Septic encephalopathy( SE) is defined as diffuse brain dysfunction that related to systemic inflammatory response and without clinical or laboratory evidence of central nervous system infection.SE is one of the most common complications of patients with severe sepsis.Its pathogenesis is very complex and is not yet clear.The clinical manifestations of SE varied and have no specificity, mainly for changes of con-sciousness and mental status.The performance of electroencephalography is more sensitive,and is associated with prognosis.Somatosensory evoked potentials could display the damage of cortical or subcortical path-ways.Brain MRI has a more accurate assessment of the nature and the extent of brain damage.SE is a diag-nosis of exclusion.Before the diagnosis of SE,we need to exclude other encephalitis and encephalopathy.The morbidity and mortality of SE are high.It requires close attention,early detection and timely treatment.

The central nervous system ( CNS) complications and peripheral nervous system ( PNS) complications are common complications of sepsis. CNS complications include sepsis associated encephalopathy (SAE),septic-metastatic encephalitis (SME),and septic-embolic encephalitis (SEE). Of them,the SAE is the most common complication of CNS complications. PNS complications include critical illness polyneuropathy (CIP) and critical illness myopathy (CIM). CIP and CIM could lead to muscle weakness and dependence on mechanical ventilation of patients with sepsis. Sepsis related neurological complications are closely related to the mortality and prognosis of patients,so early recognition and timely intervention are important.

Objective To study the clinical presentations,radiologic features,prognosis,and possible causes of liver injury associated umbilical venous catheter (UVC ).Methods We reviewed database of our NICU from December 2012 to November 2015 and identified preterm infants with liver injury while UVC in place.The gestational age,birth weight,gender,days of UVC in place prior to liver injury, the depth of UVC (cm),UVC tip position,ultrasound findings of liver injury,laboratory tests,treatment and outcomes were collected.Results During study period,322 infants received UVC insertion.Ten cases (3.1 %)of liver injury associated with UVC were diagnosed.Of the ten infants,nine were diagnosed as extravasation of fluid to liver parenchyma,and one was diagnosed as liver hematoma.The mean birth weight was (1184 ±207)g,mean gestational age was (28.5 ±1.7)weeks.The UVC tip was at thoracic (T) vertebrae T 9 to T 11 level,nine UVC tips were below the right diaphragm level.All UVCs were used for total parenteral nutrition before liver injury was diagnosed.Six infants had clinical manifestations,and four infants were free of clinical manifestations.The most prominent clinical signs were abdominal distension (n =5),weak bowel sounds (n =5)and hepatomegaly (n =4).One infant who was diagnosed with liver hematoma also had progressive hemoglobin decline (minimum 34 g/L)and shock.Abdominal ultrasound showed well-limited,irregular,hyperechoic rimmed lesions with heterogeneously hypoechoic centers or anechoic liquid dark space.Extravasation of fluid to liver parenchyma will restorate within one week after the UVC was remove.One baby who was diagnosed as liver hematoma passed away at 9 days of life.Liver ultrasound of eight returned to normal in 52 days to 3.5 months,and one had calcified lesions at 9 months of age.Conclusions Liver injury is an unusual complication of UVC insertion and usage.Proper positioning of the UVC tip may help to avoid this complication.Early recognition,prompt diagnosis with liver ultrasound examination and timely treatment can lead to better outcome in newborns with extravasation of fluid to liver parenchyma.

