The impact of pandemic restrictive visiting policies on infant wellbeing in a NICU.

BACKGROUND: Given the countrywide lockdown in the first pandemic period and the respective Hospital restrictive policies, we aimed to investigate if the SARS-COV-2 pandemic was associated to a reduced parental presence in the NICU and in which form this had an impact on infant wellbeing. METHODS: Retrospective cohort study about altered NICUs parental presence (measured by number of visits and kangaroo care time) due to pandemic restrictive policies and its impact on infant wellbeing (measured through The Neonatal Pain Agitation and Sedation scale and nurses' descriptive documentation). RESULTS: Presence of both parents at the same time was significantly lower during pandemic. Contrary, maternal presence only and total kangaroo-care time were higher within the pandemic (163.36 ± 94.07 vs 122.71 ± 64.03; p = 0.000). Lower NPASS values were documented during the lookdown (1.28 ± 1.7 vs 1.78 ± 2.2; p = 0.000). CONCLUSION: Data collected through the pandemic confirm the importance of parental presence for infants' wellbeing in a NICU setting. IMPACT: Parental support is an extremely important aspect for infants hospitalized in an intensive care unit. Their presence was limited in many NICUs worldwide during the SARS-COV-2 pandemic. This study confirm the importance of parental presence for infants' wellbeing also in a pandemic situation. Our results support a family-centered newborn individualized developmental care approach in the NICU.

Pain and neurodevelopmental outcomes of infants born very preterm.

AIM: To investigate the impact of the level of pain experienced by infants born preterm on neurodevelopmental outcomes during their stay in a neonatal intensive care unit. METHOD: In this retrospective data analysis we included all surviving infants born preterm with a gestational age between 23 and 32 weeks from 2011 to 2015, who were assessed using the Neonatal Pain, Agitation, and Sedation Scale and examined at 1 year of age using the Bayley Scales of Infant Development. We excluded all infants who had suffered severe neurological morbidities and undergone surgical interventions. RESULTS: A total of 196 infants born preterm were included in the analyses: 105 in the 'no pain group' and 91 in the 'pain group'. Significant differences between the groups were detected for both mental and motor development (p = 0.003, 95% confidence interval [CI] 2.23-10.92; p = 0.025, 95% CI 0.64-9.78). The results remained significant after controlling for other important medical conditions (p = 0.001, 95% CI -19.65 to -5.40; p = 0.010, 95% CI -16.18 to -2.29). INTERPRETATION: Neonatal pain exposure was associated with altered neurodevelopmental outcomes of infants born very preterm at a corrected age of 12 months. This observation highlights the importance of adequate pain management to reduce the risk of poor neurodevelopmental outcomes in these vulnerable patients.

Routine Use of Cerebral Magnetic Resonance Imaging in Infants Born Extremely Preterm.

OBJECTIVE: To describe cerebral abnormalities and their risk factors in a contemporary cohort of infants born extremely premature after the introduction of routine cerebral magnetic resonance imaging (cMRI) at term-equivalent age. STUDY DESIGN: All cMRI examinations performed during November 2017 and November 2020, based on a standardized neonatal cMRI protocol, were included into analysis. Pathologies were retrospectively classified into 3 categories: intraventricular hemorrhage (IVH), white matter disease, and cerebellar injuries. RESULTS: A total of 198 cMRI examinations were available for analyses; 93 (47%) showed abnormalities, most frequently IVH (n = 65, 33%), followed by cerebellar injuries (n = 41, 21%), and white matter disease (n = 28, 14%). Severe abnormalities were found in 18% of patients (n = 36). Significant clinical risk factors for abnormalities on cMRI were lower Apgar scores, lower umbilical artery and first neonatal pH, asphyxia, blood culture-proven sepsis (especially late-onset), and prolonged need of respiratory support and supplemental oxygen. CONCLUSIONS: After routine cMRI, without preconfirmed pathology by cranial ultrasonography, low-grade IVH, noncystic white matter disease, and cerebellar injuries were the most frequently found abnormalities. The clinical value and long-term benefit of the detection of these low-grade pathologies have yet to be confirmed.

Development of a 3D printed patient-specific neonatal brain simulation model using multimodality imaging for perioperative management.

