In utero lead exposure and auditory neural myelination in premature infants.

Oral motor interventions are used in the neonatal intensive care unit (NICU) to support the development of oral feeding skills in preterm infants. Literature regarding which oral motor interventions are used, when they are implemented, and how parents are involved is lacking. To benchmark evidence of oral motor interventions in the literature against reported use in NICU practice, understand the type and timing of oral motor interventions used, and understand perspectives about implementation of oral motor interventions and family involvement in these interventions. Qualitative study. Virtual, recorded focus groups and interviews on Zoom. Convenience sampling was used to recruit neonatal therapists with at least 2 yr experience in the NICU who were currently implementing oral motor interventions. Duringsemistructured focus groups and interviews, we specifically probed the type of oral motor interventions that clinicians use in the NICU to benchmark them against the interventions found in the literature. Recordings were transcribed and uploaded to NVivo for descriptive analysis. Twelve neonatal therapists participated in focus groups or interviews. Applied oral motor stimulation, nonnutritive sucking, NTrainer®intervention, swallowing exercises, and milk drops were reported to be used in practice and most often were started as early as 29 wk postmenstrual age. Parents were primarily involved in facilitating nonnutritive sucking and providing milk drops. Oral motor interventions reported to be used in NICU clinical practice mirror those available in the evidence. This supports ongoing implementation of evidence-based practice for neonatal therapists. Plain-Language Summary: A recent integrative review was completed to understand oral motor interventions used to support the development of feeding skills in preterm infants. The interventions identified in the scientific literature were nonnutritive sucking, applied oral motor stimulation, swallowing exercises, and use of a pacifier that elicits a sensory pulse to stimulate sucking called the NTrainer. However, there was no published information regarding which (if any) of these interventions are actually used in clinical practice and how parents may be involved in their implementation at the neonatal intensive care unit bedside. Therefore, this qualitative study sought to understand oral motor interventions used in clinical practice. All of the interventions identified in the scientific literature were identified as being used in practice, although at varying rates. An additional intervention, milk drops, was also identified. Nonnutritive sucking and milk drops were most commonly used by the study participants and were most commonly taught to parents to implement with their infants. These findings support neonatal therapists' implementation of evidence-based practice in the clinical setting.

Evaluation of timing of repair of congenital diaphragmatic hernia in premature infants.

Effects of Remote Patient Monitoring on Neonatal Intensive Care Unit Patients Discharged with Nasogastric Tube Feeding.

Automated estimation of EEG maturity in preterm neonates and its association with long-term outcome.

Preterm infants are at risk of growth faltering at term age. Our primary objective is to assess post-discharge growth patterns in these infants and investigate the association between growth faltering and neurodevelopment. We divided the sample into two groups according to growth during the initial hospital stay: infants who suffered from growth faltering (GF, loss of >1 weight z-score from birth to 36 weeks postmenstrual age, n = 115) and infants who did not suffer from GF (non-growth faltering, NGF, n = 85). The NFG group weight z-score was significantly lower at 36 postmenstrual ages (PMA) compared to birth (p < 0.001), at 1-year corrected age (CA), it was significantly higher than at birth (p = 0.0026), and by 2 years CA, there were no differences compared to the birth z-scores. In the GF infants' group, statistical differences were found at all time points. At 3 and 6 months, CA GF infants were still in weight z-score values lower than -1 point compared to the birth median value. At 12 and 24 months CA, they still had not achieved birth z-score values (p < 0.001). In the Parent Report of Children's Abilities-Revised (PARCA), NGF infants had a higher score in the language development scale at 2 years than GF infants (88.5 [78.5; 96.5] vs. 84.5 [69.5; 91.5], p = 0.03). The Bayley-III test was available for 35 infants. We found a significant difference in motor development, with a higher score in the NGF group (94 [88; 100] vs. 85 [79; 91], p = 0.03). In this cohort study, GF is associated with growth differences till 2 years CA, and with lower scores in neurodevelopment assessment.

