Disparities in unintentional drowning fatalities in Texas, USA, 1999-2020.

BACKGROUND: Drowning is the third-leading cause of unintentional injury death worldwide. Although the USA as a whole bears a heavy burden, with approximately 4000 drowning fatalities annually, Texas stands out as a high-risk state for drowning due to its large population, suitable climate for year-round aquatic activities and availability of water-related recreational opportunities. METHODS: Using mortality data from the Centers for Disease Control and Prevention Wide-ranging ONline Data for Epidemiologic Research online database, this retrospective, cross-sectional study overviews the magnitude and patterns of fatal unintentional drownings among Texans from 1999 to 2020. RESULTS: Over the 22-year period, 7737 Texans died from unintentional drowning. An average of 352 drowning deaths occurred annually, with a rate of 1.4 deaths per 100 000 population. The highest proportion of unintentional drownings occurred in natural water settings (eg, lakes, ponds or rivers), accounting for 40% of fatal drownings. Children aged 1-4 years had the highest drowning death rate compared with all other age groups. Male Texans had a drowning death rate three times higher than that of female Texans. Black Texans had a higher drowning death rate than White Texans and Asian or Pacific Islander Texans. CONCLUSIONS: Drowning remains a significant public health issue in Texas. Data on high-risk groups and settings should be used to strengthen drowning prevention efforts and policy initiatives and encourage more research to address the multifaceted factors contributing to drowning.

Traumatic Head Injuries from Ceiling Fans at a Pediatric Level I Trauma Center in the United States.

BACKGROUND/OBJECTIVE: Several studies describe traumatic head injuries caused by ceiling fans in Australia, the Middle East, and Malaysia. Some injuries required neurosurgical intervention, especially those caused by metallic ceiling fans. This study describes traumatic head injuries caused by ceiling fans at a single pediatric level 1 trauma center in the Southern USA. METHODS: Medical records were retrospectively reviewed for patients under 18 years of age who presented with a traumatic injury to the head from a ceiling fan from January 1, 2008, through December 31, 2021. The cohort of patients meeting all inclusion criteria was identified by querying multiple free-text fields derived from the electronic medical record, followed by a manual record review. RESULTS: Of 60 children treated for traumatic head injury from a ceiling fan, the median age was 5.7 years and 53% were female. Laceration was the most common injury (80%), followed by scalp swelling/hematoma (20%), contusion (8%), and skull fracture (7%). Two patients (3%) with intracranial hemorrhage and fracture underwent neurosurgery. One neurosurgical case involved a metal ceiling fan and the other involved an outdoor ceiling fan. Nearly half of the injuries involved bunk or loft beds (47%) and young children were often injured while being lifted up by a caregiver (18%). CONCLUSION: Although most pediatric traumatic head injuries from ceiling fans resulted in minor injuries, our center saw a similar proportion of cases with skull fractures to what has been reported in Australia (5%). The effects of fan construction and blade material on the severity of head injury may warrant further study. Understanding the most common mechanisms for these injuries may guide injury prevention efforts.

Incomplete Fractures of Distal Long Bones Among Children in Early Stages of Mobility: Few Concerns for Physical Child Abuse at a Single Pediatric Level I Trauma Center.

OBJECTIVES: This study aimed to describe the characteristics, reported mechanism of injury, fracture morphology, and level of physical abuse concern among children in the early stages of mobility diagnosed with an incomplete distal extremity long bone fracture. METHODS: A retrospective chart review was performed for all children aged 10 to 12 months with an incomplete fracture of a distal forearm or distal lower limb who were reportedly pulling up, cruising, or ambulating, and who were evaluated by the child abuse pediatrics team at a single pediatric level I trauma center. RESULTS: Of the 29 patients who met inclusion criteria, the child abuse pediatrics team had concerns about physical child abuse for 3 children. Not every case with an unknown or discrepant history of injury was deemed concerning for abuse, but all 3 for whom the team determined that concern was warranted had an unknown or discrepant history. All 3 of these children had distal forearm fractures; 1 child had multiple concomitant fractures (including a scapular fracture), and 2 had evidence of bone healing at initial presentation. Each of these observations raises concern for abusive injury based on current evidence. Both-bone buckle fractures of the radius/ulna and tibia/fibula were the most common type of incomplete distal fracture. CONCLUSIONS: This age group presents a unique challenge when designing evidence-based algorithms for the detection of occult injuries in emergency departments. Incomplete fractures of a distal limb are commonly related to a fall and may be considered "low specificity" for physical abuse. However, some publications conclude they should prompt universal physical abuse screening. Our small study indicated that the presence or absence of certain risk factors may provide additional information which could help guide the need for a more thorough evaluation for occult injury in early-mobile children with incomplete distal extremity long bone fractures. Ongoing research is warranted.