BACKGROUND: Medical-imaging-based three-dimensional (3D) printed models enable improvement in skills training, surgical planning, and decision-making. This pilot study aimed to use multimodality imaging and to add and compare 3D ultrasound as a future standard to develop realistic neonatal brain models including the ventricular system. METHODS: Retrospective computed tomography (CT), magnetic resonance imaging (MRI), and 3D ultrasound-based brain imaging protocols of five neonatal patients were analyzed and subsequently segmented with the aim of developing a multimodality imaging-based 3D printed model. The ventricular anatomy was analyzed to compare the MRI and 3D ultrasound modalities. RESULTS: A realistic anatomical model of the neonatal brain, including the ventricular system, was created using MRI and 3D ultrasound data from one patient. T2-weighted isovoxel 3D MRI sequences were found to have better resolution and accuracy than 2D sequences. The surface area, anatomy, and volume of the lateral ventricles derived from both MRI and 3D ultrasound were comparable. CONCLUSIONS: We created an ultrasound- and MRI-based 3D printed patient-specific neonatal brain simulation model that can be used for perioperative management. To introduce 3D ultrasound as a standard for 3D models, additional dimensional correlations between MRI and ultrasound need to be examined. IMPACT: We studied the feasibility of implementing 3D ultrasound as a standard for 3D printed models of the neonatal brain. Different imaging modalities were compared and both 3D isotropic MRI and 3D ultrasound imaging are feasible for printing neonatal brain models with good dimensional accuracy and anatomical replication. Further dimensional correlations need to be defined to implement it as a standard to produce 3D printed models.

Long-term impact of systematic pain and sedation management on cognitive, motor, and behavioral outcomes of extremely preterm infants at preschool age.

BACKGROUND: Providing optimal pain relief is a challenging task when caring for premature infants. The aim of this study was to compare the long-term cognitive, motor, and behavioral outcomes of preterm infants before and after the implementation of a pain and sedation protocol. In addition, we investigated whether the increased opiate administration resulting after the implementation process had an impact on these outcomes. METHODS: Cognitive outcomes were evaluated using the Kaufman Assessment Battery for Children (KABC), neuromotor examinations were based on Amiel-Tison, and behavioral outcomes were assessed using the parent-reported Child Behavior Checklist (CBCL). RESULTS: One hundred extremely preterm infants were included in the study (control group, n = 53; intervention group, n = 47). No significant differences were found in cognitive and motor outcomes at preschool age. However, every increase in the cumulative opiate exposure for each 100 mg/kg was weakly significantly associated with a higher risk for autism spectrum features (adjusted odds ratio (aOR) = 1.822, 95% confidence interval (CI) [1.231-2.697]; P = 0.03) and withdrawn behavior (aOR = 1.822, 95% CI [1.231-2.697]; P = 0.03) at preschool age. CONCLUSION: Increased neonatal cumulative opiate exposure did not alter cognitive and motor outcomes but may represent a risk factor for autism spectrum and withdrawn behavior at preschool age. IMPACT: The implementation of a protocol for the management of pain and sedation in preterm infants resulted in increased cumulative opiate exposure. Our study adds further evidence that increased neonatal opiate exposure did  not alter cognitive and motor outcomes but may yield a potential risk factor for autism spectrum disorders and withdrawn behavior at preschool age. A vigilant use of opiates is recommended. Further studies are needed looking for novel pain management strategies and drugs providing optimal pain relief with minimal neurotoxicity.

Readiness for and Response to Coronavirus Disease 2019 Among Pediatric Healthcare Providers: The Role of Simulation for Pandemics and Other Disasters.

OBJECTIVES: Early preparation for the training and education of healthcare providers, as well as the continuation or modification of routine medical education programs, is of great importance in times of the coronavirus disease 2019 pandemic or other public health emergencies. The goal of this study was to characterize these self-reported efforts by the pediatric simulation community. DESIGN: This was a global, multicenter survey developed via a Delphi process. SETTING: International survey study. SUBJECTS: The survey was sent to 555 individual members of the three largest international pediatric simulation societies (The International Pediatric Simulation Society, International Network for Simulation-based Pediatric Innovation, Research & Education, and Netzwerk Kindersimulation e.V.) between April 27, 2020, and May 18, 2020. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Description of coronavirus disease 2019 pandemic simulation-based preparation activities of pediatric acute and critical care healthcare providers. The Delphi process included 20 content experts and required three rounds to reach consensus. The survey was completed by 234 participants (42.2%) from 19 countries. Preparation differed significantly between the geographic regions, with 79.3% of Anglo-American/Anglo-Saxon, 82.6% of Indian, and 47.1% of European participants initiating specifically coronavirus disease 2019-related simulation activities. Frequent modifications to existing simulation programs included the use of telesimulation and virtual reality training. Forty-nine percent of institutions discontinued noncoronavirus disease 2019-related simulation training. CONCLUSIONS: The swift incorporation of disease-specific sessions and the transition of standard education to virtual or hybrid simulation training modes occurred frequently. The approach used, however, depended heavily on local requirements, limitations, and circumstances. In particular, the use of telesimulation allowed education to continue while maintaining social distancing requirements.

