The main objective of this report is to help the study that analyses the use of lower segment uterine scar thickness as a tool to assess the likelihood of having a successful vaginal birth after caesarean section.
The data analysis describes in the report will check if having the scar thickness measured is an effective screening tool and if it increases the rate of vaginal deliveries. Beyond that, the report will demonstrate statistical methods to discover if the variables used to create the report graphs have a significant correlation between them.
The dataset analysed in this report has a total of 341 observations and 17 variables. The variables examined were the scar thickness, the hospital abbreviation and scar grade.
First of all, a table will illustrate the number and percentage of missing values for all variable of the dataset.
## variable n_miss pct_miss
## 1 myometrium thickness 340 99.706745
## 2 Complications 340 99.706745
## 3 Indication details 286 83.870968
## 4 scar thickness new 246 72.140762
## 5 Scar grade 236 69.208211
## 6 scar thickness 234 68.621701
## 7 CsUrgency 165 48.387097
## 8 gestation at the scan 25 7.331378
## 9 MaternalIndicationCs 8 2.346041
## 10 ObstetricIndication 7 2.052786
## 11 FetalIndicationCs 7 2.052786
## 12 Condition of the CS scar 4 1.173021
## 13 1 0 0.000000
## 14 Hospital Abbreviation 0 0.000000
## 15 ModeBirth 0 0.000000
## 16 uterus closed in 0 0.000000
## 17 US at 36 weeks 0 0.000000Secondly, it is analyzed the number of observations for each response of the variable hospital abbreviation:
Finally, it was considered the following statements for the data analysis performed by this report:
The initial graph created will illustrate the difference between the number of FVBAC and VBAC scar thickness measured.
It can be noticed with the FVBAC vs VBAC Scar Thickness Measured Graph that the indended mode of birth VBAC has more scar thickness not measure. Furthermoe, a Chi-Square test will be applyied to check if there is a significant correlation between the graph varaibles indended mode of birth and scar thickness measure (yes/no).
Chi-Square test is a statistical method which used to determine if two categorical variables have a significant correlation between them. The test is based on the p-value, if the p-value that comes out in the result is less than a predetermined significance level, which is 0.05 usually, then we reject the null hypothesis and conclude that the two variables are in fact dependent.
##
## Pearson's Chi-squared test with Yates' continuity correction
##
## data: scar$`Hospital Abbreviation` and scar$`thickness measured`
## X-squared = 8.9841, df = 1, p-value = 0.002723The p-value obtain by the Chi-Square test is less than significance level of 0.05, therefore the two variables are related to each other.
The second graph of this report will illustrate the scar thickness measure according to the hospital abbreviation. The variables used to reproduce this graph were scar thickness and Hospital Abbreviation. The dashed line (yintercept) represents the mean of the scar thickness for the dataset.
It was decided to use the variable scar thickness because it has less missing value comparing to the varaible scar thickness new. In order to use the variable scar thickness as a numeric variable, it was consider the folllowing statments:
To evaluate the correlation between the variables from Scar Thickness Graph it will be used a heterogeneous correlation matrix. This matrix based on Pearson product-moment correlations between variables, it is a statistical technique recommended to evaluate the relationship between a categorical variable (Intended Mode of Birth) and a numerical variable (Scar Thickness).
hetcor(scar$`scar thickness`, scar$`Hospital Abbreviation`)##
## Two-Step Estimates
##
## Correlations/Type of Correlation:
## scar$`scar thickness` scar..Hospital.Abbreviation.
## scar$`scar thickness` 1 Polyserial
## scar..Hospital.Abbreviation. -0.1181 1
##
## Standard Errors:
## scar$`scar thickness` scar..Hospital.Abbreviation.
## 0.1057
## Levels: 0.1057
##
## n = 107
##
## P-values for Tests of Bivariate Normality:
## scar$`scar thickness` scar..Hospital.Abbreviation.
## 2.317e-08
## Levels: 2.317e-08The p-value obtained by the Tests of Bivariate Normality is 0.0007, this value is less than the significance level of 0.05. Therefore, the two variables are related to each other.
To conclude, the report final data analysis will check if measured scar thickness has a correlation with scar grade. The scar grade can was grouped as grade I in one group and II, III, IV in another group called grade II.
It can be noticed with Scar Grade Graph that the FVBAC intended mode of birth has a higher number of scar grade ranked as II. Furthermore, a Chi-Square test will be applied to check the significant correlation between the graph variables.
##
## Pearson's Chi-squared test
##
## data: scar$`Hospital Abbreviation` and scar$`Scar grade`
## X-squared = 6.5687, df = 2, p-value = 0.03747The p-value obtain by the Chi-Square test is less than the significance level of 0.05, therefore the two variables related to each other.