Small Change, BIG Impact: Proposal of the Brain Injury Guidelines for Kids (kBIG) Annika B. Kay a, Sommer L. Glasgow b, Anastasia M. Kahan b, *, Robert A. Swendiman b, Zachary J. Kastenberg b, Christopher M. Roach b, Hsuan-Yu Wan c, Robert J. Bollo d, Rajiv R. Iyer d, Vijay M. Ravindra d, David S. Morris a, Brian K. Yorkgitis e, Bellal Joseph f, Katie W. Russell b a Intermountain Health, Intermountain Medical Center, Salt Lake City, UT, USA b University of Utah, Division of Pediatric Surgery, Salt Lake City, UT, USA c Intermountain Health, Primary Children's Hospital, Salt Lake City, UT, USA d University of Utah, Department of Neurosurgery, Salt Lake City, UT, USA e Indiana University, Department of Surgery, Division of Trauma, and Acute Care Surgery, Indianapolis, IN, USA f The University of Arizona Medical Center, Division of Trauma, and Acute Care Surgery, Phoenix, AZ, USA article info Article history: Received 4 April 2025 Received in revised form 7 May 2025 Accepted 14 May 2025 Keywords: Pediatric trauma Brain injury Brain injury guidelines abstract Background: The adult Brain Injury Guidelines (BIG) stratify patients with traumatic intracranial hem- orrhage into one of three risk categories and supports management of lower risk traumatic brain injury (TBI) with minimal resources. While the BIG have been widely adopted by adult trauma centers, application to pediatric populations remains unvalidated. The objective of this study was to determine the efficacy of BIG in a pediatric trauma population and develop adapted criteria based on injury patterns that may be unique to children with TBI. Methods: BIG was applied to a retrospective cohort of pediatric TBI patients, de fined as <18 years old with skull fracture or intracranial hemorrhage on CT scan, from January 2018 to April 2024. Misclassi- fication of BIG was de fined as patients categorized as low- or moderate-risk who required neurosurgical intervention. Based on the performance of BIG, modi fications were made to derive a pediatric-specifi c guideline, BIG for Kids (kBIG). Results: 1894 patients were analyzed. The misclassi fication rate of BIG 1 and 2 was 0 % and 1.4 %, respectively. We derived kBIG, with modi fications to the categorization of minor skull fracture, epidural hematoma, neurologic exam and injury mechanism, which improved the misclassi fication rate of kBIG 2 to 0.8 % and maintained a rate of 0 % for all low-risk patients. Conclusions: We propose BIG for Kids, a modi fied version of the adult BIG for the management of TBI in pediatric patients. The kBIG was conservatively designed to ensure safety and to reduce unnecessary radiation exposure and resource utilization. Level of evidence: Cohort study/IV. © 2025 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license ( http://creativecommons.org/licenses/by-nc-nd/4.0/). 1. Background Traumatic brain injury (TBI) is the most common cause of pe- diatric injury and death worldwide [ 1]. Routine practice for pedi- atric patients with traumatic intracranial hemorrhage (ICH) includes transfer to a trauma center, repeat head computed to- mography (RHCT), and neurosurgery consultation (NSC). With resource allocation at the forefront, there is an increasing focus on the judicious use of healthcare resources in managing TBI across all ages. In children, suf ficient evidence exists to guide initial imaging decisions after head trauma [ 2] and there is a growing body of literature evaluating potentially avoidable transfers in mild TBI [3e7]. However, consensus on the strati fication and management of mild to moderate pediatric TBI is lacking. In 2013, the Brain Injury Guidelines (BIG) were proposed in adult trauma patients [ 8], which has since undergone multicenter validation [ 9]. Many trauma centers across the country have adopted these guidelines, and efforts to implement them across health systems and trauma networks are ongoing [ 10e15]. The BIG * Corresponding author. University of Utah, Division of Pediatric Surgery, Salt Lake City, UT, USA E-mail address: sasha.kahan@hsc.utah.edu (A.M. Kahan). Contents lists available at ScienceDirect Journal of Pediatric Surgery journal homepage: www.sciencedirect.com/journal/ journal-of-pediatric-surgery https://doi.org/10.1016/j.jpedsurg.2025.162372 0022-3468/© 2025 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license ( http://creativecommons.org/licenses/by-nc-nd/4.0/). Journal of Pediatric Surgery 60 (2025) 162372
