Os dados do banco UIS (Survival Analysis Stata UIS.

dta)
O objetivo dos dados do UIS é modelar o tempo até o retorno ao uso de drogas para pacientes inscritos em dois
programas diferentes de tratamento residencial que diferem na duração (treat = 0 é o programa curto e treat = 1
é o programa longo). Os pacientes foram distribuídos aleatoriamente em dois locais diferentes (site = 0 é site A e
site = 1 é site B). A variável age indica a idade de admissão, herco indica uso de heroína ou cocaína nos últimos
três meses (herco = 1 indica uso de heroína e cocaína, herco = 2 indica uso de heroína ou cocaína e herco = 3
não indica consumo de heroína e cocaína) e ndrugtx indica o número de tratamentos medicamentosos
anteriores. A variável time contém o tempo até o retorno ao uso de drogas e a variável censor indica se o sujeito
retornou ao uso de drogas (censor = 1 indica retorno ao uso de drogas e censor = 0 caso contrário).
Vejamos as 10 primeiras observações do conjunto de dados do UIS. Observe que o sujeito 5 é censurado e não
experimentou um evento enquanto estava no estudo. Observe também que a codificação do censor é bastante
contra-intuitiva, já que o valor 1 indica um evento e 0 indica censura. Talvez seja mais apropriado chamar essa
variável de “evento”.
list id time censor age ndrugtx treat site herco in 1/10, nodisplay

id time censor age ndrugtx treat site herco

1. 1 188 1 39 1 1 0 3
2. 2 26 1 33 8 1 0 3
3. 3 207 1 33 3 1 0 2
4. 4 144 1 32 1 0 0 3
5. 5 551 0 24 5 1 0 2
6. 6 32 1 30 1 1 0 1
7. 7 459 1 39 34 1 0 3
8. 8 22 1 27 2 1 0 3
9. 9 210 1 40 3 1 0 2
10. 10 184 1 36 7 1 0 2

Explorando os dados: Análise Univariada

Em qualquer análise de dados, é sempre uma boa ideia fazer uma análise univariada antes de prosseguir para
modelos mais complicados. Na análise de sobrevivência, é altamente recomendável olhar para as curvas de
Kaplan-Meier para todos os preditores categóricos. Isto proporcionará uma visão da forma da função de
sobrevivência para cada grupo e dará uma ideia se os grupos são ou não proporcionais (isto é, as funções de
sobrevivência são aproximadamente paralelas). Também consideramos os testes de igualdade entre estratos
para explorar se devemos ou não incluir o preditor no modelo final. Para as variáveis categóricas, usaremos o
teste log-rank de igualdade entre estratos, que é um teste não paramétrico. Para as variáveis contínuas,
usaremos uma regressão de risco proporcional univariada de Cox, que é um modelo semi-paramétrico.
Consideraremos incluir o preditor se o teste tiver um valor p de 0,2 a 0,25 ou menos. Estamos usando este
esquema de eliminação porque todos os preditores no conjunto de dados são variáveis que podem ser
relevantes para o modelo. Se o preditor tiver um valor p maior que 0,25 em uma análise univariada, é altamente
improvável que ele contribua com qualquer coisa para um modelo que inclua outros preditores.
O teste de log-rank de igualdade entre os estratos para o preditor de tratamento tem um valor de p de 0,0091,
assim, o tratamento será incluído um potencial candidato para o modelo final. A partir do gráfico, vemos que a
função de sobrevivência para cada grupo de tratamento não é perfeitamente paralela, mas separada, exceto no
início e no final. A sobreposição no final não deve causar muita preocupação, porque é determinado por apenas
um número muito pequeno de sujeitos censurados de uma amostra com 628 indivíduos. Em geral, o teste log-
rank coloca mais ênfase nas diferenças nas curvas em valores de tempo maiores. É por isso que obtemos um
valor p tão pequeno, embora as duas curvas de sobrevivência pareçam estar muito próximas umas das outras
por um tempo inferior a 100 dias.
Primeiro passo, declarar o banco como survival analysis:
stset time, failure(censor=1)

sts test treat, logrank

sts graph, by(treat)

failure _d: censor

analysis time _t: time

Log-rank test for equality of survivor functions

| Events Events
treat | observed expected
------+-------------------------
0 | 265 235.80
1 | 243 272.20
------+-------------------------
Total | 508 508.00

