Início / coding / Perguntas / 77489747
Accepted
firmo23
Asked: 2023-11-16 01:18:56 +0800 CST

Extraia valores de intervalo de confiança de 95% do modelo Kaplan Meier

  • 772

Depois de calcular um modelo Kaplan Meier, obtenho a km_fitlista abaixo. Como posso extrair valores de intervalo de confiança de 95% agora?

# Plot Kaplan-Meier curves
sg1_survival <- Surv(
  time = sg1$`Time from breast cancer diagnosis to bone metastasis (months) (NA = unknown)`,
  event = sg1$`status (=1 as all pts have bone metastasis and BC)`
)


km_fit <- survfit(sg1_survival ~ sg1$`Molecular type (0=hr+her2+, 1=hr+her2-, 2=hr-her2+, 3=TNBC)`, data = sg1)

# Calculate 95% confidence intervals? fails with "no applicable method"
ci <- confint(km_fit, level = 0.95)

2 respostas

  1. Por favor, experimente o pacote de vassoura abaixo e verifiqueconf.high & conf.low

    new_km_fit <- broom::tidy(km_fit)


# A tibble: 5 × 9
   time n.risk n.event n.censor estimate std.error conf.high conf.low strata                         
  <dbl>  <dbl>   <dbl>    <dbl>    <dbl>     <dbl>     <dbl>    <dbl> <chr>                          
1  28.4      3       1        0    0.667     0.408         1   0.300  sg1$`Molecular type (0=hr+her2…
2  34.2      2       1        0    0.333     0.816         1   0.0673 sg1$`Molecular type (0=hr+her2…
3 137.       1       1        0    0       Inf            NA  NA      sg1$`Molecular type (0=hr+her2…
4  14.0      2       1        0    0.5       0.707         1   0.125  sg1$`Molecular type (0=hr+her2…
5  75.0      1       1        0    0       Inf            NA  NA      sg1$`Molecular type (0=hr+her2…
    • 0
  2. Best Answer

    Está tudo dentro km_fit.

    
km_fit <- survfit(sg1_survival ~ sg1$`Molecular type (0=hr+her2+, 1=hr+her2-, 2=hr-her2+, 3=TNBC)`, data = sg1)
c <- data.frame(km_fit$time, km_fit$lower, km_fit$upper) 
c


     sum.time  sum.lower sum.upper
1  28.37260 0.29950713         1
2  34.15890 0.06727839         1
3 137.32603         NA        NA
4  14.00548 0.12504883         1
5  75.02466         NA        NA
    • 0

relate perguntas