27 Survival analysis: Possible typing errors

[AleksandarMedarevic]

In the paragraph at the end of the page you describe

The model results indicate there is a negative association between onset to admission duration and all-cause mortality. The expected hazard is 0.9 times lower in a person who who is one day later admitted than another, holding gender constant. Or in a more straightforward explanation, a one unit increase in the duration of onset to admission is associated with a 10.7% (coef *100) decrease in the risk of death.

Results show also a positive association between the source of infection and the all-cause mortality. Which is to say there is an increased risk of death (1.21x) for patients that got a source of infection other than funerals.

data from table

Call:
coxph(formula = Surv(futime, event) ~ gender + age_years + source + 
    days_onset_hosp, data = linelist_surv)

  n= 2772, number of events= 1180 
   (1767 observations deleted due to missingness)

                     coef exp(coef)  se(coef)      z Pr(>|z|)    
genderm          0.004710  1.004721  0.060827  0.077   0.9383    
age_years       -0.002249  0.997753  0.002421 -0.929   0.3528    
sourceother      0.178393  1.195295  0.084291  2.116   0.0343 *  
days_onset_hosp -0.104063  0.901169  0.014245 -7.305 2.77e-13 ***
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

                exp(coef) exp(-coef) lower .95 upper .95
genderm            1.0047     0.9953    0.8918    1.1319
age_years          0.9978     1.0023    0.9930    1.0025
sourceother        1.1953     0.8366    1.0133    1.4100
days_onset_hosp    0.9012     1.1097    0.8764    0.9267

Concordance= 0.566  (se = 0.009 )
Likelihood ratio test= 71.31  on 4 df,   p=1e-14
Wald test            = 59.22  on 4 df,   p=4e-12
Score (logrank) test = 59.54  on 4 df,   p=4e-12

Could you please check if the data from table correlates with text number?

I also suggest you that use dplyr in front of function:

data1 = bmt %>% dplyr::select(id, T1, delta1)
data2 = bmt %>% dplyr::select(id, T1, delta1, TA, deltaA)