R_Survival Lung
Binh Thang Tran
11/22/2019
****if deces: 1=death; 0 alive)*************
1 Prepare: packages, data
library(foreign, lib.loc = "C:/Program Files/R/R-3.6.1/library")
library(survival)
library(ggplot2)
library(survminer)## Loading required package: ggpubr## Loading required package: magrittr2 read data and group for variable
2.0.1 Read Data
data=read.csv("C:/Users/BINH THANG-TRAN/Dropbox/PhD/Data/BS TRi Ca/Data/data.csv")
head(data)## ID dossier chirurgien age sexe tabagisme Apparition ASIA_preop Frankel_preop
## 1 1 928667 ZW 55 1 1 2 D D
## 2 2 5473345 ZW 71 1 0 2 E E
## 3 3 5466545 GB 68 2 1 2 D D
## 4 4 553346 DS 71 1 1 1 D D
## 5 5 734329 DS 63 1 1 1 E E
## 6 6 390270 GB 65 2 1 1 D D
## ambulation_preop fctSphincter approcheChx dateChx ambulationPO
## 1 2 2 1 10/6/2018 2
## 2 1 2 1 1/7/2018 1
## 3 2 1 1 1/3/2018 2
## 4 2 2 1 16/12/2017 2
## 5 1 2 1 16/12/2017 1
## 6 2 2 2 28/07/2016 1
## amelioration_ambulance ASIA_PO FrankelPO duree_sejour ATCD_RoRx RoRx_PO
## 1 0 D D 13 0 1
## 2 0 E E 4 1 0
## 3 0 D D 12 0 1
## 4 0 D D 25 1 1
## 5 0 E E 5 1 1
## 6 1 E D 16 1 0
## TypeSurgery chimiotxPO Tokuhashi Tokuhashi_cat SINS SINS_cat
## 1 CO 1 5 1 12 2
## 2 CO 0 9 2 8 2
## 3 CO 1 6 1 10 2
## 4 LA 0 6 1 6 1
## 5 CO 0 6 1 11 2
## 6 CO 1 5 1 10 2
## ImprovementPainPO deces date_deces SurviePosOpMOIS LastFollowUp BloodLoss
## 1 1 0 31/10/2018 4.77 31/10/2018 1500
## 2 1 0 31/10/2018 4.07 15/08/2018 500
## 3 1 0 31/10/2018 8.13 1/4/2018 300
## 4 1 1 12/2/2018 1.93 9/2/2018 200
## 5 1 1 1/2/2018 1.57 30/01/2018 2000
## 6 1 1 12/8/2017 12.67 18/07/2017 200
## DurationSurgery Histology
## 1 299 1
## 2 124 1
## 3 108 1
## 4 110 1
## 5 126 1
## 6 120 1names(data) #name of variables## [1] "ID" "dossier" "chirurgien"
## [4] "age" "sexe" "tabagisme"
## [7] "Apparition" "ASIA_preop" "Frankel_preop"
## [10] "ambulation_preop" "fctSphincter" "approcheChx"
## [13] "dateChx" "ambulationPO" "amelioration_ambulance"
## [16] "ASIA_PO" "FrankelPO" "duree_sejour"
## [19] "ATCD_RoRx" "RoRx_PO" "TypeSurgery"
## [22] "chimiotxPO" "Tokuhashi" "Tokuhashi_cat"
## [25] "SINS" "SINS_cat" "ImprovementPainPO"
## [28] "deces" "date_deces" "SurviePosOpMOIS"
## [31] "LastFollowUp" "BloodLoss" "DurationSurgery"
## [34] "Histology"2.1 explaination
###subset dataset -Independent vars:
data1 <- subset(data, select=c(ID, sexe, age, tabagisme,Histology, ASIA_preop, ASIA_PO, Tokuhashi_cat,ambulation_preop, ambulationPO, RoRx_PO,chimiotxPO, SurviePosOpMOIS, deces))Independent variables
data1$age_gr <- ifelse(data1$age >= 60,c("0"), c("1")) #Age: 60+; <60; (age) -need to group
data1$sexe = factor(data1$sexe) # Sex: male - female (sexe)
data1$tabagisme = factor(data1$tabagisme) #Tobacco use: Yes - No (tabagisme)
