ObjectiveTo analyze the infection characteristics of human respiratory syncytial virus (HRSV) among children hospitalized with acute lower respiratory tract infection (ALRTI) in a specialized pediatric hospital in Shanghai, so as to provide evidence-based support for optimizing the prevention and control strategies and clinical diagnosis and treatment of respiratory tract infections in children in this region. MethodsA retrospective analysis was performed to the clinical and etiological data of 29 260 children hospitalized for ALRTI in Shanghai Children’s Hospital from January 2021 to December 2024. HRSV and 12 other common respiratory pathogens were detected with multiplex polymerase chain reaction (PCR) and capillary electrophoresis. Demographic and clinical data were collected for statistical analyses. A total of2 412 cases with positive HRSV were divided into the severe group and the non-severe group. Clinical characteristics between the two groups were compared using the Mann-Whitney U test and the chi- square (χ²) test. Additionally, the related influencing factors of severe HRSV infection were explored. ResultsThe overall positivity rate of HRSV from 2021 to 2024 was 8.24% (2 412/29 260), with statistically significant differences observed across the four years (χ²=389.42, P<0.001). The highest positivity rate was in 2021 (14.76%), with a high prevalence throughout the year. In 2022, when non-pharmaceutical interventions (NPIs) were implemented, the HRSV positivity rate was the lowest (4.93%), with a winter-dominant epidemic pattern. In 2023, after the NPIs were lifted, the HRSV positivity rate showed a slight rebound (8.14%), presenting a double-peak pattern. In 2024, the HRSV positivity rate slightly decreased compared to that in 2023 (6.29%), exhibiting a winter and spring-dominant epidemic pattern. Among the hospitalized children with ALRTI, the HRSV positivity rate in males (8.85%) was higher than that in females (7.51%), and the difference was statistically significant (χ²=17.33, P<0.001). Age distribution showed that 82.26% (1 984/2 412) of HRSV infections occurred in children aged 3 years old and below. Besides, as age increased, the infection rate of HRSV showed a gradually decreasing trend (P<0.001). Among the 2 412 children with HRSV infection, the proportion of severe cases was 22.31% (538/2 412), while the non-severe cases accounted for 77.69% (1 874/2 412). Compared with non-severe cases, severe cases were more frequently presented with high fever, longer duration of wheezing, as well as higher rates of underlying diseases or co-infection with Mycoplasma pneumoniae (P<0.001). ConclusionThe prevalence intensity of HRSV varied yearly from 2021 to 2024. After the removal of NPIs in 2023, a slight rebound with a double-peak epidemic pattern was observed. HRSV remained a common pathogen in children hospitalized for ARLTI, and children aged 3 years old and below constituted the highest proportion for infection. Compared with non-severe cases, those with severe HRSV infections were more prone to presenting with high fever and a longer duration of wheezing. Children with positive HRSV who had underlying diseases or co-infection with Mycoplasma pneumonia were more likely to develop severe conditions.

Objective:To evaluate the diagnostic performance of fecal calprotectin (FC) in predicting endoscopic activity of ulcerative colitis (UC), and to compare it with high-sensitivity C reactive protein (hsCRP) and erythrocyte sedimentation rate (ESR) .Methods:A cross-sectional stydy was conducted. UC patients diagnosed at Peking Union Medical College Hospital between May 2023 and July 2025 were retrospective enrolled. Patients were divided into the endoscopically active group and endoscopic remission group according to endoscopic activity. FC levels were measured using latex-enhanced turbidimetric immunoassay (LETIA). Receiver operating characteristic (ROC) curves and logistic regression models were used to assess diagnostic efficacy. Subgroup analyses were conducted according to disease extent.Results:A total of 166 UC patients were enrolled, including 92 males and 74 females with the age of 40.00 (32.00, 52.00) years old and disease course 5.00 (2.00, 10.75) years. Forty-six patients were assigned to the active group, while the remaining 120 were assigned to the remission group. FC levels were significantly higher in the active group than in the remission group (620.72 μg/g vs. 29.00 μg/g, P < 0.001), with an AUC of 0.894 at a cutoff value of 122.54 μg/g. hsCRP and ESR had lower AUC (0.712 and 0.736, respectively). The combination of FC, hsCRP, and ESR slightly improved specificity (AUC 0.898). FC was strongly correlated with the endoscopic activity ( r =0.669, P < 0.001) but not with disease extent. Conclusions:FC measured by latex-enhanced turbidimetric immunoassay had comparable diagnostic accuracy to ELISA-based methods commonly used abroad, and provided a reference cutoff value of 122.54 μg/g. FC outperforms hsCRP and ESR in assessing intestinal inflammation in UC and it is less affected by disease extent, making it a reliable non-invasive biomarker for UC monitoring.