Urinalysis in Suspected Child Abuse Evaluation in the Emergency Department.

BACKGROUND: Intra-abdominal injury (IAI) is the second leading cause of mortality in abused children. It is challenging to identify in young patients due to their limited verbal skills, delayed symptoms, less muscular abdominal wall, and limited bruising. METHODS: We conducted a retrospective cohort study of children aged 0 to 12 months who were evaluated in the emergency department for suspected child abuse with a skeletal survey and urinalysis between January 1, 2015, and December 31, 2017. Our primary objective was to identify the proportion of IAI cases identified by urinalysis alone (>10 RBC/HPF) and not by examination findings or other laboratory results. A secondary objective was to quantify potential delay in disposition while waiting for urinalysis results, calculated as the length of time between receiving skeletal survey and laboratory results and receiving urinalysis results. RESULTS: Six hundred thirteen subjects met our inclusion criteria; two subjects had hematuria, one of whom had a urinary tract infection. The other was determined to have blood from a catheterized urine specimen. One subject was found to have an IAI. We further found that urinalysis was delayed for 78% of subjects and took a median of 93 [interquartile range, 46-153] minutes longer than imaging and/or laboratories. CONCLUSIONS: No subjects were diagnosed with abdominal trauma based on urinalysis during evaluation in the emergency department who would not have been identified by other standard testing. In addition, patients' disposition was delayed while waiting for urinalysis.

Universal Suicidality Screening in a Pediatric Emergency Department to Improve Mental Health Safety Risk.

INTRODUCTION: Suicide is the second leading cause of death for youth 12 to 18 years of age. Suicidal ideation can be predictive of suicide attempt, so screening for suicidal ideation by emergency nurses can help identify those at risk and facilitate timely intervention. This study evaluates the use of a universal suicide screening using the Patient Safety Screener 3 and the Columbia Suicide Severity Rating Scale to identify youth ages 12 to 18 years experiencing suicide risk and assess factors predictive of suicide risk level. METHODS: We conducted a retrospective cohort study using data from patients presenting to the emergency department at an acute care hospital that uses a universal screening program for suicide risk. We determined the frequency of positive screens and performed multivariate analyses to identify predictive factors of scoring high on the Columbia Suicide Severity Rating Scale. RESULTS: Notably, 9.1% of patients were experiencing some level of suicide risk; 10% of those with positive scores had no mental health history and were not presenting for a mental health reason. After controlling for other independent variables, insurance status, mental health presentation, and known mental health history were significantly associated with Columbia Suicide Severity Rating Scale score. DISCUSSION: Universal screening for suicide risk in pediatric emergency departments by nurses is critical for all patients older than 12 years, given that we identified patients at risk of suicide who presented for non-mental health reasons. These patients may not have been identified or referred to treatment if they were not screened for suicidality increasing risk of future suicide attempt.

Risk Factors for Upper Extremity Refractures in Children.

BACKGROUND: Multiple descriptive studies have been published on refracture patterns, particularly for forearm fractures. However, few large cohorts have been analyzed quantitatively including the odds of refracture, and with a comprehensive assessment of the possible predictive factors associated with refracture. This study aimed to assess the frequency and timing of upper extremity refracture in a large pediatric orthopaedics practice, and to evaluate the strength of association of various patient-level and fracture-related factors with refracture. METHODS: Medical records were reviewed retrospectively for patients 1 to 18 years of age with at least 1 upper extremity fracture (ICD-9 codes 810 to 819) between June 1, 2010 and May 31, 2011. Characteristics of patients and fractures were assessed for the association with refracture using bivariate analysis and multivariable logistic regression. RESULTS: Among 2793 patients with a total of 2902 upper extremity fractures, 2% were treated for refracture within 2 years, at a median of 6 months (188 d) after the initial injury. Midshaft location, and characterization of the fracture as angulated or buckle, were associated with being more likely to refracture. Eighty percent of refractures were the result of a fall, with almost 25% involving a high-energy mechanism and about 15% from monkey bars or other playground equipment. The adjusted odds of refracture were 4 times higher if noncompliance with treatment recommendations was documented, when controlling for insurance type and number of days before orthopaedic evaluation. Forearm fractures were almost 4 times more likely to refracture compared with other bones, controlling for midshaft location, days immobilized, and buckle or torus characterization of the fracture. CONCLUSIONS: Our practice saw a refracture occurrence in 2% of patients, with median time to refracture of ~6 months. The factors most strongly associated with refracture were midshaft fracture location, forearm fracture as opposed to clavicle or humerus, and noncompliance as defined in the study. Falls and high energy activities, such as use of wheeled devices, skis, or trampolines, were important mechanisms of refracture. LEVEL OF EVIDENCE: This study is a Level II prognostic study. It is a retrospective study that evaluates the effect of patient and fracture characteristics on the outcome of upper extremity refracture.