data1$Histology = factor(data1$Histology) #Histologic type: ADC; Non-ADC; SCLC (Histology)
data1$ASIA_preop = factor(data1$ASIA_preop) #Pre. ASIA score: A, C, D, E (ASIA_preop)
data1$ASIA_PO = factor(data1$ASIA_PO) #Post. ASIA score: C, D, E (ASIA_PO)
data1$Tokuhashi_cat = factor(data1$Tokuhashi_cat) #Revised Tokuhashi score: 0-8; 9-11 (Tokuhashi_cat)
data1$ambulation_preop = factor(data1$ambulation_preop) #Pre. ambulatory status: No, With help; Independent (ambulation_preop)
data1$ambulationPO = factor(data1$ambulationPO) #Post. ambulatory status: No, With help; Independent (ambulationPO)
data1$RoRx_PO = factor(data1$RoRx_PO) #Post. Radiotherapy: No- Yes (RoRx_PO)
data1$chimiotxPO = factor(data1$chimiotxPO) #Post. Chemotherapy: No- Yes (chimiotxPO)####View new dataset again
head(data1)## ID sexe age tabagisme Histology ASIA_preop ASIA_PO Tokuhashi_cat
## 1 1 1 55 1 1 D D 1
## 2 2 1 71 0 1 E E 2
## 3 3 2 68 1 1 D D 1
## 4 4 1 71 1 1 D D 1
## 5 5 1 63 1 1 E E 1
## 6 6 2 65 1 1 D E 1
## ambulation_preop ambulationPO RoRx_PO chimiotxPO SurviePosOpMOIS deces age_gr
## 1 2 2 1 1 4.77 0 1
## 2 1 1 0 0 4.07 0 0
## 3 2 2 1 1 8.13 0 0
## 4 2 2 1 0 1.93 1 0
## 5 1 1 1 0 1.57 1 0
## 6 2 1 0 1 12.67 1 0View(data1)3 Survival Outcome:
SurviePosOpMOIS: time-to-event deces: 0: censorted (alive); 1 event (die)
baseline = Surv(data1$SurviePosOpMOIS, data1$deces==1)
km = survfit(baseline ~ 1)
summary(km)## Call: survfit(formula = baseline ~ 1)
##
## time n.risk n.event survival std.err lower 95% CI upper 95% CI
## 0.33 87 1 0.9885 0.0114 0.96636 1.000
## 0.43 86 1 0.9770 0.0161 0.94602 1.000
## 0.83 85 1 0.9655 0.0196 0.92793 1.000
## 0.87 84 1 0.9540 0.0225 0.91101 0.999
## 1.00 83 1 0.9425 0.0250 0.89487 0.993
## 1.07 82 1 0.9310 0.0272 0.87928 0.986
## 1.10 81 1 0.9195 0.0292 0.86412 0.979
## 1.13 80 1 0.9080 0.0310 0.84931 0.971
## 1.17 79 1 0.8966 0.0327 0.83479 0.963
## 1.37 78 1 0.8851 0.0342 0.82051 0.955
## 1.57 77 2 0.8621 0.0370 0.79257 0.938
## 1.60 75 3 0.8276 0.0405 0.75190 0.911
## 1.63 72 1 0.8161 0.0415 0.73861 0.902
## 1.67 71 1 0.8046 0.0425 0.72545 0.892
## 1.70 70 1 0.7931 0.0434 0.71239 0.883
## 1.77 69 1 0.7816 0.0443 0.69944 0.873
## 1.83 68 1 0.7701 0.0451 0.68659 0.864
## 1.93 67 1 0.7586 0.0459 0.67383 0.854
## 1.97 66 1 0.7471 0.0466 0.66115 0.844
## 2.17 65 3 0.7126 0.0485 0.62362 0.814
## 2.30 62 1 0.7011 0.0491 0.61127 0.804
## 2.33 61 1 0.6897 0.0496 0.59898 0.794
## 2.37 60 1 0.6782 0.0501 0.58677 0.784
## 2.40 59 1 0.6667 0.0505 0.57462 0.773
## 2.67 58 1 0.6552 0.0510 0.56254 0.763
## 2.70 57 1 0.6437 0.0513 0.55052 0.753
## 2.77 56 1 0.6322 0.0517 0.53856 0.742
## 2.90 55 2 0.6092 0.0523 0.51483 0.721
## 3.37 53 2 0.5862 0.0528 0.49134 