Objective:To evaluate the predictive efficacy of fecal calprotectin (FC) for endoscopic activity in patients with Crohn's disease (CD) .Methods:A cross-sectional study was conducted and patients diagnosed as CD at Peking Union Medical College Hospital from June 2023 to September 2025 were enrolled consecutively. Data was collected including general information, laboratory tests [hemoglobin (HGB), platelet (PLT), FC, high-sensitivity C-reactive protein (hsCRP), erythrocyte sedimentation rate (ESR) and so on], and endoscopic results. FC levels were measured by latex-enhanced turbidimetric immunoassay (LETIA). Endoscopic activity was defined as the simplified endoscopic score for Crohn's disease (SES-CD) > 2. Patients were divided into the endoscopically active group and endoscopic remission group according to endoscopic activity, and the differences in clinical data between the two groups were compared. Spearman correlation analysis was used to assess the correlation between FC and endoscopic activity, and receiver operating characteristic (ROC) curve was used to evaluate the predictive efficacy of FC, hsCRP and ESR for endoscopic activity, and the differences were compared.Results:A total of 90 CD patients were enrolled, including 65 males and 25 females with the age of 30 (22, 41) years old and disease course 4.0 (0.5, 8.0) years. Seventy-one patients (78.9%) had ileocolonic disease involvement (L3), and 55 patients (61.1%) were using biologics. Sixty-nine patients in endoscopic active phase were assigned to the endoscopically active group, while the remaining 21 were assigned to the endoscopic remission group. There were no statistically significant differences in general characteristics such as age and gender between the two groups (all P > 0.05). Compared with endoscopic remission group, HGB was significantly lower in the endoscopically active group, while PLT, hsCRP, ESR, and FC were moderataly higher (all P < 0.05). Among the 90 CD patients, FC levels were moderatly correlated with endoscopic activity (ρ = 0.494). ROC curve analysis indicated that the area under the curve for FC in predicting endoscopic activity was 0.836 (95% CI: 0.737-0.935), with a sensitivity of 0.725, specificity of 0.952, and accuracy of 0.778 at the optimal FC cutoff value of 153.8 μg/g. FC outperformed hsCRP and ESR. Conclusion:FC measured by LETIA demonstrates certain efficacy in predicting endoscopic activity in CD and will assist in efficient clinical monitoring of CD patients.

Objective:To investigate the clinical characteristics of children with severe human metapneumovirus (HMPV) infection and identify the risk factors associated with critical illness.Methods:A retrospective cohort study was conducted, enrolling 157 hospitalized children with severe HMPV infection, who tested positive for HMPV nucleic acid via PCR-capillary electrophoresis fragment analysis of nasopharyngeal secretions at Shanghai Children′s Hospital from January 2021 to December 2023.Clinical features, co-infections, treatment, and outcomes were collected. Based on the diagnostic criteria for severe HMPV infection, the patients were categorized into a critical illness group and a non-critical illness group. Intergroup comparisons were performed using the χ2 test or the Mann-Whitney U test. Multivariate Logistic regression analysis was employed to identify risk factors for critical HMPV infection and to establish a predictive model.The performance of the model was evaluated using receiver operating characteristic (ROC) curve analysis and calibration curves. Results:Among the 157 cases of severe HMPV infection, there were 67 males and 90 females, with an onset age of 39.0 (20.0, 55.5) months. Single-pathogen infection was observed in 125 cases (79.6%), while mixed infections accounted for 32 cases (20.4%).Severe pneumonia was diagnosed in 136 cases (86.6%).The predominant manifestations of severe HMPV infection included fever 152 cases (96.8%), cough 151 cases (96.2%), and wheezing 94 cases (59.9%).Sixty-eight patients (43.3%) required non-invasive respiratory support, 58 cases (36.9%) were admitted