leuk: Survival Times and White Blood Counts for Leukaemia Patients

In SMPracticals: Practicals for Use with Davison (2003) Statistical Models

Description

A data frame of data from 33 leukaemia patients.

Usage

data(leuk)

Format

A data frame with 33 observations on the following 3 variables.

wbc

white blood cell count

ag

a test result, '"present"' or '"absent"'

time

survival time in weeks

Details

Survival times are given for 33 patients who died from acute myelogenous leukaemia. Also measured was the patient's white blood cell count at the time of diagnosis. The patients were also factored into 2 groups according to the presence or absence of a morphologic characteristic of white blood cells. Patients termed AG positive were identified by the presence of Auer rods and/or significant granulation of the leukaemic cells in the bone marrow at the time of diagnosis.

Source

Feigl, P. and Zelen, M. (1965) Estimation of exponential survival probabilities with concomitant information. Biometrics, 21, 826–838.

References

Cox, D. R. and Oakes, D. (1984) Analysis of Survival Data. Chapman & Hall, p. 9.

Davison, A. C. (2003) Statistical Models. Cambridge University Press. Page 542.

Examples

data(leuk)
library(survival)
plot(survfit(Surv(time) ~ ag, data = leuk), lty = 2:3, col = 2:3)
# fit of exponential model
summary(glm(time~ag+log10(wbc),data=leuk,family=Gamma(log)),dispersion=1)
# now Cox models
leuk.cox <- coxph(Surv(time) ~ ag + log(wbc), leuk)
summary(leuk.cox)

Example output

Loading required package: ellipse

Attaching package: 'survival'

The following objects are masked from 'package:SMPracticals':

    aml, pbc


Call:
glm(formula = time ~ ag + log10(wbc), family = Gamma(log), data = leuk)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.1746  -1.2663  -0.4251   0.4962   1.9048  

Coefficients:
            Estimate Std. Error z value Pr(>|z|)    
(Intercept)   5.8154     1.2932   4.497 6.89e-06 ***
agpresent     1.0176     0.3492   2.914  0.00357 ** 
log10(wbc)   -0.7009     0.3036  -2.308  0.02097 *  
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

(Dispersion parameter for Gamma family taken to be 1)

    Null deviance: 58.138  on 32  degrees of freedom
Residual deviance: 40.319  on 30  degrees of freedom
AIC: 301.49

Number of Fisher Scoring iterations: 8

Call:
coxph(formula = Surv(time) ~ ag + log(wbc), data = leuk)

  n= 33, number of events= 33 

             coef exp(coef) se(coef)      z Pr(>|z|)   
agpresent -1.0691    0.3433   0.4293 -2.490  0.01276 * 
log(wbc)   0.3677    1.4444   0.1360  2.703  0.00687 **
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

          exp(coef) exp(-coef) lower .95 upper .95
agpresent    0.3433     2.9126     0.148    0.7964
log(wbc)     1.4444     0.6923     1.106    1.8857

Concordance= 0.726  (se = 0.065 )
Rsquare= 0.377   (max possible= 0.994 )
Likelihood ratio test= 15.64  on 2 df,   p=0.0004014
Wald test            = 15.06  on 2 df,   p=0.0005365
Score (logrank) test = 16.49  on 2 df,   p=0.0002629

SMPracticals documentation built on May 2, 2019, 11:12 a.m.