Package 'shinyCox'

Generate Cox-model predicted Kaplan-Meier plots

Description

The main purpose of this function is to be used to create plots within the shiny app created by shine_coxph(). For this reason the argument it takes, KM.hat, is created through a process delineated in the example. This can make the function more complicated if you want to use it outside of the shiny app, although it is fully possible to do so.

Usage

cox_KM_plots(KM.hat, clrs = NULL, confint, ylab = "Prob")

Arguments

KM.hat	Time and survival probability created by predict_one_coxfit()
clrs	color of lines
confint	logical value to determine if confidence intervals should be plotted
ylab	text label for y-axis

Value

Plot of predicted survival curve(s)

Examples

library(survival)
# First colon is split into three treatment arms to compare predicted
# survival across arms
split_colon <- split(colon, colon$rx)

colon_arm1 <- split_colon$Obs
colon_arm2 <- split_colon$Lev
colon_arm3 <- split_colon$`Lev+5FU`

# One coxph model is fit for each treatment

colon1ph <- coxph(Surv(time, status) ~sex +  age + obstruct + nodes,
                  colon_arm1, x = TRUE, model = TRUE)

colon2ph <- coxph(Surv(time, status) ~ sex + age + obstruct + nodes,
                  colon_arm2, x = TRUE, model = TRUE)

colon3ph <- coxph(Surv(time, status) ~ sex + age + obstruct + nodes,
                  colon_arm3, x = TRUE, model = TRUE)

# Creating list of models
cox.fit.list <- vector("list", 3)
cox.fit.list[[1]] <- prep_coxfit(colon1ph)
cox.fit.list[[2]] <- prep_coxfit(colon2ph)
cox.fit.list[[3]] <- prep_coxfit(colon3ph)

# Creating new data row for predictions
new.data <- colon[1, ]
# Creating KM.hat object
n.models=length(cox.fit.list)
KM.hat=vector('list',n.models)
lp=rep(NA,n.models)
names(KM.hat)=names(cox.fit.list)
for (i in 1:n.models)
{
 km.hat=predict_one_coxfit(cox.fit.list[[i]],new.data)
 lp[i]=attr(km.hat,'lp')
 sfit=list(time=km.hat$time,surv=km.hat$surv)
 class(sfit)='survfit'
 KM.hat[[i]]=sfit
}
# Plot
cox_KM_plots(KM.hat)

---

Create table of Cox-model predicted probabilities

Description

Generates tables of predicted probabilities at specified time or vector of times. The KM.hat object contains time and predicted survival probability information as a list of survfit objects.

Usage

cox_times_table(KM.hat, fixTimes = NULL)

Arguments

KM.hat	List of survfit objects
fixTimes	character or vector of characters representing times for which predicted survival probability is given

Details

The main purpose of this function is to be used within the shiny app for the purpose of creating predicted probability tables for user-inputted times. For this reason it is not expressly recommended to use this function outside the context of the shiny app, but it is still possible to do so if desired. The time or vector of times are inputted as characters due to the use of this function in the shiny app, where times are inputted as numbers separated by a comma

Value

Table of predicted probabilities, one column for each time, and one row for each curve

Examples

library(survival)
library(shinyCox)
# First colon is split into three treatment arms to compare predicted
# survival across arms
split_colon <- split(colon, colon$rx)

colon_arm1 <- split_colon$Obs
colon_arm2 <- split_colon$Lev
colon_arm3 <- split_colon$`Lev+5FU`

# One coxph model is fit for each treatment

colon1ph <- coxph(Surv(time, status) ~sex +  age + obstruct + nodes,
                  colon_arm1, x = TRUE, model = TRUE)

colon2ph <- coxph(Surv(time, status) ~ sex + age + obstruct + nodes,
                  colon_arm2, x = TRUE, model = TRUE)

colon3ph <- coxph(Surv(time, status) ~ sex + age + obstruct + nodes,
                  colon_arm3, x = TRUE, model = TRUE)

# Creating list of models
cox.fit.list <- vector("list", 3)
cox.fit.list[[1]] <- prep_coxfit(colon1ph)
cox.fit.list[[2]] <- prep_coxfit(colon2ph)
cox.fit.list[[3]] <- prep_coxfit(colon3ph)

# Creating new data row for predictions
new.data <- colon[1, ]
# Creating KM.hat object
n.models=length(cox.fit.list)
KM.hat=vector('list',n.models)
lp=rep(NA,n.models)
names(KM.hat)=names(cox.fit.list)
for (i in 1:n.models)
{
 km.hat=predict_one_coxfit(cox.fit.list[[i]],new.data)
 lp[i]=attr(km.hat,'lp')
 sfit=list(time=km.hat$time,surv=km.hat$surv)
 class(sfit)='survfit'
 KM.hat[[i]]=sfit
}

# Function takes KM.hat object and a time or vector of times
cox_times_table(KM.hat, fixTimes = "100")

---

Get confidence intervals for predicted survival curves

Description

Creates confidence levels for plotting predicted survival curves.