to the intensive care unit, and 22 cases (14.0%) underwent mechanical ventilation. Of the total, 149 cases (94.9%) were discharged with improvement, 8 cases (5.1%) were discharged against medical advice, and there were no fatal cases. The cohort was further stratified into a critical illness group 31 cases and a non-critical illness group 126 cases. Compared to the non-critical illness group, the critical illness group exhibited significantly higher rates of respiratory distress, lethargy, and intercostal retractions, along with a higher proportion of underlying comorbidities, and elevated levels of C-reactive protein and procalcitonin (all P<0.05).Conversely, albumin and hemoglobin levels were significantly lower in the critical illness group (both P<0.05). ROC curve analysis revealed that the optimal cutoff value for the duration of fever in predicting severe HMPV infection was 4.5 days.The multivariate binary Logistic regression analysis revealed that prolonged fever duration (>4.5 days) ( OR=28.00, 95% CI 5.09-153.93, P<0.001), anorexia ( OR=11.72, 95% CI 1.26-108.75, P=0.030), and immune dysfunction ( OR=36.71, 95% CI 1.55-867.31, P=0.026) were independent risk factors for severe HMPV infection. A predictive model for critical illness was constructed based on these independent risk factors. ROC curve analysis demonstrated excellent discriminative ability, with an area under the curve of 0.96 (95% CI 0.92-1.00, P<0.001). The optimal predictive probability threshold was 0.17, yielding a sensitivity of 0.93 and specificity of 0.92. The calibration curve closely approximated the ideal curve, indicating good model calibration ( P=0.157). Conclusions:Severe HMPV infection is predominantly observed as a single infection and is prone to progress to severe pneumonia, with fever, cough, and wheezing as the main clinical manifestations. A subset of cases progresses to critical illness, though the overall prognosis is favorable. Prolonged fever duration (>4.5 days), anorexia, and immune dysfunction were independent risk factors for critical illness.The risk prediction model constructed for pediatric critical HMPV infection demonstrated robust discriminative ability with excellent calibration.

Objective:To evaluate the diagnostic performance of fecal calprotectin (FC) in predicting endoscopic activity of ulcerative colitis (UC), and to compare it with high-sensitivity C reactive protein (hsCRP) and erythrocyte sedimentation rate (ESR) .Methods:A cross-sectional stydy was conducted. UC patients diagnosed at Peking Union Medical College Hospital between May 2023 and July 2025 were retrospective enrolled. Patients were divided into the endoscopically active group and endoscopic remission group according to endoscopic activity. FC levels were measured using latex-enhanced turbidimetric immunoassay (LETIA). Receiver operating characteristic (ROC) curves and logistic regression models were used to assess diagnostic efficacy. Subgroup analyses were conducted according to disease extent.Results:A total of 166 UC patients were enrolled, including 92 males and 74 females with the age of 40.00 (32.00, 52.00) years old and disease course 5.00 (2.00, 10.75) years. Forty-six patients were assigned to the active group, while the remaining 120 were assigned to the remission group. FC levels were significantly higher in the active group than in the remission group (620.72 μg/g vs. 29.00 μg/g, P < 0.001), with an AUC of 0.894 at a cutoff value of 122.54 μg/g. hsCRP and ESR had lower AUC (0.712 and 0.736, respectively). The combination of FC, hsCRP, and ESR slightly improved specificity (AUC 0.898). FC was strongly correlated with the endoscopic activity ( r =0.669, P < 0.001) but not with disease extent. Conclusions:FC measured by latex-enhanced turbidimetric immunoassay had comparable diagnostic accuracy to ELISA-based methods commonly used abroad, and provided a reference cutoff value of 122.54 μg/g. FC outperforms hsCRP and ESR in assessing intestinal inflammation in UC and it is less affected by disease extent, making it a reliable non-invasive biomarker for UC monitoring.

