TaviScore

Clean repository of my Bachelor's Thesis about Developing a Developing a One-Year Risk Score for Transcatheter Aortic Valve Implantation

Main stuff is in 02_models

00 Data preparation

1. Execute create_time2event.R
2. For follow-up distribution plots, execute FUP_plots.R

3. Basic parameter plots are in parameter_plots_basic.R

4. Parameter plots stratified by risk (STS scores <=4, >8) are in´parameter_plots_riskstratified.R

01 Statistical analysis

Execute statistical_analyses.R. T-test results:

name	p.value	p.adjustBH	significantBH	p.adjustbonferroni	significantbonferroni
age	0.0125234257768774	0.0268359124	1	0.188	1
height	0.482935928103082	0.4829359281	0	1	0
weight	0.227478736745308	0.2938234259	0	1	0
bmi	0.0671981859432215	0.1119969766	0	1	0
bsa	0.31569220140739	0.3382416444	0	1	0
scoreii_log	0.000396318907582548	0.0019815945	1	0.006	1
calc_sts	2.1602992412237e-05	0.0001620224	1	0	1
creatinine	0.00190886280521956	0.0071582355	1	0.029	1
gfr	0.0274026651295069	0.0513799971	0	0.411	0
hb	1.22251732928169e-06	1.83378e-05	1	0	1
thrombo	0.300436060446901	0.3382416444	0	1	0
ck	0.0854377356608728	0.1281566035	0	1	0
hrate	0.235058740756237	0.2938234259	0	1	0
gradient_mean	0.00278155195625097	0.0083446559	1	0.042	1
lvef	0.00762988047802821	0.0190747012	1	0.114	1

Some Fisher's exact test results:

name	p.value	p.adjustBH	significantBH	p.adjustbonferroni	significantbonferroni
sex	0.0183086768861968	0.0651196902	0	1	0
diab	0.0330957099873782	0.1049932869	0	1	0
hyper	0.509051870376169	0.6787358272	0	1	0
dyslip	0.573318859356573	0.724717356	0	1	0
copd	0.00105187120823079	0.0120965189	1	0.097	0
cerebro	1	1	0	1	0
cerebro_strokebl	0.860437851780319	0.909888303	0	1	0
pacemaker	0.21658203256523	0.3985109399	0	1	0
valvulo	0.694732127232969	0.7890784655	0	1	0
cad	0.398692710469582	0.6113288227	0	1	0
pci	0.377984155566603	0.6102106167	0	1	0
mi	0.0102112330033713	0.0391430598	1	0.939	0
ad	2.45400259120383e-06	0.0001128841	1	0	1
csurgery	0.431826404033876	0.6306036376	0	1	0

02 Models

Model plots: Problem: STS and EuroSCORE II cannot distinguish between intermediate and low risk patients

We decided on Lasso_smaller.R for all patients, and Lasso_intermed_low.R for patients with STS scores <= 8.

Small model: n= 1434, number of events= 124

	coef	exp(coef)	se(coef)	z	Pr(>|z|
sex	0.613995	1.847799	0.192114	3.196	0.001394	**
age	0.044255	1.045249	0.013778	3.212	0.001318	**
ad	0.735021	2.085526	0.204445	3.595	0.000324	***
copd	0.635212	1.887422	0.224876	2.825	0.004732	**
medi_diuretic	0.473712	1.605944	0.219912	2.154	0.031233	*
hb	-0.018002	0.982159	0.004919	-3.660	0.000252	***
regurg_mitral34	0.606873	1.834686	0.190928	3.179	0.001480	**

Measure	Value
Concordance	0.732 (se = 0.023 )
Likelihood ratio test	79.04 on 7 df, p=2e-14
Wald test	79.31 on 7 df, p=2e-14
Score (logrank) test	84.79 on 7 df, p=1e-15
10 fold CV	0.7185
Mean permutation test	0.5861

Intermediate/low model: n= 1268, number of events= 93

	coef	exp(coef)	se(coef)	z	Pr(>|z|
sex	0.699172	2.012087	0.231422	3.021	0.002518	**
copd	0.490690	1.633443	0.272610	1.800	0.071866	.
ad	0.485754	1.625400	0.250047	1.943	0.052059	.
medi_diuretic	0.491276	1.634400	0.243235	2.020	0.043409	*
hb	-0.018414	0.981754	0.005561	-3.312	0.000928	***
ccs_stratified	-0.536056	0.585051	0.265223	-2.021	0.043264	*
regurg_tricuspid34	0.724856	2.064434	0.298633	2.427	0.015214	*
regurg_mitral34	0.211762	1.235854	0.250415	0.846	0.397752

Measure	Value
Concordance	0.713 (se = 0.028 )
Likelihood ratio test	50.1 on 8 df, p=4e-08
Wald test	52.68 on 8 df, p=1e-08
Score (logrank) test	55.91 on 8 df, p=3e-09
10 fold CV	0.688
Mean permutation test	0.5747

Use model to predict "new" data from high risk patients

Just predicting it does not work so well

Confusion Matrix and Statistics

	Reference=0	Reference=1
Prediction=0	133	30
Prediction=1	2	1

Patients with high STS scores also have higher linear predictors but the range of STS scores within each hazard category is high. The new data gets sorted into the high and intermediate hazard categories.

	N	Observed	Expected	(O-E)^2/E	(O-E)^2/V
hazard=high	103	26	19.41	2.23	6.04
hazard=intermediate	62	4	11.41	4.81	7.69
hazard=low	1	1	0.18	3.74	3.78

03 Test set: cohort from other geographical location

The analysis is in the R notebook