0.699
## 3.63 51 1 0.5747 0.0530 0.47968 0.689
## 3.70 50 1 0.5632 0.0532 0.46807 0.678
## 3.80 49 1 0.5517 0.0533 0.45652 0.667
## 3.83 48 2 0.5287 0.0535 0.43359 0.645
## 3.87 46 1 0.5172 0.0536 0.42221 0.634
## 3.97 45 1 0.5057 0.0536 0.41088 0.623
## 4.10 43 1 0.4940 0.0536 0.39930 0.611
## 4.13 42 1 0.4822 0.0536 0.38778 0.600
## 4.17 41 1 0.4705 0.0536 0.37632 0.588
## 4.47 40 1 0.4587 0.0535 0.36492 0.577
## 4.57 39 1 0.4469 0.0534 0.35358 0.565
## 4.67 38 1 0.4352 0.0533 0.34230 0.553
## 4.73 37 1 0.4234 0.0531 0.33107 0.542
## 4.90 35 1 0.4113 0.0530 0.31954 0.529
## 5.10 34 1 0.3992 0.0528 0.30807 0.517
## 5.43 33 2 0.3750 0.0523 0.28535 0.493
## 5.67 31 1 0.3629 0.0520 0.27409 0.481
## 5.70 30 2 0.3387 0.0513 0.25180 0.456
## 7.00 28 1 0.3266 0.0508 0.24076 0.443
## 7.40 27 1 0.3145 0.0504 0.22981 0.431
## 7.60 26 1 0.3024 0.0499 0.21893 0.418
## 7.67 25 1 0.2903 0.0493 0.20813 0.405
## 7.77 24 1 0.2782 0.0487 0.19741 0.392
## 7.93 23 1 0.2661 0.0481 0.18679 0.379
## 8.30 21 1 0.2535 0.0474 0.17565 0.366
## 9.03 20 1 0.2408 0.0467 0.16462 0.352
## 10.03 19 2 0.2155 0.0451 0.14293 0.325
## 10.27 17 1 0.2028 0.0442 0.13227 0.311
## 10.50 16 1 0.1901 0.0432 0.12175 0.297
## 11.17 15 1 0.1774 0.0422 0.11138 0.283
## 12.20 14 1 0.1648 0.0410 0.10116 0.268
## 12.67 13 1 0.1521 0.0398 0.09111 0.254
## 12.90 12 1 0.1394 0.0384 0.08124 0.239
## 13.07 11 1 0.1267 0.0370 0.07157 0.224
## 13.10 10 1 0.1141 0.0354 0.06212 0.209
## 14.87 9 1 0.1014 0.0336 0.05292 0.194
## 19.60 8 1 0.0887 0.0317 0.04402 0.179
## 20.80 7 1 0.0760 0.0296 0.03544 0.163
## 22.00 6 1 0.0634 0.0273 0.02727 0.147
## 29.27 4 1 0.0475 0.0246 0.01722 0.131
## 32.00 3 2 0.0158 0.0153 0.00238 0.105
## 97.63 1 1 0.0000 NaN NA NA4 Kaplan – Meier plot:in black and white
plot(km, xlab="Time to death", ylab="Prob of survival")
4.1 Kaplan – Meier plot: plot in collor
ggsurvplot( km, # survfit object with calculated statistics.
data = data1, # data used to fit survival curves.
risk.table = TRUE, # show risk table.
pval = TRUE, # show p-value of log-rank test.
conf.int = TRUE, # show confidence intervals for
# point estimates of survival curves.
xlim = c(0,100), # present narrower X axis, but not affect
# survival estimates.
xlab = "Time in month", # customize X axis label.
break.time.by = 6, # break X axis in time intervals by 500.
ggtheme = theme_light(), # customize plot and risk table with a theme.
risk.table.y.text.col = T, # colour risk table text annotations.
risk.table.y.text = FALSE # show bars instead of names in text annotations
# in legend of risk table
)## Warning in .pvalue(fit, data = data, method = method, pval = pval, pval.coord = pval.coord, : There are no survival curves to be compared.
