Liver Function Tests in Child-Pugh Scale



Figure 2.9 . Liver damage 3/6 according to Ishak (Patient N20-0204706)


Figure 2.10 . Liver damage grade 4/6 according to Ishak (Patient A10-0047088)


Figure 2.11 . Liver damage 5/6 according to Ishak (Patient N18-0396611)



Figure 2.12 . Liver damage 6/6 according to Ishak (Patient N19-0208006)


2.6.4 Data collection

Data were collected according to sample medical records (Appendix).

ICG clearance and preserved liver volume were measured and calculated by the same group and according to a unified procedure.

2.7 Research process


Data collection: from October 2016 to the end of March 2021, in which, secondary data was collected from October 2016 to before July 9, 2019, primary data was collected from July 9, 2019 to the end of March 2021.

Data processing: from April 2021 to the end of June 2021.

Writing thesis and two thesis-related articles: July 2021 to December 2021.

2.8 Methods of data presentation and analysis


Data were processed using IBM SPSS 26.0 and R 4.0.5 software.

Continuous variables will be presented in quartiles with Q1 being the 25th percentile, Q2 being the median, Q3 being the 75th percentile, minimum and maximum values: age,


body surface area (BSA), ICG-R15, ICG-PDR, serum albumin, serum bilirubin, INR, RLV/SLV, RLV/P, number of days in hospital. When comparing between groups, variables will be presented using Q2 quartiles (Q1 - Q3).

- If the variable has a normal distribution, the difference between two groups will be tested by T-test, between multiple groups by One-Way ANOVA test and post hoc analysis by Tukey's HSD method.

- If the variable is not normally distributed but the logarithm is normally distributed, the difference between two groups will be tested by T-test, between multiple groups by One-Way ANOVA test for the logarithm of the variable. This method will find more differences between groups than non-parametric tests.

- If the variable is not normally distributed and the logarithm is also not normally distributed, the difference between two groups will be tested by Mann-Whitney U test, between multiple groups by Kruskal-Wallis test.

Qualitative or ordinal variables: gender, Child-Pugh score, Child-Pugh classification, extent of liver resection, method of liver resection, degree of cirrhosis on histopathology, nature of liver tumor on histopathology, surgical complications, complications of liver failure, liver failure classification, will be presented as proportions, tested for differences by Chi-Square test or Fisher's Exact test.

In the first research objective , we found the correlation between the continuous variable ICG clearance (ICG-R15) and the ordinal variable Ishak score cirrhosis level using Spearman correlation.

In the second study objective , we compared ICG clearance and Child-Pugh score in assessing liver function before liver resection based on two criteria:

- First criterion: compare ICG clearance with Child-Pugh score in predicting the degree of liver cirrhosis on histopathology according to Ishak score.


- Second criterion: compare ICG clearance with Child-Pugh score in predicting liver failure complications after liver resection.

In the third research objective , we investigated the relationship between two preoperative liver function assessment tools, ICG clearance (ICG-R15) and preserved liver volume (RLV/SLV), and complications of liver dysfunction after major hepatectomy. From there, we built a model to estimate the risk of liver failure after major hepatectomy based on a logistic regression model.

In the study, we built two prognostic models to estimate two outcome variables: the degree of liver fibrosis on pathology according to the Ishak scale and the rate of liver dysfunction after major hepatectomy.

- Independent variables were selected to be included in the prognostic model based on clinical practice regarding the application of these variables in predicting outcome variables.

- Model to estimate the degree of liver cirrhosis on pathology according to the Ishak scale:

This model was built based on an ordinal logistic regression model for the outcome variable. Based on two demographic variables (gender, age) and preoperative liver function tests (ICG-R15, serum bilirubin concentration, INR, platelet count), three models to estimate the Ishak score were built as follows:

+ Univariate model based on ICG-R15 values ​​only.

+ Multivariate model based on all 6 variables above.