Objective:To evaluate the predictive efficacy of fecal calprotectin (FC) for endoscopic activity in patients with Crohn's disease (CD) .Methods:A cross-sectional study was conducted and patients diagnosed as CD at Peking Union Medical College Hospital from June 2023 to September 2025 were enrolled consecutively. Data was collected including general information, laboratory tests [hemoglobin (HGB), platelet (PLT), FC, high-sensitivity C-reactive protein (hsCRP), erythrocyte sedimentation rate (ESR) and so on], and endoscopic results. FC levels were measured by latex-enhanced turbidimetric immunoassay (LETIA). Endoscopic activity was defined as the simplified endoscopic score for Crohn's disease (SES-CD) > 2. Patients were divided into the endoscopically active group and endoscopic remission group according to endoscopic activity, and the differences in clinical data between the two groups were compared. Spearman correlation analysis was used to assess the correlation between FC and endoscopic activity, and receiver operating characteristic (ROC) curve was used to evaluate the predictive efficacy of FC, hsCRP and ESR for endoscopic activity, and the differences were compared.Results:A total of 90 CD patients were enrolled, including 65 males and 25 females with the age of 30 (22, 41) years old and disease course 4.0 (0.5, 8.0) years. Seventy-one patients (78.9%) had ileocolonic disease involvement (L3), and 55 patients (61.1%) were using biologics. Sixty-nine patients in endoscopic active phase were assigned to the endoscopically active group, while the remaining 21 were assigned to the endoscopic remission group. There were no statistically significant differences in general characteristics such as age and gender between the two groups (all P > 0.05). Compared with endoscopic remission group, HGB was significantly lower in the endoscopically active group, while PLT, hsCRP, ESR, and FC were moderataly higher (all P < 0.05). Among the 90 CD patients, FC levels were moderatly correlated with endoscopic activity (ρ = 0.494). ROC curve analysis indicated that the area under the curve for FC in predicting endoscopic activity was 0.836 (95% CI: 0.737-0.935), with a sensitivity of 0.725, specificity of 0.952, and accuracy of 0.778 at the optimal FC cutoff value of 153.8 μg/g. FC outperformed hsCRP and ESR. Conclusion:FC measured by LETIA demonstrates certain efficacy in predicting endoscopic activity in CD and will assist in efficient clinical monitoring of CD patients.

Objective:To analyze the distribution and epidemiological characteristics of non-bacterial pathogens in hospitalized children with acute respiratory infections at a tertiary pediatric hospital in Shanghai during 2023.Methods:A retrospective study was conducted on 10 591 children with acute respiratory tract infections who were hospitalized in Shanghai Children's Hospital from January to December 2023. A multiplex PCR combined with capillary electrophoresis platform was used to detect 11 common non-bacterial respiratory pathogens（including viruses and atypical pathogens）. Statistical analysis was carried out using SPSS 29.0 software. Qualitative data were presented as numbers and percentages，and the Chi-square test was employed to make comparisons between groups，aiming to analyze the differences in the distribution of different pathogens according to gender，age group，and season. Additionally，based on the severity of the disease，patients were calssified into a severe pneumonia group and a non-severe pneumonia group to further explore the characteristics of the pathogen spectrum of severe pneumonia.Results:The total detection rate of pathogens was 54.39%（5 760/10 591），and the proportion of mixed infections was 12.76%（735/5 760）. The dominant pathogens and their proportions were as follows： Mycoplasma pneumoniae（19.20%，2 034/10 591），human rhinovirus（12.16%，1 288/10 591），influenza A virus（8.31%，880/10 591），and respiratory syncytial virus（8.14%，862/10 591）. Epidemiological characteristics showed that：（1）In terms of age： Mycoplasma pneumoniae was more common in older children（29.55%，901/3 049，in the school-age group，χ 2 = 653.67， P<0.001）. Influenza A virus had a high incidence in the adolescent group（11.34%，45/397，χ 2=48.69， P<0.001）. Respiratory syncytial virus was most susceptible in the infant group（20.94%，280/1 337，χ 2=739.92， P<0.001）. Human rhinovirus showed the characteristic of general susceptibility across all ages.（2）Monthly and seasonal distribution： Mycoplasma pneumoniae had a seasonal epidemic in summer and autumn（it began to rise in May and peaked in October at 34.22%，439/1 283）；influenza A virus had a bimodal distribution in spring and winter（the peak was 37.15% in March，315/848）；respiratory syncytial virus had a dominant epidemic in spring and summer（the detection rate was 21.24% in May，206/970），and human rhinovirus was prevalent throughout the year.（3）Clinical correlation：The detection rate of pathogens in the severe pneumonia group was significantly higher than that in the non-severe group：84.19%（426/506） vs 2.89%（5 334/10 085），χ 2=56.23， P<0.001. Conclusions:In 2023，the pathogen spectrum of hospitalized children with acute respiratory infections in the Shanghai area exhibits an epidemic pattern dominated by Mycoplasma pneumoniae，and its transmission dynamics are significantly age-dependent. This study delineates the pathogen-host-environment tripartite interactions，establishing an evidence-based foundation for formulating precision diagnostic-therapeutic algorithms and seasonal nosocomial infection prevention frameworks.

