Screening Of Traditional Chinese And Western Medicine Predictors Of Outcome Of Early Threatened Abortion And Establishment Of Clinical Predictive Model

Objective: Through analyzing the clinical data of patients with early threatened abortion,screening the independent influencing factors with the efficacy of predicting the abortion outcome,and then constructing the clinical prediction model,to provide the basis for clinical workers to identify high-risk patients among patients with early threatened abortion,so as to achieve early,effective and targeted treatment,and improve the success rate of abortion.Materials and methods: A total of 120 patients diagnosed with early threatened abortion from January 2019 to November 2022 in the Affiliated Hospital of Shandong University of Traditional Chinese Medicine were included.According to the pregnancy outcome,they were divided into a success group(n=72)and a failure group(n=48).Clinical data of patients were collected including age,previous adverse pregnancy history,vaginal bleeding during pregnancy,abdominal pain during pregnancy,waist during pregnancy,lower abdomen falling during pregnancy,nausea during pregnancy,vomiting during pregnancy,levels of serum progesterone,estradiol,human chorionic gonadotropin and TCM syndrome type.First,univariate analysis was performed on the above variables to screen out the influencing factors that had significant differences between the two groups.Subsequently,through LASSO regression analysis,the multicollinearity and over-fitting problems of the variables were excluded.Finally,independent predictors of abortion outcomes in early threatened abortion were obtained and predictive models were established.The area under the ROC curve,C index,calibration curve,goodness of fit test,bootstrap internal verification method,and decision curve were used to analyze,evaluate and validate the model.Results:(1)Among 120 cases of early threatened abortion,72 cases(60%)were successfully insured;There were 48 cases of fetal failure(40%).Univariate analysis showed that age had a correlation with the outcome of abortion(P<0.05),and the success rate of abortion was significantly reduced when the patient was older than 35 years old.The times of spontaneous abortion,induced abortion and medical abortion had no significant effect on the abortion outcome(P>0.05).Patients with large and long-lasting vaginal bleeding,soreness of the waist,lower abdomen pain,and severe lower abdomen falling had a high rate of abortion failure,while nausea and mild vomiting pointed to a better pregnancy outcome(P<0.05).Serum progesterone and human chorionic gonadotropin levels as well as serum estradiol levels from 7 to 8 weeks of gestation in the abortion failure group were significantly lower than those in the success group(P<0.05),while serum estradiol at 6 weeks of gestation showed no difference between the two groups(P>0.05).The distribution of TCM syndromes in patients with early threatened abortion were kidney deficiency and blood stasis type(39.2%),spleen and kidney deficiency type(35.8%),kidney deficiency type(25.0%),and syndrome differentiation for kidney deficiency and blood stasis type of patients,the highest abortion failure rate.(P<0.05)(2)After dimension reduction treatment was performed on the variables obtained from single factor analysis by LASSO regression analysis,finally age,vaginal bleeding volume,lower abdomen pain,waist soreness,lower abdomen falling,nausea,mild vomiting,serum progesterone content at six weeks of gestation and TCM syndrome types were included in the outcome of early threatened abortion prediction model as independent predictors,and the visualization of the model and quantitative scoring of risk factors were achieved in the form of nomogram.Nomogram showed that in the elderly,patients with excessive vaginal bleeding,severe lower abdominal pain,severe waist soreness,and severe lower abdomen falling without nausea or mild vomiting.Serum P content was low at six weeks of pregnancy.When the TCM syndrome was kidney deficiency and blood stasis,the patients had high risk score and the probability of fetal failure was higher.The area under the curve obtained by plotting the ROC curve of the model was 0.995,the C-index obtained by bootstrap internal verification method was 0.982,and the result of goodness of fit test P>0.05.The calibration diagram showed that the calibration line of the model highly coincided with the standard line.The optimal clinical net benefit rate was about 28% based on the cut-off value of ROC curve and clinical decision-making curve.The above data indicated that the model had relatively ideal accuracy,discrimination and relatively significant clinical effectiveness.Conclusion: The clinical prediction model based on age,vaginal bleeding,abdominal pain,waist soreness,abdominal drop,nausea,mild vomiting,serum P content at six weeks of pregnancy and TCM syndrome types has a relatively accurate prediction ability for the abortion outcome of patients with early threatened abortion,and it has been verified that the model has excellent efficiency and stability.Combining the predictive model with risk assessment for patients with early threatened abortion can help clinical workers to identify,intervene and treat patients with high abortion probability early,which is of great clinical significance for improving the abortion outcome or timely terminating pregnancy to reduce the damage to the mother.