chi2(1) = 6.80
Pr>chi2 = 0.0091

O teste de log-rank de igualdade entre estratos para o local do preditor tem um valor de p de 0,1240, portanto, o
site será incluído como um potencial candidato para o modelo final porque esse valor de p ainda é menor do que
nosso limite de 0,2. A partir do gráfico, vemos que as curvas de sobrevivência não são todas paralelas e que
existem dois períodos ([0, 100] e [200, 300]) onde as curvas estão muito próximas. Isso explicaria o valor p
bastante alto do teste de log-rank.
sts test site, logrank
sts graph, by(site)

failure _d: censor

analysis time _t: time

Log-rank test for equality of survivor functions

| Events Events
site | observed expected
------+-------------------------
0 | 364 347.94
1 | 144 160.06
------+-------------------------
Total | 508 508.00

chi2(1) = 2.37
Pr>chi2 = 0.1240

O teste log-rank de igualdade entre os estratos para o preditor herco tem um valor de p de 0,1473, assim
herco será incluído como candidato potencial para o modelo final. A partir do gráfico, vemos que os três
grupos não são paralelos e que, especialmente, os grupos herco = 1 e herco = 3 se sobrepõem para a maior
parte do gráfico. Essa falta de paralelismo pode representar um problema quando incluímos esse preditor no
modelo de risco proporcional de Cox, já que uma das premissas é a proporcionalidade dos preditores.
sts test herco
sts graph, by(herco) noborder

failure _d: censor

analysis time _t: time

Log-rank test for equality of survivor functions

| Events Events
herco | observed expected
------+-------------------------
1 | 228 242.14
2 | 100 84.19
3 | 180 181.67
------+-------------------------
Total | 508 508.00

chi2(2) = 3.83
Pr>chi2 = 0.1473

Não é possível calcular uma curva de Kaplan-Meier para os preditores contínuos, pois haveria uma curva para
cada nível do preditor e um preditor contínuo teria muitos níveis diferentes. Em vez disso, consideramos o
modelo de risco proporcional de Cox com um único preditor contínuo. Infelizmente, não é possível produzir um
gráfico ao usar o comando stcox. Em vez disso, consideramos o teste Qui-quadrado para ndrugtx que tem
um valor de p de 0,0003, portanto, ndrugtx é um candidato potencial para o modelo final, pois o valor p é
menor que nosso valor de corte de 0,2. Especificamos a opção nohr para indicar que não queremos ver a
taxa de risco, mas queremos ver os coeficientes.
stcox ndrugtx, nohr

Cox regression -- Breslow method for ties

No. of subjects = 611 Number of obs = 611

No. of failures = 496
Time at risk = 143002
LR chi2(1) = 13.35
Log likelihood = -2868.299 Prob > chi2 = 0.0003
------------------------------------------------------------------------------
_t |
_d | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
ndrugtx | .029372 .0074979 3.92 0.000 .0146763 .0440676
------------------------------------------------------------------------------
Neste modelo, o teste qui-quadrado de age também tem um valor de p inferior a 0,2 e, portanto, é um
candidato em potencial para o modelo final.

stcox age, nohr

Cox regression -- Breslow method for ties

No. of subjects = 623 Number of obs = 623

No. of failures = 504
Time at risk = 146816
LR chi2(1) = 3.24
Log likelihood = -2931.4929 Prob > chi2 = 0.0719
------------------------------------------------------------------------------
_t |
_d | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
age | -.0128641 .0071888 -1.79 0.074 -.0269539 .0012256
------------------------------------------------------------------------------
Model Building
Para o nosso modelo de construção, consideraremos primeiro o modelo que incluirá todos os preditores que
tivessem um valor p inferior a 0,2 - 0,25 nas análises univariadas, o que, nessa análise específica, significa
que incluiremos todos os preditores em nosso modelo. O preditor categórico herco tem três níveis e, portanto,
incluiremos esse preditor usando a variável dummy com o grupo herco = 1 como o grupo de referência.
Podemos criar essas variáveis dummy rapidamente usando o comando xi com stcox.

