SVM Survival 072524
Karl Brown
2024-07-25
Table 1
|
BA (n=2341)
|
WA (n=1697)
|
|||
|---|---|---|---|---|
| Mean | SD | Mean | SD | |
| Age | 68.2 | 3.5 | 69.7 | 4.2 |
| % male | 94.3 |
|
95.9 |
|
| SVM risk score | -0.56 | 0.41 | -0.49 | 0.38 |
| % dementia incidence | 4.9 |
|
2.9 |
|
| % censored and alive | 78.3 |
|
77.2 |
|
| % censored and dead | 16.8 |
|
20.0 |
|
6-month outcome summaries
## # A tibble: 2 × 2
## Race overall
## <chr> <int>
## 1 AA 1293
## 2 EA 868## # A tibble: 6 × 5
## # Groups: Race [2]
## Race event_status2 total overall proportion
## <chr> <fct> <int> <int> <dbl>
## 1 AA censored 1035 1293 80.0
## 2 AA dementia 64 1293 4.95
## 3 AA death 194 1293 15.0
## 4 EA censored 672 868 77.4
## 5 EA dementia 28 868 3.23
## 6 EA death 168 868 19.4
1-year outcome summaries
## # A tibble: 2 × 2
## Race overall
## <chr> <int>
## 1 AA 2341
## 2 EA 1697## # A tibble: 6 × 5
## # Groups: Race [2]
## Race event_status2 total overall proportion
## <chr> <fct> <int> <int> <dbl>
## 1 AA censored 1834 2341 78.3
## 2 AA dementia 114 2341 4.87
## 3 AA death 393 2341 16.8
## 4 EA censored 1310 1697 77.2
## 5 EA dementia 50 1697 2.95
## 6 EA death 337 1697 19.9
Competing risks analysis (score quintiles)
6-month data
## Call:
## coxph(formula = Surv(time_to_event, event_status2) ~ score_quintile *
## Race, data = surv_6m, id = X)
##
## n= 2161, number of events= 454
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## score_quintile_1:2 0.52417 1.68906 0.09696 0.10840 4.836 1.33e-06
## RaceEA_1:2 -1.77283 0.16985 0.95355 1.00515 -1.764 0.0778
## score_quintile:RaceEA_1:2 0.28569 1.33068 0.22116 0.23434 1.219 0.2228
## score_quintile_1:3 0.30726 1.35969 0.05142 0.05494 5.592 2.24e-08
## RaceEA_1:3 0.06364 1.06571 0.30891 0.31119 0.205 0.8380
## score_quintile:RaceEA_1:3 0.02386 1.02414 0.08110 0.08215 0.290 0.7715
##
## score_quintile_1:2 ***
## RaceEA_1:2 .
## score_quintile:RaceEA_1:2
## score_quintile_1:3 ***
## RaceEA_1:3
## score_quintile:RaceEA_1:3
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## score_quintile_1:2 1.6891 0.5920 1.36576 2.089
## RaceEA_1:2 0.1699 5.8875 0.02369 1.218
## score_quintile:RaceEA_1:2 1.3307 0.7515 0.84063 2.106
## score_quintile_1:3 1.3597 0.7355 1.22088 1.514
## RaceEA_1:3 1.0657 0.9383 0.57910 1.961
## score_quintile:RaceEA_1:3 1.0241 0.9764 0.87183 1.203
##
## Concordance= 0.647 (se = 0.013 )
## Likelihood ratio test= 132.1 on 6 df, p=<2e-16
## Wald test = 108.1 on 6 df, p=<2e-16
## Score (logrank) test = 129.8 on 6 df, p=<2e-16, Robust = 110.1 p=<2e-16
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).