Accuracy of ICD-10-CM coding for physical child abuse in a paediatric level I trauma centre.

This retrospective study examined the accuracy of the International Classification of Diseases, Clinical Modification (ICD-10-CM) coding for physical child abuse among patients less than 18 years of age who were evaluated due to concern for physical abuse by a multidisciplinary child protection team (MCPT) during 2016-2017 (N=312) in a paediatric level I trauma centre. Sensitivity, specificity, predictive values and diagnostic OR for ICD-10-CM coding were calculated and stratified by admission status, using as a reference standard the abuse determination of the MCPT recorded in a hospital registry. Among inpatients, child physical abuse coding sensitivity was 55.6% (95% CI 41.4% to 69.1%) and specificity was 78.6% (95% CI 59.0% to 91.7%), with diagnostic OR of 4.58 (95% CI 1.64 to 12.70). Among outpatients, sensitivity was 22.2% (95% CI 15.5% to 30.2%) and specificity was 86.3% (95% CI 77.7% to 92.5%), with diagnostic OR of 1.80 (95% CI 0.89 to 3.64). Use of ICD-10-CM coded data sets alone for surveillance may significantly underestimate the occurrence of physical child abuse.

Predictors of Traumatic Suicide Attempts in Youth Presenting to Hospitals with Level I Trauma Centers.

BACKGROUND: Limited research exists examining the predictors of suicide attempts by mechanism. OBJECTIVE: The purpose of this study was to examine predictors of traumatic suicide attempts in youth. METHODS: Data came from patients 5-18 years of age presenting because of a suicide attempt at 2 hospitals in Central Texas with level I trauma centers. Univariate logistic regression examined the association between traumatic suicide attempts and variables describing the patient's demographic, mental health, and social information. We used the Mann-Whitney U test to examine the association between traumatic suicide attempts and the continuous variable of age. RESULTS: Of 231 patients included in this study, most were female (75.8%), non-Hispanic white (48.1%), and had a median age of 15.0 years (interquartile range 14-16). Compared with patients presenting because of an intentional overdose, patients presenting because of traumatic suicide attempts were associated with a reported criminal history (odds ratio [OR] 14.50 [95% confidence interval {CI} 3.84-54.82]), reported Child Protective Services history (OR 3.26 [95% CI 0.99-10.77]), being publicly insured or uninsured (OR 1.80 [95% CI 1.02-3.19]), male (OR 2.37 [95% CI 1.28-4.38]), and identifying as Hispanic (OR 2.01 [95% CI 1.10-3.68). CONCLUSIONS: Our findings inform targeted preventative resources and education efforts to populations of greatest need.

Predictors of Childhood Injury in Children Reported to Child Protective Services.

BACKGROUND: Adverse childhood experiences, such as child maltreatment, have been shown to result in negative health outcomes throughout an individual's life. Previous research has found that children with a prior allegation of maltreatment die due to unintentional injuries at twice the rate of children who were not reported to Child Protective Services, however, death is only one outcome of injury and many unintentional injuries do not result in death. METHODS: This secondary analysis of Longitudinal Studies of Child Abuse and Neglect (LONGSCAN) data examined predictors of injury in children whose family has been reported to CPS utilizing the ecological-developmental framework theoretical domains which guided LONGSCAN: child characteristics, family/caregiver characteristics, parental and family functioning, extrafamilial relationship skills, community ecology, child outcomes, and systems of care factors. RESULTS: Logistic regression modeling showed that the likelihood of childhood injury significantly increased for those children who did not recognize or appropriately respond to cues of ending a conversation (OR=.37), caregiver reported child always has a good place to play in their neighborhood (OR=1.57), and child utilized educational services (OR=2.06). CONCLUSION: Understanding the predictors of injury is necessary to implement injury prevention interventions targeting the unique needs of this vulnerable population and increase awareness of prevention strategies to reduce childhood injuries.