stcox age ndrugtx i.treat i.site i.herco, nohr

failure _d: censor

analysis time _t: time

Iteration 0: log likelihood = -2868.555

Iteration 1: log likelihood = -2851.6989
Iteration 2: log likelihood = -2851.0884
Iteration 3: log likelihood = -2851.0863
Refining estimates:
Iteration 0: log likelihood = -2851.0863

Cox regression -- Breslow method for ties

No. of subjects = 610 Number of obs = 610

No. of failures = 495
Time at risk = 142994
LR chi2(6) = 34.94
Log likelihood = -2851.0863 Prob > chi2 = 0.0000

------------------------------------------------------------------------------
_t | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
age | -.0237543 .0075611 -3.14 0.002 -.0385737 -.0089349
ndrugtx | .034745 .0077538 4.48 0.000 .0195478 .0499422
1.treat | -.2540169 .091005 -2.79 0.005 -.4323834 -.0756504
1.site | -.1723881 .1020981 -1.69 0.091 -.3724966 .0277205
|
herco |
2 | .2467753 .1227597 2.01 0.044 .0061706 .4873799
3 | .125668 .1030729 1.22 0.223 -.0763513 .3276873
------------------------------------------------------------------------------

test 2.herco 3.herco

( 1) 2.herco = 0
( 2) 3.herco = 0

chi2( 2) = 4.36
Prob > chi2 = 0.1130
O preditor herco claramente não é significativo e vamos descartá-lo do modelo final. O site do preditor
também não é significativo, mas a partir de pesquisas anteriores sabemos que essa é uma variável muito
importante a ter no modelo final e, portanto, não iremos eliminar o site do modelo. Assim, o modelo final dos
principais efeitos inclui: age, ndrugtx, treat e site.

stcox age ndrugtx i.treat i.site, nohr

failure _d: censor

analysis time _t: time

Iteration 0: log likelihood = -2868.555

Iteration 1: log likelihood = -2853.8641
Iteration 2: log likelihood = -2853.2393
Iteration 3: log likelihood = -2853.2371
Refining estimates:
Iteration 0: log likelihood = -2853.2371

Cox regression -- Breslow method for ties

No. of subjects = 610 Number of obs = 610

No. of failures = 495
Time at risk = 142994
LR chi2(4) = 30.64
Log likelihood = -2853.2371 Prob > chi2 = 0.0000

------------------------------------------------------------------------------
_t | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
age | -.0221289 .0075108 -2.95 0.003 -.0368499 -.007408
ndrugtx | .0350249 .0076676 4.57 0.000 .0199967 .050053
1.treat | -.2436784 .0905411 -2.69 0.007 -.4211358 -.0662211
1.site | -.1683325 .1004119 -1.68 0.094 -.3651362 .0284712
------------------------------------------------------------------------------
Interactions
Em seguida, precisamos considerar as interações. Nós não temos nenhum conhecimento prévio de interações
específicas que devemos incluir, então consideraremos todas as interações possíveis. Como o nosso modelo
é bastante pequeno, isso é administrável, mas a situação ideal é quando toda a construção de modelos,
incluindo interações, é orientada por teoria.

O termo de interação de age com ndrugtx não é significativo e não será incluído no modelo.

stcox age ndrugtx i.treat i.site c.age#c.ndrug, nohr

failure _d: censor

analysis time _t: time

Iteration 0: log likelihood = -2868.555

Iteration 1: log likelihood = -2854.6056
Iteration 2: log likelihood = -2851.8845
Iteration 3: log likelihood = -2851.8195
Iteration 4: log likelihood = -2851.8195
Refining estimates:
Iteration 0: log likelihood = -2851.8195

Cox regression -- Breslow method for ties

No. of subjects = 610 Number of obs = 610
No. of failures = 495
Time at risk = 142994
LR chi2(5) = 33.47
Log likelihood = -2851.8195 Prob > chi2 = 0.0000

------------------------------------------------------------------------------
_t | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
age | -.0110172 .0100068 -1.10 0.271 -.0306302 .0085959
ndrugtx | .1054144 .0419532 2.51 0.012 .0231875 .1876412
1.treat | -.2352811 .0906447 -2.60 0.009 -.4129416 -.0576207
1.site | -.1746173 .1004498 -1.74 0.082 -.3714953 .0222607
|
c.age#|
c.ndrugtx | -.0020967 .0012469 -1.68 0.093 -.0045406 .0003472
------------------------------------------------------------------------------
The interaction age and treat is not significant and will not be included in the model.