Neonatal doses for the off-patent antibiotics fosfomycin and flomoxef, which offer coverage against many extended-spectrum beta-lactamase (ESBL)-producing organisms, are based on limited data. We performed a pharmacokinetic (PK) and safety study of fosfomycin and flomoxef to confirm proposed neonatal dosing before further investigation in a trial (NeoSep1, ISRCTN48721236). Neonates with suspected sepsis, weighing more than 1,000 g, were sequentially enrolled into three antibiotic treatment cohorts: fosfomycin and amikacin (Cohort 1), flomoxef and amikacin (Cohort 2), and flomoxef and fosfomycin (Cohort 3), and followed for 28 days. Plasma samples were taken for PK assessment, with population PK modeling and simulations performed. Sixty-two neonates (48/62 [77%] preterm; 48/62 [77%] ≤7 days postnatal age [PNA]) received at least one dose of study antibiotics. Fosfomycin and flomoxef plasma concentrations were best described by a two-compartment and a one-compartment model, respectively, with postmenstrual age and PNA significantly influencing clearance. The probability of target attainment for fosfomycin was 100% for minimum inhibitory concentrations (MICs) of up to 8 mg/L, and for flomoxef, it was 100% for MICs of up to 0.5 mg/L. Adverse events (AEs) were common in this critically ill cohort. Thirteen (21%) neonates developed 19 trial antibiotic-related AEs (17 with grade ≤2, and 2 of grade 3), none of which required modification or discontinuation of allocated treatment. Seven neonates (11.6%) died. In this predominately preterm population, fosfomycin and flomoxef were safe, with drug exposures similar to published studies supporting the proposed doses for the larger, randomized NeoSep1 trial.This study is registered with ISRCTN48721236.

BackgroundGrowth of head circumference is critically associated with neurodevelopmental outcomes. Extrauterine growth restriction of head circumference from birth to term-equivalent age is linked to impaired neurodevelopment. This study examined whether a proinflammatory state and metabolic dysregulation characterize the association between delayed feeding progression and extrauterine restricted head growth in extremely preterm infants.MethodsThis cohort study included infants born ≤ 28 weeks’ gestation between 2019 and 2021. Feeding progression trajectories, categorized as improvement or delayed improvement based on daily enteral feeding milk volumes during the first 8 weeks, were analyzed using kmlShape. Plasma metabolomics were assessed at 36 weeks postmenstrual age, and head growth and brain MRI were evaluated at term-equivalent age.ResultsAmong the 98 extremely preterm infants, 62 (63%) demonstrated improvement in feeding progression, while 36 (37%) had delayed improvement. Compared to the feeding improvement group, the delayed feeding improvement group had higher rates of gastrointestinal morbidities, including necrotizing enterocolitis (NEC) of Bell stage II or higher (17% vs. 2%, p = 0.009) and abdominal surgery for non-NEC events (25% vs. 8%, p = 0.021) during admission, and a significantly increased risk of extrauterine growth restriction in head circumference by term-equivalent age (47% vs. 23%, p = 0.021). The multivariable analysis showed delayed feeding improvement was also a significant risk associated with the delta z-scores below − 1.5 in head circumference (adjusted odds ratio [aOR]: 5.26 [95% CI 1.66–16.65]). MRI examinations revealed significantly smaller residual brain volumes involving total brain tissue volume, brainstem, and cerebellum in the delayed improvement group. Untargeted plasma metabolomics showed elevated levels of hydroxyeicosatetraenoic acid, leukotriene B4, prostaglandins, bile acids and immune markers, and reduced levels of L-tyrosine, phenylpyruvic acid, L-tryptophan metabolism, and L-carnitine biosynthesis were found in the delayed improvement group compared to that in the improvement group.ConclusionsProinflammatory and dysregulated metabolic state following early delayed feeding progression were associated with impaired extrauterine head growth, highlighting the potential role of the immature gut-brain axis in preterm infants.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12916-025-04525-w.