Objective To evaluate the effectiveness of maternal prenatal education on promoting breastfeeding rate for preterm infants in neonatal intensive care unit (NICU).Methods Gravidas who were admitted to Peking University First Hospital for any risk of preterm delivery from November 2013 to December 2014 and preterm infants admitted to the NICU (length of stay ≥ 5 d) of the same hospital during the same period were involved.All the gravidas received prenatal education and completed questionnaires.Two hundred and ninety-five preterm infants who met the inclusion criteria were divided into two groups including education group (n=125) and non-education group (n=170) according to whether their mothers received prenatal education or not.Conditions of the preterm infants during hospitalization and after discharge and breastfeeding rates were comparatively analyzed between the two groups.T test,Chi-test and Wilcoxon rank-sum test were used for statistical analysis.Results (1) There were 380 gravidas received questionnaires.Among them,346 (91.1％)completed the questionnaires and were recruited in this study.Only 31.8％ (110/346) of these gravidas were active in learning more about breastfeeding and 46.2％ (160/346) of them lacked confidence in breastfeeding after the appearance of preterm delivery signs.There were significant improvements in their attitudes towards considering breastfeeding seriously and discussing with their family,confidence in breastfeeding,the importance of colostrum and how to breastfeed a preterm infant in hospital after maternal prenatal education (all P＜0.001).Prenatal education was thought to be helpful in 77.5％ (268/346) of the gravidas.(2) The two groups of preterm infants showed good homogeneities in gestational age,gender,birth weight and other basal conditions as well as in incidences of in-hospital complications and conditions after discharge (all P＞0.05).Proportions of breast milk intake (breast milk intake over total dairy intake) in preterm infants were higher in education group than those in non-education group within 5 d after birth [0.0 (0.0-16.5)％ vs 0.0 (0.0-2.5)％,Z=-3.422],＞5-≤ 7 d [33.7 (0.0-82.8)％ vs 0.0 (0.0-50.3)％,Z=-3.070],＞7-≤ 14 d [75.2(23.5-96.4)％ vs 47.6(0.0-92.2)％,Z=-2.345] and during hospitalization [58.4 (21.0-78.8)％ vs 31.9 (0.0-71.7)％,Z=-3.902] (all P＜0.05).Breastfeeding rates were higher in education group than those in non-education group at the age of 5 d [47.2％(59/125) vs 27.1％ (46/170),x2=12.747],7 d [70.4％ (88/125) vs 51.2％ (87/170),x2=11.031],three months [83.3％(65/78) vs 56.1％ (60/107),x2=15.297] and six months [64.5％ (49/76) vs 49.1％ (53/108),x2=4.282] (all P＜0.05).Exclusive breastfeeding rates in the first,third and sixth months after birth were higher in education group [45.7％(53/116),42.3％ (33/78) and 36.9％ (28/76)] than those in non-education group [21.3％ (32/150),28.0％ (30/107)and 22.2％ (24/108)] (22=17.847,4.091 and 4.703,all P＜0.05),respectively.Conclusions Most gravidas with risk factors of preterm delivery have no confidence on breastfeeding.Prenatal maternal education is an effective and feasible intervention to improve breastfeeding rate for preterm infants in NICU from early hospitalization till six months after birth.