stcox age ndrugtx i.treat i.site c.age#i.treat, nohr

failure _d: censor

analysis time _t: time

Iteration 0: log likelihood = -2868.555

Iteration 1: log likelihood = -2852.845
Iteration 2: log likelihood = -2852.1654
Iteration 3: log likelihood = -2852.1631
Refining estimates:
Iteration 0: log likelihood = -2852.1631

Cox regression -- Breslow method for ties

No. of subjects = 610 Number of obs = 610

No. of failures = 495
Time at risk = 142994
LR chi2(5) = 32.78
Log likelihood = -2852.1631 Prob > chi2 = 0.0000

------------------------------------------------------------------------------
_t | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
age | -.0114621 .0103995 -1.10 0.270 -.0318448 .0089205
 ndrugtx | .0357659 .0077155 4.64 0.000 .0206437 .050888
1.treat | .4483383 .4809163 0.93 0.351 -.4942403 1.390917
1.site | -.1492698 .1010768 -1.48 0.140 -.3473766 .048837
|
treat#c.age |
1 | -.021469 .0146588 -1.46 0.143 -.0501996 .0072616
------------------------------------------------------------------------------
The interaction age anf site is significant and will be included in the model.

stcox age ndrugtx i.treat i.site c.age#i.site, nohr

failure _d: censor

analysis time _t: time

Iteration 0: log likelihood = -2868.555

Iteration 1: log likelihood = -2851.487
Iteration 2: log likelihood = -2850.8935
Iteration 3: log likelihood = -2850.8915
Refining estimates:
Iteration 0: log likelihood = -2850.8915

Cox regression -- Breslow method for ties

No. of subjects = 610 Number of obs = 610

No. of failures = 495
Time at risk = 142994
LR chi2(5) = 35.33
Log likelihood = -2850.8915 Prob > chi2 = 0.0000

------------------------------------------------------------------------------
_t | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
age | -.0336943 .0092913 -3.63 0.000 -.051905 -.0154837
ndrugtx | .0364537 .0077012 4.73 0.000 .0213597 .0515478
1.treat | -.2674113 .0912282 -2.93 0.003 -.4462153 -.0886073
1.site | -1.245928 .5087349 -2.45 0.014 -2.24303 -.2488262
|
site#c.age |
1 | .0337728 .0155087 2.18 0.029 .0033764 .0641693
------------------------------------------------------------------------------
The interaction drug anf treat is not significant and will be not included in the model.

stcox age ndrugtx i.treat i.site c.ndrug#i.treat, nohr

failure _d: censor

 analysis time _t: time

Iteration 0: log likelihood = -2868.555

Iteration 1: log likelihood = -2854.1019
Iteration 2: log likelihood = -2853.0275
Iteration 3: log likelihood = -2853.0174
Iteration 4: log likelihood = -2853.0174
Refining estimates:
Iteration 0: log likelihood = -2853.0174

Cox regression -- Breslow method for ties

No. of subjects = 610 Number of obs = 610

No. of failures = 495
Time at risk = 142994
LR chi2(5) = 31.08
Log likelihood = -2853.0174 Prob > chi2 = 0.0000

------------------------------------------------------------------------------
_t | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
age | -.0220158 .0075029 -2.93 0.003 -.0367212 -.0073105
ndrugtx | .0404798 .011066 3.66 0.000 .0187909 .0621686
1.treat | -.1949252 .1166714 -1.67 0.095 -.423597 .0337465
1.site | -.1708522 .1004592 -1.70 0.089 -.3677487 .0260442
|
treat#|
c.ndrugtx |
1 | -.0099061 .0149405 -0.66 0.507 -.0391889 .0193767
------------------------------------------------------------------------------
The interaction drug and site is not significant and will not be included in the model.

stcox age ndrugtx i.treat i.site c.ndrug#i.site, nohr

failure _d: censor

analysis time _t: time

Iteration 0: log likelihood = -2868.555

Iteration 1: log likelihood = -2853.9255
Iteration 2: log likelihood = -2853.1789
Iteration 3: log likelihood = -2853.1746
Iteration 4: log likelihood = -2853.1746
Refining estimates:
Iteration 0: log likelihood = -2853.1746