Objective:To analyze the distribution and epidemiological characteristics of non-bacterial pathogens in hospitalized children with acute respiratory infections at a tertiary pediatric hospital in Shanghai during 2023.Methods:A retrospective study was conducted on 10 591 children with acute respiratory tract infections who were hospitalized in Shanghai Children's Hospital from January to December 2023. A multiplex PCR combined with capillary electrophoresis platform was used to detect 11 common non-bacterial respiratory pathogens（including viruses and atypical pathogens）. Statistical analysis was carried out using SPSS 29.0 software. Qualitative data were presented as numbers and percentages，and the Chi-square test was employed to make comparisons between groups，aiming to analyze the differences in the distribution of different pathogens according to gender，age group，and season. Additionally，based on the severity of the disease，patients were calssified into a severe pneumonia group and a non-severe pneumonia group to further explore the characteristics of the pathogen spectrum of severe pneumonia.Results:The total detection rate of pathogens was 54.39%（5 760/10 591），and the proportion of mixed infections was 12.76%（735/5 760）. The dominant pathogens and their proportions were as follows： Mycoplasma pneumoniae（19.20%，2 034/10 591），human rhinovirus（12.16%，1 288/10 591），influenza A virus（8.31%，880/10 591），and respiratory syncytial virus（8.14%，862/10 591）. Epidemiological characteristics showed that：（1）In terms of age： Mycoplasma pneumoniae was more common in older children（29.55%，901/3 049，in the school-age group，χ 2 = 653.67， P<0.001）. Influenza A virus had a high incidence in the adolescent group（11.34%，45/397，χ 2=48.69， P<0.001）. Respiratory syncytial virus was most susceptible in the infant group（20.94%，280/1 337，χ 2=739.92， P<0.001）. Human rhinovirus showed the characteristic of general susceptibility across all ages.（2）Monthly and seasonal distribution： Mycoplasma pneumoniae had a seasonal epidemic in summer and autumn（it began to rise in May and peaked in October at 34.22%，439/1 283）；influenza A virus had a bimodal distribution in spring and winter（the peak was 37.15% in March，315/848）；respiratory syncytial virus had a dominant epidemic in spring and summer（the detection rate was 21.24% in May，206/970），and human rhinovirus was prevalent throughout the year.（3）Clinical correlation：The detection rate of pathogens in the severe pneumonia group was significantly higher than that in the non-severe group：84.19%（426/506） vs 2.89%（5 334/10 085），χ 2=56.23， P<0.001. Conclusions:In 2023，the pathogen spectrum of hospitalized children with acute respiratory infections in the Shanghai area exhibits an epidemic pattern dominated by Mycoplasma pneumoniae，and its transmission dynamics are significantly age-dependent. This study delineates the pathogen-host-environment tripartite interactions，establishing an evidence-based foundation for formulating precision diagnostic-therapeutic algorithms and seasonal nosocomial infection prevention frameworks.

Objective:To investigate the clinical characteristics of children with severe human metapneumovirus (HMPV) infection and identify the risk factors associated with critical illness.Methods:A retrospective cohort study was conducted, enrolling 157 hospitalized children with severe HMPV infection, who tested positive for HMPV nucleic acid via PCR-capillary electrophoresis fragment analysis of nasopharyngeal secretions at Shanghai Children′s Hospital from January 2021 to December 2023.Clinical features, co-infections, treatment, and outcomes were collected. Based on the diagnostic criteria for severe HMPV infection, the patients were categorized into a critical illness group and a non-critical illness group. Intergroup comparisons were performed using the χ2 test or the Mann-Whitney U test. Multivariate Logistic regression analysis was employed to identify risk factors for critical HMPV infection and to establish a predictive model.The performance of the model was evaluated using receiver operating characteristic (ROC) curve analysis and calibration curves. Results:Among the 157 cases of severe HMPV infection, there were 67 males and 90 females, with an onset age of 39.0 (20.0, 55.5) months. Single-pathogen infection was observed in 125 cases (79.6%), while mixed infections accounted for 32 cases (20.4%).Severe pneumonia was diagnosed in 136 cases (86.6%).The predominant manifestations of severe HMPV infection included fever 152 cases (96.8%), cough 151 cases (96.2%), and wheezing 94 cases (59.9%).Sixty-eight patients (43.3%) required non-invasive respiratory support, 58 cases (36.9%) were admitted