to triage young children appropriately. This study also indicates the need to develop NFTI criteria specific to children.

LEVEL OF EVIDENCE

IV.

In Search of the Truth: Choice of Ground Truth for Predictive Modeling of Trauma Team Activation in Pediatric Trauma

BACKGROUND

Assigning trauma team activation (TTA) levels for trauma patients is a classification task that machine learning models can help optimize. However, performance is dependent on the "ground-truth" labels used for training. Our purpose was to investigate 2 ground truths, the Cribari matrix and the Need for Trauma Intervention (NFTI), for labeling training data.

STUDY DESIGN

Data were retrospectively collected from the institutional trauma registry and electronic medical record, including all pediatric patients (age <18 years) who triggered a TTA (January 2014 to December 2021). Three ground truths were used to label training data: (1) Cribari (Injury Severity Score >15 = full activation), (2) NFTI (positive for any of 6 criteria = full activation), and (3) the union of Cribari+NFTI (either positive = full activation).

RESULTS

Of 1,366 patients triaged by trained staff, 143 (10.47%) were considered undertriaged using Cribari, 210 (15.37%) using NFTI, and 273 (19.99%) using Cribari+NFTI. NFTI and Cribari+NFTI were more sensitive to undertriage in patients with penetrating mechanisms of injury (p = 0.006), specifically stab wounds (p = 0.014), compared with Cribari, but Cribari indicated overtriage in more patients who required prehospital airway management (p < 0.001), CPR (p = 0.017), and who had mean lower Glasgow Coma Scale scores on presentation (p < 0.001). The mortality rate was higher in the Cribari overtriage group (7.14%, n = 9) compared with NFTI and Cribari+NFTI (0.00%, n = 0, p = 0.005).

CONCLUSIONS

To prioritize patient safety, Cribari+NFTI appears best for training a machine learning algorithm to predict the TTA level.

Defining pediatric trauma center resource utilization: Multidisciplinary consensus-based criteria from the Pediatric Trauma Society

BACKGROUND

Pediatric trauma triage and transfer decisions should incorporate the likelihood that an injured child will require pediatric trauma center (PTC) resources. Resource utilization may be a better basis than mortality risk when evaluating pediatric injury severity. However, there is currently no consensus definition of PTC resource utilization that encompasses the full scope of PTC services.

METHODS

Consensus criteria were developed in collaboration with the Pediatric Trauma Society (PTS) Research Committee using a modified Delphi approach. An expert panel was recruited representing the following pediatric disciplines: prehospital care, emergency medicine, nursing, general surgery, neurosurgery, orthopedics, anesthesia, radiology, critical care, child abuse, and rehabilitation medicine. Resource utilization criteria were drafted from a comprehensive literature review, seeking to complete the following sentence: "Pediatric patients with traumatic injuries have used PTC resources if they..." Criteria were then refined and underwent three rounds of voting to achieve consensus. Consensus was defined as agreement of 75% or more panelists. Between the second and third voting rounds, broad feedback from attendees of the PTS annual meeting was obtained.

RESULTS

The Delphi panel consisted of 18 members from 15 institutions. Twenty initial draft criteria were developed based on literature review. These criteria dealt with airway interventions, vascular access, initial stabilization procedures, fluid resuscitation, blood product transfusion, abdominal trauma/solid organ injury management, intensive care monitoring, anesthesia/sedation, advanced imaging, radiologic interpretation, child abuse evaluation, and rehabilitative services. After refinement and panel voting, 14 criteria achieved the >75% consensus threshold. The final consensus criteria were reviewed and endorsed by the PTS Guidelines Committee.

CONCLUSION

This study defines multidisciplinary consensus-based criteria for PTC resource utilization. These criteria are an important step toward developing a criterion standard, resource-based, pediatric injury severity metric. Such metrics can help optimize system-level pediatric trauma triage based on likelihood of requiring PTC resources.

LEVEL OF EVIDENCE

Diagnostic Test/Criteria; Level II.

Comparison of Vital Sign Cutoffs to Identify Children With Major Trauma

Importance

Vital signs are essential components in the triage of injured children. The Advanced Trauma Life Support (ATLS) and Pediatric Advanced Life Support (PALS) physiologic criteria are frequently used for trauma assessments.

