Effect of Routine Child Physical Abuse Screening Tool on Emergency Department Efficiency.

OBJECTIVES: Physical abuse is a significant cause of morbidity and mortality for children. Routine screening by emergency nurses has been proposed to improve recognition, but the effect on emergency department (ED) workflow has not yet been assessed. We sought to evaluate the feasibility of routine screening and its effect on length of stay in a network of general EDs. METHODS: A 2-question child physical abuse screening tool was deployed for children <6 years old who presented for care in a system of 27 general EDs. Data were compared for the 6 months before and after screening was deployed (4/1/2019-10/2/2019 vs 10/3/2019-3/31/2020). The main outcome was ED length of stay in minutes. RESULTS: There were 14,133 eligible visits in the prescreening period and 16,993 in the screening period. Screening was completed for 13,404 visits (78.9%), with 116 (0.7%) screening positive. The mean ED length of stay was not significantly different in the prescreening (95.9 minutes) and screening periods (95.2 minutes; difference, 0.7 minutes; 95% CI, -1.5, 2.8). Among those who screened positive, 29% were reported to child protective services. On multivariable analysis, implementation of the screening tool did not impact overall ED length of stay. There were no significant differences in resource utilization between the prescreening and screening periods. CONCLUSIONS: Routine screening identifies children at high risk of physical abuse without increasing ED length of stay or resource utilization. Next steps will include determining rates of subsequent serious physical abuse in children with or without routine screening.

Bone Anchor Fixation in the Repair of Blunt Traumatic Abdominal Wall Hernias: A Western Trauma Association Multicenter Study.

BACKGROUND: Blunt traumatic abdominal wall hernias (TAWHs) are rare but require a variety of operative techniques to repair including bone anchor fixation (BAF) when tissue tears off bony structures. This study aimed to provide a descriptive analysis of BAF technique for blunt TAWH repair. Bone anchor fixation and no BAF repairs were compared, hypothesizing increased hernia recurrence with BAF repair. METHODS: A secondary analysis of the WTA blunt TAWH multicenter study was performed including all patients who underwent repair of their TAWH. Patients with BAF were compared to those with no BAF with bivariate analyses. RESULTS: 176 patients underwent repair of their TAWH with 41 (23.3%) undergoing BAF. 26 (63.4%) patients had tissue fixed to bone, with 7 of those reinforced with mesh. The remaining 15 (36.6%) patients had bridging mesh anchored to bone. The BAF group had a similar age, sex, body mass index, and injury severity score compared to the no BAF group. The time to repair (1 vs 1 days, P = .158), rate of hernia recurrence (9.8% vs 12.7%, P = .786), and surgical site infection (SSI) (12.5% vs 15.6%, P = .823) were all similar between cohorts. CONCLUSIONS: This largest series to date found nearly one-quarter of TAWH repairs required BAF. Bone anchor fixation repairs had a similar rate of hernia recurrence and SSI compared to no BAF repairs, suggesting this is a reasonable option for repair of TAWH. However, future prospective studies are needed to compare specific BAF techniques and evaluate long-term outcomes including patient-centered outcomes such as pain and quality of life.

A western trauma association multicenter comparison of mesh versus non-mesh repair of blunt traumatic abdominal wall hernias.