BackgroundInfants born preterm are at high risk of anemia, red blood cell transfusions, and iron deficiency, all of which may negatively influence long-term neurodevelopment. To ameliorate these complications of prematurity, we developed a Phase II trial to determine whether treatment with an erythropoietic-stimulating agent, darbepoetin (Darbe), plus a slow-release intravenous (IV) iron preparation (ferumoxytol (FMX) or low-molecular-weight iron dextran (LMW-ID)) might decrease transfusions while maintaining iron sufficiency.MethodsThis single-center study is a parallel design, prospective, randomized controlled Phase II trial of 120 infants born 24–0/7 to 31–6/7 weeks of gestation cared for in the University of Washington Neonatal Intensive Care Unit. After informed consent, infants less than 72 h of age are randomized to one of five treatment groups: (1) Oral iron (standard care), n = 40, or weekly Darbe 10 µg/kg/dose IV or SQ plus; (2) FMX — 10 mg/kg/dose IV, n = 20; (3) FMX — 20 mg/kg/dose IV, n = 20; (4) LMW-ID — 10 mg/kg/dose IV, n = 20; or (5) LMW-ID — 20 mg/kg/dose IV, n = 20. Infants will be followed to 2-year corrected age with sequential developmental testing. Our primary outcome is ferritin level at 34–36 weeks postmenstrual age. Secondary outcomes include other hematologic assessments, drug safety, evaluation of the gut microbiome, and neurodevelopment to 2 years corrected age.DiscussionThis trial will determine whether darbepoetin plus a slow-release IV iron preparation is safe, which iron preparation and dose best maintain iron sufficiency and decrease or eliminate transfusions, whether IV iron will result in a more diverse, less pathogenic microbiome when compared to oral iron supplementation, and, finally, whether these treatments affect neurodevelopment to 2 years corrected age.Trial registrationNational Clinical Trial (NCT) NCT05340465. Registered on March 1, 2022.

Test the effect of thickened liquids on preterm infants (25-35 weeks of gestation) who are struggling to achieve full oral feeds. Retrospective case-control study of preterm infants struggling to achieve full oral feeds at term, prompting provision of thickened liquids without aspiration on instrumental assessment, and controls, matched for gestational age and comorbidities, who followed a typical thin liquid feeding progression. Paired t-tests were used to test differences in overall milk transfer (%, milk consumed/prescribed) before and after thickening, with mixed linear regression to test differences between the control and thickened group. 38 infants (19 thickened) were included. Prior to thickening, infants struggling with feeds had significantly slower rates of improvement in overall transfer than controls (struggling feeders, 1.3%/day; controls, 4.5%/day) (p < .001). Overall transfer increased 13% 48 h after thickening (p = 0.01). Thickened liquids may be an effective treatment for select preterm infants who are not progressing to full oral feeds at term postmenstrual age, even if aspiration is not observed on instrumental assessment.

A Prospective, Multicenter, Observational StUdy of PREMEdication Before Laryngoscopy in NEOnates (SUPREMEneo).

Adequate nutritional intake is necessary for optimal growth and development of high-risk newborns, specifically preterm infants and newborns with congenital diaphragmatic hernia (CDH). They face challenges initiating and advancing enteral and oral feeding, originating from mechanical dysfunction of gastrointestinal motility, which probably results from delayed or deviant maturation of the autonomous nervous system (ANS). This often leads to long-term feeding difficulties and failure to thrive. Antenatally, ANS development is influenced by various forms of sensorimotor stimulation. Postnatally, development may be delayed, partly due to reduced sensory inputs. A sensorimotor stimulation programme was developed directed at improving gastrointestinal functionality: HAPTOS (Handling Adapted to Postnatal age with Tactile-kinaesthetic and Oral sensorimotor Stimulation). This multicentre, randomized clinical trial in preterm ( < 30 weeks gestation) and CDH infants compares standard care to HAPTOS in addition to standard care. Primary outcomes are postnatal age at full enteral feeding and postmenstrual age at full oral feeding. Secondary outcomes are short-term clinical outcomes, short-term autonomic regulation, short-term care and parental participation, long-term growth and wellbeing, and long-term neurodevelopment. This study evaluates HAPTOS: a novel concept of developmental care to improve functional outcomes in high-risk newborns IMPACT: High-risk newborns often present with immature gastrointestinal motility, leading to challenges initiating and progressing with enteral and oral feeding with long-term sequelae. Sensorimotor stimulation has positive effects on feeding outcomes but current protocols are unsuitable for high-risk newborns. Thus, a new sensorimotor stimulation programme was developed: HAPTOS (Handling Adapted to Postnatal age with Tactile-kinaesthetic and Oral sensorimotor Stimulation). This randomized controlled trial assesses clinical effects of HAPTOS and provides evidence for the role of autonomic regulation in functional maturation. HAPTOS intervention is a novel concept of developmental care to improve gastrointestinal outcomes and overall wellbeing in high-risk newborns.