Objective

To evaluate the performance of ATLS and PALS criteria vs empirically derived criteria for identifying major trauma in children.

Design, Setting, and Participants

This retrospective cohort study used 2021 American College of Surgeons Trauma Quality Improvement Program (TQIP) data contributed by US trauma centers. Included encounters involved pediatric patients (aged <18 years) with severe injury, excluding those who experienced out-of-hospital cardiac arrest, were receiving mechanical ventilation, or were transferred from another facility. Data were analyzed between April 9 and December 21, 2023.

Exposure

Initial hospital vital signs, including heart rate, respiratory rate, and systolic blood pressure (SBP).

Main Outcome and Measures

Major trauma, determined by the Standard Triage Assessment Tool, a composite measure of injury severity and interventions performed. Multivariable models developed from PALS and ATLS vital sign criteria were compared with models developed from the empirically derived criteria using the area under the receiver operating characteristic curve. Validation of the findings was performed using the 2019 TQIP dataset.

Results

A total of 70 748 patients (median [IQR] age, 11 [5-15] years; 63.4% male) were included, of whom 3223 (4.6%) had major trauma. The PALS criteria classified 31.0% of heart rates, 25.7% of respiratory rates, and 57.4% of SBPs as abnormal. The ATLS criteria classified 25.3% of heart rates, 4.3% of respiratory rates, and 1.1% of SBPs as abnormal. Among children with all 3 vital signs documented (64 326 [90.9%]), PALS had a sensitivity of 88.4% (95% CI, 87.1%-89.3%) and specificity of 25.1% (95% CI, 24.7%-25.4%) for identifying major trauma, and ATLS had a sensitivity of 54.5% (95% CI, 52.7%-56.2%) and specificity of 72.9% (95% CI, 72.6%-73.3%). The empirically derived cutoff vital sign z scores had a sensitivity of 80.0% (95% CI, 78.5%-81.3%) and specificity of 48.7% (95% CI, 48.3%-49.1%) and area under the receiver operating characteristic curve of 70.9% (95% CI, 69.9%-71.8%), which was similar to PALS criteria (69.6%; 95% CI, 68.6%-70.6%) and greater than ATLS criteria (65.4%; 95% CI, 64.4%-66.3%). Validation using the 2019 TQIP database showed similar performance to the derivation sample.

Conclusions and Relevance

These findings suggest that empirically derived vital sign criteria strike a balance between the sensitivity of PALS criteria and the specificity of ATLS criteria in identifying major trauma in children. These criteria may help to identify children at greatest risk of trauma-related morbidity and mortality.

Assessing Effectiveness and Efficiency of Need for Trauma Intervention (NFTI) and Modified NFTI in Identifying Overtriage and Undertriage Rates and Associated Outcomes

INTRODUCTION

Limited research has assessed the effectiveness of Need for Trauma Intervention (NFTI) and Modified NFTI (MNFTI) criteria in accurately identifying triage rates in major trauma. We aim to evaluate the predictive capability of NFTI/MNFTI in determining rates of overtriage and undertriage, as well as associated outcomes.

METHODS

A literature search was conducted utilizing PubMed, Google Scholar, EMBASE, ProQuest, and Cochrane from conception to April 13th, 2023. Studies assessing the utilization of NFTI/MNFTI in identifying over and undertriage rates were included. Additional outcomes including mortality, ICU LOS, and resource allocation were evaluated. Outcomes were compared between NFTI/MNFTI and other triage metrics.

RESULTS

A total of 8 articles, including 175,650 trauma patients, were evaluated. NFTI utilization was associated with reduced overtriage rates compared to numerous tools including trauma triage matrix (TTM) and need for emergent intervention within 6 h (NEI-6) (NFTI 32.15%, TTM 44.5%, NEI-6 42.23%). Regarding undertriage, NFTI had lower rates than the secondary triage assessment tool (STAT) and TTM (NFTI 14.0%, STAT, 22.3%, TTM 14.3%) as well as Cribari Matrix Method (CMM) (NFTI .8%, CMM 7.6%, P < .0003). Additionally, the utilization of NFTI in combination with CMM yielded a significant reduction in undertriage rates compared to either tool alone (CMM/NFTI 2.7%, NFTI 4.6%, CMM 8.2%).