Cox regression -- Breslow method for ties

No. of subjects = 610 Number of obs = 610

No. of failures = 495
Time at risk = 142994
LR chi2(5) = 30.76
Log likelihood = -2853.1746 Prob > chi2 = 0.0000

------------------------------------------------------------------------------
_t | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
age | -.0222734 .0075266 -2.96 0.003 -.0370251 -.0075216
ndrugtx | .0366438 .0088665 4.13 0.000 .0192658 .0540218
1.treat | -.2454197 .0906816 -2.71 0.007 -.4231524 -.067687
1.site | -.1417165 .1253391 -1.13 0.258 -.3873766 .1039435
|
site#|
c.ndrugtx |
1 | -.0059702 .0169939 -0.35 0.725 -.0392776 .0273373
------------------------------------------------------------------------------
The interaction treat and site is not significant and will not be included in the model.

stcox age ndrugtx i.treat i.site i.treat#i.site, nohr

failure _d: censor

analysis time _t: time

Iteration 0: log likelihood = -2868.555

Iteration 1: log likelihood = -2852.4136
Iteration 2: log likelihood = -2851.8662
Iteration 3: log likelihood = -2851.8645
Refining estimates:
Iteration 0: log likelihood = -2851.8645

Cox regression -- Breslow method for ties

No. of subjects = 610 Number of obs = 610

No. of failures = 495
Time at risk = 142994
LR chi2(5) = 33.38
Log likelihood = -2851.8645 Prob > chi2 = 0.0000
------------------------------------------------------------------------------
_t | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
age | -.0238596 .007638 -3.12 0.002 -.0388297 -.0088895
ndrugtx | .0361507 .0077457 4.67 0.000 .0209694 .0513321
1.treat | -.3404088 .107682 -3.16 0.002 -.5514616 -.129356
1.site | -.3238557 .139417 -2.32 0.020 -.597108 -.0506033
|
treat#site |
1 1 | .3335144 .2009322 1.66 0.097 -.0603054 .7273342
------------------------------------------------------------------------------
The final model including interaction. Now we can see why it was important to include site in our model as prior
research had suggested because it turns out that site is involved in the only significant interaction in the model.
We can compare the model with the interaction to the model without the interaction using the lrtest command
since the models are nested. The significant lrtest indicates that we reject the null hypothesis that the two
models fit the data equally well and conclude that the bigger model with the interaction fits the data better than
the smaller model which did not include the interaction.

stcox age ndrugtx i.treat i.site c.age#i.site, nohr

failure _d: censor

analysis time _t: time

Iteration 0: log likelihood = -2868.555

Iteration 1: log likelihood = -2851.487
Iteration 2: log likelihood = -2850.8935
Iteration 3: log likelihood = -2850.8915
Refining estimates:
Iteration 0: log likelihood = -2850.8915

Cox regression -- Breslow method for ties

No. of subjects = 610 Number of obs = 610

No. of failures = 495
Time at risk = 142994
LR chi2(5) = 35.33
Log likelihood = -2850.8915 Prob > chi2 = 0.0000

------------------------------------------------------------------------------
_t | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
age | -.0336943 .0092913 -3.63 0.000 -.051905 -.0154837
ndrugtx | .0364537 .0077012 4.73 0.000 .0213597 .0515478
1.treat | -.2674113 .0912282 -2.93 0.003 -.4462153 -.0886073
1.site | -1.245928 .5087349 -2.45 0.014 -2.24303 -.2488262
 |
site#c.age |
1 | .0337728 .0155087 2.18 0.029 .0033764 .0641693
------------------------------------------------------------------------------

estimates store m1

stcox age ndrugtx i.treat i.site, nohr

failure _d: censor

analysis time _t: time

Iteration 0: log likelihood = -2868.555

Iteration 1: log likelihood = -2853.8641
Iteration 2: log likelihood = -2853.2393
Iteration 3: log likelihood = -2853.2371
Refining estimates:
Iteration 0: log likelihood = -2853.2371

Cox regression -- Breslow method for ties

No. of subjects = 610 Number of obs = 610

No. of failures = 495
Time at risk = 142994
LR chi2(4) = 30.64
Log likelihood = -2853.2371 Prob > chi2 = 0.0000