1-year data
## Call:
## coxph(formula = Surv(time_to_event, event_status2) ~ score_quintile *
## Race, data = surv_1y, id = X)
##
## n= 4038, number of events= 894
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## score_quintile_1:2 0.49666 1.64322 0.07210 0.07938 6.257 3.94e-10
## RaceEA_1:2 -1.83683 0.15932 0.68566 0.78168 -2.350 0.0188
## score_quintile:RaceEA_1:2 0.29844 1.34775 0.16101 0.18448 1.618 0.1057
## score_quintile_1:3 0.32041 1.37769 0.03646 0.03906 8.203 2.34e-16
## RaceEA_1:3 -0.22086 0.80183 0.22335 0.23021 -0.959 0.3374
## score_quintile:RaceEA_1:3 0.08173 1.08516 0.05806 0.06022 1.357 0.1747
##
## score_quintile_1:2 ***
## RaceEA_1:2 *
## score_quintile:RaceEA_1:2
## score_quintile_1:3 ***
## RaceEA_1:3
## score_quintile:RaceEA_1:3
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## score_quintile_1:2 1.6432 0.6086 1.40645 1.9198
## RaceEA_1:2 0.1593 6.2766 0.03443 0.7373
## score_quintile:RaceEA_1:2 1.3477 0.7420 0.93881 1.9348
## score_quintile_1:3 1.3777 0.7259 1.27615 1.4873
## RaceEA_1:3 0.8018 1.2472 0.51065 1.2590
## score_quintile:RaceEA_1:3 1.0852 0.9215 0.96435 1.2211
##
## Concordance= 0.653 (se = 0.009 )
## Likelihood ratio test= 274.9 on 6 df, p=<2e-16
## Wald test = 219.8 on 6 df, p=<2e-16
## Score (logrank) test = 270.6 on 6 df, p=<2e-16, Robust = 222.3 p=<2e-16
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).
Competing risks with continuous scores
6-month data
## Call:
## coxph(formula = Surv(time_to_event, event_status2) ~ Score_std *
## Race, data = surv_6m, id = X)
##
## n= 2161, number of events= 454
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## Score_std_1:2 0.66678 1.94796 0.09356 0.09231 7.223 5.07e-13 ***
## RaceEA_1:2 -0.62379 0.53591 0.27617 0.26575 -2.347 0.0189 *
## Score_std:RaceEA_1:2 0.07009 1.07260 0.16797 0.15175 0.462 0.6442
## Score_std_1:3 0.40243 1.49545 0.05967 0.05857 6.871 6.39e-12 ***
## RaceEA_1:3 0.19357 1.21358 0.11168 0.11035 1.754 0.0794 .
## Score_std:RaceEA_1:3 -0.03477 0.96583 0.09084 0.08659 -0.402 0.6880
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## Score_std_1:2 1.9480 0.5134 1.6256 2.3343
## RaceEA_1:2 0.5359 1.8660 0.3183 0.9022
## Score_std:RaceEA_1:2 1.0726 0.9323 0.7967 1.4441
## Score_std_1:3 1.4954 0.6687 1.3333 1.6774
## RaceEA_1:3 1.2136 0.8240 0.9775 1.5066
## Score_std:RaceEA_1:3 0.9658 1.0354 0.8151 1.1445
##
## Concordance= 0.649 (se = 0.013 )
## Likelihood ratio test= 138.8 on 6 df, p=<2e-16
## Wald test = 170.4 on 6 df, p=<2e-16
## Score (logrank) test = 169.9 on 6 df, p=<2e-16, Robust = 98.92 p=<2e-16
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).
1-year data
## Call:
## coxph(formula = Surv(time_to_event, event_status2) ~ Score_std *
## Race, data = surv_1y, id = X)
##
## n= 4038, number of events= 894
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## Score_std_1:2 0.66217 1.93900 0.06928 0.06822 9.707 < 2e-16 ***
## RaceEA_1:2 -0.71982 0.48684 0.20784 0.20058 -3.589 0.000332 ***
## Score_std:RaceEA_1:2 0.14004 1.15032 0.12602 0.11751 1.192 0.233372
## Score_std_1:3 0.40929 1.50575 0.04160 0.04093 9.999 < 2e-16 ***
## RaceEA_1:3 0.08297 1.08651 0.07944 0.07859 1.056 0.291063
## Score_std:RaceEA_1:3 0.04886 1.05007 0.06293 0.06163 0.793 0.427897
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## Score_std_1:2 1.9390 0.5157 1.6963 2.2164
## RaceEA_1:2 0.4868 2.0541 0.3286 0.7213
## Score_std:RaceEA_1:2 1.1503 0.8693 0.9137 1.4483
## Score_std_1:3 1.5057 0.6641 1.3897 1.6315
## RaceEA_1:3 1.0865 0.9204 0.9314 1.2675
## Score_std:RaceEA_1:3 1.0501 0.9523 0.9306 1.1849
##
## Concordance= 0.655 (se = 0.009 )
## Likelihood ratio test= 298.7 on 6 df, p=<2e-16
## Wald test = 355.5 on 6 df, p=<2e-16
## Score (logrank) test = 372.5 on 6 df, p=<2e-16, Robust = 203.3 p=<2e-16
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).
