| Objective: Benefiting from the continuous development of perinatal medicine and neonatal critical care technology,the survival rate of ultra early and ultra-low weight premature infants is getting higher and higher.The premature birth rate in China is about 7%.About 1.2 million premature infants are born every year,and the surviving gestational age is getting smaller and smaller,but there are many problems in their quality of life in the future.MBDP is a major problem.This disease,also known as osteopenia,is the reduction of bone tissue content caused by the reduction of the number of bone trabeculae and the thinning of bone cortex,It is worth noting that the disease may be asymptomatic in the early stage.The typical clinical symptoms usually appear 6-12 weeks after birth,when the bone has been severely demineralized.The common symptoms are rickets and pathological fractures.Rib softening or fracture can affect the lungs,leading to respiratory problems or dependence on mechanical ventilation.In the long term,it will also affect the final height and bone status in adulthood.It is a disease that seriously affects the short-term and long-term prognosis of children.Because the early clinical manifestations of children are not obvious,it is difficult to identify only by symptoms and signs.At present,there is still a lack of unified early screening and diagnosis methods at home and abroad.This study attempts to explore the ultra early prediction model of metabolic bone disease in preterm infants combined with pregnancy and postpartum related high-risk factors,and try to study the joint diagnosis model with ultrasonic guided wave technology as the core.At present,most domestic institutions usually carry out regular venous blood index detection and / or radiography for hospitalized preterm infants with high risk of metabolic bone disease to assist in diagnosis.However,this is not suitable for some high-risk premature infants,very low or ultra-low birth weight infants,because the development of MBDP takes a certain time,so it is not suitable to repeat the examination with traumatic or radiation hazards for many times.The first part of this study is to analyze the clinical data of preterm infants and their mothers during pregnancy in the past 8 years,and establish a super early prediction model of metabolic bone disease in preterm infants based on clinical high-risk factors.The second part of this study explores a new joint diagnosis model of metabolic bone disease in preterm infants through ultrasonic guided wave detection technology.Methods: The first part included preterm infants and mothers born in Shanghai first maternal and infant hospital from 2014 to 2021.The clinical risk factors related to metabolic bone disease of preterm infants were analyzed by single factor and multi factor.After variable screening,the prediction model of metabolic bone disease of preterm infants was established and verified internally.The second part included premature in NICU of Shanghai first maternal and infant hospital from January to March 2022.The ultrasonic bone measurement experiment was carried out within 72 hours after birth.At the same time,the medical records were collected,and the relevant analysis was carried out on the records of all subjects,including gender,gestational age,weight,blood alkaline phosphatase and blood phosphorus.Results: The first part included 2701 premature infants and their mothers in the Shanghai first maternal and infant hospital from 2014 to 2021.There were 132 cases of MBDP premature and 2569 cases in the control group.It’s no difference in the general condition of gravida between the two groups in different age groups,pre pregnancy BMI,nationality,occupation,payment method,registered residence,birth and scar uterus(P > 0.05).In terms of nutrition during pregnancy,in the MBDP group of premature infants,the pregnant women with folic acid deficiency during pregnancy was 15.91%,which was significantly higher than that in the control group(7.59%)(P = 0.001).31.06% of pregnant women in the metabolic bone disease group of preterm infants had vitamin D deficiency,which was higher than that in the control group(22.11%)(P = 0.016).In terms of medication during pregnancy,the proportion of pregnant women with a history of calcium supplementation in the metabolic bone disease group of preterm infants was 15.91%,which was significantly higher than 9.85% in the control group(P = 0.024).In multivariate analysis,the risk of metabolic bone disease in preterm infants lacking folate during pregnancy was 2.04 times higher than that in mothers without folate deficiency(95% CI: 1.23-3.41).In the group of metabolic bone disease of premature infants and the control group,the proportion of pregnant women using dexamethasone was 45.45% and 28.42%,respectively.It’s difference between the groups(P < 0.001);The proportion of pregnant women using magnesium sulfate was72.73% and 32.35% respectively.It’s difference between two groups(P < 0.001);In multivariate analysis,the risk of MBDP in prematures of mothers with magnesium sulfate medication history was 7.01 times higher than that of mothers without magnesium sulfate medication history(95% CI: 4.46-11.02);The history of antibiotics was the protective factor of MBDP in prematures(AOR = 0.60,95% CI:0.40-0.89).In terms of pregnancy outcome and neonatal diseases,the median Apgar score(1minute)of premature infants with metabolic bone disease and the control group were8 and 9 respectively,and the difference between the groups was statistically(P <0.001);The proportion of LBW premature was 33.33% and 55.16%,the proportion of VLBW premature was 55.30% and 5.22%,and the proportion of ultra-low birth weight was 9.10% and 0.66%.It’s difference in the distribution of birth weight between the groups(P < 0.001).In the abnormal bone metabolism group and control group,the proportion of neonatal anemia was 55.30% and 5.72% respectively.It’s difference between two groups(P < 0.001);The proportion of neonatal