RESUMO
OBJECTIVE: We aim to analyze each variable affecting FHR to build objective decision basis using canonical correlation analysis METHODS: The sixty four hundred and fifty five cases of NST from 1988 to 1997 at Hanyang University hospital were collected. We used FHR interpretation softwares, HYFM-I & II those were already developed by author, for extraction of each variable of FHR. We classified the variables into 2 groups, F.A.M(FHR, Amplitude, Mean minute range) and N.S.F(NST weeks, Signal loss, Fetal movement) groups. The Canonical correlation was compared between each variable and group. RESULTS: The canonical correlation between F.A.M & N.S.F are as follows; r=0.40 (p=0.001) in preterm, r=0.34 (p=0.001) in term, and r=0.41 (p=0.004) in postterm. The high linear dependency is shown as r=0.79 (p=0.013) before 23 weeks, the irregular changes shown from 23 weeks to 30 weeks (r=0.47 in 23-24 weeks, r=0.29 in 25-26 weeks, r=0.56 in 27-28 weeks, r=0.24 in 29-30 weeks) and the stable relative changes shown (about r=0.3-0.4) after 30 weeks (r=0.32 in 31-32 weeks, r=0.33 in 33-34 weeks, r=0.37 in 35-36, r=0.33 in 37-38 weeks, r=0.37 in 39-40 weeks, r=0.35 after 41 weeks) (p<0.02). In the linear correlation of FHR, two indicies of FHR variables, AMP and MMR pertaining to N.S.F., FHR(r) is 0.06, AMP and MMR seem to be identical as r=0.30 and then the canonization in term and postterm[FHR(r)=-0.27, AMR(r)=0.23, and MMR(r)=0.35]lies in the regular pattern. Moreover, there is rather a linear correlation between F.A.M and the gestational weeks in preterm and it is changed into linearity similar to zero in term and postterm pregnancies[term(r)=0.01, postterm(r)=0.06] CONCLUSIONS: This study shows the linearity of FHR and the variable which are based on the gestational weeks organically, comprehensibly, and quantitatively. It would be utilized as the fundamental standard of the linearity. The changes from the irregular linear correlation to the regular pattern according to advancing gestational weeks means the changes from the immature and transitional level to the mature level. The closer investigation of the linear and
Assuntos
Feminino , Gravidez , Coração Fetal , Frequência Cardíaca FetalRESUMO
OBJECTIVE: We aim to statistically analyze the variables affecting FHR using principal component analysis and multiple regression analysis. It is tried to find which variables significantly affect FHR. Thus, it would help further work of building objective decision basis to analyze variables affecting FHR. METHODS: The samples are chosen from pregnant women who took a NST at Hanyang University Hospital. It consists of 5,314 data which have been collected from 1989 to 1997. For collection of data and values of each variable, our own FHR interpretation softwares, HYFM-I (DOS version, since 1989) and HYFM-II (Windows version, since 1998) were used. Then, the principal component analysis is performed to find significant variable and multiple regression analysis is performed using these variables. RESULTS: The mean baseline FHR is not influenced by a specific factor in abnormal group. The gestational week, loss of signal, variability, and 5 minutes Apgar score are chosen as main effects through principal component analysis. From the results of regression analysis, it is noticed that 5 minutes Apgar score which is one of neonate prognosis after birth causes an increase of estimated FHR score (EFS), whereas the loss of signal causes a decrease of EFS. The variability in normal group is more active increasing factor of EFS than abnormal group. CONCLUSIONS: Based on gestational week, 5 minutes Apgar score is highly related with fetal maturity in both normal and abnormal groups. Also, it is found that different patterns of each variable by groups are due to delay of fetal growth that is caused by disease of a pregnant women. Thus, the further studies to build objective decision basis are need.