Sensitivity analyses: handling death
We can treat death as a censoring event (middle panel), but as extremes we can also set the censoring time of those who die to be the maximum censoring date (left panel) or we can treat those that die as having developed dementia at the date of death (right panel)
Summarizing death-sensitivity results
|
BA group
|
WA group
|
||||||
|---|---|---|---|---|---|---|---|
| Analysis | Data | HR estimate | Lower CI | Upper CI | HR estimate | Lower CI | Upper CI |
| Dementia only | 6m | 1.948 | 1.622 | 2.340 | 2.089 | 1.589 | 2.748 |
| Dementia only | 1y | 1.922 | 1.677 | 2.204 | 2.229 | 1.813 | 2.741 |
| Dementia/death | 6m | 1.554 | 1.399 | 1.726 | 1.485 | 1.307 | 1.688 |
| Dementia/death | 1y | 1.566 | 1.453 | 1.687 | 1.625 | 1.487 | 1.775 |
| Deaths as max | 6m | 1.777 | 1.496 | 2.111 | 1.979 | 1.501 | 2.609 |
| Deaths as max | 1y | 1.757 | 1.544 | 2.000 | 2.009 | 1.643 | 2.456 |
Individual results for each model
6-month data
## Call:
## coxph(formula = Surv(time_to_event, event_status) ~ Score_std *
## Race, data = surv_6m_demonly, id = X)
##
## n= 2161, number of events= 92
##
## coef exp(coef) se(coef) z Pr(>|z|)
## Score_std 0.66678 1.94796 0.09356 7.127 1.03e-12 ***
## RaceEA -0.62379 0.53591 0.27617 -2.259 0.0239 *
## Score_std:RaceEA 0.07009 1.07260 0.16797 0.417 0.6765
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## Score_std 1.9480 0.5134 1.6216 2.3400
## RaceEA 0.5359 1.8660 0.3119 0.9208
## Score_std:RaceEA 1.0726 0.9323 0.7717 1.4908
##
## Concordance= 0.729 (se = 0.027 )
## Likelihood ratio test= 67.73 on 3 df, p=1e-14
## Wald test = 80.84 on 3 df, p=<2e-16
## Score (logrank) test = 88.98 on 3 df, p=<2e-16
1-year data
## Call:
## coxph(formula = Surv(time_to_event2, event_status2) ~ Score_std *
## Race, data = surv_1y_demonly, id = X)
##
## n= 4038, number of events= 163
##
## coef exp(coef) se(coef) z Pr(>|z|)
## Score_std 0.64844 1.91255 0.06955 9.324 < 2e-16 ***
## RaceEA -0.68238 0.50541 0.20757 -3.287 0.00101 **
## Score_std:RaceEA 0.14040 1.15074 0.12580 1.116 0.26440
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## Score_std 1.9125 0.5229 1.6688 2.1919
## RaceEA 0.5054 1.9786 0.3365 0.7592
## Score_std:RaceEA 1.1507 0.8690 0.8993 1.4725
##
## Concordance= 0.735 (se = 0.02 )
## Likelihood ratio test= 124.1 on 3 df, p=<2e-16
## Wald test = 147.1 on 3 df, p=<2e-16
## Score (logrank) test = 163.1 on 3 df, p=<2e-16
Death and dementia as the same outcome analysis
6-month data
## Call:
## coxph(formula = Surv(time_to_event2, event_status2) ~ Score_std *
## Race, data = surv_6m_both, id = X)
##
## n= 2161, number of events= 408
##
## coef exp(coef) se(coef) z Pr(>|z|)
## Score_std 0.44077 1.55390 0.05354 8.232 <2e-16 ***
## RaceEA 0.10480 1.11048 0.10712 0.978 0.328
## Score_std:RaceEA -0.04503 0.95597 0.08429 -0.534 0.593
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## Score_std 1.554 0.6435 1.3991 1.726