BACKGROUND: Blunt traumatic abdominal wall hernias (TAWH) occur in <1 % of trauma patients. Optimal repair techniques, such as mesh reinforcement, have not been studied in detail. We hypothesize that mesh use will be associated with increased surgical site infections (SSI) and not improve hernia recurrence. MATERIALS AND METHODS: A secondary analysis of the Western Trauma Association blunt TAWH multicenter study was performed. Patients who underwent TAWH repair during initial hospitalization (1/2012-12/2018) were included. Mesh repair patients were compared to primary repair patients (non-mesh). A logistic regression was conducted to assess risk factors for SSI. RESULTS: 157 patients underwent TAWH repair during index hospitalization with 51 (32.5 %) having mesh repair: 24 (45.3 %) synthetic and 29 (54.7 %) biologic. Mesh patients were more commonly smokers (43.1 % vs. 22.9 %, p = 0.016) and had a larger defect size (10 vs. 6 cm, p = 0.003). Mesh patients had a higher rate of SSI (25.5 % vs. 9.5 %, p = 0.016) compared to non-mesh patients, but a similar rate of recurrence (13.7 % vs. 10.5%, p = 0.742), hospital length of stay (LOS), and mortality. Mesh use (OR 3.66) and higher ISS (OR 1.06) were significant risk factors for SSI in a multivariable model. CONCLUSION: Mesh was used more frequently in flank TAWH and those with a larger defect size. Mesh use was associated with a higher incidence and risk of SSI but did not reduce the risk of hernia recurrence. When repairing TAWH mesh should be employed judiciously, and prospective randomized studies are needed to identify clear indications for mesh use in TAWH.

Risk factors for recurrence in blunt traumatic abdominal wall hernias: A secondary analysis of a Western Trauma association multicenter study.

BACKGROUND: Few studies have investigated risk factors for recurrence of blunt traumatic abdominal wall hernias (TAWH). METHODS: Twenty trauma centers identified repaired TAWH from January 2012 to December 2018. Logistic regression was used to investigate risk factors for recurrence. RESULTS: TAWH were repaired in 175 patients with 21 (12.0%) known recurrences. No difference was found in location, defect size, or median time to repair between the recurrence and non-recurrence groups. Mesh use was not protective of recurrence. Female sex, injury severity score (ISS), emergency laparotomy (EL), and bowel resection were associated with hernia recurrence. Bowel resection remained significant in a multivariable model. CONCLUSION: Female sex, ISS, EL, and bowel resection were identified as risk factors for hernia recurrence. Mesh use and time to repair were not associated with recurrence. Surgeons should be mindful of these risk factors but could attempt acute repair in the setting of appropriate physiologic parameters.

Discreet Values of Shock Index Pediatric Age-Adjusted (SIPA) to Predict Intervention in Children With Blunt Organ Injuries.

INTRODUCTION: Elevated shock index pediatric age-adjusted (SIPA) has been shown to be associated with the need for both blood transfusion and intervention in pediatric patients with blunt liver and spleen injuries (BLSI). SIPA has traditionally been used as a binary value, which can be classified as elevated or normal, and this study aimed to assess if discreet values above SIPA cutoffs are associated with an increased probability of blood transfusion and failure of nonoperative management (NOM) in bluntly injured children. MATERIALS AND METHODS: Children aged 1-18 y with any BLSI admitted to a Level-1 pediatric trauma center between 2009 and 2020 were analyzed. Blood transfusion was defined as any transfusion within 24 h of arrival, and failure of NOM was defined as any abdominal operation or angioembolization procedure for hemorrhage control. The probabilities of receiving a blood transfusion or failure of NOM were calculated at different increments of 0.1. RESULTS: There were 493 patients included in the analysis. The odds of requiring blood transfusion increased by 1.67 (95% CI 1.49, 1.90) for each 0.1 unit increase of SIPA (P < 0.001). A similar trend was seen initially for the probability of failure of nonoperative management, but beyond a threshold, increasing values were not associated with failure of NOM. On subanalysis excluding patients with a head injury, increased 0.1 increments were associated with increased odds for both interventions. CONCLUSIONS: Discreet values above age-related SIPA cutoffs are correlated with higher probabilities of blood transfusion in pediatric patients with BLSI and failure of NOM in those without head injury. The use of discreet values may provide clinicians with more granular information about which patients require increased resources upon presentation.

Adolescent Vaping-Associated Trauma in the Western United States.