Usage

get_confint(p, se, conf.type, conf.int, ulimit = TRUE)

Arguments

p	Vector of survival probabilities
se	Vector of standard errors
conf.type	Type of confidence interval, includes 'plain', 'log', 'log-log', 'logit', and 'arcsin'.
conf.int	The level for two-sided confidence interval on the predicted survival curve, default is 0.95.
ulimit	Should upper bound be limited to 1, default is 'TRUE'

Value

list of length two, containing the lower and upper confidence levels

Examples

library(survival)
library(shinyCox)
colondeaths <- colon[colon$etype == 2, ]
split_colon <- split(colondeaths, colondeaths$rx)

colon_arm1 <- split_colon$Obs
colon1ph <- coxph(Surv(time, status) ~ factor(extent) + nodes + strata(surg)
                  + factor(differ),
                  colon_arm1,
                  x = TRUE, model = TRUE)

new.data = cbind.data.frame(`factor(extent)` = 3,
                         `surg` = "surg=0",`factor(differ)` = 2,`nodes` = 5)


coxfit = prep_coxfit(colon1ph)
coxlist = surv_pred_info(colon1ph)

for_ci = predict_se(coxlist, coxfit, new.data)

get_confint(for_ci$surv, for_ci$std.err, conf.int = 0.95,
            conf.type = "log-log")

---

Wrapper to create survival::coxph() object suitable for shine_coxph()

Description

Performs survival::coxph() with model = TRUE and x = TRUE as defaults. Checks that Cox model is appropriate for use with shine_coxph().

Usage

make_coxph(formula, data, ...)

Arguments

formula	a formula object, with the response on the left of a ~ operator, and the terms on the right. The response must be a survival object as returned by the Surv function.
data	a data.frame in which to interpret the variables named in the formula, or in the subset and the weights argument.
...	other arguments which will be passed to coxph(). Note that x = TRUE and model = TRUE are the default arguments (and required by shine_coxph()), you do not need to include them here.

Value

Object of class "coxph" representing the fit

Examples

library(survival)
ovarianph <- make_coxph(Surv(futime, fustat) ~ age + strata(rx),
data = ovarian)

---

Compute Cox-model predicted survival function

Description

Computes Cox-model predicted survival function for one new data row using coxfit list object created by prep_coxfit().

Usage

predict_one_coxfit(coxfit, newdata)

Arguments

coxfit	This is an object returned by prep_coxfit()
newdata	vector of new data

Value

data.frame of predicted survival probabilities over time, one column is time, one is probability

Note

This function's primary use is within the shiny app, where a coxph object is not available. It can be used outside of that context but that is the main purpose of this function, and why it only accepts the return object of prep_coxfit(). In the context of the shiny app, the new data is taken from user inputs.

Examples

# First, fit model using coxph
library(survival)
bladderph <- coxph(Surv(stop, event) ~ rx + number + size, bladder,
model = TRUE, x = TRUE)
# Use coxph object with function
bladderfit <- prep_coxfit(bladderph)
# Take first row of bladder as 'new data'
newdata <- bladder[1, ]
predictions <- predict_one_coxfit(bladderfit, newdata)

---

Creates predicted survival and standard errors for confidence intervals

Description

Adapted from parts of survival::survfit.coxph(), computes predictions for standard errors based on surv_pred_info() output and newdata from the shiny app.

Usage

predict_se(listsurv, coxfit, newdata)

Arguments

listsurv	Output from surv_pred_info() function
coxfit	coxfit object created for predictions. Used to find strata
newdata	Data used to make predicted standard errors

Value

a list of number of subjects for each curve, times at which the curve has a step, number at risk for each time, number of events at each time, number censored at each time (no event but exit risk set), estimated survival, cumulative hazard at each transition, and standard error of the cumulative hazard.