## RaceEA 1.110 0.9005 0.9002 1.370
## Score_std:RaceEA 0.956 1.0461 0.8104 1.128
##
## Concordance= 0.638 (se = 0.014 )
## Likelihood ratio test= 93.63 on 3 df, p=<2e-16
## Wald test = 106.6 on 3 df, p=<2e-16
## Score (logrank) test = 108.8 on 3 df, p=<2e-16
1-year data
## Call:
## coxph(formula = Surv(time_to_event2, event_status2) ~ Score_std *
## Race, data = surv_1y_both, id = X)
##
## n= 4038, number of events= 803
##
## coef exp(coef) se(coef) z Pr(>|z|)
## Score_std 0.44834 1.56571 0.03812 11.760 <2e-16 ***
## RaceEA 0.02823 1.02863 0.07686 0.367 0.713
## Score_std:RaceEA 0.03688 1.03757 0.05897 0.625 0.532
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## Score_std 1.566 0.6387 1.4530 1.687
## RaceEA 1.029 0.9722 0.8848 1.196
## Score_std:RaceEA 1.038 0.9638 0.9243 1.165
##
## Concordance= 0.647 (se = 0.01 )
## Likelihood ratio test= 218.7 on 3 df, p=<2e-16
## Wald test = 256.8 on 3 df, p=<2e-16
## Score (logrank) test = 263 on 3 df, p=<2e-16
Setting all those who died to the maximum censoring time
6-month data
## Call:
## coxph(formula = Surv(time_to_event, event_status) ~ Score_std *
## Race, data = surv_6m_deathmax, id = X)
##
## n= 2161, number of events= 92
##
## coef exp(coef) se(coef) z Pr(>|z|)
## Score_std 0.57480 1.77678 0.08791 6.538 6.22e-11 ***
## RaceEA -0.63575 0.52954 0.27595 -2.304 0.0212 *
## Score_std:RaceEA 0.10786 1.11389 0.16614 0.649 0.5162
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## Score_std 1.7768 0.5628 1.4956 2.1109
## RaceEA 0.5295 1.8884 0.3083 0.9095
## Score_std:RaceEA 1.1139 0.8978 0.8043 1.5426
##
## Concordance= 0.717 (se = 0.027 )
## Likelihood ratio test= 56.96 on 3 df, p=3e-12
## Wald test = 68.62 on 3 df, p=8e-15
## Score (logrank) test = 73.65 on 3 df, p=7e-16
1-year data
## Call:
## coxph(formula = Surv(time_to_event, event_status) ~ Score_std *
## Race, data = surv_1y_deathmax, id = X)
##
## n= 4038, number of events= 164
##
## coef exp(coef) se(coef) z Pr(>|z|)
## Score_std 0.56362 1.75703 0.06597 8.544 < 2e-16 ***
## RaceEA -0.69613 0.49851 0.20578 -3.383 0.000717 ***
## Score_std:RaceEA 0.13383 1.14320 0.12197 1.097 0.272553
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## Score_std 1.7570 0.5691 1.5439 1.9995
## RaceEA 0.4985 2.0060 0.3331 0.7462
## Score_std:RaceEA 1.1432 0.8747 0.9001 1.4519
##
## Concordance= 0.72 (se = 0.02 )
## Likelihood ratio test= 105 on 3 df, p=<2e-16
## Wald test = 124.5 on 3 df, p=<2e-16
## Score (logrank) test = 135.6 on 3 df, p=<2e-16
Grace period sensitivity analysis
Grace period method 1 involves subtracting the immortal time (data creation date (9/12/18) - index date) from each time to event
9 people excluded from 6 month analysis (all died)
57 people excluded from 1 year analysis (all died)
Grace period method 2 invloves subtracting one year from every time to event
61 people excluded from 6 month analysis (4 developed dementia, 57 died
died)
119 people excluded from 1 year analysis (4 developed dementia, 115
died)
|
Dementia incidence
|
Death
|
||||||
|---|---|---|---|---|---|---|---|