------------------------------------------------------------------------------
_t | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
age | -.0221289 .0075108 -2.95 0.003 -.0368499 -.007408
ndrugtx | .0350249 .0076676 4.57 0.000 .0199967 .050053
1.treat | -.2436784 .0905411 -2.69 0.007 -.4211358 -.0662211
1.site | -.1683325 .1004119 -1.68 0.094 -.3651362 .0284712
------------------------------------------------------------------------------

lrtest . m1

Likelihood-ratio test LR chi2(1) = 4.69

(Assumption: . nested in m1) Prob > chi2 = 0.0303
The final model and interpretation of the hazard ratios.

stcox age ndrugtx i.treat i.site c.age#i.site

 failure _d: censor
analysis time _t: time

Iteration 0: log likelihood = -2868.555

Iteration 1: log likelihood = -2851.487
Iteration 2: log likelihood = -2850.8935
Iteration 3: log likelihood = -2850.8915
Refining estimates:
Iteration 0: log likelihood = -2850.8915

Cox regression -- Breslow method for ties

No. of subjects = 610 Number of obs = 610

No. of failures = 495
Time at risk = 142994
LR chi2(5) = 35.33
Log likelihood = -2850.8915 Prob > chi2 = 0.0000

------------------------------------------------------------------------------
_t | Haz. Ratio Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
age | .966867 .0089835 -3.63 0.000 .9494191 .9846355
ndrugtx | 1.037126 .0079871 4.73 0.000 1.021589 1.052899
1.treat | .7653582 .0698223 -2.93 0.003 .6400459 .9152049
1.site | .2876737 .1463497 -2.45 0.014 .1061364 .7797155
|
site#c.age |
1 | 1.03435 .0160414 2.18 0.029 1.003382 1.066273
------------------------------------------------------------------------------
From looking at the hazard ratios (also called relative risks) the model indicates that as the number of previous
drug treatment (ndrugtx) increases by one unit, and all other variables are held constant, the rate of relapse
increases by 3.7%. If the treatment length is altered from short to long, while holding all other variables
constant, the rate of relapse decreases by (100% – 76.5%) = 23.5%. As treatment is moved from site A to site
B and age is equal to zero, and all other variables are held constant, the rate of relapse decreases by (100% –
28.8%) = 71.2%. These results are all based on the output using Hazard ratios. To discuss the variables that
are involved in an interaction term, such as age and site in our model, we need to use the raw coefficients and
here they are listed below just for convenience.

stcox, nohr

Cox regression -- Breslow method for ties

No. of subjects = 610 Number of obs = 610

No. of failures = 495
Time at risk = 142994
 LR chi2(5) = 35.33
Log likelihood = -2850.8915 Prob > chi2 = 0.0000

------------------------------------------------------------------------------
_t | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
age | -.0336943 .0092913 -3.63 0.000 -.051905 -.0154837
ndrugtx | .0364537 .0077012 4.73 0.000 .0213597 .0515478
1.treat | -.2674113 .0912282 -2.93 0.003 -.4462153 -.0886073
1.site | -1.245928 .5087349 -2.45 0.014 -2.24303 -.2488262
|
site#c.age |
1 | .0337728 .0155087 2.18 0.029 .0033764 .0641693
------------------------------------------------------------------------------
Comparing 2 subjects within site A (site=0), an increase in age of 5 years while all other variables are held
constant yields a hazard ratio equal to exp(-0.03369*5) = .84497351. Thus, the rate of relapse is decreased by
(100% – 84.5%) = 15.5% with an increase of 5 years in age. Comparing 2 subjects within site B, an increase
in age of 5 years while holding all other variables constant, yields a hazard ratio equal to exp(-0.03369*5 +
0.03377*5) = 1.0004. Thus, the rate of relapse stays fairly flat for subjects at site B since 1.0004 if so close to
1.

Proportionality Assumption

One of the main assumptions of the Cox proportional hazard model is proportionality. There are several
methods for verifying that a model satisfies the assumption of proportionality. We will check proportionality by
including time-dependent covariates in the model by using the tvc and the texp options in
the stcox command. Time dependent covariates are interactions of the predictors and time. In this analysis we
choose to use the interactions with log(time) because this is the most common function of time used in time-
dependent covariates but any function of time could be used. If a time-dependent covariate is significant this
indicates a violation of the proportionality assumption for that specific predictor.