Impact of neonatal pain and opiate administration in animal models: A meta-analysis concerning pain threshold.

BACKGROUND AND AIM: Neonatal intensive care treatment, including frequently performed painful procedures and administration of analgesic drugs, can have different effects on the neurodevelopment. This systematic review and meta-analysis aimed to investigate the influence of pain, opiate administration, and pre-emptive opiate administration on pain threshold in animal studies in rodents, which had a brain development corresponding to preterm and term infants. METHODS: A systematic literature search of electronic data bases including CENTRAL (OVID), CINAHL (EBSCO), Embase.com, Medline (OVID), Web of Science, and PsycInfo (OVID) was conducted. A total of 42 studies examining the effect of pain (n = 38), opiate administration (n = 9), and opiate administration prior to a painful event (n = 5) in rodents were included in this analysis. RESULTS: The results revealed that pain (g = 0.42, 95%CI 0.16-0.67, p = 0.001) increased pain threshold leading to hypoalgesia. Pre-emptive opiate administration had the opposite effect, lowering pain threshold, when compared to pain without prior treatment (g = -1.79, 95%CI -2.71-0.86, p = 0.0001). Differences were found in the meta regression for type of stimulus (thermal: g = 0.66, 95%CI 0.26-1.07, p = 0.001; vs. mechanical: g = 0.13, 95%CI -0.98-1.25, p = 0.81) and gestational age (b = -1.85, SE = 0.82, p = 0.027). In addition, meta regression indicated an association between higher pain thresholds and the amount of cumulative pain events (b = 0.06, SE = 0.03, p = 0.05) as well as severity of pain events (b = 0.94, SE = 0.28, p = 0.001). CONCLUSION: Neonatal exposure to pain results in higher pain thresholds. However, caution is warranted in extrapolating these findings directly to premature infants. Further research is warranted to validate similar effects in clinical contexts and inform evidence-based practices in neonatal care.

Impact of a virtual reality-based simulation training for shoulder dystocia on human and technical skills among caregivers: a randomized-controlled trial.

This study analyzed the adherence to the modified Advanced Life Support in Obstetrics (ALSO) algorithm (HELP-RER) for handling shoulder dystocia (SD) using a virtual reality (VR) training modality. Secondary outcomes were improvements in the post-training diagnosis-to-delivery time, human skills factors (HuFSHI), and perceived task-load index (TLX). Prospective, case-control, single-blind, 1:1 randomized crossover study. Participants were shown a 360° VR video of SD management. The control group was briefed theoretically. Both groups underwent HuFSHI and HELP-RER score assessments at baseline and after the manikin-based training. The TLX questionnaire was then administered. After a washout phase of 12 weeks, we performed a crossover, and groups were switched. There were similar outcomes between groups during the first training session. However, after crossover, the control group yielded significantly higher HELP-RER scores [7 vs. 6.5; (p = 0.01)], with lower diagnosis-to-delivery-time [85.5 vs. 99 s; (p = 0.02)], and TLX scores [57 vs. 68; (p = 0.04)]. In the multivariable linear regression analysis, VR training was independently associated with improved HELP-RER scores (p = 0.003). The HuFSHI scores were comparable between groups. Our data demonstrated the feasibility of a VR simulation training of SD management for caregivers. Considering the drawbacks of common high-fidelity trainings, VR-based simulations offer new perspectives.

Bi-allelic variants in CELSR3 are implicated in central nervous system and urinary tract anomalies.

CELSR3 codes for a planar cell polarity protein. We describe twelve affected individuals from eleven independent families with bi-allelic variants in CELSR3. Affected individuals presented with an overlapping phenotypic spectrum comprising central nervous system (CNS) anomalies (7/12), combined CNS anomalies and congenital anomalies of the kidneys and urinary tract (CAKUT) (3/12) and CAKUT only (2/12). Computational simulation of the 3D protein structure suggests the position of the identified variants to be implicated in penetrance and phenotype expression. CELSR3 immunolocalization in human embryonic urinary tract and transient suppression and rescue experiments of Celsr3 in fluorescent zebrafish reporter lines further support an embryonic role of CELSR3 in CNS and urinary tract formation.