Risk Factors Associated With Emergency Department Return Visits Following Trauma System Discharge.

OBJECTIVES: Little evidence exists in the pediatric trauma literature regarding what factors are associated with re-presentation to the hospital for patients discharged from the emergency department (ED). METHODS: This was a retrospective cohort study of trauma system activations at a pediatric trauma center from June 30, 2007, through June 30, 2013, who were subsequently discharged from the ED or after a brief inpatient stay. Returns within 30 days were reviewed. χ, Student t test, and univariate logistical regression were used to compare predictive factors for those returning and not. RESULTS: One thousand eight hundred sixty-three patient encounters were included in the cohort. Seventy-two patients (3.9%) had at least 1 return visit that was related to the original trauma activation. Age, sex, language, race/ethnicity, ED length of stay, arrival mode, level of trauma activation, and transfer from an outside hospital did not vary significantly between the groups. Patients with public insurance were almost 2 times more likely to return compared with those with private insurance (odds ratio, 1.92; 95% confidence interval, 1.11-3.35). Income by zip code was associated with the risk of a return visit, with patients in neighborhoods at less than the 50th percentile income twice as likely to return to the ED (odds ratio, 2.15; 95% confidence interval, 1.30-3.54). CONCLUSIONS: Patients with public insurance and those from low-income neighborhoods were significantly more likely to return to the ED after trauma system activation. These data can be used to target interventions to decrease returns in high-risk trauma patients.

Designing optimal mortality risk prediction scores that preserve clinical knowledge.

Many in-hospital mortality risk prediction scores dichotomize predictive variables to simplify the score calculation. However, hard thresholding in these additive stepwise scores of the form "add x points if variable v is above/below threshold t" may lead to critical failures. In this paper, we seek to develop risk prediction scores that preserve clinical knowledge embedded in features and structure of the existing additive stepwise scores while addressing limitations caused by variable dichotomization. To this end, we propose a novel score structure that relies on a transformation of predictive variables by means of nonlinear logistic functions facilitating smooth differentiation between critical and normal values of the variables. We develop an optimization framework for inferring parameters of the logistic functions for a given patient population via cyclic block coordinate descent. The parameters may readily be updated as the patient population and standards of care evolve. We tested the proposed methodology on two populations: (1) brain trauma patients admitted to the intensive care unit of the Dell Children's Medical Center of Central Texas between 2007 and 2012, and (2) adult ICU patient data from the MIMIC II database. The results are compared with those obtained by the widely used PRISM III and SOFA scores. The prediction power of a score is evaluated using area under ROC curve, Youden's index, and precision-recall balance in a cross-validation study. The results demonstrate that the new framework enables significant performance improvements over PRISM III and SOFA in terms of all three criteria.

Penetrating knitting needle through the mediastinum in a child.

Thoracic injuries are second only to central nervous system injuries as the leading cause of traumatic death in patients of all ages. Penetrating chest injury is very uncommon in children, but it comes with significant morbidity and mortality. Presentation of penetrating thoracic injury in the child is unique with inherent opportunities for learning. The purpose of this case report was to disseminate information regarding rare thoracic trauma in the pediatric patient.

Threat of paediatric hyperthermia in an enclosed vehicle: a year-round study.

OBJECTIVE: To describe temperature change throughout the workday in an enclosed vehicle in Austin, Texas across the calendar year while accounting for heat index. METHODS: In this observational study, vehicular temperature was measured 1 day per month during 2012 in Austin, Texas. Data were recorded at 5-min intervals via an EL-USB-1-PRO digital temperature sensor from 8:00 to 16:00. Selected days were primarily cloud-free (with 'clear' or 'few clouds') with a predicted ambient temperature high within ±20°F of the 30-year normal high. Referent temperature and 30-year normal data were collected via the nearest National Weather Service (NWS) weather station. The NWS heat index and corresponding hazard levels were used as a guideline for this study. RESULTS: Per NWS guidelines, the enclosed vehicle temperature rose to 'danger' levels of ≥105°F (41°C) in all months except January and December and to 'extreme caution' levels of ≥90°F (32°C) in every month of the year. In June, the vehicle rose to ≥105°F (41°C) by 9:25. The hottest vehicular temperature achieved was 137°F (58°C). In 9 months of the year, the vehicle reached ≥90°F (32°C) by noon. We also found that an ambient temperature as low as 68°F was associated with vehicular temperatures ≥105°F (41°C). CONCLUSIONS: Infants and children in states that experience mild winter temperatures face the threat of vehicular hyperthermia disability and death across the calendar year. Prevention efforts that focus on awareness of a childhood heat vulnerability, parental perception of susceptibility to forgetting a child in a vehicle and universal availability of vehicular safety devices may reduce paediatric vehicular hyperthermia death.