Background/Objectives: Respiratory morbidity in preterm infants has been widely studied, with evidence showing that individuals born prematurely experience more frequent respiratory symptoms, airflow obstruction, and radiological lung abnormalities throughout life. Methods: This study included 150 children aged 6 to 11 years, divided into two groups. The preterm group (n = 90) consisted of children born before 32 weeks of gestation, while the control group (n = 60) included term-born children. All participants underwent spirometry and completed a respiratory health questionnaire. Results: A significantly higher proportion of preterm children exhibited respiratory morbidity compared to term-born peers (χ2 = 7.035; p = 0.030). However, no significant differences were found between preterm children with and without bronchopulmonary dysplasia (BPD) defined at day 28 (BPD28) (χ2 = 0.002; p = 0.968) or BPD defined at 36 weeks postmenstrual age (BPD36) (χ2 = 0.029; p = 0.864). Lung function parameters—forced vital capacity (FVC), forced expiratory volume in 1 s (FEV1), FEV1/IVC, maximal expiratory flow at 25%, 50% and 75% of FVC (MEF25, MEF50, and MEF75) were significantly lower in the preterm group (p = 0.004 for FVC and p &lt; 0.001 for all other parameters). No significant differences were observed between BPD and non-BPD subgroups in any lung function parameter. BPD36 was found to be a stronger predictor of respiratory morbidity (OR = 1.214) than BPD28 (OR = 1.093), although neither BPD28 nor BPD36 were statistically significant predictors. Conclusions: This study underscores the long-term respiratory consequences of prematurity and challenges the traditional view of BPD as the primary determinant of poor respiratory outcomes. Our findings suggest that prematurity itself, rather than BPD, may play a more central role.

ABSTRACT Objective Most antiseizure medications (ASMs) are prescribed off label for neonates. Lacosamide's efficacy in infants and availability in intravenous formulation suggest potential utility for neonates. We evaluated the safety and efficacy of lacosamide for neonatal seizures. Methods This 10‐center, retrospective study of neonates with seizures and lacosamide treatment initiated by ≤ 48 weeks postmenstrual age collected clinical data from medical records and electroencephalogram recordings. Lacosamide efficacy was determined by changes in seizure burden with lacosamide treatment and seizure cessation by hospital discharge. Potential adverse events were reviewed. Results Among 62 eligible neonates, 33 had acute provoked seizures while 29 had neonatal‐onset epilepsy; there was no difference in seizure type or baseline seizure severity between groups. There were high rates of pretreatment status epilepticus (48%) and treatment‐resistant seizures, with 93% receiving ≥ 3 ASMs before lacosamide. Most received intravenous lacosamide, with a median loading dose of 5.0 mg/kg and median daily dose of 7.3 mg/kg. Seizure cessation occurred in 37% of neonates; 21% had no additional ASM administered after lacosamide. Seizure burden, measured in seizure minutes per hour, was lower at both 4 h and 7 days following lacosamide administration. In addition, there was a median reduction in seizure frequency of 30 seizures per day at 7 and 30 days posttreatment ( p &lt; .05). Lacosamide was continued at discharge in most neonates (72%). Seventy adverse events were reported in 35 (56%) neonates. Four transient events with possible or unknown relationship to lacosamide were likely multifactorial in origin; none were cardiac arrhythmias. Summary Despite high rates of treatment‐resistant seizures in this neonatal cohort, 37% experienced seizure cessation and most remained on lacosamide at hospital discharge. Most adverse events were not attributed to lacosamide. These results favor use of lacosamide and provide a rationale for future prospective studies.

Extremely preterm infants may use medical technology after discharge from neonatal intensive care. The aim of the study was to determine which inpatient morbidities have the strongest associations with technology at toddler-age follow-up. Retrospective cohort analysis of 3904 extremely preterm infants born 22.0-26.6 weeks' gestation from 2014 to 2019 who survived to 36 weeks' postmenstrual age and had data on medical technology at 22-26 months' corrected gestational age. 18.8% of children used medical technology; 10.1% used one and 8.7% ≥ 2. Use of a gastrostomy tube was most common (12.8%), followed by pulse oximeter (8.2%), oxygen (5.9%), tracheostomy (3.9%), shunt for hydrocephalus (3.6%), ventilator/continuous positive airway pressure (2.2%), apnea monitor (1.4%), and total parenteral nutrition (0.3%). After adjusting for significant maternal and infant characteristics, Grade 2 or 3 bronchopulmonary dysplasia (BPD) was most strongly associated with medical technology (aOR (95% CI): 3.20 (2.65, 3.87)), followed by serious brain injury (SBI) 3.06 (2.55, 3.66) and surgical NEC (sNEC) 2.67 (1.84, 3.87). In this cohort of extremely preterm infants, BPD, SBI and sNEC were most associated with medical technology use at toddler-age. These findings provide information for counseling of families and support during discharge planning. Generic Database: NCT00063063. In this cohort of extremely preterm infants <27 weeks' gestation at birth, nearly 1 in 5 children used medical technology at 22-26 months' corrected age. The inpatient morbidities of bronchopulmonary dysplasia, serious brain injury, and surgical necrotizing enterocolitis were most associated with persistent medical technology use. These findings provide important information for counseling families of children with these morbidities during the hospital stay and call for increased support of these families after discharge.