## This is a null model.
#survival by gender
baseline = Surv(data1$SurviePosOpMOIS, data1$deces==1)
km1 = survfit(Surv(data1$SurviePosOpMOIS, data1$deces==1) ~ data1$sexe)
summary(km1)## Call: survfit(formula = Surv(data1$SurviePosOpMOIS, data1$deces ==
## 1) ~ data1$sexe)
##
## data1$sexe=1
## time n.risk n.event survival std.err lower 95% CI upper 95% CI
## 0.33 45 1 0.9778 0.0220 0.9356 1.000
## 0.43 44 1 0.9556 0.0307 0.8972 1.000
## 1.00 43 1 0.9333 0.0372 0.8632 1.000
## 1.13 42 1 0.9111 0.0424 0.8316 0.998
## 1.17 41 1 0.8889 0.0468 0.8017 0.986
## 1.57 40 2 0.8444 0.0540 0.7449 0.957
## 1.60 38 2 0.8000 0.0596 0.6913 0.926
## 1.77 36 1 0.7778 0.0620 0.6653 0.909
## 1.83 35 1 0.7556 0.0641 0.6399 0.892
## 1.93 34 1 0.7333 0.0659 0.6149 0.875
## 2.17 33 2 0.6889 0.0690 0.5661 0.838
## 2.37 31 1 0.6667 0.0703 0.5422 0.820
## 2.40 30 1 0.6444 0.0714 0.5187 0.801
## 2.70 29 1 0.6222 0.0723 0.4955 0.781
## 2.77 28 1 0.6000 0.0730 0.4727 0.762
## 2.90 27 2 0.5556 0.0741 0.4278 0.721
## 3.37 25 2 0.5111 0.0745 0.3841 0.680
## 3.63 23 1 0.4889 0.0745 0.3626 0.659
## 3.80 22 1 0.4667 0.0744 0.3415 0.638
## 3.97 21 1 0.4444 0.0741 0.3206 0.616
## 4.17 19 1 0.4211 0.0738 0.2987 0.594
## 4.67 18 1 0.3977 0.0733 0.2771 0.571
## 5.10 16 1 0.3728 0.0728 0.2542 0.547
## 5.43 15 2 0.3231 0.0711 0.2099 0.497
## 5.70 13 1 0.2982 0.0698 0.1885 0.472
## 7.40 12 1 0.2734 0.0683 0.1676 0.446
## 7.60 11 1 0.2485 0.0664 0.1472 0.420
## 7.67 10 1 0.2237 0.0643 0.1274 0.393
## 7.77 9 1 0.1988 0.0618 0.1082 0.365
## 9.03 8 1 0.1740 0.0588 0.0897 0.338
## 10.03 7 2 0.1243 0.0515 0.0552 0.280
## 10.27 5 1 0.0994 0.0468 0.0395 0.250
## 10.50 4 1 0.0746 0.0412 0.0253 0.220
## 12.20 3 1 0.0497 0.0341 0.0129 0.191
## 12.90 2 1 0.0249 0.0245 0.0036 0.172
## 97.63 1 1 0.0000 NaN NA NA
##
## data1$sexe=2
## time n.risk n.event survival std.err lower 95% CI upper 95% CI
## 0.83 42 1 0.9762 0.0235 0.9312 1.000
## 0.87 41 1 0.9524 0.0329 0.8901 1.000
## 1.07 40 1 0.9286 0.0397 0.8539 1.000
## 1.10 39 1 0.9048 0.0453 0.8202 0.998
## 1.37 38 1 0.8810 0.0500 0.7883 0.985
## 1.60 37 1 0.8571 0.0540 0.7576 0.970
## 1.63 36 1 0.8333 0.0575 0.7279 0.954
## 1.67 35 1 0.8095 0.0606 0.6991 0.937
## 1.70 34 1 0.7857 0.0633 0.6709 0.920
## 1.97 33 1 0.7619 0.0657 0.6434 0.902
## 2.17 32 1 0.7381 0.0678 0.6164 0.884
## 2.30 31 1 0.7143 0.0697 0.5899 0.865
## 2.33 30 1 0.6905 0.0713 0.5639 0.845
## 2.67 29 1 0.6667 0.0727 0.5383 0.826
## 3.70 28 1 0.6429 0.0739 0.5131 0.805
## 3.83 27 2 0.5952 0.0757 0.4639 0.764
## 3.87 25 1 0.5714 0.0764 0.4398 0.743
## 4.10 24 1 0.5476 0.0768 0.4160 0.721
## 4.13 23 1 0.5238 0.0771 0.3926 0.699
## 4.47 22 1 0.5000 0.0772 0.3695 0.677
## 4.57 21 1 0.4762 0.0771 0.3468 0.654
## 4.73 20 1 0.4524 0.0768 0.3243 0.631
## 4.90 19 1 0.4286 0.0764 0.3022 0.608
## 5.67 18 1 0.4048 0.0757 0.2805 0.584
## 5.70 17 1 0.3810 0.0749 0.2591 0.560
## 7.00 16 1 0.3571 0.0739 0.2380 0.536
## 7.93 15 1 0.3333 0.0727 0.2173 0.511
## 8.30 13 1 0.3077 0.0715 0.1951 0.485
## 11.17 12 1 0.2821 0.0700 0.1734 0.459
## 12.67 11 1 0.2564 0.0682 0.1523 0.432
## 13.07 10 1 0.2308 0.0660 0.1317 0.404
## 13.10 9 1 0.2051 0.0635 0.1119 0.376
## 14.87 8 1 0.1795 0.0605 0.0927 0.347
## 19.60 7 1 0.1538 0.0570 0.0744 0.318
## 20.80 6 1 0.1282 0.0530 0.0570 0.288
## 22.00 5 1 0.1026 0.0482 0.0408 0.258
## 29.27 3 1 0.0684 0.0426 0.0202 0.232
## 32.00 2 2 0.0000 NaN NA NA#graph by gender
library(ggplot2)
ggsurvplot(
km1, # survfit object with calculated statistics.
