Performance Evaluation of a Supervised Machine Learning Pain Classification Model Developed by Neonatal Nurses.

BACKGROUND: Early-life pain is associated with adverse neurodevelopmental consequences; and current pain assessment practices are discontinuous, inconsistent, and highly dependent on nurses' availability. Furthermore, facial expressions in commonly used pain assessment tools are not associated with brain-based evidence of pain. PURPOSE: To develop and validate a machine learning (ML) model to classify pain. METHODS: In this retrospective validation study, using a human-centered design for Embedded Machine Learning Solutions approach and the Neonatal Facial Coding System (NFCS), 6 experienced neonatal intensive care unit (NICU) nurses labeled data from randomly assigned iCOPEvid (infant Classification Of Pain Expression video) sequences of 49 neonates undergoing heel lance. NFCS is the only observational pain assessment tool associated with brain-based evidence of pain. A standard 70% training and 30% testing split of the data was used to train and test several ML models. NICU nurses' interrater reliability was evaluated, and NICU nurses' area under the receiver operating characteristic curve (AUC) was compared with the ML models' AUC. RESULTS: Nurses weighted mean interrater reliability was 68% (63%-79%) for NFCS tasks, 77.7% (74%-83%) for pain intensity, and 48.6% (15%-59%) for frame and 78.4% (64%-100%) for video pain classification, with AUC of 0.68. The best performing ML model had 97.7% precision, 98% accuracy, 98.5% recall, and AUC of 0.98. IMPLICATIONS FOR PRACTICE AND RESEARCH: The pain classification ML model AUC far exceeded that of NICU nurses for identifying neonatal pain. These findings will inform the development of a continuous, unbiased, brain-based, nurse-in-the-loop Pain Recognition Automated Monitoring System (PRAMS) for neonates and infants.

Impact of Parent Presence and Engagement on Stress in NICU Infants.

BACKGROUND: Infants in neonatal intensive care units (NICUs) are exposed to frequent stressors that impact their neurodevelopmental outcomes. Parent presence and engagement are considered critical to improving infant outcomes, yet associations between cumulative NICU parent presence, engagement, and infant stress are infrequently examined. PURPOSE: To examine associations between NICU infant stress and the amount (hours per week) or frequency (days per week) of parent presence and skin-to-skin care (SSC). METHODS: A secondary analysis of a data set representing 78 NICU families was conducted. Infant acuity was measured using Neonatal Medical Index (NMI) scores. Parent presence and SSC data were collected from electronic medical records. Infant stress was measured using resting salivary cortisol levels collected at NICU discharge (median = 33 days of life). RESULTS: More cumulative SSC was associated with lower discharge cortisol in NICU infants for SSC measured in hours per week (P = .03) or days per week (P = .05). Cumulative parent presence was not significantly associated with infant cortisol at discharge. Hierarchical regression analyses examining timing of parent presence supported a model including admission cortisol, NMI score, and parent presence during weeks 1 to 4 of life for explaining infant stress at discharge (R2 = 0.44, P = .004). Analyses examining timing of SSC supported a model including admission cortisol, NMI score, and frequency of SSC during week 1 for explaining infant stress at discharge (R2 = 0.21, P = .04). IMPLICATIONS FOR PRACTICE AND RESEARCH: Early, frequent SSC to mitigate stress in NICU infants was supported. Results suggested that timing of parent presence impacts NICU infant stress; however, additional study is recommended.

Understanding family-centered care in the NICU: a scoping review protocol.

OBJECTIVE: This scoping review will aim to identify and categorize the definitions of neonatal intensive care unit (NICU) family-centered care (FCC) and its associated concepts. It also aims to identify and categorize the practices and interventions that comprise NICU FCC, and catalog the metrics used to evaluate NICU FCC. INTRODUCTION: FCC has been identified as an important element of care for neonates and infants admitted to the NICU, and there is clear evidence that the incorporation of families in care improves clinical outcomes. However, FCC has been linked to numerous associated terms and concepts and lacks a unifying definition or framework, thus limiting the ability to categorize, prioritize, and identify practices and interventions to optimize both institutional approaches for individual centers and for the field at large. INCLUSION CRITERIA: Studies that include or apply at least one FCC concept or its associated terms will be considered eligible for inclusion. Studies not related exclusively to the NICU will be excluded. METHODS: The review will follow the JBI methodology for scoping reviews and will be reported using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR). Several electronic databases and sources of gray literature will be searched from 1992 to the present day. The review will include only full-text studies in English and will be independently screened by a minimum of 2 authors. Data will be extracted using a modified JBI data extraction tool and presented using narrative summaries; concept mapping; and categorization of practices, interventions, and metrics.