INTRODUCTION: Electronic cigarettes (e-cigarettes) are handheld, battery-powered vaporizing devices. It is estimated that more than 25% of youth have used these devices recreationally. While vaping-associated lung injury is an increasingly recognized risk, little is known about the risk of traumatic injuries associated with e-cigarette malfunction. METHODS: A multi-institutional retrospective study was performed by querying the electronic health records at nine children's hospitals. Patients who sustained traumatic injuries while vaping from January 2016 through December 2019 were identified. Patient demographics, injury characteristics, and the details of trauma management were reviewed. RESULTS: 15 children sustained traumatic injuries due to e-cigarette explosion. The median age was 17 y (range 13-18). The median injury severity score was 2 (range 1-5). Three patients reported that their injury coincided with their first vaping experience. Ten patients required hospital admission, three of whom required intensive care unit admission. Admitted patients had a median length of stay of 3 d (range 1-6). The injuries sustained were: facial burns (6), loss of multiple teeth (5), thigh and groin burns (5), hand burns (4), ocular burns (4), a radial nerve injury, a facial laceration, and a mandible fracture. Six children required operative intervention, one of whom required multiple operations for a severe hand injury. CONCLUSIONS: In addition to vaping-associated lung injury, vaping-associated traumatic injuries are an emerging and worrisome injury pattern sustained by adolescents in the United States. This report highlights another means by which e-cigarettes pose an increasing risk to a vulnerable youth population.

To give or not to give? Blood for pediatric trauma patients prior to pediatric trauma center arrival.

PURPOSE: This study evaluates the indications, safety and clinical outcomes associated with the administration of blood products prior to arrival at a pediatric trauma center (prePTC). METHODS: Children (≤ 18 years) who were highest level activations at an ACS level 1 pediatric trauma center (PTC) from 2009-2019 were divided into groups:(1) patients with transport times < 4 h who received blood prePTC(preBlood) versus (2) age matched controls with transport times < 4 h who only received crystalloid prePTC (preCrystalloid). RESULTS: Of 1269 trauma activations, 38 met preBlood and 38 met preCrystalloid inclusion criteria. A similar volume of prePTC crystalloid infusion was observed between cohorts (p = 0.311). PreBlood patients evidenced greater hemodynamic instability as demonstrated by higher prePTC pediatric age-adjusted shock index (SIPA) scores. PreBlood patients showed improvement in lactate (p = 0.038) and hemoglobin (p = 0.041) levels upon PTC arrival. PreBlood patients received less crystalloid within 12 h of PTC admission (p = 0.017). No significant differences were found in blood transfusion volumes within six (p = 0.293) and twenty-four (p = 0.575) hours of admission, nor in mortality between cohorts (p = 0.091). CONCLUSIONS: The administration of blood to pediatric trauma patients prior to arrival at a PTC is safe, transiently improves markers of shock, and was not associated with worse outcomes.

Not as simple as ABC: Tools to trigger massive transfusion in pediatric trauma.

BACKGROUND: Early and accurate identification of pediatric trauma patients who will receive massive transfusion (MT) is not well established. We developed the ABCD (defined as penetrating mechanism, positive focused assessment with sonography for trauma, shock index, pediatric age-adjusted [SIPA], lactate, and base deficit [BD]) and BIS scores (defined as a combination of BD, international normalized ratio [INR], and SIPA) and hypothesized that the BIS score would perform best in the ability to predict the need for MT in children. METHODS: Pediatric trauma patients (≤18 years old) admitted to our trauma center between 2008 and 2019 were identified. Using a receiver operator curve, we defined cutoff points for lactate (≥3.2), BD (≤-6.9), and INR (≥1.4). ABCD scores were calculated by combining penetrating mechanism; positive focused assessment with sonography for trauma examination; SIPA; lactate; and BD. BIS scores were calculated by combining BD, INR, and SIPA. The sensitivity, specificity, and accuracy of each score were calculated based on receiving MT. RESULTS: Seven hundred seventy-two patients were included, of which 59 (7.6%) underwent MT. The best predictor of receiving MT was achieved by a BIS score of ≥2 that was 98% sensitive and 23% specific with an area under the curve of 0.81. The ABCD score of ≥2 was 97% sensitive and 20% specific with an area under the curve of 0.77. CONCLUSION: The BIS score, which takes into account derangements in acidosis, coagulopathy, and SIPA, is accurate and easy to perform and can be incorporated into a simple bedside screening tool for triggering MT in pediatric trauma patients. LEVEL OF EVIDENCE: Diagnostic Tests or Criteria, Level IV.