data = data1, # data used to fit survival curves.
risk.table = TRUE, # show risk table.
pval = TRUE, # show p-value of log-rank test.
conf.int = TRUE, # show confidence intervals for
# point estimates of survival curves.
xlim = c(0,100), # present narrower X axis, but not affect
# survival estimates.
xlab = "Time in month", # customize X axis label.
break.time.by = 6, # break X axis in time intervals by 500.
ggtheme = theme_light(), # customize plot and risk table with a theme.
risk.table.y.text.col = T, # colour risk table text annotations.
risk.table.y.text = FALSE # show bars instead of names in text annotations
# in legend of risk table
)
5 table 1: general chracteristics of participants
#install package "moonBook", ztable
library(moonBook)
require(ztable)## Loading required package: ztable## Welcome to package ztable ver 0.2.0require(magrittr)
options(ztable.type="html")#table 1: general chracteristics of participants
mytable(data1)##
## Descriptive Statistics
## -----------------------------------
## N Total
## -----------------------------------
## ID 87 44.0 ± 25.3
## sexe 87
## - 1 45 (51.7%)
## - 2 42 (48.3%)
## age 87 61.3 ± 8.8
## tabagisme 87
## - 0 35 (40.2%)
## - 1 52 (59.8%)
## Histology 87
## - 0 22 (25.3%)
## - 1 58 (66.7%)
## - 2 7 (8.0%)
## ASIA_preop 87
## - A 1 (1.1%)
## - C 5 (5.7%)
## - D 39 (44.8%)
## - E 42 (48.3%)
## ASIA_PO 87
## - C 2 (2.3%)
## - D 27 (31.0%)
## - E 58 (66.7%)
## Tokuhashi_cat 87
## - 1 72 (82.8%)
## - 2 15 (17.2%)
## ambulation_preop 87
## - 0 22 (25.3%)
## - 1 41 (47.1%)
## - 2 24 (27.6%)
## ambulationPO 87
## - 0 4 (4.6%)
## - 1 48 (55.2%)
## - 2 35 (40.2%)
## RoRx_PO 87
## - 0 28 (32.2%)
## - 1 59 (67.8%)
## chimiotxPO 87
## - 0 54 (62.1%)
## - 1 33 (37.9%)
## SurviePosOpMOIS 87 7.5 ± 12.0
## deces 87
## - 0 4 (4.6%)
## - 1 83 (95.4%)
## age_gr 87
## - 0 51 (58.6%)
## - 1 36 (41.4%)
## -----------------------------------6 Univariate cox-model
library(survival)
#model 1: sex
cox1 = coxph(Surv(data1$SurviePosOpMOIS, data1$deces==1) ~ data1$sexe)
summary(cox1)## Call:
## coxph(formula = Surv(data1$SurviePosOpMOIS, data1$deces == 1) ~
## data1$sexe)
##
## n= 87, number of events= 83
##
## coef exp(coef) se(coef) z Pr(>|z|)
## data1$sexe2 -0.3841 0.6810 0.2291 -1.677 0.0936 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## data1$sexe2 0.681 1.468 0.4347 1.067
##
## Concordance= 0.541 (se = 0.032 )
## Likelihood ratio test= 2.82 on 1 df, p=0.09
## Wald test = 2.81 on 1 df, p=0.09
## Score (logrank) test = 2.84 on 1 df, p=0.09#model 2: age group
cox2 = coxph(Surv(data1$SurviePosOpMOIS, data1$deces==1) ~ data1$age_gr)
summary(cox2)## Call:
## coxph(formula = Surv(data1$SurviePosOpMOIS, data1$deces == 1) ~
## data1$age_gr)
##
## n= 87, number of events= 83
##
## coef exp(coef) se(coef) z Pr(>|z|)
## data1$age_gr1 -0.4134 0.6614 0.2358 -1.753 0.0795 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## data1$age_gr1 0.6614 1.512 0.4167 1.05
##
## Concordance= 0.542 (se = 0.031 )
## Likelihood ratio test= 3.16 on 1 df, p=0.08
## Wald test = 3.07 on 1 df, p=0.08
## Score (logrank) test = 3.12 on 1 df, p=0.08#model 3: tobacco use
cox3 = coxph(Surv(data1$SurviePosOpMOIS, data1$deces==1) ~ data1$tabagisme)
summary(cox3)## Call:
## coxph(formula = Surv(data1$SurviePosOpMOIS, data1$deces == 1) ~
## data1$tabagisme)
##
## n= 87, number of events= 83
##
## coef exp(coef) se(coef) z Pr(>|z|)
## data1$tabagisme1 0.6532 1.9216 0.2445 2.672 0.00755 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## data1$tabagisme1 1.922 0.5204 1.19 3.103
##
## Concordance= 0.561 (se = 0.033 )
## Likelihood ratio test= 7.42 on 1 df, p=0.006
## Wald test = 7.14 on 1 df, p=0.008
## Score (logrank) test = 7.34 on 1 df, p=0.007#Model 4: Histology
cox4 = coxph(Surv(data1$SurviePosOpMOIS, data1$deces==1) ~ data1$Histology)
summary(cox4)## Call:
## coxph(formula = Surv(data1$SurviePosOpMOIS, data1$deces == 1) ~
## data1$Histology)
##
## n= 87, number of events= 83
##
## coef exp(coef) se(coef) z Pr(>|z|)
## data1$Histology1 -0.23118 0.79359 0.25730 -0.898 0.369
## data1$Histology2 0.01456 1.01467 0.44202 0.033 0.974
##
## exp(coef) exp(-coef) lower .95 upper .95
## data1$Histology1 0.7936 1.2601 0.4793 1.314
## data1$Histology2 1.0147 0.9855 0.4267 2.413
##
## Concordance= 0.53 (se = 0.029 )
## Likelihood ratio test= 0.99 on 2 df, p=0.6
## Wald test = 1.02 on 2 df, p=0.6
## Score (logrank) test = 1.02 on 2 df, p=0.6#model 5: Pre. ASIA score
cox5 = coxph(Surv(data1$SurviePosOpMOIS, data1$deces==1) ~ data1$ASIA_preop)
summary(cox5)## Call:
## coxph(formula = Surv(data1$SurviePosOpMOIS, data1$deces == 1) ~
## data1$ASIA_preop)
##
## n= 87, number of events= 83
##
## coef exp(coef) se(coef) z Pr(>|z|)
## data1$ASIA_preopC -1.1574 0.3143 1.1153 -1.038 0.2994
## data1$ASIA_preopD -1.6480 0.1924 1.0426 -1.581 0.1140
## data1$ASIA_preopE -2.1236 0.1196 1.0479 -2.027 0.0427 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## data1$ASIA_preopC 0.3143 3.181 0.03532 2.7971
## data1$ASIA_preopD 0.1924 5.196 0.02494 1.4850
## data1$ASIA_preopE 0.1196 8.361 0.01534 0.9326
##
## Concordance= 0.555 (se = 0.033 )
## Likelihood ratio test= 7.62 on 3 df, p=0.05
## Wald test = 8.84 on 3 df, p=0.03
## Score (logrank) test = 10.03 on 3 df, p=0.02#model 6: Post. ASIA score
cox6 = coxph(Surv(data1$SurviePosOpMOIS, data1$deces==1) ~ data1$ASIA_PO)
summary(cox6)## Call:
## coxph(formula = Surv(data1$SurviePosOpMOIS, data1$deces == 1) ~
## data1$ASIA_PO)
##
## n= 87, number of events= 83
##
## coef exp(coef) se(coef) z Pr(>|z|)
## data1$ASIA_POD 0.2683 1.3078 0.7380 0.364 0.716
## data1$ASIA_POE -0.5017 0.6055 0.7267 -0.690 0.490
##
## exp(coef) exp(-coef) lower .95 upper .95
## data1$ASIA_POD 1.3078 0.7647 0.3078 5.556
## data1$ASIA_POE 0.6055 1.6515 0.1457 2.516
##
## Concordance= 0.575 (se = 0.028 )
## Likelihood ratio test= 8.25 on 2 df, p=0.02
## Wald test = 8.79 on 2 df, p=0.01
## Score (logrank) test = 9.2 on 2 df, p=0.01#model 7: Revised Tokuhashi score
cox7 = coxph(Surv(data1$SurviePosOpMOIS, data1$deces==1) ~ data1$Tokuhashi_cat)