Application of a human-centered design for embedded machine learning model to develop data labeling software with nurses: Human-to-Artificial Intelligence (H2AI).

BACKGROUND: Nurses are essential for assessing and managing acute pain in hospitalized patients, especially those who are unable to self-report pain. Given their role and subject matter expertise (SME), nurses are also essential for the design and development of a supervised machine learning (ML) model for pain detection and clinical decision support software (CDSS) in a pain recognition automated monitoring system (PRAMS). Our first step for developing PRAMS with nurses was to create SME-friendly data labeling software. PURPOSE: To develop an intuitive and efficient data labeling software solution, Human-to-Artificial Intelligence (H2AI). METHOD: The Human-centered Design for Embedded Machine Learning Solutions (HCDe-MLS) model was used to engage nurses. In this paper, HCDe-MLS will be explained using H2AI and PRAMS as illustrative cases. FINDINGS: Using HCDe-MLS, H2AI was developed and facilitated labeling of 139 videos (mean = 29.83 min) with 3189 images labeled (mean = 75 s) by 6 nurses. OpenCV was used for video-to-image pre-processing; and MobileFaceNet was used for default landmark placement on images. H2AI randomly assigned videos to nurses for data labeling, tracked labelers' inter-rater reliability, and stored labeled data to train ML models. CONCLUSIONS: Nurses' engagement in CDSS development was critical for ensuring the end-product addressed nurses' priorities, reflected nurses' cognitive and decision-making processes, and garnered nurses' trust for technology adoption.

Implementing Guidelines for NICU Parent Presence: Effects on Parent and Infant Stress.

BACKGROUND: Relationships between parents and infants are essential for mitigating stressors encountered in neonatal intensive care units (NICUs) and are supported by parent presence and engagement. PURPOSE: The purpose of this study was to compare NICU parent and infant outcomes pre- and postimplementation of an intervention aimed at increasing parent presence and engagement in the NICU. This family-centered care intervention consisted of communicating specific guidelines for parent presence. METHODS: Data related to parent presence, skin-to-skin care, and breastfeeding; parental stress; infant outcomes including weight gain, length of stay, feeding status at discharge, and stress; and unit-level outcomes were collected from a convenience sample of 40 NICU families recruited preimplementation and compared with data for 38 NICU families recruited postimplementation of specific guidelines for parent presence. To establish comparability of groups, infants were assigned scores using the Neonatal Medical Index. RESULTS: Parent presence, engagement in skin-to-skin care, and breastfeeding rates were not significantly different between groups. Stress-related outcomes were significantly decreased in NICU mothers, fathers, and infants, and infant feeding outcomes were improved in the postintervention group. CONCLUSIONS: Specific guidelines for parent presence may represent an invitation for parents to engage with their NICU infants and may positively impact parent and infant stress.

Setting the stage for successful oral feeding: the impact of implementing the SOFFI feeding program with medically fragile NICU infants.

To evaluate the impact of implementing the Supporting Oral Feeding in Fragile Infants (SOFFI) program in a tertiary-level neonatal intensive care unit (NICU) on the oral feeding, growth, and length of stay outcomes of a heterogeneous population of medically fragile infants at discharge and feeding and growth outcomes postdischarge at 3 to 5 months postterm. Data related to feeding, growth, and length of stay from a convenience sample of 81 infants recruited pre-SOFFI training were compared with data for 75 infants recruited post-SOFFI training of the NICU staff. Subjects were not excluded on the basis of level of illness or medical diagnoses. To establish comparability of subject groups, infants were assigned scores using the Neonatal Medical Index. At 3 to 5 months postterm, semistructured parent phone interviews related to feeding and growth at home were conducted (n = 128). Post-SOFFI infants born at less than 37 weeks' gestation achieved full oral feedings in significantly fewer days than pre-SOFFI infants (P = .01). Time to achieve full oral feedings was not significantly different in post-SOFFI infants born at 37 or more weeks' gestation. Growth and length of stay were not significantly different at discharge. At follow-up, parents of post-SOFFI infants reported significantly fewer feeding problems overall (P = .01), less arching (P = .003), less vomiting (P = .006), and fewer infants seeing feeding specialists (P = .03). Results of the study support that NICU implementation of the SOFFI feeding program positively influences feeding outcomes before and following discharge.