The use and timing of angioembolization in pediatric blunt liver and spleen injury.

BACKGROUND: Nonoperative management (NOM) is the standard of care for the management of blunt liver and spleen injuries (BLSI) in the stable pediatric patient. Angiography with embolization (AE) is used as an adjunctive therapy in the management of adult BLSI patients, but it is rarely used in the pediatric population. In this planned secondary analysis, we describe the current utilization patterns of AE in the management of pediatric BLSI. METHODS: After obtaining IRB approval at each center, cohort data was collected prospectively for children admitted with BLSI confirmed on CT at 10 Level I pediatric trauma centers (PTCs) throughout the United States from April 2013 to January 2016. All patients who underwent angiography with or without embolization for a BLSI were included in this analysis. Data collected included patient demographics, injury details, organ injured and grade of injury, CT finding specifics such as contrast blush, complications, failure of NOM, time to angiography and techniques for embolization. RESULTS: Data were collected for 1004 pediatric patients treated for BLSI over the study period, 30 (3.0%) of which underwent angiography with or without embolization for BLSI. Ten of the patients who underwent angiography for BLSI failed NOM. For patients with embolized splenic injuries, splenic salvage was 100%. Four of the nine patients undergoing embolization of the liver ultimately required an operative intervention, but only one patient required hepatorrhaphy and no patient required hepatectomy after AE. Few angiography studies were obtained early during hospitalization for BLSI, with only one patient undergoing angiography within 1 hour of arrival at the PTC, and 7 within 3 hours. CONCLUSION: Angioembolization is rarely used in the management of BLSI in pediatric trauma patients with blunt abdominal trauma and is generally used in a delayed fashion. However, when implemented, angioembolization is associated with 100% splenic salvage for splenic injuries. LEVEL OF EVIDENCE: Therapeutic/Care Management; Level IV.

Implementing Screening, Brief Interventions, and Referral to Treatment at Pediatric Trauma Centers: A Step Wedge Cluster Randomized Trial.

BACKGROUND: Pediatric trauma centers have had challenges meeting the American College of Surgeons criteria for screening and intervening for alcohol with adolescent trauma patients. The study objective was to conduct an implementation trial to evaluate the effectiveness of the Science to Service Laboratory (SSL) implementation strategy in improving alcohol and other drugs (AOD) screening, brief intervention, and referral to treatment (SBIRT) delivery at pediatric trauma centers. METHODS: Using a stepped wedge cross-over cluster randomized design, 10 US pediatric trauma centers received the SSL implementation strategy to deliver SBIRT with admitted adolescent (12-17 years old) trauma patients. The strategy adapted three core SSL elements: didactic training, performance feedback, and facilitation. The main outcome measured was SBIRT reach. Data were collected from each center's electronic health record (EHR) during pre- and post-implementation wedges (2018-2022). RESULTS: EHR data from 8461 adolescent patients were extracted. Aggregated across all sites, the reach of screening with a validated AOD screening tool increased significantly from 25.2% (95% CI: 23.9, 26.5%) of adolescents during pre-implementation to 47.7% (95% CI: 46.3%, 49.2%) post-implementation. There was variability of change across centers. Brief interventions continued to be delivered at high levels to identified adolescents. Referral to primary care providers for further AOD discussion or referral to specialty service for adolescents with high risk use did not improve post-implementation and remained low. CONCLUSIONS: The SSL implementation strategy can be successfully utilized by pediatric trauma centers to improve AOD screening, but challenges exist in connecting adolescents for continuation of AOD discussions after discharge. LEVEL OF EVIDENCE: Level II, Therapeutic.

Adolescents' Perceptions of Screening, Brief Intervention, and Referral to Treatment Service at Pediatric Trauma Centers.