Child physical abuse and COVID-19: Trends from nine pediatric trauma centers.

BACKGROUND: Economic, social, and psychologic stressors are associated with an increased risk for abusive injuries in children. Prolonged physical proximity between adults and children under conditions of severe external stress, such as witnessed during the COVID-19 pandemic with "shelter-in-place orders", may be associated with additional increased risk for child physical abuse. We hypothesized that child physical abuse rates and associated severity of injury would increase during the early months of the pandemic as compared to the prior benchmark period. METHODS: We conducted a nine-center retrospective review of suspected child physical abuse admissions across the Western Pediatric Surgery Research Consortium. Cases were identified for the period of April 1-June 30, 2020 (COVID-19) and compared to the identical period in 2019. We collected patient demographics, injury characteristics, and outcome data. RESULTS: There were no significant differences in child physical abuse cases between the time periods in the consortium as a whole or at individual hospitals. There were no differences between the study periods with regard to patient characteristics, injury types or severity, resource utilization, disposition, or mortality. CONCLUSIONS: Apparent rates of new injuries related to child physical abuse did not increase early in the COVID-19 pandemic. While this may suggest that pediatric physical abuse was not impacted by pandemic restrictions and stresses, it is possible that under-reporting, under-detection, or delays in presentation of abusive injuries increased during the pandemic. Long-term follow-up of subsequent rates and severity of child abuse is needed to assess for unrecognized injuries that may have occurred.

Pre-hospital and emergency department shock index pediatric age-adjusted (SIPA) "cut points" to identify pediatric trauma patients at risk for massive transfusion and/or mortality.

BACKGROUND: Shock index pediatric age-adjusted (SIPA) is a validated measure to identify severely injured children. Previous literature categorized SIPA as normal or elevated, but the relationship between specific SIPA values and outcomes has not been determined. We sought to determine specific SIPA cut points in the pre-hospital and Emergency Department (ED) settings to identify patients at risk for massive transfusion (MT) and/or mortality. METHODS: Patients ≤ 18 years old admitted to our Level I pediatric trauma center following trauma activation were included. Youdin J index was used to define pre-hospital and ED SIPA cut points to identify those at risk of MT and/or mortality for the following age groups: < 1 year, 1-6 years, 7-12 years, and > 12 years old. Sensitivity, specificity, accuracy, and area under the curve (AUC) were calculated to determine SIPA threshold values associated with MT and/or mortality. RESULTS: Of 1,072 patients, 6.3% (n = 68) required MT and 8.4% (n = 90) died. For predicting MT, pre-hospital SIPA cut points performed best in the > 12 year-old age group (AUC = 0.86) and ED SIPA cut points performed best in the 6-12 year-old age group (AUC = 0.87). For predicting mortality, pre-hospital (AUC = 0.78) and ED SIPA cut points (AUC = 0.84) performed best in the > 12 year-old age group. CONCLUSION: Pre-hospital and ED SIPA cut points performed better at predicting MT and/or mortality in older pediatric patients compared to very young children. Age remains an important factor when determining the validity of SIPA to predict outcomes in pediatric trauma patients. STUDY TYPE/LEVEL OF EVIDENCE: Level III, Retrospective Cohort Study.

Novel tool (BIS) heralds the need for blood transfusion and/or failure of non-operative management in pediatric blunt liver and spleen injuries.

