Development and validation of a risk prediction model for respiratory tract Ureaplasma urealyticum infection in preterm infants Authors: Jinju Shi, Huiyue Zhang, Yanni Xu, Hongzhu Cai, Shaoru Zheng, Cuimin Su Published: 2026 ABSTRACT BackgroundUreaplasma urealyticum (UU) infection in preterm infants lacks typical clinical manifestations. This study aims to analyze the risk factors for UU infection in preterm infants, construct and validate a nomogram prediction model, and provide evidence for early clinical risk assessment.MethodsA total of 430 preterm infants admitted to the Department of Neonatology of Jinjiang Municipal Hospital from September 2022 to December 2024 were selected as the research subjects. Polymerase chain reaction (PCR)-fluorescent probe method was used to detect UU-DNA in respiratory secretions within 48 hours after admission. According to the test results, the preterm infants were divided into the UU-positive group (n=92) and the UU-negative group (n=338). Multivariable binary logistic regression analysis was performed to identify independent risk factors for UU infection. R software was used to construct a nomogram prediction model for respiratory UU infection in preterm infants, and the receiver operating characteristic (ROC) curve was used to evaluate the efficacy of the model.ResultsMultivariable logistic regression analysis showed that maternal vaginal delivery [odds ratio (OR) =0.07, 95% confidence interval (CI): 0.033–0.149], premature rupture of membranes (OR =6.04, 95% CI: 3.023–12.065), chorioamnionitis (OR =3.498, 95% CI: 1.545–7.921), UU infection of the genital tract (OR =13.949, 95% CI: 6.858–28.373), and white blood cell count in preterm infants (OR =1.11, 95% CI: 1.033–1.192) were independent risk factors for UU infection in preterm infants. The above risk factors were introduced into R software to construct a nomogram model. The area under the ROC curve was 0.931 (95% CI: 0.905–0.956), the sensitivity was 90.2%, and the specificity was 82.8%, which had good prediction efficiency.ConclusionsThe nomogram model based on the five independent risk factors of vaginal delivery, premature rupture of membranes, chorioamnionitis, UU infection of the genital tract, and white blood cell count of preterm infants has a high predictive ability. For preterm infants with high-risk factors, UU detection should be performed in time to achieve early diagnosis and intervention, reduce the incidence of serious complications, and improve the prognosis of children. Highlight box Key findings • Five independent influencing factors for preterm infants Ureaplasma urealyticum (UU) infection were verified. • The established nomogram model has robust predictive performance. • The model supports early clinical prevention and intervention. What is known and what is new? • Preterm infants with UU infection present atypical symptoms and can easily lead to adverse prognoses. Multiple maternal and neonatal factors are correlated with UU infection; yet clinically, there is still a lack of intuitive quantitative tools for early high-risk screening. • This study identified five independent risk factors for preterm UU infection and constructed a highly efficient nomogram model. It provides a reliable quantitative tool for early clinical screening and intervention to improve the prognosis of preterm infants. What is the implication, and what should change now? • This high-performance nomogram achieves precise quantitative risk evaluation for preterm UU infection, offering a reliable and standardized tool for early clinical risk assessment. • Clinicians should adopt this model for routine risk screening. High-risk preterm infants require timely UU testing and early intervention to reduce complications. Introduction Background Due to the incomplete development of the immune system and the immature function of various organs, the risk of infection in preterm infants is significantly higher than that in full-term infants, among which Ureaplasma urealyticum (UU) infection is more common in clinical practice (1-3). UU is the smallest known self-replicating, free-living prokaryotic microorganism and belongs to the genus Ureaplasma within the family of Mycoplasmataceae. UU adheres to a variety of host cells, including epithelial and germ cells, and commonly colonizes the adult urogenital tract and the neonatal respiratory tract. It is generally regarded as an opportunistic pathogen (4). Epidemiological studies have shown that Ureaplasma spp. colonize the genital tract of approximately 40–80% of women of reproductive age (5). High colonization rates have also been reported among pregnant women (6), supporting the role of maternal genital tract colonization as an important source of vertical transmission. UU infection in pregnant women is closely related to adverse pregnancy outcomes, which can lead to complications such as premature delivery, premature rupture of membranes, and chorioamnionitis (7-9). However, whether asymptomatic women or men should undergo routine screening and treatment for genital Ureaplasma or Mycoplasma remains controversial. Current international expert recommendations do not support routine screening because available evidence has not demonstrated that screening and treatment improve maternal or