OBJECTIVE Evaluate the effectiveness of an algorithm using urine sodium (Na) concentration to guide dietary Na supplementation compared to standard care to promote growth in preterm infants. METHODS Randomized trial conducted in four neonatal units in Indianapolis, Indiana involving infants between 250/7- 296/7 weeks gestation with birthweight &amp;gt;500 grams. Study arms consisted of standard care or use of algorithm beginning at 14-17 days postnatal age. The primary outcomes were somatic growth (weight, length, head circumference) measured by change in Z-score between 2 weeks of age and 36 weeks postmenstrual age (PMA). Secondary outcomes included total body water and body composition. RESULTS Of 260 eligible infants, 90 were consented and randomized; 45 allocated to each arm, with 43 in each arm completing the study. Thirty-eight (88%) infants randomized to study algorithm had a low urine Na which qualified for Na supplementation (4 mEq/kg/d). Algorithm infants received more Na (5.75 +/- 0.96 mEq/kg/d) than standard care infants (3.62 +/- 1.25 mEq/kg/d); caloric intake did not differ. Change in weight Z-score from study entry until 36 weeks PMA was not statistically different between groups (0.16 +/- 0.41 vs. 0.04 +/-0.63, algorithm vs. control, p=0.16). Algorithm infants displayed a significantly more rapid increase in weight between weeks 0 to 7 of study than standard care infants. Total body water and fat-free mass did not differ at 32 weeks PMA. CONCLUSIONS Na supplementation guided by urine Na concentration did not significantly improve weight gain in infants 250/7 - 296/7 weeks gestation at 36 weeks PMA.

To analyze the serum metabolomic changes of preterm infants with bronchopulmonary dysplasia (BPD) at postmenstrual age (PMA) 36 weeks, screen potential biomarkers and associated metabolic pathways, and assess their relationship with short-term respiratory outcomes. A retrospective case-control study was conducted. Infants with gestational age 28-32 weeks admitted to the Children's Hospital Affiliated to Zhengzhou University from January to December 2024 were included. Twenty infants with BPD and 20 gestational age-, birth weight-, and sex-matched non-BPD preterm infants were included. Serum collected at PMA 36 weeks was subjected to untargeted metabolomics analysis, and associations with short-term respiratory outcomes were analyzed. Thirteen potential biomarkers distinguishing BPD were identified (area under the curve >0.75, P<0.05). Eight biomarkers-including terephthalic acid, phosphatidylinositol, fumarate, and lysophosphatidic acid-were significantly upregulated (FC≥1.5), while five biomarkers, such as 7α-hydroxy-3-oxo-4-cholestenoate ester and phosphatidylcholine, were significantly downregulated (FC≤1/1.5). Pathway analysis indicated five pathways associated with BPD, including glycerophospholipid metabolism and phenylalanine metabolism. Dysregulation of glycerophospholipid and bile acid metabolism may affect adverse short-term respiratory outcomes in infants with BPD. The 13 significantly different metabolites may serve as biomarkers for the diagnosis of BPD. Glycerophospholipid metabolism is associated with the occurrence of BPD and with adverse short-term respiratory outcomes.