BACKGROUND: Non-operative management (NOM) is the standard of care for the majority of children with blunt liver and spleen injuries (BLSI). The shock index pediatric age-adjusted (SIPA) was previously shown to predict the need for blood transfusions in pediatric trauma patients with BLSI. We combined SIPA with base deficit (BD) and International Normalized Ratio (INR) to create the BIS score. We hypothesized that the BIS score would predict the need for blood transfusions and/or failure of NOM in pediatric trauma patients with BLSI. METHODS: Patients (≤ 18 years) who presented to our Level I pediatric trauma center with BLSI from 2009 to 2019 were identified. BIS scores were calculated by giving 1 point for each of the following: base deficit ≤ -8.8, INR ≥ 1.5, or elevated SIPA. Receiver operating characteristic curves (ROC) were generated for BIS scores ≥ 1, ≥ 2, and ≥ 3. Area under the curve (AUC), sensitivity, and specificity of each score were calculated for ability to predict need for blood transfusions and/or failure of NOM. RESULTS: Of 477 children included, 19.9% required a blood transfusion and 6.7% failed NOM. A BIS score ≥ 1 was the best predictor of the need for blood transfusions with an AUC of 0.81 and a sensitivity of 96.0%. A BIS score ≥ 1 was also the best predictor of failure of NOM with an AUC of 0.72 and a sensitivity of 97.0%. CONCLUSION: The BIS score is a highly sensitive tool that identifies pediatric patients with BLSI at risk for blood transfusions and/or failure of NOM. LEVEL OF EVIDENCE: Level III. TYPE OF STUDY: Retrospective comparative study.

A surgeon's predicament: Clinical predictors of surgery and mortality in neutropenic enterocolitis.

BACKGROUND: Neutropenic enterocolitis is uncommon but potentially life-threatening, with the cornerstone of treatment being medical management (MM), and surgical intervention reserved for clinical deterioration or bowel perforation. We hypothesized that the Shock Index Pediatric Age-Adjusted (SIPA) is elevated in patients who are at greatest risk for surgical intervention and mortality. We also sought to identify computed tomography (CT) findings associated with surgical intervention and mortality. METHODS: A single-center cancer registry was reviewed for neutropenic enterocolitis patients from 2006 -2018. Survival models compared patients with normal versus elevated SIPA throughout their hospitalizations for the time to surgical management (SM), as well as in-hospital mortality. RESULTS: Seventy-four patients with neutropenic enterocolitis were identified; 7 underwent surgery. In-hospital mortality was 12% in MM and 29% in SM; mortality among patients with elevated SIPA was 4.7 times higher compared to those with normal SIPA (95% CI: 1.1, 19.83, p = 0.04). CT findings of bowel obstruction, pneumatosis, and a greater percentage of large bowel involvement were associated with surgical intervention (all ps < 0.05). CONCLUSION: Select pre-operative CT findings were associated with need for operative management. Elevated SIPA was associated with increased mortality. Elevated SIPA in pediatric cancer patients with neutropenic enterocolitis may help to identify those with more severe disease and expedite beneficial interventions.

Management of blunt traumatic abdominal wall hernias: A Western Trauma Association multicenter study.

BACKGROUND: Blunt traumatic abdominal wall hernias (TAWH) occur in approximately 15,000 patients per year. Limited data are available to guide the timing of surgical intervention or the feasibility of nonoperative management. METHODS: A retrospective study of patients presenting with blunt TAWH from January 2012 through December 2018 was conducted. Patient demographic, surgical, and outcomes data were collected from 20 institutions through the Western Trauma Association Multicenter Trials Committee. RESULTS: Two hundred and eighty-one patients with TAWH were identified. One hundred and seventy-six (62.6%) patients underwent operative hernia repair, and 105 (37.4%) patients underwent nonoperative management. Of those undergoing surgical intervention, 157 (89.3%) were repaired during the index hospitalization, and 19 (10.7%) underwent delayed repair. Bowel injury was identified in 95 (33.8%) patients with the majority occurring with rectus and flank hernias (82.1%) as compared with lumbar hernias (15.8%). Overall hernia recurrence rate was 12.0% (n = 21). Nonoperative patients had a higher Injury Severity Score (24.4 vs. 19.4, p = 0.010), head Abbreviated Injury Scale score (1.1 vs. 0.6, p = 0.006), and mortality rate (11.4% vs. 4.0%, p = 0.031). Patients who underwent late repair had lower rates of primary fascial repair (46.4% vs. 77.1%, p = 0.012) and higher rates of mesh use (78.9% vs. 32.5%, p < 0.001). Recurrence rate was not statistically different between the late and early repair groups (15.8% vs. 11.5%, p = 0.869). CONCLUSION: This report is the largest series and first multicenter study to investigate TAWHs. Bowel injury was identified in over 30% of TAWH cases indicating a significant need for immediate laparotomy. In other cases, operative management may be deferred in specific patients with other life-threatening injuries, or in stable patients with concern for bowel injury. Hernia recurrence was not different between the late and early repair groups. LEVEL OF EVIDENCE: Therapeutic/care management, Level IV.