neonatal outcomes, while asymptomatic colonization is common, eradication is often incomplete, recolonization frequently occurs, and unnecessary antibiotic use may contribute to antimicrobial resistance (10,11). Nevertheless, maternal genital colonization remains an important source of vertical transmission, particularly among pregnancies complicated by preterm birth. In recent years, studies at home and abroad have shown that UU can infect newborns through mother-infant vertical transmission, causing bronchopulmonary dysplasia (BPD), neonatal pneumonia, neonatal sepsis, and periventricular-intraventricular hemorrhage (PVH-IVH) (12-15), which seriously affect the quality of life and long-term prognosis of neonates. Because routine maternal screening is not currently recommended and cannot reliably identify all neonates at risk, early postnatal risk stratification remains clinically important. Developing an individualized prediction model based on neonatal clinical characteristics may facilitate early recognition of high-risk infants and provide a practical strategy for optimizing antimicrobial management and neonatal care. Rationale and knowledge gap There are few reports on risk prediction models for respiratory UU infection in preterm infants. As an intuitive graphical prediction tool, nomogram has the advantages of simple operation and non-invasive application. It can predict the probability of disease or adverse events by quantifying the weight of various influencing factors and then assist clinicians in formulating individualized treatment plans. At present, there is still a lack of specific prediction tools for respiratory UU infection in preterm infants based on multi-dimensional clinical variables. Objective This study intends to explore the risk factors for respiratory UU infection in preterm infants by retrospectively analyzing the clinical data of 430 preterm infants admitted to the Department of Neonatology of Jinjiang Municipal Hospital from September 2022 to December 2024. Independent risk factors were screened, and a nomogram prediction model was constructed. The results of this study can provide an evidence-based basis for the early development of targeted prevention and intervention strategies and have important clinical significance for lowering the infection rate of UU in preterm infants, reducing infection-related complications, and improving the prognosis of preterm infants. We present this article in accordance with the TRIPOD reporting checklist (available at https://tp.amegroups.com/article/view/10.21037/tp-2026-0504/rc). Methods Participants Preterm infants hospitalized in the Department of Neonatology of Jinjiang Municipal Hospital between September 2022 and December 2024 were consecutively enrolled in this retrospective study. The inclusion criteria were as follows: (I) preterm infants admitted to the Department of Neonatology within 24 hours after birth; (II) completion of respiratory UU testing within 48 hours after admission; and (III) complete clinical data available for analysis. The exclusion criteria were as follows: (I) congenital inherited metabolic diseases, gastrointestinal malformations, cyanotic congenital heart disease, or other major congenital malformations; and (II) infants who were lost to follow-up, died within 28 days after birth, or were discharged against medical advice before assessment of BPD outcomes. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the Ethics Committee of Jinjiang Municipal Hospital (Approval No. jjsyyyxll-2022008). Informed consent was taken from all the patients’ parents or legal guardians. During the informed consent process, the specimen collection procedure, its potential risks and discomforts, the study objectives, and confidentiality measures were explained. All clinical data were anonymized before analysis to protect participants’ privacy. Only the investigators had access to the anonymized study data. Specimen collection and UU testing Upon admission, endotracheal aspirates or oropharyngeal secretions were collected under sterile conditions by trained personnel using a sterile sputum collection device. UU was detected using a polymerase chain reaction (PCR) fluorescence probe assay (ABI 7300 Real-Time PCR System; UU nucleic acid amplification fluorescence quantitative detection kit, Daan Gene Co., Ltd., Guangzhou, China). A UU-DNA level of ≤500 copies/mL was considered negative, whereas a level of >500 copies/mL was considered positive. Block method According to the UU-DNA test results of respiratory secretions, they were divided into the UU-positive group (n=92) and the UU-negative group (n=338). According to gestational age, preterm infants were divided into three groups: 28–31+6 weeks group (n=30), 32–33+6 weeks group (n=69), and 34–36+6 weeks group (n=331). Collection of clinical data The data were collected from the electronic medical record system of the neonatology department and the Obstetrics Department of Jinjiang Municipal Hospital and double-checked before the data were entered into the computer to ensure their completeness and accuracy. Clinical data of neonates (I) General information: gender, gestational age, birth weight, length of hospital stay, history of resuscitation at birth; (II) laboratory indicators on admission: white blood cell count, absolute neutrophil count, platelet count, C-reactive protein. Perinatal data of the mothers Gravidity, parity, vaginal delivery, history of adverse pregnancy and childbirth, premature rupture of membranes, chorioamnionitis, prenatal use of glucocorticoids, amniotic fluid turbidity, fetal distress, gestational hypertension, gestational diabetes mellitus, and UU culture results of the reproductive tract. Statistical analysis Statistical analyses were performed using SPSS version 27.0 and R software version 4.2.3. SPSS version 27.0 was used for data entry and statistical analyses. Continuous variables were tested for normality. Normally distributed data were expressed as the mean ± standard deviation, whereas non-normally distributed data were presented as the median and interquartile range (IQR) and were compared using the Mann-Whitney U test. Categorical variables were expressed as frequencies and percentages and compared using the Chi-squared test or Fisher’s exact test, as appropriate. A two-sided P<0.05 was considered statistically significant. Variables with statistical significance in the univariate analysis were further entered into a multivariable binary logistic regression model. R software (version 4.2.3) was used to construct the nomogram prediction model and perform internal validation using the bootstrap resampling method. The predictive performance of the nomogram was evaluated using receiver operating characteristic (ROC) curve analysis and the area under the curve (AUC). Calibration of the model was assessed using the Hosmer-Lemeshow goodness-of-fit test. Results Comparison of preterm infants and maternal perinatal data between the UU-positive group and the UU-negative group The gestational age at birth in the UU-positive group was smaller than that in the UU-negative group, and the incidence of vaginal delivery, premature rupture of membranes, chorioamnionitis, and UU infection of the genital tract was higher than that in the UU-negative group, and the difference was statistically significant (P<0.05). The proportion of gestational hypertension and gestational diabetes mellitus in the UU-negative group was higher than that in the UU-positive group, and the difference was statistically significant (P<0.05). However, there were no significant differences in gender, birth weight, history of resuscitation at birth, mother’s first pregnancy, history of adverse pregnancy and delivery, prenatal use of glucocorticoids, amniotic fluid opacity, and fetal intrauterine distress between the two groups (P>0.05), as shown in Table 1. Table 1Comparison of preterm infants and maternal perinatal data between the UU-positive group and the UU-negative groupVariablesUU-negative group (n=338)UU-positive group (n=92)χ2/ZPPerinatal characteristics of preterm infants Male190 (56.21)53 (57.61)0.057†0.81 Birth weight (g)2,352.5 (2,100.0–2,600.0)2,305.0 (1,935.0–2,550.0)−1.722‡0.09 Gestational age (week)35.14 (34.3–36.0)34.86 (33.0–35.8)−2.677†0.007** 28–31+6 weeks group16 (4.73)14 (15.22)19.851†<0.001*** 32–33+6 weeks group47 (13.91)22 (23.91) 34–36+6 weeks group275 (81.36)56 (60.87) History of resuscitation at birth17 (5.03)4 (4.35)0.072†0.79Perinatal characteristics of mothers First pregnancy104 (30.77)33 (35.87)0.867†0.35 History of adverse pregnancy and childbirth152 (44.97)34 (36.96)1.892†0.17 Vaginal delivery74 (21.89)66 (71.74)81.828†<0.001*** Premature rupture of membranes95 (28.11)64 (69.57)53.338†<0.001*** Chorioamnionitis50 (14.79)26 (28.26)9.015†0.003** Use of antenatal corticosteroids235 (69.53)58 (63.04)1.4†0.24 Cloudy amniotic fluid7 (2.07)4 (4.35)1.504†0.22 Fetal distress in the uterus25 (7.40)3 (3.26)2.032†0.15 Hypertension in pregnancy80 (23.67)12 (13.04)4.854†0.03* Gestational diabetes mellitus114 (33.73)15 (16.30)10.454†0.001** UU infection of the genital tract66 (19.53)70 (76.09)106.982†<0.001***Data are presented as n (%) or median (interquartile range). *, P<0.05; **, P<0.01; ***, P<0.001. †, Chi-squared test. ‡, Mann-Whitney U test. UU, Ureaplasma urealyticum. Comparison of preterm infants and maternal perinatal data between the UU-positive group and the UU-negative group Data are presented as n (%) or median (interquartile range). *, P<0.05; **, P<0.01; ***, P<0.001. †, Chi-squared test. ‡, Mann-Whitney U test. UU, Ureaplasma urealyticum. Comparison of laboratory indexes between the UU-positive group and the UU-negative group at admission Compared with the UU-negative group, the UU-positive group had higher white blood cell counts and C-reactive protein on admission, and the difference was statistically significant (P<0.05). There was no significant difference in absolute neutrophil count and platelet count between the two groups at admission (P>0.05), as shown in Table 2. Table 2Comparison of laboratory parameters at admission between the UU-positive group and the UU-negative groupVariablesUU-negative group (n=338)UU-positive group (n=92)ZPWhite blood cell count (109/L)12.89 (11.0–15.7)14.64 (11.8–18.5)−3.4620.001**Absolute neutrophil count (%)8.28 (6.4–10.6)8.89 (6.2–13.3)−1.5850.11Blood platelet count (109/L)317.0 (272.8–376.5)343.5 (263.0–399.0)−0.8110.42C-reactive protein (mg/L)0.81 (0.5–1.8)1.66 (0.7–6.8)−4.709<0.001***Data are presented as median (interquartile