| Analysis | Data | HR estimate | Lower CI | Upper CI | HR estimate | Lower CI | Upper CI |
| Primary | 6m | 1.948 | 1.626 | 2.334 | 1.495 | 1.333 | 1.677 |
| Primary | 1y | 1.939 | 1.696 | 2.216 | 1.506 | 1.390 | 1.632 |
| Grace method 1 | 6m | 1.949 | 1.626 | 2.336 | 1.494 | 1.327 | 1.682 |
| Grace method 1 | 1y | 1.952 | 1.705 | 2.234 | 1.502 | 1.378 | 1.637 |
| Grace method 2 | 6m | 1.960 | 1.620 | 2.373 | 1.503 | 1.316 | 1.716 |
| Grace method 2 | 1y | 1.947 | 1.697 | 2.234 | 1.506 | 1.372 | 1.653 |
Individual model results
Grace method 1, 6-month data
## Call:
## coxph(formula = Surv(time_to_event_grace, event_status2) ~ Score_std *
## Race, data = surv_6m_grace_v1, id = X)
##
## n= 2152, number of events= 445
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## Score_std_1:2 0.66715 1.94868 0.09365 0.09250 7.212 5.50e-13 ***
## RaceEA_1:2 -0.61796 0.53904 0.27593 0.26567 -2.326 0.0200 *
## Score_std:RaceEA_1:2 0.06579 1.06800 0.16763 0.15205 0.433 0.6653
## Score_std_1:3 0.40123 1.49366 0.06088 0.06052 6.630 3.36e-11 ***
## RaceEA_1:3 0.20266 1.22465 0.11343 0.11204 1.809 0.0705 .
## Score_std:RaceEA_1:3 -0.03748 0.96321 0.09210 0.08855 -0.423 0.6721
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## Score_std_1:2 1.9487 0.5132 1.6256 2.3360
## RaceEA_1:2 0.5390 1.8551 0.3202 0.9073
## Score_std:RaceEA_1:2 1.0680 0.9363 0.7928 1.4388
## Score_std_1:3 1.4937 0.6695 1.3266 1.6818
## RaceEA_1:3 1.2247 0.8166 0.9832 1.5254
## Score_std:RaceEA_1:3 0.9632 1.0382 0.8097 1.1458
##
## Concordance= 0.649 (se = 0.013 )
## Likelihood ratio test= 136.1 on 6 df, p=<2e-16
## Wald test = 164.7 on 6 df, p=<2e-16
## Score (logrank) test = 166.8 on 6 df, p=<2e-16, Robust = 95.76 p=<2e-16
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).
Grace method 1, 1-year data
## Call:
## coxph(formula = Surv(time_to_event_grace, event_status2) ~ Score_std *
## Race, data = surv_1y_grace_v1, id = X)
##
## n= 3981, number of events= 837
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## Score_std_1:2 0.66879 1.95187 0.06962 0.06891 9.705 < 2e-16 ***
## RaceEA_1:2 -0.70754 0.49286 0.20770 0.20027 -3.533 0.000411 ***
## Score_std:RaceEA_1:2 0.13006 1.13890 0.12592 0.11754 1.107 0.268497
## Score_std_1:3 0.40673 1.50190 0.04383 0.04408 9.228 < 2e-16 ***
## RaceEA_1:3 0.09726 1.10215 0.08231 0.08131 1.196 0.231608
## Score_std:RaceEA_1:3 0.03114 1.03163 0.06653 0.06504 0.479 0.632073
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## Score_std_1:2 1.9519 0.5123 1.7053 2.2341
## RaceEA_1:2 0.4929 2.0290 0.3329 0.7298
## Score_std:RaceEA_1:2 1.1389 0.8780 0.9046 1.4339
## Score_std_1:3 1.5019 0.6658 1.3776 1.6374
## RaceEA_1:3 1.1022 0.9073 0.9398 1.2926
## Score_std:RaceEA_1:3 1.0316 0.9693 0.9082 1.1719
##
## Concordance= 0.654 (se = 0.01 )
## Likelihood ratio test= 276.4 on 6 df, p=<2e-16
## Wald test = 325.6 on 6 df, p=<2e-16
## Score (logrank) test = 344.1 on 6 df, p=<2e-16, Robust = 185.9 p=<2e-16
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).