It is time for TEG in pediatric trauma: unveiling meaningful alterations in children who undergo massive transfusion.

BACKGROUND: Hemorrhage is the leading cause of preventable death in pediatric trauma patients. In adults, goal-directed thrombelastography (TEG) has been shown to reduce mortality when used to guide massive transfusion (MT) resuscitation. There remains a paucity of data on the utility of TEG in directing resuscitation of pediatric trauma patients. We hypothesize that abnormalities on admission TEG will differ in pediatric trauma patients who undergo MT, compared to those who do not. METHODS: Pediatric patients (≤ 18 years) who were highest level trauma activations at two trauma centers from 2015 to 2018 were analyzed. We included patients who had admission TEGs and excluded those who did not. Patients were stratified into two groups: those who received MT (> 40 cc/kg total blood product within 6 h of admission) and those who did not. We defined TEG abnormalities based on each institution's normative values and compared TEG abnormalities between the groups. RESULTS: Of 117 children included, 39 had MT. MT patients had higher injury severity scores (30 vs. 23, p = 0.0004), lactates levels (7.0 vs. 3.5, p < 0.001), base deficit levels ( - 12.2 vs. - 5.8, p < 0.001), and INR values (1.8 vs. 1.3, p < 0.001). MT patients had significantly shortened alpha-angles (35.9% vs. 15.4%, p = 0.023), maximum amplitude (MA) values (43.6% vs. 10.3%, p < 0.001), and significantly lower platelet counts (165 vs. 281, p < 0.001) compared to those who did not receive MT. There was no difference in the trends in R-time, LY30 (lysis or shutdown), or fibrinogen concentration between the groups. Logistic regression identified a decreased MA as a significant predictor for MT [OR 3.68 (CI 1.29-10.52)] CONCLUSIONS: Pediatric trauma patients who undergo MT are more likely to have lower alpha-angles and MA values, as well as lower platelet counts. These findings support the use of TEG to identify hemorrhaging pediatric trauma patients, who may benefit from cryoprecipitate and/or platelet transfusions. TEG provides real-time information on coagulation status, which may expedite the delivery of specific blood products during trauma resuscitation. LEVEL OF EVIDENCE: LEVEL III: Type of study: Retrospective comparative study.

The novel fetal MRI O/E CLV versus O/E LHR in predicting prognosis in congenital diaphragmatic hernias: can we teach an old dog new tricks?

PURPOSE: In congenital diaphragmatic hernia (CDH), ultrasound (U/S) measurements of the contralateral lung commonly provide the observed-to-expected lung-to-head ratio (O/E LHR) and are used to determine the severity of pulmonary hypoplasia. Fetal magnetic resonance imaging (MRI) measurement of the observed-to-expected total lung volume (O/E TLV) has been used as an adjunct to O/E LHR in predicting outcomes. Since O/E LHR only measures the contralateral lung, we sought to investigate if MRI measurements of the contralateral lung volume (O/E CLV) can accurately predict outcomes in CDH. We hypothesize that O/E CLV is a better predictor of CDH outcomes than O/E LHR. METHODS: We identified all infants with a prenatal diagnosis of CDH at our fetal center who had both MRI and U/S measurements. Using lung volume ratios of right-left 55:45, we calculated O/E CLV from O/E TLV. We used receiver-operating characteristic (ROC) curves to calculate the area under the curve (AUC) to compare the predictive accuracy of O/E CLV to O/E LHR for ECMO support, as well as survival to both discharge and 1 year. RESULTS: Seventy-four patients had complete prenatal imaging with 39% requiring ECMO support. The median O/E CLV was 48.0% and the median O/E LHR was 42.3%. O/E CLV was a better predictor of the need for ECMO support (AUC 0.81 vs. 0.74). O/E CLV was a better predictor of survival to discharge (AUC 0.84 vs. 0.64) and 1-year survival (AUC 0.83 vs. 0.63) than O/E LHR. CONCLUSION: O/E LHR is a well-validated standard for predicting outcomes and guiding prenatal counseling in CDH. We provide evidence that fetal MRI measurements of the contralateral lung volume corrected for gestational age were more accurate in predicting the need for ECMO and survival. Future prospective studies validating O/E CLV regarding outcomes and ECMO utilization are warranted. LEVEL OF EVIDENCE: Level III, retrospective comparative study.