CONCLUSION

Implementation of NFTI/MNFTI resulted in more accurately capturing over and undertriage rates. Similar trends were identified when NFTI was used in combination with CMM. When compared to other triage tools, NFTI outperformed CMM, TTM, STAT, and NEI-6 in overtriage and/or undertriage rates.

Refining triage practices by predicting the need for emergent care following major trauma: the experience of a level 1 adult trauma center

PURPOSE

We examined the predictability of selected parameters for establishing the need for urgent care following multi-trauma as a means to warrant the highest level of trauma activation and potentially improve over- and under-triage rates.

METHODS

In this retrospective cohort study of multi-trauma patients aged ≥ 16 years performed at a level 1 trauma center, trauma activation criteria and additional characteristics were examined with respect to treatment urgency, defined as: a direct disposition to the operating room or intensive care unit, initiating acute intervention in the trauma room, and in-hospital death within 7 days of admission.

RESULTS

We enrolled 1373 patients (median age 36.0 years). The following parameter were inserted into the final multivariable model: age > 75 years, male sex, Charlson comorbidity index, trauma circumstances and mechanism, signs of respiratory distress, systolic BP ≤ 110 and GCS ≤ 13. Adjusted independent predictors of acute care requirement were as follows: GCS ≤ 13 (aOR 5.27 [95% CI 3.45-8.05], p < 0.001), systolic BP ≤ 110 mmHg (aOR 2.15 [95% CI 1.45-3.21], p < 0 .001), respiratory distress (aOR 2.05 [95% CI 1.53-2.77], p < 0.001), and age ≥ 75 years (aOR 1.90 [95% CI 1.18-3.08], p = 0.008).

CONCLUSION

A GCS ≤ 13, systolic BP < 110 mmHg, signs of respiratory distress, and age > 75 years best predicted the need for acute care following multisystem trauma. Prospective studies are warranted to confirm the predictability of these criteria and to assess the extent to which their implementation will refine over- and under-triage rates.

Use of reverse shock index times Glasgow coma scale (rSIG) to determine need for transfer of pediatric trauma patients to higher levels of care

INTRODUCTION

Most children in the US live more than one hour from a Level 1 PTC. The Need For Trauma Intervention (NFTI) score was developed to assess trauma triage criteria and is dependent on whether someone requires one of six urgent interventions (NFTI+). We sought to determine if a novel scoring tool, rSIG, could predict NFTI and facilitate the transfer decision making process.

METHODS

Children 1-18 years old transferred to our level 1 PTC from 2010 - 2020 with complete vital signs and Glasgow Coma Scale (GCS) score at the transferring facility were included. rSIG was calculated as previously described [(SBP/HR) x GCS], and the following cutoffs were used for each age group: ≤13.1, ≤16.5, and ≤20.1 for 1-6, 7-12, and 13-18 years, respectively. Clinical outcomes upon arrival to the PTC were collected to determine if patients met any NTFI criteria.

RESULTS

A total of 456 patients met inclusion criteria. The proportion of patients with an abnormal rSIG was 60.1% (274) and 37.0% (169) were NFTI+. Patients with an abnormal rSIG had an odds ratio of 6.18 (95% CI: 3.90, 10.07), p < 0.001 of being NFTI+ compared to those with a normal rSIG.

CONCLUSION

Children with an abnormal rSIG are more likely to be NFTI+ and require higher levels of care, indicating this scoring tool can identify pediatric trauma patients who may benefit from expedited transfer. Incorporating rSIG into initial evaluation and triage of traumatically injured children may expedite the transfer decision making process and limit delays in transport to a PTC.

TYPE OF STUDY

Retrospective Comparative Study LEVEL OF EVIDENCE: III.

Evaluation of activation criteria in paediatric multi-trauma

Objective

To explore the optimal set of trauma activation criteria predicting paediatric patients' need for acute care following multi-trauma, with particular attention to Glasgow Coma Scale (GCS) cut-off value.