range). **, P<0.01; ***, P<0.001. UU, Ureaplasma urealyticum. Comparison of laboratory parameters at admission between the UU-positive group and the UU-negative group Data are presented as median (interquartile range). **, P<0.01; ***, P<0.001. UU, Ureaplasma urealyticum. Multivariate logistic regression analysis of risk factors for UU infection Variables with statistically significant differences in the univariate analysis of maternal and neonatal perinatal characteristics and laboratory indicators at admission were entered into a multivariable binary logistic regression model as independent variables. The results showed that maternal vaginal delivery, premature rupture of membranes, chorioamnionitis, UU infection of the genital tract, and a high white blood cell count in preterm infants were independent risk factors for respiratory UU infection in preterm infants, as shown in Table 3. Table 3Multivariate logistic regression analysis of risk factors for UU infection in preterm infantsInfluence factorβStandard errorWaldPOR95% CIVaginal delivery−2.6570.38348.133<0.0010.070.033–0.149Premature rupture of membranes1.7980.35325.947<0.0016.043.023–12.065Chorioamnionitis1.2520.4179.0190.0033.4981.545–7.921UU infection of the genital tract2.6350.36252.922<0.00113.9496.858–28.373White blood cell count (109/L)0.1040.0368.1760.0041.111.033–1.192Constant−3.9260.64437.184<0.001––CI, confidence interval; OR, odds ratio; UU, Ureaplasma urealyticum. Multivariate logistic regression analysis of risk factors for UU infection in preterm infants CI, confidence interval; OR, odds ratio; UU, Ureaplasma urealyticum. Construction of a risk prediction model for UU infection The five variables “maternal vaginal delivery, premature rupture of membranes, chorioamnionitis, UU infection of the genital tract, and white blood cell count in preterm infants” screened by multivariate logistic regression analysis were included in the nomogram to construct a prediction model for respiratory UU infection in preterm infants, and the nomogram was drawn for visualization, as shown in Figure 1. Each predictor variable could project a vertical score to the scoring axis, and then the scores of each predictor variable were added to obtain a total score, and the corresponding position was found on the total score axis. The value projected vertically to the risk axis was the probability of respiratory UU infection in preterm infants. Higher total score values indicate a higher risk of respiratory UU infection in preterm infants. Nomogram model for predicting the risk of UU respiratory infection in preterm infants. For maternal vaginal delivery, “0” represents non-vaginal delivery and “1” represents vaginal delivery; for premature rupture of membranes, chorioamnionitis, and UU infection of the genital tract, “0” represents absence and “1” represents presence. UU, Ureaplasma urealyticum. Validation of the nomogram prediction model for UU infection Bootstrap resampling was used for internal validation of the prediction model. The ROC curve of the model was drawn, and the AUC was 0.931 [95% confidence interval (CI): 0.905–0.956], the sensitivity was 90.2%, and the specificity was 82.8%, indicating that the prediction ability of the model was strong, as shown in Figure 2. The calibration curve was drawn, and the Hosmer-Lemeshow goodness-of-fit test results showed that χ2=5.526, P=0.79, suggesting that the predicted probability of the model was in good agreement with the actual incidence and had good accuracy, as shown in Figure 3. ROC curve of the UU infection risk prediction model. AUC, area under the curve; ROC, receiver operating characteristic; UU, Ureaplasma urealyticum. Calibration curve of a risk prediction model for UU infection. UU, Ureaplasma urealyticum. Clinical application value of the nomogram prediction model for UU infection Decision curve analysis (DCA) revealed the net benefits provided by the model at different risk thresholds, and the extreme strategies of all preterm infants with respiratory UU infection (gray line) or none of preterm infants with respiratory UU infection (black line) were compared, see Figure 4. The model provided large net benefits between the risk threshold of 0.2–0.6, indicating that the model had strong clinical application value. DCA curve of the prediction model. DCA, decision curve analysis. Discussion Perinatal UU infection is mainly transmitted to the fetus and newborn through maternal vertical transmission, including in-utero infection and birth infection (16). Pathogens can colonize the respiratory tract, digestive tract, central nervous system, and blood, which seriously threaten the life and health of newborns. At present, there is a lack of rapid and simple diagnostic methods for early identification of respiratory UU infection in preterm infants. Although the traditional culture method is the gold standard for the diagnosis of UU infection, it has the limitations of a long detection cycle and a low positive detection rate. Although PCR has higher sensitivity and specificity, its clinical application is limited because UU has no cell wall, grows in animal serum-dependent culture medium, and requires professional laboratory conditions. In this study, multivariate logistic regression analysis showed that vaginal