stratifies patients into one of three categories based on neurologic exam, preinjury anticoagulation, and radiographic findings. Low- (BIG 1) and moderate-risk (BIG 2) patients can be safely managed without RHCT, NSC, and, in some cases, hospital admission or transfer, representing a signi ficant departure from historical prac- tice. Recent single-institution reports have found that BIG may safely reduce RHCT, NSC, and admission in pediatric TBI pop- ulations [16e21]. We aimed to retrospectively apply the BIG, as described for adults, to a pediatric TBI population and hypothesized that it would safely reduce RHCT, NSC, and unnecessary hospital admissions in mild to moderate TBI. Additionally, we aimed to develop adapted criteria for kids based on injury patterns that may be unique to children with TBI. 2. Methods 2.1. Study design and population Following exemption from the Institutional Review Board (#00176573), we performed a retrospective review of pediatric (<18 years old) trauma patients with TBI seen at a level 1 pediatric trauma center from January 2018 to April 2024. Initial data was queried from the trauma registry, followed by a comprehensive chart review. All patients with a blunt mechanism of injury and a head abbreviated injury scale (AIS) score /C21 1 were captured. Pa- tients were included in the analysis if there was a radiographic abnormality on head CT, de fined as skull fracture and/or traumatic ICH. While not in the original BIG, isolated skull fracture (ISF) pa- tients were included. Patients with a penetrating mechanism, no radiographically evident TBI, or /C21 18 years were excluded. 2.2. Data points Trained personnel abstracted data including age, sex, race, ethnicity, mechanism, transfer status, AIS scores, injury severity score (ISS), initial emergency department (ED) Glasgow Coma Scale (GCS), pupillary response, intoxication, and preinjury anti- coagulation. Patients with AIS scores /C20 1 in all non-head body re- gions were classi fied as isolated TBI. Medical records and CT head images for TBI-related data were reviewed, including ICH type, measurement in millimeters or sulci, and presence of skull fracture. Skull fracture was classi fied as minor (non-displaced or minimally displaced /C20 1 cranium width) or displaced ( >1 cranium width). Extra-axial hemorrhage was considered an epidural hematoma (EDH) if associated with skull fracture and subdural hematoma (SDH) if without. All CTs were initially read by and when necessary reviewed by board-certi fied pediatric radiologists. 2.3. Guideline application We applied BIG (Table 1) to the study cohort, with the following working definitions [8]: abnormal neurologic exam was de fined as ED GCS score /C20 12, abnormal pupillary response, and focal neuro- logic examination. Preinjury anticoagulation included warfarin, direct oral anticoagulants, aspirin, or clopidogrel. The presence of multiple ICH types was categorized as BIG 3. Progression on RHCT was defined as either the appearance of new ICH or an increase in the size of initial ICH. Operative neurosurgical procedures included craniotomy, craniectomy, and/or external ventricular drain (EVD) or intracranial pressure monitor placement. We utilized a modi fica- tion for subarachnoid hemorrhage (SAH) similar to previous re- ports to enhance reproducibility [14], which further defined “tra ce ” as /C20 3 sulci total, “localized” as >3 sulci and single hemisphere, and “scattered” as >3 sulci and bi-hemispheric. Patients were first strati fied by BIG score to determine the ef- ficacy of categorization and management. The primary outcome was neurosurgical intervention in BIG 1 and 2 patients and the misclassification rate of the guideline. Secondary outcomes included RHCT, NSC, hospital admission, hospital transfer, length of stay (LOS), and return-to-ED visits. In addition, we considered how management of minor skull fracture with and without ICH was addressed by BIG, accounting for the increased incidence of skull fracture observed in children compared to their adult counterparts. Based on the initial performance of BIG and with input from pe- diatric neurosurgery, pediatric surgery, trauma surgery, and pedi- atric radiology, classi fication modi fications were devised to generate Big for Kids, or kBIG ( Table 3). 2.4. Statistical analysis Data were analyzed using R software (version 4.4.1, R Foundation for Statistical Computing, Vienna, Austria). Descriptive statistics were performed on patient demographics, baseline characteristics, and clinical outcomes. Continuous variables were reported as medians with interquartile ranges, while categorical variables were presented as frequency and percentage. To test the hypothesis and evaluate whether the guideline could classify low-risk patients as not requiring hospitalization, the negative predictive value (NPV) and the misclassification rate were calculated based on need for neurosur- gical intervention. Misclassification was defined as cases where pa- tients categorized as low- or moderate-risk required neurosurgical intervention, potentially indicating inadequacies in the risk Table 1 Original brain injury guidelines (BIG). Variables BIG 1 BIG 2 BIG 3 Loss of consciousness Yes/No Yes/No Yes/No Neurologic examination Normal Normal Abnormal Intoxication No Yes/No Yes/No Anticoagulation/antiplatelet No No Yes Skull fracture No Non-displaced Displaced Epidural hematoma /C20 4m m 5e7m m /C21 8m m Subdural hematoma /C20 4m m 5e7m m /C21 8m m Intraparenchymal hematoma /C20 4 mm, 1 location 3e7 mm, 2 locations /C21 8 mm, multiple locations Subarachnoid hemorrhage Trace Localized Scattered Intraventricular hemorrhage No No Yes Therapeutic plan Hospitalization Observation (6 h) Yes Yes Repeat head computed tomography No No Yes Neurosurgery consultation No No Yes A.B. Kay, S.L. Glasgow, A.M. Kahan et al. / Journal of Pediatric Surgery 60 (2025) 1623722