NeoCheck: A New Checklist to Assess Performance during Newborn Life Support-A Validation Study.

BACKGROUND: The aim of this study was to design and validate a new checklist and standardized scenario for assessing providers' performance during Newborn Life Support (NLS). METHODS: We invited twelve experts in Neonatology to take part in a three-step Delphi process. They rated the importance of each item of a newly designed assessment tool to evaluate participants' performance during Newborn Life Support independently on a numeric rating scale from 1 to 5 (1 = lowest; 5 = highest) and were able to give additional comments. All items achieving a mean rating below four after the third round were deleted. For the reliability of the checklist, we calculated interrater reliability. RESULTS: Using a standardized Delphi process, we revised the initial checklist according to the experts' ratings and comments. The final assessment tool includes 38 items covering all relevant steps during NLS. The mean expert rating of all items was 4.40. Interrater reliability showed substantial agreement between the two raters in the first draft (κ = 0.80) as well as in the final draft of the checklist (κ = 0.73). CONCLUSION: We designed a feasible assessment tool for evaluating performance during NLS. We proved the checklist to be valid and reasonable using a Delphi validation process and calculating interrater reliability.

Leri-Weill Dyschondrosteosis Caused by a Leaky Homozygous SHOX Splice-Site Variant.

SHOX deficiency is a common genetic cause of short stature of variable degree. SHOX haploinsufficiency causes Leri-Weill dyschondrosteosis (LWD) as well as nonspecific short stature. SHOX haploinsufficiency is known to result from heterozygous loss-of-function variants with pseudo-autosomal dominant inheritance, while biallelic SHOX loss-of-function variants cause the more severe skeletal dysplasia, Langer mesomelic dyschondrosteosis (LMD). Here we report for the first time the pseudo-autosomal recessive inheritance of LWD in two siblings caused by a novel homozygous non-canonical, leaky splice-site variant in intron 3 of SHOX: c.544+5G>C. Transcript analyses in patient-derived fibroblasts showed homozygous patients to produce approximately equal amounts of normally spliced mRNA and mRNA with the abnormal retention of intron 3 and containing a premature stop codon (p.Val183Glyfs*31). The aberrant transcript was shown to undergo nonsense-mediated mRNA decay, and thus resulting in SHOX haploinsufficiency in the homozygous patient. Six healthy relatives who are of normal height are heterozygous for this variant and fibroblasts from a heterozygote for the c.544+5G>C variant produced wild-type transcript amounts comparable to healthy control. The unique situation reported here highlights the fact that the dosage of SHOX determines the clinical phenotype rather than the Mendelian inheritance pattern of SHOX variants. This study extends the molecular and inheritance spectrum of SHOX deficiency disorder and highlights the importance of functional testing of SHOX variants of unknown significance in order to allow appropriate counseling and precision medicine for each family individual.

Real-Time Intubation and Ventilation Feedback: A Randomized Controlled Simulation Study.

BACKGROUNDS: This study aimed to determine the best educational application of a respiratory function monitor and a video laryngoscope. METHODS: This study was a randomized controlled simulation-based trial, including 167 medical students. Participants had to execute ventilation and intubation maneuvers on a newborn manikin. Participants were randomized into 3 groups. In group A (no-access), the feedback devices were not visible but recording. In group B (supervisor-access), the feedback devices were visible to the supervisor only. In group C (full-access), both the participant and the supervisor had visual access. RESULTS: The two main outcome variables were the percentage of ventilations within the tidal volume target range (4-8mL/kg) and the number of intubation attempts. Group C achieved the highest percentage of ventilations within the tidal volume target range (full-access 63.6%, supervisor-access 51.0%, no-access 31.1%, P < .001) and the lowest mask leakage (full-access 34.9%, supervisor-access 46.6%, no-access 61.6%; A to B: P < .001, A to C: P < .001, B to C: P = .003). Overall, group C achieved superior ventilation quality regarding primary and secondary outcome measures. The number of intubation attempts until success was lowest in the full-access group (full-access: 1.29, supervisor-access: 1.77, no-access: 2.43; A to B: P = .001, A to C: P < .001, B to C: P = .015). CONCLUSIONS: Our findings confirm that direct visual access to feedback devices for supervisor and trainees alike considerably benefits outcomes and can contribute to the future of clinical education.

The consequences of neonatal pain, stress and opiate administration in animal models: An extensive meta-analysis concerning neuronal cell death, motor and behavioral outcomes.