The conclusion is that all of the time-dependent variables are not significant either collectively or individually
thus supporting the assumption of proportional hazard.

stcox age ndrugtx i.treat i.site c.age#i.site, nohr tvc(age ndrugtx treat site)
texp(ln(_t))

failure _d: censor

analysis time _t: time

Iteration 0: log likelihood = -2868.555

Iteration 1: log likelihood = -2850.4619
Iteration 2: log likelihood = -2849.8647
Iteration 3: log likelihood = -2849.8626
Iteration 4: log likelihood = -2849.8626
Refining estimates:
Iteration 0: log likelihood = -2849.8626
Cox regression -- Breslow method for ties

No. of subjects = 610 Number of obs = 610

No. of failures = 495
Time at risk = 142994
LR chi2(9) = 37.38
Log likelihood = -2849.8626 Prob > chi2 = 0.0000

------------------------------------------------------------------------------
_t | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
main |
age | -.0322788 .0340846 -0.95 0.344 -.0990834 .0345258
ndrugtx | .0173789 .0321568 0.54 0.589 -.0456473 .0804052
1.treat | -.6671007 .4114915 -1.62 0.105 -1.473609 .1394078
1.site | -1.637207 .6801889 -2.41 0.016 -2.970353 -.3040617
|
site#c.age |
1 | .033723 .015548 2.17 0.030 .0032495 .0641965
-------------+----------------------------------------------------------------
tvc |
age | -.0004057 .007119 -0.06 0.955 -.0143587 .0135473
ndrugtx | .0042828 .0069637 0.62 0.539 -.0093658 .0179314
treat | .0860457 .0863163 1.00 0.319 -.0831312 .2552226
site | .084347 .0974399 0.87 0.387 -.1066317 .2753258
------------------------------------------------------------------------------
Note: variables in tvc equation interacted with ln(_t)
Another method of testing the proportionality assumption is by using the Schoenfeld and scaled Schoenfeld
residuals which must first be saved through the stcox command. In the stphtest command we test the
proportionality of the model as a whole and by using the detail option we get a test of proportionality for each
predictor. By using the plot option we can also obtain a graph of the scaled Schoenfeld assumption. If the
tests in the table are not significance (p-values over 0.05) then we can not reject proportionality and we
assume that we do not have a violation of the proportional assumption. A horizontal line in the graphs is
further indication that there is no violation of the proportionality assumption. The stphplot command uses log-
log plots to test proportionality and if the lines in these plots are parallel then we have further indication that
the predictors do not violate the proportionality assumption.

The predictor treat might warrant some closer examination since it does have a significant test and the curve
in the graph is not completely horizontal. The graph from the stphplot command does not have completely
parallel curves. However, we choose to leave treat in the model unaltered based on prior research.

quietly stcox age ndrugtx treat site c.age#i.site, schoenfeld(sch*) scaledsch(sca*)

stphtest, detail
stphtest, plot(age) msym(oh)
stphtest, plot(ndrugtx) msym(oh)
stphtest, plot(treat) msym(oh)
stphtest, plot(site) msym(oh)
stphtest, plot(c.age#1.site) msym(oh)

stphplot, by(treat) plot1(msym(oh)) plot2(msym(th))

stphplot, by(site) plot1(msym(oh)) plot2(msym(th))

drop sch1-sch5 sca1-sca5

Test of proportional hazards assumption

Time: Time
----------------------------------------------------------------
| rho chi2 df Prob>chi2
------------+---------------------------------------------------
age | 0.01210 0.07 1 0.7912
ndrugtx | 0.05563 1.47 1 0.2260
treat | 0.10598 5.61 1 0.0179
site | 0.02336 0.25 1 0.6150
age_site | -0.01350 0.08 1 0.7722
------------+---------------------------------------------------
global test | 8.27 5 0.1419
----------------------------------------------------------------
 If one of the predictors were not proportional there are various solutions to consider. One solution is to include
the time-dependent variable for the non-proportional predictors. Another solution is to stratify on the non-
proportional predictor. The following is an example of stratification on the predictor treat. Note that treat is no
longer included in the model statement instead it is specified in the strata statement.