stratification process. True negatives were defined as BIG 1 patients who did not require neurosurgery, while false negatives were pa- tients incorrectly classi fied as low-risk because of a subsequent neurosurgical intervention. The misclassification rate was calculated as false negatives divided by total low-risk sample size. We repeated the analysis on moderate-risk patients (BIG 2) and finally on kBIG, with low-risk defined as kBIG 0 and kBIG 1 and moderate-risk as kBIG 2. We implemented the Strengthening the Reporting of Observa- tional Studies in Epidemiology (STROBE) guidelines for reporting observational studies. 3. Results 3.1. Overall characteristics A total of 5271 records were identi fied with TBI and 3377 were excluded due to the absence of positive CT findings. Therefore, 1894 patients were included in the analysis. The overall study population was on average 5.5 years old, 61 % male, 78 % Caucasian, and 72 % non-Hispanic. The mean ISS was 10, and the most common mechanism was fall (49 %), followed by motor vehicle collision Table 2 Patient characteristics and outcomes based on the original BIG. All (N ¼ 1455) BIG 1 (n ¼ 197) BIG 2 (n ¼ 566) BIG 3 (n ¼ 692) Patient characteristics ED GCS 3e8 268 (18) 0 (0) 0 (0) 268 (39) 9e12 73 (5) 0 (0) 0 (0) 73 (11) 13e14 177 (12) 19 (10) 78 (14) 80 (12) 15 937 (64) 178 (90) 488 (86) 271 (39) Pupillary response Both 1397 (96) 196 (99) 566 (100) 635 (92) One 17 (1) 0 (0) 0 (0) 17 (2) None 39 (3) 0 (0) 0 (0) 39 (6) Intoxication Yes 2 (0) 0 (0) 1 (0) 1 (0) No 32 (2) 3 (2) 4 (1) 25 (4) Anticoagulation Yes 2 (0) 0 (0) 0 (0) 2 (0) No 1451 (100) 197 (100) 566 (100) 688 (99) Skull fracture Yes 914 (63) 0 (0) 484 (86) 430 (62) No 541 (37) 197 (100) 82 (14) 262 (38) Intracerebral hematoma/hemorrhage Epidural hematoma 401 (28) 11 (6) 163 (29) 227 (33) Subdural hematoma 780 (54) 119 (60) 311 (55) 350 (51) Intraparenchymal hematoma 154 (11) 12 (6) 22 (4) 120 (17) Subarachnoid hemorrhage 318 (22) 55 (28) 70 (12) 193 (28) Intraventricular hemorrhage 6 (0) 0 (0) 0 (0) 6 (1) Isolated TBI 1118 (77) 160 (81) 497 (88) 461 (67) ISS 10 (9, 17) 9 (6, 11) 10 (9, 12) 17 (10, 26) Transferred Yes 1068 (73) 145 (74) 405 (72) 518 (75) No 387 (27) 52 (26) 161 (28) 174 (25) Patient outcomes Repeat head CT Yes 688 (47) 42 (21) 213 (38) 433 (63) No 767 (53) 155 (79) 353 (62) 259 (37) Progression on repeat head CT Yes 177 (12) 1 (1) 30 (5) 146 (21) No 511 (35) 41 (21) 183 (32) 287 (41) Neurosurgical consultation Yes 1396 (96) 190 (96) 536 (95) 670 (97) No 59 (4) 7 (4) 30 (5) 22 (3) OR neurosurgery Yes 255 (18) 0 (0) 8 (1) 247 (36) No 1200 (82) 197 (100) 558 (99) 445 (64) Neurosurgical intervention Craniectomy 37 (3) 0 (0) 0 (0) 37 (5) Craniotomy 123 (8) 0 (0) 4 (1) 119 (17) ICP monitor (EVD, BOLT) 125 (9) 0 (0) 4 (1) 121 (17) Admission ICU 601 (41) 17 (9) 90 (16) 494 (71) Floor 854 (59) 175 (89) 466 (82) 193 (28) None (discharged from ED) 20 (1) 5 (3) 10 (2) 5 (1) Ventilator day 0 (0, 0) 0 (0, 0) 0 (0, 0) 0 (0, 2) LOS, day 2 (1, 5) 1 (1, 2) 1 (1, 3) 4 (2, 10) ED bounce-back Yes 14 (1) 3 (2) 6 (1) 5 (1) No 1441 (99) 194 (98) 560 (99) 687 (99) All data are presented as frequency (%) except that ISS, Ventilator day and LOS are presented as median (interquartile range). ED: emergency department; GCS: Glasgow coma scale; ISS: injury severity score; CT: computed tomography; OR: operating room; ICP: intracranial pressure; EVD: external ventricular drainage; BOLT: subdural screw; ICU: intensive care unit; LOS: length of hospital stay. A.B. Kay, S.L. Glasgow, A.M. Kahan et al. / Journal of Pediatric Surgery 60 (2025) 162372 3
(26 %). The majority (73 %) were transferred in from referring facilities. 3.2. BIG application The BIG score was applied to 1455 patients, excluding 439 pa- tients with ISF ( Table 2). No BIG 1 patient underwent neurosurgical intervention. There were 8 (1.4 %) BIG 2 patients who required neurosurgical intervention. Two patients with EDH and 5 with SDH underwent craniotomy. One patient with SAH had an EVD placed. All SDH patients had a non-accidental trauma (NAT) mechanism. No low-risk (BIG 1) patients required intervention, resulting in a NPV of 100 % and a misclassi fication rate of 0 %. The BIG missed 8 moderate-risk (BIG 2) patients, with a NPV of 98.6 % and a misclassification rate of 1.4 %. Evaluating skull fractures, 802 (55 %) patients had minor frac- tures with concomitant ICH. Of those, 375 (47 %) were GCS 13 e15 with diminutive ICH that otherwise met BIG 1 criteria and were categorized as