Using deep learning and natural language processing models to detect child physical abuse.

BACKGROUND: The recognition of child physical abuse can be challenging and often requires a multidisciplinary assessment. Deep learning models, based on clinical characteristics, laboratory studies, and imaging findings, were developed to facilitate unbiased identification of children who may have been abused. METHODS: Level 1 pediatric trauma center registry data from 1/1/2010-1/31/2020 were queried for abused children and matched participants with non-abusive trauma. Observations were de-identified and divided into training and validation sets. Model 1 used patient demographics (age, gender, and insurance type) and clinical characteristics (vital signs, shock index pediatric age-adjusted, Glasgow Coma Score, lactate, base deficit, and international normalized ratio). Model 2 used the same features as Model 1, but with the text of the radiology reports of head computed tomography, brain MRIs, and skeletal surveys. Google's latest BERT Natural Language Processing (NLP) model, which was pre-trained on a large corpus, was used for fine-tuning Model 2. Accuracy, sensitivity, specificity, F1 scores, and positive predictive values were used to assess performance. RESULTS: Of 1,312 patients, 737 (56.2%) were abused. Model 1 had an accuracy of 86.3%, sensitivity of 87.2%, specificity of 85.1%, F1 score of 0.86, and positive predictive value (PPV) of 88.7% for the validation set with an area under the receiver Operating Curve (ROC AUC) of 0.86. NLP based Model 2 had an accuracy of 93.4%, sensitivity 92.5%, specificity of 94.6%, F1 score of 0.93, and PPV of 95.9% for the validation set, with a ROC AUC of 0.94. Most features had weak individual correlations with abuse (r < 0.3). CONCLUSIONS: Deep learning models accurately distinguished child physical abuse from non-abuse, and NLP further improved the accuracy of the models. Such models could be developed to run in real-time in the electronic medical record and alert clinicians when certain criteria are met, which would prompt them to pursue the diagnosis of abuse. LEVEL OF EVIDENCE: III STUDY TYPE: Diagnostic.

Comparison of non-invasive physiological assessment tools between simple and perforated appendicitis in children.

PURPOSE: The role of non-invasive measures of physiologic reserve, specifically the Compensatory reserve index (CRI) and the Shock index pediatric age-adjusted (SIPA), is unknown in the management of children with acute appendicitis. CRI is a first-in-class algorithm that uses pulse oximetry waveforms to continuously monitor central volume status loss. SIPA is a well-validated, but a discontinuous measure of shock that has been calibrated for children. METHODS: Children with suspected acute appendicitis (2-17 years old) were prospectively enrolled at a single center from 2014 to 2015 and monitored with a CipherOx CRI™ M1 pulse oximeter. CRI values range from 1 (normovolemia) to 0 (life-threatening hypovolemia). SIPA is calculated by dividing heart rate by systolic blood pressure and categorized as normal or abnormal, based on age-specific cutoffs. Univariate and multivariable regression models were developed with simple versus perforated appendicitis as the outcome. RESULTS: Almost half the patients (45/94, 48%) had perforated appendicitis. On univariate analysis, the median admission CRI value was significantly higher (0.60 versus 0.33, p < 0.001) and the ED SIPA values were significantly lower (0.90 versus 1.10, p = 0.002) in children with simple versus perforated appendicitis. In a multivariable model, only CRI significantly detected differences in the physiologic state between patients with simple and perforated appendicitis. CONCLUSIONS: CRI is a non-invasive measure of physiologic reserve that may be used to accurately guide early management of children with acute simple versus perforated appendicitis.