delivery of the mother, premature rupture of membranes, chorioamnionitis, UU infection of the genital tract, and white blood cell count in preterm infants were independent risk factors for UU infection. Based on the above factors, a nomogram prediction model for respiratory tract UU infection in preterm infants was constructed, with an AUC of 0.931 (95% CI: 0.905–0.956), showing good prediction efficiency. The Hosmer-Lemeshow goodness of fit test showed that the calibration curve was highly consistent with the ideal curve (χ2=5.526, P=0.79), which confirmed that the model had good discrimination and consistency in predicting the probability of UU infection. Further drawing of DCA showed that the model had high clinical practicability and was expected to provide an effective tool for early risk assessment of respiratory UU infection in preterm infants. The results of this study showed that the vaginal delivery rate of preterm infants with respiratory UU infection was significantly higher than that of preterm infants without UU infection (71.74% vs. 21.89%), which was highly consistent with the results of Ma et al. (17). Through a single-center retrospective case-control study, Sun et al. (18) performed UU detection on respiratory secretions, gastric juice, and urine of 291 premature infants within 48 hours after birth and systematically analyzed clinical characteristics, blood changes, and disease outcome, confirming that vaginal delivery was an independent risk factor for UU infection [odds ratio (OR) =0.18, 95% CI: 0.09–0.35]. Romero et al. (19) showed that intra-amniotic infections mostly result from ascending infection, suggesting that preterm infants delivered vaginally are at higher risk of exposure to UU infection. Mother-to-child vertical transmission is the main route of UU infection. Newborns exposed to maternal secretions containing UU pathogens through the birth canal during delivery will significantly increase their risk of postnatal infection. The above evidence together reveals the close association between the mode of delivery and respiratory UU infection in preterm infants, and provides an important basis for early clinical risk prevention and control. The results of this study showed that the incidence of premature rupture of membranes in the UU-positive group was higher than that in the UU-negative group (P<0.05), which was consistent with the results of many previous studies, suggesting that UU can infect fetuses and neonates through the reproductive tract. Chu et al. (20) retrospectively analyzed the clinical data of UU infection-related complications and treatment strategies and found that the vertical transmission rate of women with premature rupture of membranes at delivery was positively correlated with the duration of premature rupture of membranes. To explore the mechanism, premature rupture of membranes can lead to a weak alkaline vaginal pH in pregnant women, which creates favorable conditions for the survival and proliferation of pathogens and then causes fetal infection through vertical transmission. During vaginal delivery, the newborn is in direct contact with the birth canal, carrying the pathogen, further increasing the risk of vertical transmission. In addition, the immune response triggered by infection activates the immune system and promotes the massive release of cytokines and prostaglandins, which induces uterine contraction, cervical dilatation, and eventually premature rupture of membranes (21). Chorioamnionitis is a pathological condition characterized by inflammation of the amniotic sac and placenta. Its diagnosis depends on clinical symptoms, histopathological examination, and postpartum placental microbial culture analysis. UU, as an important pathogenic microorganism in the perinatal period, can cause chorioamnionitis by invading the amniotic membrane and placental tissue. Studies have shown that the invasion of microorganisms into the amniotic cavity is the main cause of inflammation, among which UU, Mycoplasma parvosus, and Mycoplasma hominis have the highest detection rates in amniotic fluid and placenta isolates (22). The results of this study showed that mothers of UU-infected preterm infants had a higher proportion of chorioamnionitis, which is consistent with previous literature (23,24). Stol et al. (25) showed that the proportion of chorioamnionitis in UU-infected placentas was higher than that in placentas infected with other microorganisms (68.4% vs. 26.7%). The pathogenic mechanism of UU is that UU decomposes urea to produce ammonia, which leads to an increase in the pH value in the local environment. This alkaline environment not only destroys the integrity of the fetal membrane, but also activates the inflammatory cascade, eventually leading to the occurrence and development of chorioamnionitis. Chorioamnionitis, as an independent risk factor for preterm birth, can significantly increase the risk of preterm birth by releasing pro-inflammatory factors to activate uterine contractions and degrade cervical collagen. Although maternal genital colonization with Ureaplasma has consistently been associated with adverse pregnancy and neonatal outcomes, current international guidelines do not recommend routine screening or treatment of asymptomatic women or