BIG 2 solely for skull fracture. None underwent neurosurgical intervention. Isolated TBI was present in 227 of the skull fracture patients. All had a LOS of /C20 2 days and 180 (79 %) had a LOS /C20 1 day. The majority, 164 (72 %), did not get additional imaging. There were 255 patients who required neurosurgical interven- tion, 247 (97 %) of which were BIG 3. Looking at hemorrhage type, there were 104 EDH patients (26 % of all EDH), and 72 SDH patients (9 % of all SDH) that required intervention. Of the 1 10 patients with displaced skull fractures, 45 (41 %) had an operative intervention. In the 122 patients with suspected or con firmed NAT, 30 (25 %) required operative intervention. Initial GCS 13e15 was documented in 109 (43 %) of the operative cases. 3.3. kBIG derivation Based on these results, six major modi fications were made to derive kBIG ( Table 3, Fig. 1): 1. A kBIG 0 category was created for minor ISF in neurologically normal patients, with similar management recommendations to kBIG 1. 2. Minor skull fracture was downgraded to kBIG 1. 3. Small EDH de fined as < 8 mm was upgraded to kBIG 2. 4. Suspected or con firmed NAT was added as kBIG 3. 5. The de finition of normal neurologic exam required for kBIG 0 and 1 was changed to GCS 15 or baseline, and without requiring admission for headache or persistent nausea and emesis. 6. The 6-h observation period was removed from kBIG 0 and 1 management pathways. 3.4. kBIG application After derivation, kBIG was applied to the entire study population with the inclusion of ISF (n ¼ 1894) (Table 4). No kBIG 0 or 1 pa- tients required neurosurgical intervention. There were three kBIG 2 patients who required neurosurgical intervention. Two patients with EDH and minor skull fracture underwent craniotomy: one was a 4-year-old with a 5 mm EDH that progressed to 12 mm on hos- pital day 1 and the other was a 10-year-old with 7 mm EDH that progressed to 1 1 mm by hospital day 4. The third operative case was an 8-year-old with a nondisplaced ISF involving the temporal bone with pneumocephalus. The neurologic exam included GCS 14 and intractable nausea and vomiting. A hospital day 1 event note re- ported a syncopal episode and fall, which preceded a neurologic decline. RHCT demonstrated a new 2 cm EDH requiring emergent craniotomy. The kBIG accurately identi fied low-risk patients (kBIG 0 and 1) in whom resources can be safely minimized, with a NPV of 100 % and a misclassi fication rate of 0 %. Compared to BIG, kBIG reduced the misclassi fication rate for moderate-risk (kBIG 2) patients from 1.4 % to 0.8 %. 3.5. Resource utilization The rates of RHCT and NSC were 41 % and 88 %, respectively. The application of kBIG could have potentially avoided 282 RHCT, or a 36 % total reduction. Across kBIG 0, 1 and 2, 873 patients received neurosurgical consultation, representing a potential 52 % total reduction. To evaluate potentially preventable admissions and transfers, 1467 (77 %) isolated TBI patients were considered. Of those, 558 (38 %) were kBIG 0 e1 patients, 45 of whom were dis- charged from the ED. Thus, 513 patients could have potentially avoided admission, resulting in a 36 % admissions reduction. Of those, 367 could have potentially avoided transfer, representing a 26 % potential transfer reduction. 4. Discussion We applied BIG to a pediatric TBI cohort to evaluate its effi cacy. Application of the original BIG accurately predicted the need for neurosurgical intervention in BIG 1 patients. BIG 2 categorization resulted in 8 patients who required intervention, with a misclas- sification rate of 1.4 %, higher than reported in adults [ 9]. The original BIG performed well in our pediatric cohort but required modifications to address specific gaps. These adjustments led to the development of the Brain Injury Guidelines for Kids, or kBIG. With neurosurgical intervention as the primary safety outcome, ensuring Table 3 Proposed brain injury guidelines for kids (kBIG). Variables kBIG 0 kBIG 1 kBIG 2 kBIG 3 Mechanism of injury Blunt TBI excluding NAT Blunt TBI excluding NAT Blunt TBI excluding NAT Blunt TBI including NAT Initial ED GCS 15 15 13 e14 /C20 12 Intoxication No No Yes Yes/No Anticoagulation/antiplatelet No No No Yes Skull fracture Nondisplaced/mildly displaced Nondisplaced/mildly displaced Nondisplaced/mildly displaced Displaced (with or without bleed) Epidural hematoma No No <8m m /C21 8m m Subdural hematoma No /C20 4m m 5e7m m /C21 8m m Intraparenchymal hematoma No /C20 4m m 5e7m m /C21 8 mm, multiple locations Subarachnoid hemorrhage No /C20 3 sulci >3 sulci, single hemisphere >3 sulci, bi-hemispheric Intraventricular hemorrhage No No No Yes Therapeutic plan Hospitalization No No Yes Yes Repeat head CT No No No Per neurosurgery request Neurosurgery consultation No No No Yes TBI: traumatic brain injury; NAT: Non-accidental trauma; ED: emergency department; GCS: Glasgow coma scale; CT: computed tomography. A.B. Kay, S.L. Glasgow, A.M. Kahan et al. / Journal of Pediatric Surgery 60 (2025) 1623724