sepsis was21.21% and 1.75% respectively,and the proportion of neonatal respiratory distress syndrome was 88.64% and 23.51% respectively(P < 0.001);The proportion of neonatal pneumonia was 27.27% and 16.62% respectively.It’s difference between two groups(P = 0.002);In multivariate analysis,the risk of abnormal bone metabolism in low birth weight preterm infants was 5.87 times higher than that in non low birth weight preterm infants(95% CI: 1.75-19.64),the risk of abnormal bone metabolism in very low birth weight preterm infants was 25.23 times higher than that in non low birth weight preterm infants(95% CI: 7.16-88.99),and the risk of abnormal bone metabolism in ultra-low birth weight preterm infants was 26.29 times higher than that in non low birth weight preterm infants(95% CI: 5.82-118.72),The risk of abnormal bone metabolism in anemia preterm infants was 2.78 times higher than that in non anemia preterm infants(95% CI: 1.68-4.60),the risk of abnormal bone metabolism in septic preterm infants was 2.80 times higher than that in non septic preterm infants(95% CI: 1.44-5.42),and the risk of abnormal bone metabolism in respiratory distress syndrome preterm infants was 6.11 times higher than that in non respiratory distress syndrome preterm infants(95% CI: 3.27,11.43)The prediction model of metabolic bone disease in preterm infants was established by incorporating the above influencing factors during pregnancy,prenatal and postpartum.The results showed that the AUC value was the largest,with accuracy(0.835),sensitivity(0.973),specificity(0.828),positive predictive value(0.213)and negative predictive value(0.998).The second part included 73 subjects.18 children with metabolic bone disease;55 prematures in control group.The median gestational week of the MBPD group was 32.1 weeks,which was lower than which of control group(33.3 weeks),and it’s statistically(P < 0.001).The median birth weight of MBDP group was 1435 g,which was much lower than which of control group(1885 g),and the difference was statistically significant(P < 0.001).The median level of alkaline phosphatase in the case group was 554 IU / L,which was higher than which in control group of preterm babies(271 IU / L)(P < 0.001).The median concentration of serum phosphorus in MBDP group was 1.47 mmol / L and that in control group of premature infants was 1.82 mmol / L.The serum phosphorus concentration in MBDP group was lower than which in control group(P < 0.001).In terms of ultrasonic sound velocity,the median value of sunlight SOS in the case group was 2491 and 2914 in control group of premature infants;The median value of verasonic SOS was 3857 in the case group and 4382 in the premature control group.The two ultrasonic velocities in the case group were lower than which in each control group,and the difference was statistically(P < 0.001).It’s a positive correlation between sunlight SOS and gestational week,birth weight and serum phosphorus level.The correlation coefficients were 0.678,0.823 and 0.743 respectively,P values were less than 0.001,which was statistically significant.There was a positive correlation between verasonics SOS and gestational week,birth weight and serum phosphorus level.The correlation coefficients were 0.634,0.810 and 0.717 respectively,P values were less than 0.001,which was statistically significantThere was a strong positive correlation between sunlight SOS and verasonics SOS.Before adjusting for other factors,the correlation coefficient was 0.958,which was statistically(P < 0.001).After adjusting gestational week,birth weight,alkaline phosphatase level and serum phosphorus level,the partial correlation coefficient was0.822,which was still statistically significant(P < 0.001).For the diagnostic value of MBDP in prematures,the ROC curves of sunlight SOS and verasonics SOS in the diagnosis of MBDP in preterm infants were drawn respectively(see Figure 5).The AUC of area under ROC curve of sunlight SOS was0.950,and its 95% confidence interval was 0.895 ～ 1.000.Compared with AUC = 0.5,the discrepancy was statistically(P < 0.001).The AUC of verasonics SOS was0.970,and its 95% confidence interval was 0.933 ～ 1.000.Compared with AUC = 0.5,the discrepancy was statistically(P < 0.001).Compared with sunlight,verasonics has higher accuracy(0.959 vs 0.849),specificity(0.982 vs 0.818),while sunlight has higher sensitivity(0.944 vs 0.889).In premature infants,the cut-off point of ROC curve is 0.244,and the critical value of sound velocity for diagnosis is 2731-2773.The critical value of 2750 was selected as the best critical value for the diagnosis of metabolic bone disease in mid-term preterm infants by sunlight SOS.2550 was selected as the best critical value for the diagnosis of metabolic bone disease in early preterm infants.Verasonics SOS in preterm infants,the cut-off point of ROC curve is 0.409,and the critical value range of sound velocity is 3897-3941.The critical value of 3900 was selected as the best critical value for the diagnosis of MBDP in all preterm babies.Conclusion: There are potential risk factors of MBDP in preterm babies in three different stages of pregnancy,prenatal and postpartum,including maternal nutritional status during pregnancy,pregnancy complications / complications,medication and neonatal diseases.The preliminary establishment of ultra early prediction model is helpful to clinical early detection,early diagnosis and early treatment,and reduce the risk of MBDP in preterm babies.The influencing factors of MBDP in preterm babies and the feasibility of quantitative ultrasound technology in the evaluation of MBDP in preterm babies were explored.At the same time,through the control experiment of sunlight omnisense and verasonics,the AUC,sensitivity,specificity and other indicators under different critical values were compared,and the best critical values for the diagnosis of MBDP in premature babies were obtained,It can provide a scientific basis for the early diagnosis and treatment of metabolic bone disease. |
PDF Full Text Request