| Analyses univariées des difference de chaque variable en utilisant la donnée d= variable au temps t0 - variable au temps t1 de tous les items, ensuite voir s’il y’a une correlation avec le fait que les patients soit répondeur ou non |
| On cherche à quantifier l’association brute entre une variable à expliquer (réponse/non réponse) et une variable explicative (facteur). Ici on a pris la variable dependente : Patient répondeur/non répondeur les variables predictives definie dans l’item explanatory |
| le but etant d’identifier le facteur lié à la réponse ou non réponse du patient. |
| l’OR doit exclure la valeur 1 pour etre significatif. |
| plus l’OR s’eloigne de 1 plus l’effet ou le rapport est élevé |
| Les resulats contiennent l’OR, l’interval de confiance et la p-value |
#Les odds ratio( OR) des analyses univariées
library(funModeling)
library(tidyverse)
library(finalfit)
library(knitr)
library(rmarkdown)
library(markdown)
library(MASS)
HD<-OR_M
dependent = "P_Reponse"
explanatory1 = c("HID","volume_lobe_T","MMRC", "BODE", "VEMS", "Tiffe", "CV_ml", "CVF_ml", "VR_ml", "CPT_ml", "TLCO", "Pi_pct", "SNIP_KPA", "Ci_rest_recalc", "CI_pic_recalc", "CI_diff_recalc", "CI_isoP", "temps_sec","VO2_max", "Wmax","VT_isotimE", "vt_pic", "VRI_isotime","VO2_FC","dist_parc", "Nadir_sat")
explanatory2 = c("volume_lobe_T","VEMS","CV_ml","CVF_ml","VR_ml","CPT_ml")
res_glm_uni1<-HD%>%
glmuni(dependent, explanatory1) %>%
fit2df(estimate_suffix=" (univarié)")
kable(res_glm_uni1,row.names=FALSE, align=c("l", "l", "r", "r", "r", "r"))
| HIDOui |
0.42 (0.05-2.14, p=0.329) |
| volume_lobe_T |
1.00 (1.00-1.01, p=0.011) |
| MMRC |
1.72 (0.83-3.91, p=0.161) |
| BODE |
1.81 (1.13-3.43, p=0.030) |
| VEMS |
0.94 (0.84-0.98, p=0.096) |
| Tiffe |
0.95 (0.86-1.03, p=0.250) |
| CV_ml |
1.00 (1.00-1.00, p=0.009) |
| CVF_ml |
1.00 (0.99-1.00, p=0.005) |
| VR_ml |
1.00 (1.00-1.00, p=0.018) |
| CPT_ml |
1.00 (1.00-1.00, p=0.066) |
| TLCO |
1.00 (0.85-1.15, p=0.969) |
| Pi_pct |
0.97 (0.91-1.02, p=0.210) |
| SNIP_KPA |
0.50 (0.11-1.36, p=0.229) |
| Ci_rest_recalc |
1.00 (1.00-1.00, p=0.128) |
| CI_pic_recalc |
1.00 (0.99-1.00, p=0.133) |
| CI_diff_recalc |
1.00 (1.00-1.00, p=0.518) |
| CI_isoP |
1.00 (0.99-1.00, p=0.033) |
| temps_sec |
0.99 (0.99-1.00, p=0.118) |
| VO2_max |
0.03 (0.00-93.93, p=0.421) |
| Wmax |
0.77 (0.58-0.90, p=0.017) |
| VT_isotimE |
0.99 (0.98-1.00, p=0.022) |
| vt_pic |
1.00 (0.99-1.00, p=0.068) |
| VRI_isotime |
1.00 (0.99-1.00, p=0.288) |
| VO2_FC |
1.05 (0.69-1.55, p=0.776) |
| dist_parc |
1.00 (0.99-1.00, p=0.342) |
| Nadir_sat |
0.97 (0.87-1.01, p=0.304) |
| volume_lobe_T |
1.00 (1.00-1.01, p=0.011) |
| VEMS |
0.94 (0.84-0.98, p=0.096) |
| CV_ml |
1.00 (1.00-1.00, p=0.009) |
| CVF_ml |
1.00 (0.99-1.00, p=0.005) |
| VR_ml |
1.00 (1.00-1.00, p=0.018) |
| CPT_ml |
1.00 (1.00-1.00, p=0.066) |
###Les odds ratio( OR) des analyses multivarié
ceci permet de predire l’association entre 1 ou plusieurs variables explicativessur une variable dite à expliquer
| volume_lobe_T |
Mean (SD) |
401.4 (322.7) |
1447.1 (659.0) |
1.00 (1.00-1.01, p=0.011) |
1.03 (0.00-60107.33, p=1.000) |
| VEMS |
Mean (SD) |
60.0 (159.5) |
-305.2 (143.4) |
0.94 (0.84-0.98, p=0.096) |
0.45 (0.00-177451268695755.94, p=0.999) |
| CV_ml |
Mean (SD) |
-70.0 (533.5) |
-729.0 (518.5) |
1.00 (1.00-1.00, p=0.009) |
0.95 (0.00-4814069.98, p=1.000) |
| CVF_ml |
Mean (SD) |
-26.4 (519.9) |
-711.8 (389.7) |
1.00 (0.99-1.00, p=0.005) |
1.08 (0.00-871696.32, p=1.000) |
| VR_ml |
Mean (SD) |
38.2 (917.9) |
1078.7 (993.0) |
1.00 (1.00-1.00, p=0.018) |
1.13 (0.00-1927.80, p=0.999) |
| CPT_ml |
Mean (SD) |
44.6 (614.1) |
484.1 (595.7) |
1.00 (1.00-1.00, p=0.066) |
0.86 (0.00-174849.06, p=0.999) |