that no low-risk patients and limited moderate-risk patients required neurosurgical intervention was critical. In the kBIG deri- vation process, six signi ficant changes were made. The kBIG maintained safety for low-risk patients with a misclassi fication rate of 0 %, while improving the misclassi fication rate of moderate-risk patients from 1.4 % to 0.8 %. The application of kBIG would have Pediatric Patient with Traumatic Brain Injury (TBI) Patients with >1 intracranial injury should be automatically classified as kBIG3 Low risk isolated skull fracture and concern for child physical abuse: please call safe and healthy families. Otherwise treat as kBIG 0. IVH: Intraventricular Hemorrhage EDH: Epidural Hematoma SDH: Subdural Hematoma ICH: Intracranial Hemorrhage IPH: Intraparenchymal Hemorrhage SAH: Subarachnoid Hemorrhage Abbreviations Are ANY of the following present? Radiographic Findings IVH EDH ≥8mm SDH ≥8mm Displaced skull fracture > 1 cranium width Clinical Findings Anticoagulation or antiplatelet Non-accidental trauma (NAT) mechanism IPH ≥ 8mm or multiple SAH >3 sulci AND bi-hemispheric GCS ≤12 YES YES kBIG 3 High Risk Transfer / Admit to pediatric trauma center with neurosurgery consultation available Intoxication Are ANY of the following present? Radiographic Findings Clinical Findings Persistent nausea, vomiting EDH <8mm SAH>3 sulci, single hemisphere SDH 4.1 – 7.9mm IPH 4.1 – 7.9mm Non-displaced skull fracture or mildly displaced < 1 cranium width GCS 13-14 YES Confirm Intracranial Hemorrhage Size Radiographic Findings Clinical Findings GCS 15SDH ≤4mm IPH ≤4mm SAH ≤3 sulci (total) Non-displaced skull fracture or mildly displaced < 1 cranium width YES Isolated Skull Fracture Radiographic Findings Clinical Findings GCS 15No ICH Non-displaced skull fracture or mildly displaced < 1 cranium width If worsening: neuro exam If worsening: neuro exam If worsening: neuro exam Medium Risk Transfer / Admit to pediatric trauma center with neurosurgery consultation available NO neurosurgical consultation NO repeat head CT GCS 15 at discharge kBIG 2 Low Risk NO repeat head CT NO neurosurgical consultation Discharge from Emergency Department kBIG 1 Low risk isolated skull fracture NO repeat head CT NO neurosurgical consultation Discharge from Emergency Department kBIG 0 Fig. 1. The Brain Injury Guidelines for Kids (kBIG) clinical flow diagram is designed to facilitate scoring and management of children with traumatic brain injury. For kBIG 0 e2 patients, any decline in neurologic exam from presentation escalates the patients to kBIG 3 category. GCS, Glasgow coma scale; IVH, intraventricular hemorrhage; EDH, epidural hematoma; SDH, subdural hematoma; IPH, intraparenchymal hematoma; SAH, subarachnoid hemor- rhage; ICH, intracranial hemorrhage; FU, follow-up. A.B. Kay, S.L. Glasgow, A.M. Kahan et al. / Journal of Pediatric Surgery 60 (2025) 162372 5
saved significant resources in the form of RHCT, NSC, and hospital admissions in pediatric TBI de fined as low-to moderate-risk while protecting those at higher risk for decline. This guideline may also serve as a model to direct location of care and transport decisions at the trauma system-level. Previous evaluations of BIG in pediatric TBI have demonstrated efficacy and a potential for reduced resource utilization [ 16e21]. Others have proposed pediatric-specifi c modi fications, including the addition of minor skull fracture to BIG 1 criteria and the up- grade of small EDH to BIG 2 [ 17e19]. One group implemented a pediatric-specific TBI guideline in 139 patients and showed improvement in all outcomes, including reduced RHCT, NSC, and transfers [19]. These reports are limited by small sample sizes and retrospective single-center methodology. The kBIG score includes two changes related to skull fracture given its high incidence in pediatric patients compared to adults [ 9] and based on previous work evaluating management strategies similar to BIG 1 [ 6,17,19,23]. First, minor ISF without ICH was added as its own category, kBIG 0. ISF was common, representing 23 % of the entire cohort, and evidence suggests these kids, if neurologi- cally normal, can be managed with limited resources [ 18,19,22,23]. Second, we classi fied minor skull fracture with ICH as low-risk Table 4 Patient characteristics and outcomes based on the proposed Brain Injury Guidelines for Kids (kBIG). All (N ¼ 1894) kBIG 0 (n ¼ 318) kBIG 1 (n ¼ 336) kBIG 2 (n ¼ 393) kBIG 3 (n ¼ 847) Patient characteristics ED GCS 3e8 297 (16) 0 (0) 0 (0) 0 (0) 297 (35) 9e12 81 (4) 0 (0) 0 (0) 0 (0) 81 (10) 13e14 218 (12) 0 (0) 0 (0) 125 (32) 93 (11) 15 1298 (69) 318 (100) 336 (100) 268 (68) 376 (44) Pupillary response Both 1396 (74) 0 (0) 335 (100) 355 (90) 706 (83) One 18 (1) 0 (0) 0 (0) 0 (0) 18 (2) None 40 (2) 0 (0) 0 (0) 0 (0) 40 (5) Intoxication Yes 2 (0) 0 (0) 0 (0) 1 (0) 1 (0) No 32 (2) 0 (0) 2 (1) 4 (1) 26 (3) Anticoagulation