Grace method 2, 6-month data
## Call:
## coxph(formula = Surv(time_to_event_grace, event_status2) ~ Score_std *
## Race, data = surv_6m_grace_v2, id = X)
##
## n= 2100, number of events= 393
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## Score_std_1:2 0.67308 1.96027 0.09619 0.09741 6.910 4.85e-12 ***
## RaceEA_1:2 -0.61018 0.54325 0.28169 0.27268 -2.238 0.0252 *
## Score_std:RaceEA_1:2 0.06585 1.06806 0.17165 0.15853 0.415 0.6779
## Score_std_1:3 0.40734 1.50282 0.06711 0.06782 6.006 1.90e-09 ***
## RaceEA_1:3 0.28875 1.33476 0.12056 0.11951 2.416 0.0157 *
## Score_std:RaceEA_1:3 -0.05940 0.94233 0.10008 0.09618 -0.618 0.5368
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## Score_std_1:2 1.9603 0.5101 1.6196 2.3726
## RaceEA_1:2 0.5433 1.8408 0.3183 0.9271
## Score_std:RaceEA_1:2 1.0681 0.9363 0.7828 1.4573
## Score_std_1:3 1.5028 0.6654 1.3158 1.7165
## RaceEA_1:3 1.3348 0.7492 1.0560 1.6871
## Score_std:RaceEA_1:3 0.9423 1.0612 0.7804 1.1378
##
## Concordance= 0.649 (se = 0.014 )
## Likelihood ratio test= 126.1 on 6 df, p=<2e-16
## Wald test = 149.2 on 6 df, p=<2e-16
## Score (logrank) test = 154.1 on 6 df, p=<2e-16, Robust = 86.59 p=<2e-16
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).
Grace method 2, 1-year data
## Call:
## coxph(formula = Surv(time_to_event_grace, event_status2) ~ Score_std *
## Race, data = surv_1y_grace_v2, id = X)
##
## n= 3919, number of events= 775
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## Score_std_1:2 0.66640 1.94721 0.07031 0.07020 9.492 < 2e-16 ***
## RaceEA_1:2 -0.71301 0.49017 0.21023 0.20358 -3.502 0.000461 ***
## Score_std:RaceEA_1:2 0.13850 1.14855 0.12757 0.12043 1.150 0.250132
## Score_std_1:3 0.40943 1.50595 0.04658 0.04763 8.595 < 2e-16 ***
## RaceEA_1:3 0.14348 1.15429 0.08557 0.08488 1.690 0.090936 .
## Score_std:RaceEA_1:3 0.01975 1.01994 0.06999 0.06859 0.288 0.773416
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## Score_std_1:2 1.9472 0.5136 1.6969 2.2344
## RaceEA_1:2 0.4902 2.0401 0.3289 0.7305
## Score_std:RaceEA_1:2 1.1485 0.8707 0.9071 1.4543
## Score_std_1:3 1.5060 0.6640 1.3717 1.6533
## RaceEA_1:3 1.1543 0.8663 0.9774 1.3632
## Score_std:RaceEA_1:3 1.0199 0.9804 0.8916 1.1667
##
## Concordance= 0.651 (se = 0.01 )
## Likelihood ratio test= 259.8 on 6 df, p=<2e-16
## Wald test = 305.5 on 6 df, p=<2e-16
## Score (logrank) test = 324 on 6 df, p=<2e-16, Robust = 171.5 p=<2e-16
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).