Objective: Screening, brief intervention, and referral to treatment (SBIRT) for adolescent alcohol and drug (AOD) use is recommended to occur with adolescents admitted to pediatric trauma centers. Most metrics on SBIRT service delivery only reference medical record documentation. In this analysis we examined changes in adolescents' perception of SBIRT services and concordance of adolescent-report and medical record data, among a sample of adolescents admitted before and after institutional SBIRT implementation. Methods: We implemented SBIRT for adolescent AOD use using the Science to Service Laboratory implementation strategy and enrolled adolescents at 9 pediatric trauma centers. The recommended clinical workflow was for nursing to screen, social work to provide adolescents screening positive with brief intervention and referral to their PCP for continued AOD discussions with those. Adolescents screening as high-risk also referred to specialty services. Adolescents were enrolled and contacted 30 days after discharge and asked about their perception of any SBIRT services received. Data were also extracted from enrolled patient's medical record. Results: There were 430 adolescents enrolled, with 424 that were matched to their EHR data and 329 completed the 30-day survey. In this sample, EHR documented screening increased from pre-implementation to post-implementation (16.3%-65.7%) and brief interventions increased (27.1%-40.7%). Adolescents self-reported higher rates of being asked about alcohol or drug use than in EHR data both pre- and post-implementation (80.7%-81%). Both EHR data and adolescent self-reported data demonstrated low referral back to PCP for continued AOD discussions. Conclusions: Implementation of SBIRT at pediatric trauma centers was not associated with change in adolescent perceptions of SBIRT, despite improved documentation of delivery of AOD screening and interventions. Adolescents perceived being asked about AOD use more often than was documented. Referral to PCP or specialty care for continued AOD discussion remains an area of needed attention. Trial registration: Clinicaltrials.gov NCT03297060.

Perceptions of injury and prevention practices among pregnant and parenting teenagers.

Injury is a leading cause of death for young children, and the children of teenaged parents may be at increased risk. This qualitative study explored pregnant and parenting teenagers' child safety beliefs and practices related to 4 topics: preventing accidental suffocation via safe sleeping practices, motor vehicle collision safety, prevention of inflicted head trauma, and drowning prevention. Twenty-four focus groups were held with 93 pregnant and/or parenting teenagers. Participants reported variation in their sleeping arrangements, transportation methods, caregivers, and childcare settings. Confusion over safety information was common. Child safety practices were influenced by boyfriends/husbands, parents, grandparents, and teachers.

Sleep behaviors of infants and young children: associated demographic and acculturation characteristics among Hispanic teen mothers.

Accidental suffocation and strangulation in bed is a leading cause of preventable infant death. Bed sharing, teen motherhood, and Hispanic ethnicity have been associated with infant sleep suffocation death. Fifty-five Hispanic teen mothers were surveyed regarding acculturation/demographic characteristics and their infants' sleep behaviors. Most participants had 2 foreign-born parents from Latin America. Participants with 2 US-born parents were less likely to bed share than their less-acculturated peers. Many participants reported not always placing their infant in a supine sleep position. There is a significant need to reach out to Hispanic teen mothers, particularly from newer immigrant families, with culturally and linguistically appropriate multigenerational clinical messaging on the risks of infant bed sharing and nonsupine sleep positioning.

Caregiver water safety knowledge and views of toddler water competency.

BACKGROUND: Drowning is the leading cause of death for toddlers. When caregivers are knowledgeable about water safety, they can provide the best protection against drowning. The aim of this study is to survey caregivers of toddlers to better understand factors associated with water safety knowledge, attitudes about pool supervision, and toddler water competency skills. METHODS: An online survey of 650 parents/caregivers of 1-4-year-old toddlers asked about the caregiver's water safety and swimming background and views on pool supervision. Surveys included a true/false section of ten basic water safety knowledge questions. Caregivers also reported on toddler swim lesson history and whether their toddler could perform six standard water competency skills. Linear regression identified factors predictive of water safety knowledge. RESULTS: On average, caregivers selected the correct answer on six out of ten water safety knowledge questions. Water safety knowledge was predicted by the relationship of the caregiver to the toddler, gender, race, education, prior CPR training, caregiver swim capability, and reported pool supervision style. On average, caregivers reported that their toddler could perform half of the water competency skills. The majority of the toddlers had taken swimming lessons. One third of caregivers believed that after a toddler has had swimming lessons, they don't need to be watched as closely when they are in a pool. CONCLUSIONS: Findings suggest that water safety knowledge is poor and that there are misconceptions about toddler supervision needs. Efforts are needed to improve water safety knowledge and to change perceptions about supervision among caregivers of toddlers.