Yes 2 (0) 0 (0) 0 (0) 0 (0) 2 (0) No 1890 (100) 318 (100) 336 (100) 393 (100) 843 (100) Skull fracture Yes 1355 (72) 318 (100) 184 (55) 328 (83) 525 (62) No 539 (28) 0 (0) 152 (45) 65 (17) 322 (38) Intracerebral hematoma/hemorrhage Epidural hematoma 401 (21) 0 (0) 0 (0) 150 (38) 251 (30) Subdural hematoma 780 (41) 0 (0) 242 (72) 140 (35) 398 (47) Intraparenchymal hematoma 153 (8) 0 (0) 13 (4) 21 (5) 119 (14) Subarachnoid hemorrhage 317 (17) 0 (0) 80 (24) 44 (11) 193 (23) Intraventricular hemorrhage 5 (0) 0 (0) 0 (0) 0 (0) 5 (1) Isolated TBI 1467 (77) 271 (85) 287 (85) 335 (85) 575 (68) ISS 10 (9, 17) 9 (5, 10) 9 (9, 10) 10 (9, 14) 17 (10, 25) Transferred Yes 1388 (73) 239 (75) 240 (71) 281 (72) 628 (74) No 506 (27) 79 (25) 96 (29) 112 (28) 219 (26) Patient outcomes Repeat head CT Yes 780 (41) 50 (16) 72 (21) 160 (41) 498 (59) No 1114 (59) 268 (84) 264 (79) 233 (59) 349 (41) Progression on repeat head CT Yes 182 (10) 1 (0) 1 (0) 29 (7) 151 (18) No 599 (32) 49 (15) 71 (21) 131 (33) 348 (41) Neurosurgical consultation Yes 1676 (88) 193 (61) 318 (95) 362 (92) 803 (95) No 218 (12) 125 (39) 18 (5) 31 (8) 44 (5) OR neurosurgery Yes 277 (15) 0 (0) 0 (0) 3 (1) 274 (32) No 1617 (85) 318 (100) 336 (100) 390 (99) 573 (68) Neurosurgical intervention Craniectomy 38 (2) 0 (0) 0 (0) 0 (0) 38 (4) Craniotomy 135 (7) 0 (0) 0 (0) 3 (1) 132 (16) ICP monitor (EVD, BOLT) 131 (7) 0 (0) 0 (0) 0 (0) 131 (15) Admission ICU 652 (34) 6 (2) 22 (7) 72 (18) 552 (65) Floor 1181 (62) 274 (86) 307 (91) 315 (80) 285 (34) None (discharged from ED) 61 (3) 38 (12) 7 (2) 6 (2) 10 (1) Ventilator day 0 (0, 0) 0 (0, 0) 0 (0, 0) 0 (0, 0) 0 (0, 1) LOS, day 2 (1, 4) 1 (1, 1) 1 (1, 2) 1 (1, 3) 4 (2, 9) ED bounce-back Yes 20 (1) 5 (2) 5 (1) 3 (1) 7 (1) No 1874 (99) 313 (98) 331 (99) 390 (99) 840 (99) All data are presented as frequency (%) except that ISS, Ventilator day and LOS are presented as median (interquartile range). ED: emergency department; GCS: Glasgow coma scale; ISS: injury severity score; CT: computed tomography; OR: operating room; ICP: intracranial pressure; EVD: external ventricular drainage; BOLT: subdural screw; ICU: intensive care unit; LOS: length of hospital stay. A.B. Kay, S.L. Glasgow, A.M. Kahan et al. / Journal of Pediatric Surgery 60 (2025) 1623726
(kBIG 1) rather than moderate-risk as long as all other BIG 1 criteria were met [ 17]. A signi ficant portion of the study population with minor skull fractures was categorized as BIG 2 solely due to pres- ence of skull fracture. Most patients underwent a hospital LOS of /C20 1 day, with no additional imaging, and none required neurosurgical intervention. Based on this and aligning with our previous de fini- tion of potentially avoidable transfer [ 4], we concluded these could be downgraded to low-risk. While modi fications to skull fracture stratification have been supported in pediatric guidelines for TBI [17,19], no evaluation has included both the incorporation of ISF and the risk downgrade of minor skull fracture with concomitant ICH. Our amendments expand the population of pediatric head injuries that can be safely observed and discharged. Of note, displaced skull fracture with or without ICH continues to be categorized as high- risk, given the high potential for operative intervention. Another change was the escalation of small EDH <8 mm kBIG 2. The original BIG defines a subset of small EDH (/C20 4 mm) as low-risk, supporting observation without RHCT, NSC, or hospital admission, which has been demonstrated to be safe in two retrospective pe- diatric reports [ 20,21]. In adults, a modi fied BIG (mBIG) proposed that EDH be categorized as BIG 3 regardless of size based on two BIG 2 EDH patients who required craniotomy [ 14]. Our proposal aligns with another published pediatric modi fication which moved small EDH from the lowest-risk category to moderate-risk. They reported no neurosurgical interventions in this moderate-risk group but poor guideline compliance suggesting discomfort man- aging any EDH without RHCT or NSC [ 18,19]. Similarly, others have cautioned against limiting resource utilization in small to moderate EDH [21]. While no BIG 1 size EDH patients in the study population underwent neurosurgical intervention, two kBIG 2 patients did for moderate sized EDH. Furthermore, with the movement of minor skull fracture to kBIG 1 we decided to adopt a more conservative approach with EDH until larger numbers are attained. More data is needed to better elucidate the risk of minor EDH with and without skull fracture in children. Injury mechanism has not been addressed in previous iterations of the BIG. It was imperative that these guidelines would not inadvertently increase cases of missed child physical abuse or NAT. In the study population, 25 % of NAT underwent neurosurgical intervention, including all the patients who met original BIG 2 criteria and required operations that were not EDH. NAT often in- volves repeated episodes resulting