This systematic review and meta-analysis aimed to investigate the association of neonatal exposure to pain, stress, opiate administration alone, as well as opiate administration prior to a painful procedure on neuronal cell death, motor, and behavioral outcomes in rodents. In total, 36 studies investigating the effect of pain (n = 18), stress (n = 15), opiate administration (n = 13), as well as opiate administration prior to a painful event (n = 7) in rodents were included in our meta-analysis. The results showed a large effect of pain (g = 1.37, 95% CI 1.00-1.74, p < .001) on neuronal cell death. Moreover, higher number of neonatal pain events were significantly associated with increased neuronal cell death, increased anxiety (b = -1.18, SE = 0.43, p = .006), and depressant-like behavior (b = 1.74, SE = 0.51, p = .027) in rodents. Both opiates and pain had no impact on motor function (g = 0.26, 95% CI 0.18-0.70, p = .248).

Saliva Sampling for Prospective SARS-CoV-2 Screening of Healthcare Professionals.

OBJECTIVE: The objective of this study was to analyze the feasibility and acceptance of a non-invasive, daily and proactive screening program for SARS-CoV-2 infection employing serial saliva testing, in combination with a digital questionnaire among healthcare providers (HCPs) in a multi-professional setting. DESIGN: This was a prospective cohort study involving HCPs from different units at a single tertiary care center, over a pilot phase of 4 weeks during the first wave of the COVID-19 pandemic from April 18th to June 6th, 2020. SETTING: Pediatric tertiary patient care units, Comprehensive Center for Pediatrics, Medical University of Vienna. SUBJECTS: HCPs from different units, including physicians, nurses, midwives, and administrative staff (with patient contact) were considered eligible for the study. Study participants were working in different settings in our center at varying levels of risk exposure. INTERVENTIONS: Saliva collection from mouth gargle and electronic symptom and exposure monitoring (eSEM) was performed by participants at the onset of each regular clinical shift (day or night shift), using an anonymous ID for matching the results. MEASUREMENTS: RT-PCR of all saliva samples, eSEM, as well as feasibility and acceptance thereof. RESULTS: Two hundred and seventy-five volunteers collected 1,865 saliva samples and responded 1,378 times in the eSEM during a 4-week period. 1,331 (96.7%) responses were that the testing was feasible and acceptable. The most common severe symptom during the 4-week period mentioned by HCPs was headache, reported 54 times (3.9%). Two SARS-CoV-2 positive samples-one of them being associated with symptoms-were identified. The acceptance rate among HCPs was 96.6%. CONCLUSION: Serial saliva screening was a well-accepted and feasible method for monitoring SARS-CoV-2 infectious state in health care professionals. Combination of regular SARS-CoV-2 tests with sequential saliva collection and storage could potentially represent a highly efficient strategy to identify and trace virus positive staff for employee and patient safety.

NeoAct: A Randomized Prospective Pilot Study on Communication Skill Training of Neonatologists.

Background: This randomized interventional study evaluated the impact of a 1-day experiential communication skills training on neonatologists' performance in doctor-parents-communication. Methods: 17 neonatologists with different levels of professional experience from the Medical University of Vienna were randomized into one of two study groups: The intervention group (IG) as opposed to the control group (CG) participated in a 1-day experiential communication training. Eight weeks after the training, participants' communication skills were assessed during an objective structured clinical examination (OSCE). Neonatologists were assessed in a simulated conversation by how effectively they performed when conveying complex health-related information to parents of ill infants. Participants in the control group (CG) were assessed first during the OSCE and received their communication training later on. Self-assessment questionnaires before and after the workshop and OSCE were completed. Results: The study determined that neonatologists in the IG subjectively perceived that their competence level regarding their communication skills had increased after the workshop, while this was not reflected by their performance during the OSCE assessment. Discussion: A 1-day experiential communication skills training significantly increased physicians' self-evaluation concerning their communicative competence. This perceived competence did not manifest itself in increased communication skills during the OSCE. Conclusion: Repeated training is needed.

Impact of a Multifactorial Educational Training on the Management of Preterm Infants in the Central-Eastern European Region.