to the intensive care unit, and 22 cases (14.0%) underwent mechanical ventilation. Of the total, 149 cases (94.9%) were discharged with improvement, 8 cases (5.1%) were discharged against medical advice, and there were no fatal cases. The cohort was further stratified into a critical illness group 31 cases and a non-critical illness group 126 cases. Compared to the non-critical illness group, the critical illness group exhibited significantly higher rates of respiratory distress, lethargy, and intercostal retractions, along with a higher proportion of underlying comorbidities, and elevated levels of C-reactive protein and procalcitonin (all P<0.05).Conversely, albumin and hemoglobin levels were significantly lower in the critical illness group (both P<0.05). ROC curve analysis revealed that the optimal cutoff value for the duration of fever in predicting severe HMPV infection was 4.5 days.The multivariate binary Logistic regression analysis revealed that prolonged fever duration (>4.5 days) ( OR=28.00, 95% CI 5.09-153.93, P<0.001), anorexia ( OR=11.72, 95% CI 1.26-108.75, P=0.030), and immune dysfunction ( OR=36.71, 95% CI 1.55-867.31, P=0.026) were independent risk factors for severe HMPV infection. A predictive model for critical illness was constructed based on these independent risk factors. ROC curve analysis demonstrated excellent discriminative ability, with an area under the curve of 0.96 (95% CI 0.92-1.00, P<0.001). The optimal predictive probability threshold was 0.17, yielding a sensitivity of 0.93 and specificity of 0.92. The calibration curve closely approximated the ideal curve, indicating good model calibration ( P=0.157). Conclusions:Severe HMPV infection is predominantly observed as a single infection and is prone to progress to severe pneumonia, with fever, cough, and wheezing as the main clinical manifestations. A subset of cases progresses to critical illness, though the overall prognosis is favorable. Prolonged fever duration (>4.5 days), anorexia, and immune dysfunction were independent risk factors for critical illness.The risk prediction model constructed for pediatric critical HMPV infection demonstrated robust discriminative ability with excellent calibration.

Objective:To explore the clinical characteristics and prognostic risk factors of accidental injuries in pediatric intensive care unit(PICU).Methods:The children with accidental injuries admitted to the PICU at Shanghai Children′s Hospital from January 2017 to December 2019 were analyzed retrospectively.The patients were divided into survival group and death group according to outcome.We collected the patients′ clinical data and laboratory indexes, and analyzed the differences between two groups.Multivariate Logistic regression was used to screen the risk factors of death and receiver-operating characteristic(ROC)curve was used to determine the threshold.Results:(1)A total of 253 children were included, accounting for 5.2% of PICU hospitalization during the same period.There were 137 males(54.2%)and 116 females(45.8%), with an average age of 45(19, 96)months.The top three causes of accidental injuries were fall injury in 82 cases(32.4%), traffic accident in 70 cases(27.7%)and foreign body in 39 cases(15.4%). There were 22 cases of deaths, with a fatality rate of 8.7%.(2)The mean arterial pressure at admission in the death group was lower than that in the survival group( P<0.05), and the total length of stay was longer( P<0.001). There were significant differences in Glasgow coma scale, Pediatric Risk of Mortality Scoring Ⅲ(PRISM Ⅲ)between the two groups( P<0.001). In the death group, and the blood lactic acid(LA) significantly increased( P<0.001), hemoglobin and fibrinogen were lower than those in the survival group, the activated partial thromboplastin time and prothrombin time(PT)were significantly prolonged( P<0.001). (3)Multivariate Logistic regression analysis showed that PRISM Ⅲ score, LA and PT were independent risk factors for the death of accidental injuries.The area under curve(AUC)of PRISM Ⅲ score was 0.987(95% CI 0.976-0.999, P<0.001)and the cut-off value was 11 with a sensitivity of 100% and a specificity of 95.7%.The AUC of LA was 0.886(95% CI 0.810-0.961, P<0.001)and the cut-off value was 3.1 mmol/L with a sensitivity of 85.7% and a specificity of 77.3%.The AUC of PT was 0.835(95% CI 0.730-0.941, P<0.001)and the cut-off value was 13.9 s with a sensitivity of 86.6% and a specificity of 72.7%. Conclusion:The main accidental injuries in PICU of our hospital in the past three years are falling injuries, traffic accidents and foreign body injuries.PRISM Ⅲ score, LA and PT at admission are independent risk factors for predicting death.