men because evidence demonstrating improved maternal or neonatal outcomes remains insufficient. This apparent discrepancy highlights the important distinction between epidemiological association and evidence supporting population-based screening. Therefore, rather than advocating universal maternal screening, our study focuses on postnatal risk stratification of preterm infants. By integrating maternal and neonatal clinical risk factors into a prediction model, our approach complements current clinical practice and may facilitate earlier identification of infants at increased risk for respiratory Ureaplasma infection, thereby supporting individualized clinical management and more judicious antimicrobial use. Intrauterine UU infection is mainly the result of ascending infection after vaginal colonization. UU infection of the genital tract of pregnant women will migrate through the cervical canal, resulting in intra-amniotic infection, increasing the risk of premature delivery, and posing a threat to maternal and child health during the perinatal period. The results of this study showed that the UU infection rate of the maternal genital tract in the UU-positive group was higher than that in the UU-negative group (P<0.05), which was consistent with the results of Judith Rittenschober-Boehm et al. (26). Previous studies have reported that UU infection can lead to abnormal leukocyte test results in neonates. Zhang et al. (27) conducted a retrospective observational study of 291 preterm infants with a gestational age of <32 weeks and reported that UU-positive preterm infants had significantly higher white blood cell counts than UU-negative preterm infants. This study further confirmed that preterm infants with UU infection had a higher white blood cell count after birth than those without UU infection (OR =1.11, 95% CI: 1.033–1.192), which was consistent with the results of previous studies (28). This suggests that, in clinical practice, when neonates have an unexplained increase in white blood cell count, the possibility of UU infection should be considered. However, there are relatively few reports on the abnormal blood cell count caused by UU infection. The time dynamic characteristics of blood picture changes after infection and the clinical value of the white blood cell count as a diagnostic indicator of UU infection still need to be further verified by multi-center and large-sample studies. The risk prediction model for UU infection in preterm infants constructed based on the risk factors screened from the perinatal characteristics of mothers and children was verified by the ROC curve, and the AUC value was 0.931. This model can quickly assess the risk of UU infection in preterm infants, which is helpful for clinicians to early judge the possibility of UU infection in preterm infants. The study by Viscardi et al. (29) also provided strong evidence for this. By collecting data from five independent cohorts and analyzing maternal and neonatal-related factors, we developed three risk scoring systems for predicting the probability of lower respiratory tract UU infection in preterm infants. The results showed that the prediction model had good prediction performance, and the AUC value was stable in the range of 0.73 to 0.77, which was highly consistent with the core conclusion of this study that the risk prediction model of UU infection in preterm infants had good performance. This study further confirmed the practicability of the risk prediction model based on perinatal characteristics in the field of prevention and control of UU infection in preterm infants. Despite the promising performance of the present nomogram, this study was conducted at a single center and therefore requires further validation in prospective multicenter cohorts with larger and more diverse populations. Future multicenter studies will be essential to evaluate the model’s generalizability, stability, and predictive accuracy across different clinical settings. Following external validation and appropriate optimization, the nomogram may be incorporated into routine neonatal clinical practice as an easy-to-use risk assessment tool. Because the model is based on routinely available maternal and neonatal clinical variables, it could be readily integrated into electronic medical record systems or clinical decision support platforms, thereby facilitating early identification of preterm infants at high risk for respiratory UU infection, supporting individualized clinical management, and optimizing the rational use of antimicrobial therapy. Conclusions In conclusion, we developed and internally validated a clinically applicable nomogram for predicting respiratory UU infection in preterm infants using readily available maternal and neonatal clinical variables. The model demonstrated good discrimination and calibration, suggesting that it may serve as a practical tool for early risk stratification and individualized clinical management. Early identification of high-risk infants may facilitate timely monitoring and targeted interventions, thereby improving neonatal outcomes and promoting more rational use of antimicrobial therapy. Supplementary The article’s supplementary files as Available at https://tp.amegroups.com/article/view/10.21037/tp-2026-0504/prf