Background: Differences in management and outcomes of extremely preterm infants have been reported across European countries. Implementation of standardized guidelines and interventions within existing neonatal care facilities can improve outcomes of extremely preterm infants. This study evaluated whether a multifactorial educational training (MET) course in Vienna focusing on the management of extremely preterm infants had an impact on the management of extremely preterm infants in Central-Eastern European (CEE) countries. Methods: Physicians and nurses from different hospitals in CEE countries participated in a two-day MET in Vienna, Austria with theoretical lectures, bedside teaching, and simulation trainings. In order to evaluate the benefit of the workshops, participants had to complete pre- and post-workshop questionnaires, as well as follow-up questionnaires three and twelve months after the MET. Results: 162 participants from 15 CEE countries completed the two-day MET at our department. Less invasive surfactant administration (LISA) was only used by 39% (63/162) of the participants. After the MET, 80% (122/152) were planning to introduce LISA, and 66% (101/152) were planning to introduce regular simulation training, which was statistically significantly increased three and twelve months after the MET. Thirty-six percent and 57% of the participants self-reported improved outcomes three and twelve months after the MET, respectively. Conclusion: Our standardized training in Vienna promoted the implementation of different perinatal concepts including postnatal respiratory management using LISA as well as regular simulation trainings at the participants' home departments. Moreover, our MET contributed to dissemination of guidelines, promoted best-practice, and improved self-reported outcomes.

Pain and Sedation Scales for Neonatal and Pediatric Patients in a Preverbal Stage of Development: A Systematic Review.

Importance: Because children in a preverbal stage of development are unable to voice their feelings, they completely depend on their caregiving team for the interpretation and management of their pain and discomfort. Thus, accurately validated scales to assess pain and sedation levels are crucial. Objective: To provide clinicians a complete overview on the validity and reliability of the existing pain and sedation scales for different target populations (preterm infants, term infants, and toddlers) and in different clinical contexts. Evidence Review: BIOSIS Previews, Cumulative Index to Nursing and Allied Health Literature, Cochrane Central Register of Controlled Trials, Cochrane Database of Systematic Reviews, Embase, MEDLINE, PsycCRITIQUES, PsycINFO, PSYNDEXplus Literature and Audiovisual Media, and PSYNDEXplus Tests were the databases screened from their inception to August 2018. All studies examining the validity or reliability of a given pain or sedation scale for patients in a preverbal stage of development were included in this systematic review. Those scales that were tested for at least construct validity, internal consistency, and interrater reliability were subsequently scored using the consensus-based standards for the selection of health measurement instruments (COSMIN) checklist. Findings: In total, 89 validation articles comprising 65 scales were included. Fifty-seven scales (88%) were useful for assessing pain, 13 scales (20%) for assessing sedation, and 4 scales (6%) for assessing both conditions. Forty-two (65%) were behavioral scales, and 23 (35%) were multidimensional scales. Eleven scales (17%) were validated for infants on mechanical ventilation. Thirty-seven scales (57%) were validated for preterm infants, 24 scales (37%) for term and preterm infants, 7 scales (11%) for term-born children, 7 scales (11%) for preterm infants, term infants, and toddlers, and 17 scales (26%) for term infants and toddlers. Twenty-eight scales (43%) considered construct validity, internal consistency, and interrater reliability. Conclusions and Relevance: Clinicians should consider using scales that are validated for at least construct validity, internal consistency, and interrater reliability, combining this information with the population of interest and the construct the scale is intended to measure.

Effects of Feedback on Chest Compression Quality: A Randomized Simulation Study.

OBJECTIVES: Our aim for this study was to test whether visual and verbal feedback compared with instructor-led feedback improve the quality of pediatric cardiopulmonary resuscitation (CPR). METHODS: There were 653 third-year medical students randomly assigned to practice pediatric CPR on 1 of 2 manikins (infant and adolescent; n = 344 and n = 309, respectively). They were further randomly assigned to 1 of 3 feedback groups: The instructor feedback (IF) group (n = 225) received traditional, instructor-led feedback without any additional feedback device. The device feedback (DF) group (n = 223) received real-time visual feedback from a feedback device. The instructor and device feedback (IDF) group (n = 205) received verbal feedback from an instructor who continuously reviewed the trainees' performance using the feedback device. After the training, participants' CPR performance was assessed on the same manikin while no feedback was being provided. RESULTS: For the primary outcome of total compression score, participants in the DF and IDF groups performed similarly, with both groups showing scores significantly (P < .001) better than those of the IF group. The same findings held for correct hand position and the proportion of complete release. For compression rate, the DF group was at the higher end of the guideline for 100 to 120 chest compressions per minute compared with the IF and IDF groups (both P < .001). No effect of feedback on compression depth was found. CONCLUSIONS: Chest compression performance significantly improved with both visual and verbal feedback compared with instructor-led feedback. Feedback devices should be implemented during pediatric resuscitation training to improve resuscitation performance.