KM report.Rmd

---
date: "`r Sys.Date()`"
author: "Kristy Robledo"
title: "Kaplan Meier session in CC"
output: 
  officedown::rdocx_document:
    mapstyles:
      Normal: ['First Paragraph']
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE, fig.cap = TRUE)
library(officedown)
library(officer)

fp <- fp_par(
  text.align = "center", 
  padding.bottom = 20, padding.top = 120, 
  border.bottom = fp_border())

ft <- fp_text(shading.color='#EFEFEF', bold = TRUE)

library(tidyverse)
library(survminer)
library(survival)
library(patchwork)
library(Cairo)
library(gtsummary)
library(flextable)
library(glue)
```


\newpage

`r fp`

## Table of content

<!---BLOCK_TOC--->

## List of figures

<!---BLOCK_TOC{seq_id: 'fig'}--->

## List of tables

<!---BLOCK_TOC{seq_id: 'tab'}--->


\newpage

# KM plot

Summary of data:

```{r}
glimpse(lung)

table(lung$status)
```
KM curve:

```{r, fig.cap="Kaplan Meier curve by gender", fig.dim=c(9,6)}

lung %>%
  mutate(months = time/30.4, 
         dead = case_when(
           status==2 ~ 1, 
           TRUE ~ 0)) ->lung2

m1<-survfit(Surv(months, dead) ~ sex, data=lung2)

ggsurvplot(m1, 
           data=lung2, 
           legend.labs = c("Males", "Females"), 
           break.time.by = 12/2, 
           risk.table = TRUE, 
           risk.table.height= 0.2, 
           tables.theme = theme_cleantable(), 
           legend="none", 
           palette = c("red", "darkblue"), 
           conf.int = TRUE
           )

p1<-ggsurvplot(m1, 
           data=lung2, 
           legend.labs = c("Males", "Females"), 
           break.time.by = 12/2, 
           risk.table = TRUE, 
           risk.table.height= 0.2, 
           tables.theme = theme_cleantable(), 
           legend="none", 
           palette = c("red", "darkblue"))

```



# output plot into pdf for submission

```{r, eval=FALSE}
p1<-ggsurvplot(m1, 
           data=lung2, 
           legend.labs = c("Males", "Females"), 
           break.time.by = 12/2, 
           risk.table = TRUE, 
           risk.table.height= 0.2, 
           tables.theme = theme_cleantable(), 
           legend="none", 
           palette = c("red", "darkblue")
           )

p2<-ggsurvplot(m1, 
           data=lung2, 
           legend.labs = c("Males", "Females"), 
           break.time.by = 12/2, 
           risk.table = TRUE, 
           risk.table.height= 0.2, 
           tables.theme = theme_cleantable(), 
           legend="none", 
           palette = c("red", "black")
           )

plots<-list()

plots[[1]]<-p1
plots[[2]]<-p2

p<-ggarrange(plot, print=TRUE, ncol=1, risk.table.height=0.3)

ggsave(plot=p, "Fig1.pdf")
```


# output plot into pptx for presentation

```{r, eval=FALSE}

tum_graph <- rvg::dml(code = print(p1, newpage = FALSE))


doc <- read_pptx()
doc <- add_slide(doc, layout = "Title and Content", master = "Office Theme")
doc <- ph_with(doc, tum_graph, location = ph_location(width = 8, height=6) )

print(doc, target = "plots in pp.pptx")


```



# logrank test

```{r}
survdiff(Surv(months, dead) ~ sex, data=lung2)

```

new reproducible publication ready:

```{r}
tbl_survfit(m1, 
            times=c(6, 12)) %>%
  add_p() %>%
  add_nevent() %>%
  add_n() %>%
  as_flex_table() %>%
  set_caption(caption = "Survival probabilities at 6 and 12 months")


tbl_survfit(m1, 
            probs=c(0.5))
```



# cox regression

```{r}
ph<-summary(coxph(Surv(months, dead)~sex, data=lung2))

HR<-round(ph$coefficients[2], digits=2)
HRl<-round(ph$conf.int[3], digits=2)
HRu<-round(ph$conf.int[4], digits=2)
p<-signif(ph$waldtest[3], digits=1)

stats<-glue('HR = {HR} (95% CI: {HRl}-{HRu}), p-value={p}')

```


# stick information on the curve

```{r}
p1$plot +
  annotate("text", 
           x=12, y= 0.9, 
           hjust=0, 
           label=stats) 
  
```