in acute on chronic head injury and requires neurosurgical attention. Given the concern for child physical abuse in pediatrics, we recommend the highest level of resources for any suspected case. Because of these reasons, sus- pected and confirmed NAT was added as a discrete kBIG 3 criterion. Finally, we propose a more conservative de finition for normal neurologic exam in children than has been utilized in adults [ 8,14], whereby GCS 15 is required for low-risk TBI (kBIG 0 e1). GCS 13e14 mandates kBIG 2 and thus admission, as does intractable headache, nausea, and vomiting [ 18,19]. GCS is widely used to assess con- sciousness, but significant variability exists in its interpretation and GCS is particularly unreliable in young trauma patients [ 24,25]. Furthermore, GCS 13 e15 does not preclude the need for neuro- surgical intervention [ 26], as 43 % of the operative cases in our cohort had an initial GCS of 13e15. There was a single patient in the current study with a nondisplaced temporal bone ISF who declined at 30 h and required a craniotomy for a new EDH. Exam was sig- nificant for initial GCS of 14, and severe nausea and vomiting. While this patient may have been categorized as low-risk, potentially failing the algorithm if discharged home, discharge was not clini- cally appropriate. With no room for error, kBIG is designed to protect kids at risk for decline while allowing those at neurological baseline with minor CT findings to be discharged home safely. Importantly, any change in neurologic exam identi fied during observation escalates care. Requiring neurologic baseline for low- risk categorization, the 6-h observation period was removed from kBIG 0 e1 management because this would not have changed outcomes and may be a barrier to implementation [ 27]. Detailed return precautions must be emphasized. Potentially preventable hospital transfers in pediatric patients with mild TBI are common [3e7]. Uniform guidelines do not exist to support interfacility transfer decisions, and referring providers are often uncomfortable with pediatric trauma patients. We looked at this previously and found that 27 % of pediatric trauma patients transferred to our level 1 pediatric trauma center were considered preventable [ 4]. Evaluating the isolated TBI group in the current study, the application of kBIG could have potentially reduced transfers by 26 % and admissions by 36 %. With generalizability in mind, we intend for this kBIG algorithm to stimulate reevaluation of site and region-speci fic resource utilization for the care of pe- diatric TBI across the country. There are limitations inherent to the retrospective design, where the proposed guideline is based on a historical cohort at a single center and expert opinion. We identi fied radiographic find- ings that warrant future evaluation for their association with neurologic decline or neurosurgical intervention including midline shift, cranial suture diastasis, and pneumocephalus. We may have missed return-to-ED patients outside our system, but our system does capture greater than 70 % of patients in the state. Patients with low-risk TBI who were discharged from the ED may not have met trauma registry inclusion and thus may have been unaccounted for. Importantly, this study does not address the long-term sequelae of TBI. 5. Conclusions BIG functions well when applied to injured children, but important limitations were identi fied. We propose an enhanced version of BIG for the management of pediatric TBI called BIG for Kids, or kBIG. The kBIG was conservatively designed primarily to ensure safety and secondarily to reduce unnecessary radiation exposure and resource utilization with a patient-centered approach. While we intend to adopt this locally, multicenter and prospective kBIG validation efforts are underway to support broader implementation. Conflict of interest There are no financial disclosures from any of the authors. All JTACS disclosure statement forms have been supplied and provided as supplemental digital content. Acknowledgements W e would like to acknowledge our trauma program manager Kacey L Barnes, MSN for her help with data and our research coordinator Kezlyn E Larsen, BS for handling all regulation. References [1] Cunningham RM, Walton MA, Carter PM. The major causes of death in chil- dren and adolescents in the United States. N Engl J Med 2018 Dec 20;379(25): 2478. 2475 . [2] Kuppermann N, Holmes JF, Dayan PS, et al. Identi fication of children at very low risk of clinically-important brain injuries after head trauma: a prospective cohort study. Lancet 2009 Oct 3;374(9696):1160 e70. [3] Sorensen MJ, von Recklinghausen FM, Fulton G, Burchard KW. Secondary overtriage: the burden of unnecessary interfacility transfers in a rural trauma system. JAMA Surg 2013 Aug;148(8):763 e8. A.B. Kay, S.L. Glasgow, A.M. Kahan et al. / Journal of Pediatric Surgery 60 (2025) 162372 7