+ The multivariate model is reduced by the backward stepwise variable selection method based on the AIC (Akaike Information Creation) selection criterion.


- Model to estimate the rate of liver failure after major liver resection:

This model was built based on the logistic regression model. Based on demographic variables (age, sex), clearance (ICG-R15), pathological cirrhosis degree (Ishak score) and preserved liver volume (RLV/SLV), two models to estimate the risk of liver failure after liver resection were built as follows:

+ Multivariate model based on all 5 variables above.

+ The multivariate model is reduced by the backward stepwise variable selection method based on the AIC (Akaike Information Creation) selection criterion.

- The reduced multivariate model is built with the aim of making the model easy to apply in clinical practice while still maintaining the same efficiency as the full multivariate model.

- The model performance indicators are validated and optimized using bootstrap resampling with 1000 iterations.

2.9 Ethics in research


ICG has been approved for use in patients by the Vietnamese Ministry of Health since May 2016 (Appendix). Ho Chi Minh City University of Medicine and Pharmacy Hospital is the only hospital in the country that is authorized to use ICG in assessing liver function before liver resection.

All patients were explained the benefits of ICG clearance in assessing liver function, possible side effects and allergic reactions, as well as the cost of the test. Only when the patient agreed, the test was performed. Patient information was recorded and included in the study only with the patient's consent.


The topic was approved by the Ethics Council in Biomedical Research of Ho Chi Minh City University of Medicine and Pharmacy on July 9, 2019, code number 316/DHYD-HDDD (Appendix).


Chapter 3. RESULTS


During the period from October 2016 to the end of March 2021, 340 patients meeting the inclusion criteria were included in the study.

3.1 Sample population characteristics


3.1.1 Gender, age, skin area

The male:female ratio is 4.6 (279:61).

Patient age is normally distributed, median is 59 years old, Q1 is 51 years old, Q3 is 66 years old, youngest is 17 years old, oldest is 86 years old.

Skin area has a normal distribution, median is 1.66m 2 , Q1 is 1.56, Q3 is 1.75, smallest is 1.18, largest is 2.03.

3.1.2 Hepatitis

Table 3.1. Hepatitis status of patients


Hepatitis

Quantity

%

Hepatitis B

206

60.6

Hepatitis C

51

15.0

Co-infection with hepatitis B and C

1

0.3

No hepatitis

82

24.1

Total

340

100

Maybe you are interested!


3.1.3 Preoperative liver function

3.1.3.1 Liver function according to Child-Pugh scale

Serum albumin has a normal distribution.

Serum bilirubin has a right-skewed distribution, logarithm has a normal distribution. INR has a right-skewed distribution, logarithm has a non-normal distribution.


Table 3.2. Liver function tests in the Child-Pugh scale


Blood tests

Median

Q1

Q3

Smallest

Biggest

Albumin (g/L)

40.60

37.80

42.90

28.40

49.70

Bilirubin (mmol/L)

12.20

9.50

14.74

3.59

35.00

INR

1.05

1.00

1.11

0.89

1.79


Table 3.3. Liver function according to Child-Pugh scale


Classify

Point

Quantity

%


A

5

306

90.0

6

31

9.1

B

7

3

0.9

Total

340

100


3.1.3.2 Platelet count

Platelet count has right-order distribution, logarithm has normal distribution, median is 214 G/L, Q1 is 164 G/L, Q3 is 275 G/L, minimum is 46 G/L, maximum is 719 G/L.

3.1.3.3 ICG Purity

ICG-PDR has a normal distribution.

ICG-R15 has a right skewed distribution, logarithm has a normal distribution.

Table 3.4. ICG clearance


ICG Purity

Median

Q1

Q3

Smallest

Biggest

ICG-PDR (%)

19.50

16.60

19.50

4.20

43.80

ICG-R15 (%)

5.40

3.30

8.30

0.10

53.30

Comment


Agree Privacy Policy *