Objective To study the early physical growth pattern,catch-up growth situation,and the influencing factors of early growth in small for gestational age (SGA) preterm infants.Method Our study was a single center,retrospective study.Criteria for infant inclusion were prematurity,SGA (birth weight less than the 10th percentile of related gender and gestational week,according to Fenton curve 2013),born between January 2012 to October 2015,admitted to our neonatal intensive care unit (NICU) within 24 h after birth,hospitalization more than 7 days,and discharged with complete oral feeding.Corrected age (CA) was used to evaluate growth.According to our follow up plan,anthropometric data (weight,length,head circumference) were collected at corrected full term (40 ± 4 weeks),CA (3 ± 1.5) months and CA (6 ± 1.5) months.Catch-up growth was defined as ΔZ greater than 0.67 compared with that at birth,successful catch-up was defined as anthropometric data higher than 10th percentile in target population.The characteristics and influencing factors were compared between infants with and without catch-up growth.Result Eighty-one SGA preterm infants were involved,45 boys and 36 girls.The average gestational age was (34.6 ± 1.7) weeks,birthweightwas(1617 ± 348) g,birthlengthwas(41.0 ±3.2)cm and head circumference was (29.7 ± 2.0) cm.At corrected gestational age (40 ± 4) weeks,CA (3 ± 1.5) months and CA (6 ± 1.5) months,follow-up rate was 86.4％,66.7％ and 58.0％;catch-up growth in weight was 32.9％,55.6％ and 66.0％;successful catch-up growth in weight was 52.9％,64.8％ and 66.0％.At CA (40 ±4) weeks,there were more boys,sooner recover birth weight,and less patent ductus arteriosus (PDA) in catch-up infants (P ＜ 0.05).At CA (3 ± 1.5) months,catch-up infants had large gestational age,and they were longer at discharge,shorter hospital stay,less PDA,and greater body weight at CA 40 weeks,the difference was statistically significant (P ＜ 0.05).At CA (6 ± 1.5) months,there were difference in hospitalization days,percentile of body weight at CA 40 weeks and percentile of all three anthropometrics at CA (3 ± 1.5) months between catch-up and no catch-up growth infants (P ＜ 0.05).Multiple factor analysis showed that percentile of weight at CA 3 months was the independent risk factor of catch-up growth in weight at CA 6 months (P =0.002,OR =1.221,95％ CI 1.076 ～ 1.385).For every 5 percentile increase in body weight percentile at CA (3 ± 1.5) months of age,the likelihood of complete body weight catch-up growth at CA (6 ± 1.5) months increased 2.965 times (95％ CI 1.480 ～ 5.942).Conclusion Both weight and length of SGA preterm infants showed a trend of rapid gain between corrected gestational age (40 ± 4) weeks to CA (3 ± 1.5) months.The factors that influencing the completion of catch-up growth are different at different age.The weight,length,and head circumference percentile at CA about 3 months are good predictors of growth pattern and situation at CA 6 months for the SGA preterm infants.

Objective To study the influence of continuous blood purification(CBP) on cardiac out-put of pediatric patients using bioreactance. Methods Patients underwent CBP in PICU and nephrology ward from March 2014 were prospectively enrolled after approval by ethics committee. CBP therapies were all performed by Fresenius Medical Care hemodialysis machine. Cardiac output values were obtained using the non-invasive cardiac output monitoring ( NICOM) device ( Cheetah Medical). Blood pressure, heart rate, cardiac index(CI) and stroke volume index(SVI) were recorded before the therapy,at the beginning of ther-apy,during the course of therapy,and at the end of each therapy. Results Twenty-one pediatric patients (from 1. 0 year to 15. 5 years) were recruited and 69 treatments were recorded from March 2014 to Decem-ber 2016. The basic CI was 3. 4 (2. 4,6. 1) L/(min·m2),basic SVI was 43 (26,75) ml/(m2·beat). Dur-ing the beginning of therapy,mean arterial pressure(MAP),CI and SVI all dropped from the baseline ( P<0. 001),whereas heart rate increased. During the course of CBP,CI and SVI (were both recorded every 4 hours) kept on dropping and stayed at a relatively lower level. Course CI was 3. 0 (2. 4,4. 6) L/(min·m2) and course SVI was 28 (21,57) ml/(m2·beat). At the end of therapy,CI was 3. 4 (2. 5,5. 3) L/(min· m2),with no significant difference from the baseline CI (P＝0. 073). However,the SVI at the end of therapy was 35 (25,67) ml/(m2·beat),higher than the course SVI but still lower than the basic SVI,the differences were statistically significant ( P<0. 05). Conclusion CI and SVI continue to decline at the beginning of CBP treatment and remain at a lower level throughout the course of treatment. After the therapy, CI has returned to the basic level whereas SVI has not recovered.