[4] Fenton SJ, Lee JH, Stevens AM, et al. Preventable transfers in pediatric trauma: a 10-year experience at a level I pediatric trauma center. J Pediatr Surg 2016 Apr;51(4):645e8. [5] Mowbray F, Arora R, Shukla M, Shihan H, Kannikeswaran N. Factors associated with avoidable interhospital transfers for children with a minor head injury. Am J Emerg Med 2021 Jul;45:208 e12. [6] Fenton SJ, Swendiman RA, Eyre M, Larsen K, Russell KW. The Utah Pediatric Trauma Network, a statewide pediatric trauma collaborative can safely help nonpediatric hospitals admit children with mild traumatic brain injury. J Trauma Acute Care Surg 2023 Sep 1;95(3):376 e82. [7] Haag K, Duke D, Piatt J. Overtriage of transfers to the pediatric trauma center: the importance of minor head injury. J Neurosurg Pediatr 2024 Apr 19;34(10): 1e8. [8] Joseph B, Friese RS, Sadoun M, et al. The BIG (brain injury guidelines) project: defining the management of traumatic brain injury by acute care surgeons. J Trauma Acute Care Surg 2014;76(4):965 e9. [9] Joseph B, Obaid O, Dultz L, et al. Validating the brain injury guidelines: results of an American association for the surgery of trauma prospective multi- institutional trial. J Trauma Acute Care Surg 2022 August;93(2):157 e65. [10] Capron GK, Voights MB, Moore HR, Wall DB. Not every trauma patient with radiographic head injury requires transfer for neurosurgical evaluation: application of the brain injury guidelines to patients transferred to a level 1 trauma center. Am J Surg 2017 Dec;214(6):1182 e5. [11] Martin GE, Carroll CP, Plummer ZJ, et al. Safety and ef ficacy of brain injury guidelines at a Level III trauma center. J Trauma Acute Care Surg 2018;84(3): 483e9. [12] Tourigny JN, Boucher V, Paquet V, et al. External validation of the updated Brain Injury Guidelines for complicated mild traumatic brain injuries: a retrospective cohort study. J Neurosurg 2022 Jan;25:1 e7. [13] Kay AB, Malone SA, Bledsoe JR, Majercik S, Morris DS. First steps toward a BIG change: a pilot study to implement the Brain Injury Guidelines across a 24- hospital system. Am J Surg 2023 Dec;226(6):845 e50. [14] Khan AD, Elseth AJ, Brosius JA, Moskowitz A, et al. Multicenter assessment of the Brain Injury Guidelines and a proposal of guideline modi fications. Trauma Surg Acute Care Open 2020 May 28;5(1):e000483 . [15] Khan AD, Lee J, Galicia K, Billings JD, Dobaria V, Patel PP, McIntyre RC, Gonzalez RP, Schroeppel TJ. A multicenter validation of the modi fied brain injury guidelines: are they safe and effective? J Trauma Acute Care Surg 2022 Jul;93(1):106e12. [16] Azim A, Jehan FS, Rhee P, et al. Big for small: validating brain injury guidelines in pediatric traumatic brain injury. J Trauma Acute Care Surg 2017 Dec;83(6): 1200e4. [17] Schwartz J, Crandall M, Hsu A, Tepas JJ, Joseph B, Yorkgitis BK. Applying pe- diatric brain injury guidelines at a level I adult/pediatric safety-net trauma center. J Surg Res 2020 Nov;255:106 e10. [18] McNickle AG, Jones SA, Yacoub M, et al. BIG Kids: application of a modi fied brain injury guideline in a pediatric trauma center. J Pediatr Surg 2023 Mar;58(3):552e7. [19] McNickle AG, Bailey D, Yacoub M, Chang S, Fraser DR. A pediatric brain injury guideline allows safe management of traumatic brain injuries by trauma surgeons. J Pediatr Surg 2024 Nov;59(11):161644 . [20] Zeller S, Khan A, Chung JY, et al. Application of brain injury guidelines at a pediatric level 1 trauma center predicts reliability, safety, and improved resource utilization. Childs Nerv Syst 2024 Sep;40(9):2769 e74. [21] Yu N, Castillo J, Kohler JE, et al. Validating the brain injury guidelines in a pediatric population with mild traumatic brain injury and intracranial injury at a level I trauma center. J Neurotrauma 2024 Nov 7. https://doi.org/10.1089/ neu.2024.0130 [Online ahead of print]. [22] Powell EC, Atabaki SM, Wootton-Gorges S, et al. Isolated linear skull fractures in children with blunt head trauma. Pediatrics 2015 Apr;135(4):e851 e7. [23] Kommaraju K, Haynes JH, Ritter AM. Evaluating the role of a neurosurgery consultation in management of pediatric isolated linear skull fractures. Pediatr Neurosurg 2019;54(1):21 e7. [24] DiBrito SD, Cerullo M, Goldstein SD, Ziegfeld S, Stewart D, Nasr IW. Reliability of Glasgow coma score in pediatric trauma patients. J Pediatr Surg 2018 Sep;53(9):1789e94. [25] Drews JD, Shi J, Papandria D, Wheeler KK, Sribnick EA, Thakkar RK. Prehospital versus trauma center Glasgow coma scale in pediatric traumatic brain injury patients. J Surg Res 2019 Sep;241:112 e8. [26] Bellal J, Pandit V, Aziz H, et al. Mild traumatic brain injury de fined by Glasgow Coma Scale: is it really mild? Brain Inj 2015;29(1):11 e6. [27] Lee JS, Billings J, McIntyre, et al. To observe or not to observe: evaluation of the modified brain injury guideline management of small volume intracranial hemorrhage. Am J Surg 2023 Dec;226(6):808 e12. A.B. Kay, S.L. Glasgow, A.M. Kahan et al. / Journal of Pediatric Surgery 60 (2025) 1623728