Lessons Learned From the First Pilot Study of the Compensatory Reserve Index After Congenital Heart Surgery Requiring Cardiopulmonary Bypass.

BACKGROUND: Early warning systems that utilize dense physiologic data and machine learning may aid prediction of decompensation after congenital heart surgery (CHS). The Compensatory Reserve Index (CRI) analyzes changing features of the pulse waveform to predict hemodynamic decompensation in adults, but it has never been studied after CHS. This study sought to understand the feasibility, safety, and potential utility of CRI monitoring after CHS with cardiopulmonary bypass (CPB). METHODS: A single-center prospective pilot cohort of patients undergoing pulmonary valve replacement was studied. Compensatory Reserve Index was continuously measured from preoperative baseline through the first 24 postoperative hours. Average CRI values during selected procedural phases were compared between patients with an intensive care unit (ICU) length of stay (LOS) <3 days versus LOS ≥3 days. RESULTS: Twenty-three patients were enrolled. On average, 17,445 (±3,152) CRI data points were collected and 0.33% (±0.40) of data were missing per patient. There were no adverse events related to monitoring. Five (21.7%) patients had an ICU LOS ≥3 days. Compared to the ICU LOS <3 days group, the ICU LOS ≥3 days group had a greater decrease in CRI from baseline to immediately after CPB (-0.3 ± 0.1 vs -0.1 ± 0.2, P = .003) and were less likely to recover to baseline CRI during the monitoring period (20% vs 83%, P = .017). CONCLUSIONS: Compensatory Reserve Index monitoring after CHS with CPB seems feasible and safe. Early changes in CRI may precede meaningful clinical outcomes, but this requires further study.

Gastrostomy Button Placement in Infants With Cyanotic Versus Acyanotic Congenital Heart Disease.

BACKGROUND: Infants with congenital heart disease (CHD) may exhibit increased metabolic demands, and many will undergo placement of a gastrostomy to achieve adequate nutritional intake. There is a paucity of data, however, comparing the operative risks and overall complications of gastrostomy placement in cyanotic versus acyanotic infants with CHD. We hypothesized that patients with cyanotic CHD would have a higher rate of gastrostomy-associated complications than infants with acyanotic CHD. METHODS: We retrospectively reviewed patients who underwent gastrostomy button placement after cardiac surgery for CHD between 2013 and 2018. Patients were stratified into cyanotic CHD and acyanotic CHD cohorts. Patient data were extracted from the Society of Thoracic Surgeons database and merged with clinical data related to gastrostomy placement and complications from chart review. Unadjusted analyses were used to find covariates associated with cyanotic CHD and acyanotic CHD, using a t-test or Wilcoxon rank-sum test for continuous data, depending on normalcy, and χ2 or Fisher's exact tests for categorical data depending on the distribution. RESULTS: There were 257 infants with CHD who underwent gastrostomy placement during the study period, of which 86 had cyanotic CHD. There were no significant differences in baseline weight or preoperative albumin levels between the two groups. Patients with cyanotic CHD had a lower incidence of comorbid syndromes (P = 0.0001), higher Society of Thoracic Surgeons-European Association for Cardio-Thoracic Surgery scores (P < 0.0001), and higher postoperative mortality rate (P = 0.0189). There was a higher rate of granulation tissue formation in patients with acyanotic CHD (48.5% versus 22.1%, P < 0.0001). There were no differences in other gastrostomy button-related complications, including leakage, wound infection, or dislodgement. CONCLUSIONS: Patients with acyanotic CHD demonstrated a higher incidence of granulation tissue. We found no difference in gastrostomy-specific complication rates between the two groups, with the notable exception of granulation tissue formation. Based on this study, the diagnosis of cyanotic CHD does not increase the risk of gastrostomy-related complications.