stcox age ndrugtx site c.age#i.site, nohr strata(treat)

failure _d: censor

analysis time _t: time

Iteration 0: log likelihood = -2525.8022

Iteration 1: log likelihood = -2511.524
Iteration 2: log likelihood = -2510.8632
Iteration 3: log likelihood = -2510.8606
Refining estimates:
Iteration 0: log likelihood = -2510.8606

Stratified Cox regr. -- Breslow method for ties

No. of subjects = 610 Number of obs = 610

No. of failures = 495
Time at risk = 142994
LR chi2(4) = 29.88
Log likelihood = -2510.8606 Prob > chi2 = 0.0000

------------------------------------------------------------------------------
_t | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
age | -.0347463 .0092884 -3.74 0.000 -.0529512 -.0165413
ndrugtx | .0363728 .0076961 4.73 0.000 .0212888 .0514569
site | -1.251301 .5085532 -2.46 0.014 -2.248047 -.2545555
|
site#c.age |
1 | .0339858 .0155057 2.19 0.028 .0035951 .0643765
------------------------------------------------------------------------------
Stratified by treat

Graphing Survival Functions

Each covariate pattern will have a different survival function. The default survival function is for the covariate
pattern where each predictor is set equal to zero. However, for many predictors this value is not meaningful
because this value falls outside of the data such as age=0. It would be much more useful to specify an exact
covariate pattern and generate a survival function for subjects with that specific covariate pattern.

In the following example we want to graph the survival function for a subject who is 30 years old (age=30), has
had 5 prior drug treatments (ndrugtx=5), and is currently getting the long treatment (treat=1) at site A (site=0
and agesite=30*0 = 0). We first output the baseline survival function for the covariate pattern where all
predictors are set to zero. Then we raise the baseline survival function to the exponential to the linear
combination of the coefficients and the values of the covariates in the covariate pattern of interest. Thus, in this
particular instance the linear combination would be: -0.0336943*30+0.0364537*5 – 0.2674113*1 – 1.245928*0
– .0337728*0.

stcox age ndrugtx treat site c.age#i.site, nohr basesurv(surv0)

generate surv1 = surv0^exp( (-0.0336943*30+0.0364537*5 - 0.2674113))
line surv1 _t, sort ylab(0 .1 to 1) xlab(0 200 to 1200)
 Looking at the survival function for one covariate pattern is sometimes not sufficient. It is often very useful to
have a graph where we can compare the survival functions of different groups. In the following example we
generate a graph with the survival functions for the two treatment groups where all the subjects are 30 years
old (age=30), have had 5 prior drug treatments (ndrugtx=5) and are currently being treated at site A (site=0
and agesite=30*0=0). Thus, the two covariate patterns differ only in their values for treat.

generate surv2 = surv0^exp( (-0.0336943*30+0.0364537*5))

label variable surv1 "long treatment"
label variable surv2 "short treatment"
line surv1 surv2 _t, sort ylab(0 .1 to 1) xlab(0 200 to 1200)
drop surv0-surv2
 Goodness of Fit of the Final Model

We can evaluate the fit of the model by using the Cox-Snell residuals. If the model fits the data well then the
true cumulative hazard function conditional on the covariate vector has an exponential distribution with a
hazard rate of one. This translates into fitting the model using the stcox command and specifying
the mgale option which will generate the martingale residuals. Then we use the predict command with
the csnell option to generate the Cox-Snell residuals for the model. We reset the data using
the stset command specifying the variable cs, the variable containing the Cox-Snell residuals, as the time
variable. We then use the sts generate command to create the Nelson-Aalen cumulative hazard
function. Finally, we graph the Nelson-Aalen cumulative hazard function and the cs variable so that we can
compare the hazard function to the diagonal line. If the hazard function follows the 45 degree line then we
know that it approximately has an exponential distribution with a hazard rate of one and that the model fits the
data well.

quietly stcox age ndrugtx treat site c.age#i.site, nohr mgale(mg)

predict cs, csnell

stset cs, failure(censor)

sts generate H = na
line H cs cs, sort xlab(0 1 to 4) ylab(0 1 to 4)
drop mg

We see that the hazard function follows the 45 degree line very closely except for very large values of time. It
is very common for models with censored data to have some wiggling at large values of time and it is not
something which should cause much concern. Overall we would conclude that the final model fits the data very
well.