Objective To study the video-electroencephalography (vEEG) background activities of very premature infants within the first 7 days after birth.Method The inclusion criteria were gestational age (GA) within 31 weeks and bedside vEEG monitoring within first 7 days of birth.Those with severe cranial ultrasound (CUS) findings,refused treatment,dead and with seizures on vEEG were excluded.The clinical information of included cases were reviewed and the continuity and sleep-wake cycle (SWC) on their EEG background activities were analysed.Result From June 2017 to April 2018,a total of 56 cases were included.Their GA were (29.4 ± 1.6) weeks,and birth weight were (1 306 ± 316) g.A total of 31 cases were diagnosed with asphyxia(1 min Apgar score less than 8),and no one with 5 min Apgar score below 5.Their umbilical artery blood pHs were 7.29 ± 0.07.Their first CUS and the CUS at corrected gestational age (cGA) full term were all within the normal range.The Gesell neurological assessment at cGA 6 months were all within the normal range.As the cGA increased,the continuity of EEG background activities became more continuous as the normal continuous/discontinuous pattern.The 95％ confidence interval of max interburst interval was 95 s.The SWC first occurred at GA 26 w,but was immature,then became mature as the cGA increased.Conclusion The EEG background activity analysis of relatively normal very premature infants within first week shows that interburst interval can be longer than the previous report,and the SWC first appear at GA 26 w.

Objective To study the role of sequential cranial ultrasound (cUS) in the early prognosis of neurodevelopmental outcome in the very low birth weight and extremely low birth weight preterm infants.Method Clinical and examination data of premature infants with birth weight less than 1 500 g,who were admitted to the neonatal intensive care unit of our hospital within 24 hours after birth from January 2012 to December 2016 were analyzed retrospectively.Early cranial ultrasound refers to the cUS scans during the period of 0 to 14 days after birth.Classification was applied to the whole set of early cUS scans based on the most severe lesion observed.Index 1 was periventricular-intraventricular hemorrhage (PVH-IVH),which was divided into none and grade 1 to 4 according to the severity.Index 2 was paraventricular white matter echo,which was divided into normal,slightly enhanced and significantly enhanced.The cUS scan repeated at term-equivalent age (corrected gestational age 37 ～ 44 weeks) was referred to as term cUS.Term cUS was evaluated according to paraventricular leukomalacia (PVL) and enlargement of ventricle.Data were analyzed using IBM SPSS Statistics version 16.0.Result A total of 200 premature infants were collected.The gestational age was (30.1 ± 1.9) weeks,and the birth weight was (1 203 ± 186) g.The median time of cUS scans during hospitalization was 3 times.The age at which the Gesell Development Diagnosis Scale was completed was (11.8 ±4.0) months,the corrected age was (9.7 ±4.0) months.The development quotient (DQ) of gross motor was lower than that of the other four items,and the difference was statistically signi icant.Analysis of data suggested that PVH-IVH grade 3 or grade 4,significantly enhanced paraventricular white matter echo found in early cUS,and PVL or enlargement of ventricle found in term cUS were all associated to lower gross motor DQ (P ＜ 0.05).PVL was also significantly correlated with lower DQ of adaptability,fine motor,language and personal-social (P ＜ 0.05).Conclusion Compared to adaptability,fine motor,language and personal-social,gross motor has the lowest DQ among very low or extremely low birth weight infants.PVH-IVH grade 3 or worse,significantly enhanced paraventricular white matter echo found in early cUS,and PVL or enlargement of ventricle found in term cUS are high risk factors for adverse outcome of gross motor.Among them,PVL found in term cUS suggests poor neurodevelopmental outcome.