Background/Objectives: Preterm infants often require increased caloric intake to maintain appropriate growth while in the neonatal intensive care unit (NICU). Emerging evidence suggests that alterations of the gut microbiome may play a role in infant and childhood growth patterns. The fecal microbiome patterns in infants with normal and poor growth patterns were classified in this study. Methods: We conducted a prospective trial of infants of less than 29 weeks’ gestation with an embedded case–control analysis of infants with normal or poor growth patterns. Fecal samples were collected weekly from infants on full enteral feeds and analyzed blindly using 16s rRNA next-generation sequencing. The relationship between gut microbial diversity and composition and growth pattern and trajectory were assessed. Results: A total of 115 infants were enrolled in the trial with 263 fecal samples selected from 87 enrolled infants for analysis. In total, 37 samples were available from the normal growth cohort, 56 samples from the poor growth cohort, and 170 samples were available for analysis from the very poor growth cohort. Analysis of relative abundance revealed increased representation of Veillonella, Bifidobacterium, and Clostridium in very poor growth infants compared to normal growth infants. Variation in specific taxa was also found to vary significantly across post-menstrual age depending on the degree of growth failure. Conclusions: Gut microbiome composition and variance was modified by the degree of growth failure in our cohort of preterm infants. Our study adds to the growing body of evidence that alteration of the microbiome is associated with poor growth in preterm infants. This may ultimately represent a therapeutic target for growth failure in preterm infants.

Preterm infants are at risk of brain injury due to immature cerebral autoregulation. However, reliable assessment of cerebral perfusion remains difficult: advanced imaging is rarely feasible at the bedside, and surrogate markers such as blood pressure or oxygen saturation correlate poorly with actual cerebral blood flow. This study aimed to evaluate the velocity and waveform characteristics in the internal cerebral veins (ICV) of preterm infants born before 32 weeks and to explore the relationship between these velocities, clinical parameters, and term-equivalent MRI. Prospectively, a cohort of 47 preterm infants (postmenstrual age 28.9 ± 2.2 weeks, birth weight 1304 ± 386 g) admitted to the NICU at University Hospital Brussels was studied. ICV velocities were measured at multiple time points using Doppler ultrasound as a non-invasive marker of cerebral hemodynamics. Mixed model analysis compared ICV velocities with clinical and term-equivalent MRI results. ICV velocity/weight at the time of scan was significantly associated with postmenstrual age (PMA, p<0.001), postnatal age (PA), and PA squared (both p<0.001), intraventricular hemorrhage (IVH) during admission (p=0.003), and the interaction between gestational age and IVH (p=0.005), reflecting developmental changes in cerebral circulation. IVH was observed in 5 infants. ICV velocity/weight was 1 cm/s/kg higher with IVH at 25 weeks' GA, decreasing to -0.1 cm/s/kg with IVH at 29 weeks' GA. Although the number of IVH cases was limited, these findings suggest that ICV velocity monitoring may provide useful information regarding IVH risk. ICV velocity in preterm infants reflects cerebral maturation. Infants developing IVH showed significantly higher venous velocities, corrected for body weight. Despite limited case numbers, these findings indicate that ICV velocity monitoring could serve as an early marker of hemorrhage risk. Larger studies are needed to confirm its predictive value and explore potential neuroprotective strategies. IVH = Intraventricular Hemorrhage; ICV = Internal Cerebral Vein; NIRS = near-infrared spectroscopy; cUS = Cranial Ultrasound; PMA = postmenstrual age.

To construct a non-invasive and convenient early diagnostic model by integrating multidimensional clinical data and platelet (PLT) indices, and to explore the predictive value of PLT for severe Retinopathy of Prematurity (ROP). A study included premature infants admitted to our hospital from January 2020 to September 2025. According to the results of fundus screening, subjects were divided into the ROP group (n = 190) and the normal control group (n = 142). The ROP group was further categorized into mild (n = 110) and severe (n = 15) subgroups based on treatment requirements, which included platelet data corresponding to a postmenstrual age (PMA) of 30 weeks. Clinical data on parental factors, neonatal factors, and treatment factors were collected, along with PLT results from birth to PMA of 40 weeks. Lasso regression was used to select predictive variables, and a nomogram was constructed using multivariate logistic regression, with the model’s discrimination and calibration verified. Lasso regression identified gestational age, in vitro fertilization, maternal age, and PLT as core predictive factors. The Area Under the Curve (AUC) of the Nomogram in the training and validation sets was 0.80 (95%CI: 0.74–0.85) and 0.80 (95%CI: 0.71–0.89) respectively. The PLT levels at PMA of 30 weeks in the severe ROP group were significantly lower than those in the mild group (160 × 10^9/L vs. 254 × 10^9/L, p = 0.048), with the AUC of the Nomogram based on PLT combined with clinical indicators reaching 0.86 (95%CI: 0.76–0.96) for severe ROP. The ROP prediction model in this study can assist in the non-invasive identification of high-risk infants for ROP, particularly demonstrating high predictive efficacy for severe ROP. Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-025-28293-y.