summary(cox7)## Call:
## coxph(formula = Surv(data1$SurviePosOpMOIS, data1$deces == 1) ~
## data1$Tokuhashi_cat)
##
## n= 87, number of events= 83
##
## coef exp(coef) se(coef) z Pr(>|z|)
## data1$Tokuhashi_cat2 -1.1675 0.3111 0.3433 -3.401 0.000672 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## data1$Tokuhashi_cat2 0.3111 3.214 0.1588 0.6098
##
## Concordance= 0.556 (se = 0.03 )
## Likelihood ratio test= 14.19 on 1 df, p=2e-04
## Wald test = 11.57 on 1 df, p=7e-04
## Score (logrank) test = 12.29 on 1 df, p=5e-04#model 8: Pre. ambulatory status
cox8 = coxph(Surv(data1$SurviePosOpMOIS, data1$deces==1) ~ data1$ambulation_preop)
summary(cox8)## Call:
## coxph(formula = Surv(data1$SurviePosOpMOIS, data1$deces == 1) ~
## data1$ambulation_preop)
##
## n= 87, number of events= 83
##
## coef exp(coef) se(coef) z Pr(>|z|)
## data1$ambulation_preop1 -0.6800 0.5066 0.2845 -2.391 0.0168 *
## data1$ambulation_preop2 -0.4772 0.6205 0.3090 -1.544 0.1226
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## data1$ambulation_preop1 0.5066 1.974 0.2901 0.8847
## data1$ambulation_preop2 0.6205 1.612 0.3386 1.1371
##
## Concordance= 0.544 (se = 0.033 )
## Likelihood ratio test= 5.43 on 2 df, p=0.07
## Wald test = 5.77 on 2 df, p=0.06
## Score (logrank) test = 5.94 on 2 df, p=0.05#model 9: Post. ambulatory status
cox9 = coxph(Surv(data1$SurviePosOpMOIS, data1$deces==1) ~ data1$ambulationPO)
summary(cox9)## Call:
## coxph(formula = Surv(data1$SurviePosOpMOIS, data1$deces == 1) ~
## data1$ambulationPO)
##
## n= 87, number of events= 83
##
## coef exp(coef) se(coef) z Pr(>|z|)
## data1$ambulationPO1 -2.7953 0.0611 0.5922 -4.720 2.35e-06 ***
## data1$ambulationPO2 -1.9756 0.1387 0.5810 -3.401 0.000673 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## data1$ambulationPO1 0.0611 16.368 0.01914 0.195
## data1$ambulationPO2 0.1387 7.211 0.04441 0.433
##
## Concordance= 0.626 (se = 0.03 )
## Likelihood ratio test= 21.45 on 2 df, p=2e-05
## Wald test = 27.03 on 2 df, p=1e-06
## Score (logrank) test = 37.8 on 2 df, p=6e-09#model 10: Post. Radiotherapy
cox10 = coxph(Surv(data1$SurviePosOpMOIS, data1$deces==1) ~ data1$RoRx_PO)
summary(cox10)## Call:
## coxph(formula = Surv(data1$SurviePosOpMOIS, data1$deces == 1) ~
## data1$RoRx_PO)
##
## n= 87, number of events= 83
##
## coef exp(coef) se(coef) z Pr(>|z|)
## data1$RoRx_PO1 -0.6645 0.5145 0.2474 -2.686 0.00724 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## data1$RoRx_PO1 0.5145 1.944 0.3168 0.8356
##
## Concordance= 0.565 (se = 0.028 )
## Likelihood ratio test= 6.76 on 1 df, p=0.009
## Wald test = 7.21 on 1 df, p=0.007
## Score (logrank) test = 7.47 on 1 df, p=0.006#model 11: Post. Chemotherapy
cox11 = coxph(Surv(data1$SurviePosOpMOIS, data1$deces==1) ~ data1$chimiotxPO)