Objective To compare the short-term outcomes of less invasive surfactant administration (LISA) with traditional intubate-surfactant-extubate (INSURE) method for respiratory distress syndrome (RDS) in premature infants.Method From January 2017 to December 2018,premature infants (gestational age ≤32 weeks) diagnosed with RDS who needed pulmonary surfactant (PS) administration were prospectively enrolled and randomly assigned into LISA group and INSURE group.The duration of oxygen supply,side effects during PS administration,the outcome and severe adverse events,bronchopulmonary dysplasia (BPD),necrotizing enterocolitis (NEC),3 ～ 4° periventricular intraventricular hemorrhage (PIVH),periventricular leukomalacia (PVL) and death were compared.Result A total of 67 cases were enrolled including 34 in LISA group and 33 in INSURE group.No significant differences existed between the two group on mechanical ventilation rates,oxygen therapy duration and the incidences of severe adverse events.However,the duration of noninvasive ventilation in LISA group was significantly longer than INSURE group [(10.5 ± 8.7) d vs.(7.1 ± 2.6) d,P ＜ 0.05].The LISA group had significantly higher PS reflux rate than INSURE group [41.2％ (14/34)vs.18.2％ (6/33),P ＜0.05].Conclusion LISA has similar clinical effects and severe adverse events as INSURE,but with higher rates of PS reflux and longer duration of noninvasive ventilation.

Objective:To investigate the incidence of hypothyroxinemia in very low birth weight infant (VLBWI) and its effect on early postnatal feeding and weight gain.Methods:This retrospective study analyzed 164 cases of VLBWIs admitted to the Neonatal Intensive Care Unit of Peking University First Hospital from January 2017 to December 2018. According to the gestational age, these VLBWIs were divided into <30 weeks group ( n=85) or ≥30 weeks group ( n=79), and the basic data and thyroid function were compared. According to the levels of serum tetraiodothyronine and free tetraiodothyronine at the first thyroid function test, the subjects were further assigned into normal thyroxine group and hypothyroxinemia group. The risk factors of hypothyroxinemia identified at the first detection were analyzed by single and multiple-facter analysis. The results of the second detection of thyroxine were also analyzed. On the basis of the first detection and receiving treatment or not, the <30 weeks and ≥30 weeks groups were divided into normal thyroxine, hypothyroxinemia treated and hypothyroxinemia untreated subgroups, and differences in the tolerance of early feeding and weight gain were compared between different groups. Two independent samples/paired t-test, rank sum test, Chi-square test and logistic regression were used for statistical analysis. Results:Out of the 164 VLBWIs with the gestational age of (29.7±2.0) weeks and birth weight of (1 210±210) g, 27 cases (16.5%) were extremely low birth weight infants. The age at their first detection was (10.7±3.1) d and the incidence of hypothyroxinemia was 45.1% (74/164), including 71 mild and three severe cases, with a higher incidence in the ≥30 weeks group comparing to the <30 weeks group [55.7%(44/79) vs 35.5%(30/85), χ 2= 6.883, P=0.009]. All the three severe cases were in the ≥30 weeks group. The gestational age ( OR=1.413, 95% CI:1.044-1.912, P=0.025) and male infant ( OR=2.082, 95% CI: 1.047-4.143, P=0.037) were the risk factors of hypothyroxinemia. At the second detection, the incidence of hypothyroxinemia in VLBWIs with normal thyroid function at their first test was 47.6% (39/82), which is higher in the ≥30 weeks group than in the <30 weeks group [64.5%(20/31) vs 37.3%(19/51), χ 2= 5.745, P=0.017]. Among the infants with hypothyroxinemia at the first detection, those untreated had a significantly higher incidence of hypothyroxinemia at the second detection than those treated [81.3%(26/32) vs 38.7%(12/31), χ 2= 11.905, P=0.001]. The incidence of abdominal distension within 21 days, feeding volume on day 7, 14, and 21, and neonatal weight gain within 7, 14, and 21 days were similar between normal thyroxine, hypothyroxinemia treated and hypothyroxinemia untreated subgroups within the ≥30 weeks or the <30 weeks groups (all P>0.05). Conclusions:VLBWI is at high risk of hypothyroxinemia. Two times of postnatal thyroid function tests can help to detect the delayed hypothyroxinemia. Thyroxine level and receiving treatment or not may have no significant effect on the early postnatal feeding and weight gain.