Methods

A retrospective cohort study of paediatric multi-trauma patients aged 0 to 16 years, performed at a Level 1 paediatric trauma centre. Trauma activation criteria and GCS levels were examined with respect to patients' need for acute care, defined as: direct to operating room disposition, intensive care unit admission, need for acute interventions in the trauma room, or in-hospital death.

Results

We enrolled 436 patients (median age 8.0 years). The following predicted need for acute care: GCS <14 (adjusted odds ratio [aOR] 23.0, 95% confidence interval [CI]: 11.5 to 45.9, P < 0.001), hemodynamic instability: (aOR 3.7, 95% CI: 1.2-8.1, P = 0.01), open pneumothorax/flail chest (aOR: 20.0, 95% CI: 4.0 to 98.7, P < 0.001), spinal cord injury (aOR 15.4, 95% CI; 2.4 to 97.1, P = 0.003), blood transfusion at the referring hospital (aOR: 7.7, 95% CI: 1.3 to 44.2, P = 0.02) and GSW to the chest, abdomen, neck, or proximal extremities (aOR 11.0, 95% CI; 1.7 to 70.8, P = 0.01). Using these activation criteria would have decreased over- triage by 10.7%, from 49.1% to 37.2% and under-triage by 1.3%, from 4.7% to 3.5%, in our cohort of patients.

Conclusions

Using GCS<14, hemodynamic instability, open pneumothorax/flail chest, spinal cord injury, blood transfusion at the referring hospital, and GSW to the chest, abdomen, neck of proximal extremities, as T1 activation criteria could decrease over- and under-triage rates. Prospective studies are needed to validate the optimal set of activation criteria in paediatric patients.

Rethinking pediatric trauma triage

Despite advances in the delivery of trauma care, trauma remains the leading cause of death amongst the pediatric population within the United States and is one of the leading causes of death in children worldwide.  Accurately triaging pediatric trauma patients is essential to minimize preventable mortality without burdening the system by utilizing unnecessary resources.  This article will review the accuracy of current pediatric trauma triage practices and how it will evolve in the future including moving away from mechanism of injury towards physiologic scoring tools such as the pediatric age-adjust shock index, and intervention-based systems including. Need for Surgeon Presence and Need For Trauma Intervention. This paper will also present evidence regarding over-utilization of air transport for pediatric trauma patients and the associated unnecessary costs placed on the trauma system.

Prehospital Factors Predict Outcomes in Pediatric Trauma: A Principal Component Analysis

BACKGROUND

Trauma team activation leveling decisions are complex and based on many variables. Accurate triage decisions improve patient safety and resource utilization. Our purpose was to establish proof-of-concept for using principal component analysis (PCA) to identify multivariate predictors of injury severity and to assess their ability to predict outcomes in pediatric trauma patients. We hypothesized that we could identify significant principal components (PCs) among variables used for decisions regarding trauma team activation and that PC scores would be predictive of outcomes in pediatric trauma.

METHODS

We conducted a retrospective review of the trauma registry (1/2014-12/2020) at our pediatric trauma center, including all pediatric patients (age < 18 y) who triggered a trauma team activation. Data included patient demographics, prehospital report, Injury Severity Score, and outcomes. Four significant principal components were identified using PCA. Differences in outcome variables between the highest and lowest quartile for PC score were examined.

RESULTS

1090 pediatric patients were included. The 4 significant PCs accounted for >96% of the overall date variance. The first PC was a composite of prehospital Glasgow Coma Scale and Revised Trauma Score (RTS) and was predictive of outcomes, including injury severity, length of stay, and mortality. The second PC was characterized primarily by prehospital systolic blood pressure (SBP) and high PC scores were associated with increased length of stay. The third and fourth PCs were characterized by patient age and by prehospital RTS and SBP, respectively.

CONCLUSIONS

We demonstrate that, using information available at the time of trauma team activation, PCA can be used to identify key predictors of patient outcome. While the ultimate goal is to create a machine learning-based predictive tool to support and improve clinical decision making, this study serves as a crucial step toward developing a deep understanding of the features of the model and their behavior with actual clinical data.

LEVEL OF EVIDENCE

III; Diagnostic test.