summary(cox11)## Call:
## coxph(formula = Surv(data1$SurviePosOpMOIS, data1$deces == 1) ~
## data1$chimiotxPO)
##
## n= 87, number of events= 83
##
## coef exp(coef) se(coef) z Pr(>|z|)
## data1$chimiotxPO1 -1.1749 0.3088 0.2477 -4.744 2.1e-06 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## data1$chimiotxPO1 0.3088 3.238 0.1901 0.5018
##
## Concordance= 0.663 (se = 0.022 )
## Likelihood ratio test= 24.24 on 1 df, p=9e-07
## Wald test = 22.51 on 1 df, p=2e-06
## Score (logrank) test = 24.67 on 1 df, p=7e-077 multivariate model
#model 12: multivariate model
cox12 = coxph(Surv(data1$SurviePosOpMOIS, data1$deces==1) ~ data1$RoRx_PO +data1$tabagisme+ data1$ASIA_preop +data1$Tokuhashi_cat+ data1$ambulation_preop+ data1$ambulationPO+data1$RoRx_PO+data1$chimiotxPO)
summary(cox12)## Call:
## coxph(formula = Surv(data1$SurviePosOpMOIS, data1$deces == 1) ~
## data1$RoRx_PO + data1$tabagisme + data1$ASIA_preop + data1$Tokuhashi_cat +
## data1$ambulation_preop + data1$ambulationPO + data1$RoRx_PO +
## data1$chimiotxPO)
##
## n= 87, number of events= 83
##
## coef exp(coef) se(coef) z Pr(>|z|)
## data1$RoRx_PO1 -0.56985 0.56561 0.26740 -2.131 0.03308 *
## data1$tabagisme1 0.76691 2.15311 0.29125 2.633 0.00846 **
## data1$ASIA_preopC -2.32389 0.09789 1.19271 -1.948 0.05136 .
## data1$ASIA_preopD -1.45476 0.23346 1.10891 -1.312 0.18956
## data1$ASIA_preopE -1.07091 0.34270 1.13022 -0.948 0.34337
## data1$Tokuhashi_cat2 -1.05935 0.34668 0.40871 -2.592 0.00954 **
## data1$ambulation_preop1 0.13339 1.14270 0.43212 0.309 0.75755
## data1$ambulation_preop2 -0.62944 0.53289 0.33784 -1.863 0.06244 .
## data1$ambulationPO1 -3.07851 0.04603 0.77607 -3.967 7.28e-05 ***
## data1$ambulationPO2 -2.32716 0.09757 0.74322 -3.131 0.00174 **
## data1$chimiotxPO1 -1.35077 0.25904 0.29846 -4.526 6.02e-06 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## data1$RoRx_PO1 0.56561 1.7680 0.334895 0.9553
## data1$tabagisme1 2.15311 0.4644 1.216614 3.8105
## data1$ASIA_preopC 0.09789 10.2153 0.009452 1.0139
## data1$ASIA_preopD 0.23346 4.2834 0.026565 2.0517
## data1$ASIA_preopE 0.34270 2.9180 0.037399 3.1402
## data1$Tokuhashi_cat2 0.34668 2.8845 0.155609 0.7724
## data1$ambulation_preop1 1.14270 0.8751 0.489905 2.6653
## data1$ambulation_preop2 0.53289 1.8766 0.274831 1.0333
## data1$ambulationPO1 0.04603 21.7260 0.010056 0.2107
## data1$ambulationPO2 0.09757 10.2488 0.022735 0.4187
## data1$chimiotxPO1 0.25904 3.8604 0.144319 0.4650
##
## Concordance= 0.752 (se = 0.027 )
## Likelihood ratio test= 69.22 on 11 df, p=2e-10
## Wald test = 64.1 on 11 df, p=2e-09
## Score (logrank) test = 84.87 on 11 df, p=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