Examples

library(survival)
  library(shinyCox)
  colondeaths <- colon[colon$etype == 2, ]
  split_colon <- split(colondeaths, colondeaths$rx)

  colon_arm1 <- split_colon$Obs
  colon1ph <- coxph(Surv(time, status) ~
  factor(extent) + nodes + strata(surg) + factor(differ), colon_arm1, x =
  TRUE, model = TRUE)

  new.data = cbind.data.frame(`factor(extent)` = 3, `surg` =
  "surg=0",`factor(differ)` = 2,`nodes` = 5)


  coxfit = prep_coxfit(colon1ph)
  coxlist = surv_pred_info(colon1ph)

  predict_se(coxlist, coxfit, new.data)

---

Create simplified coxph() object for shiny app

Description

Simplifies coxph() output and checks that predictions match those of the original object

Usage

prep_coxfit(coxph.result, tol = 1e-07)

Arguments

coxph.result	Result returned by coxph()
tol	numerical tolerance for prediction differences, default is 1e-7

Value

list containing baseline survival estimates, linear predictor estimates, predictor types, coefficient estimates, mean and range of numeric predictors, levels of categorical predictors, strata if any, coxph() formula, table of hazard ratios, table with proportional hazard assumption results, number of subjects, and number of events

Examples

# First, fit model using coxph
library(survival)
bladderph <- coxph(Surv(stop, event) ~ rx + number + size, bladder,
model = TRUE, x = TRUE)
# Use coxph object with function
bladderfit <- prep_coxfit(bladderph)

---

Generates a shiny app for predictions from Cox model(s)

Description

Writes a shiny app to visualize predicted survival curves from one or multiple Cox models. One feature of this function is that the shiny app, once created, will not contain any identifiable data, containing only information necessary for predictions.

Usage

shine_coxph(..., app.dir = NULL, theme = c("default", "dashboard"))

Arguments

...	Arbitrary number of Cox proportional hazard models, created by survival::coxph() or make_coxph(), which automatically ensures the models are appropriate for shine_coxph()
app.dir	Directory where shiny app is created. Specifically, a sub-folder will be made containing the app.R file as well as the .Rdata file within app.dir. If no directory is provided, execution will pause and the user will be asked to confirm whether this sub-folder may be created in the working directory or to stop the function and provide an input app.dir.
theme	Theme of shiny app. ⁠"default⁠: default theme, requires only shiny "dashboard": requires "shinydashboard" and "DT" packages

Value

A list containing Cox model information along with the shiny app code. The app is written to the directory while the function is operating.

Notes

There are some requirements in order for this function to run without error: in your original survival::coxph() function or functions, model = TRUE and x = TRUE are required arguments (used to create the simplified "coxph" object). Currently, this function does not support penalized models (e.g., as created by ridge() and pspline()). Multiple strata terms and strata by covariate interaction terms in the formula are also not currently supported, but workarounds are available by respectively using a new strata factor variable encompassing all combinations of desired stratum variable levels. Use of time-varying covariates (e.g. with tt()) and multi-state models is not supported in our function. The package is not intended to support Fine-gray models by survival::finegray() creating Cox models, but doing so will not result in an error.

Guidelines

This package is intended to visualize and present predicted survival functions for fitted Cox models. In regards to formula notation, the variable names used are ultimately what will be displayed in the application. Using functions in the formula will work, but with multiple nested functions it will fail. Using "." notation is not currently supported. The na.action is inherited from the Cox models, with omit being the only option with support at this time. For these reasons, we recommend creating all final variables (including suitable transformations) with meaningful names prior to using survival::coxph().

Examples

library(survival)

  # Data used is from survival package, renamed for legibility
  names(leukemia)[names(leukemia) == "x"] <- "treatment"
  # Make Cox model, with x = TRUE and model = TRUE
  model1 <- coxph(Surv(time, status) ~ treatment,
  leukemia, x = TRUE, model = TRUE)

  # Use shine_coxph() to create shiny app in temporary directory
  shine_coxph("Model 1" = model1)

  # Get directory for shiny app (should be first, check file list if not)
  filedir <- list.files(tempdir())[1]

  # Run shiny app from temporary directory
  shiny::runApp(paste0(tempdir(), "/", filedir))
  # Remove app from directory once finished
  unlink(paste0(tempdir(),"/",filedir), recursive = TRUE)

---

Obtains information for standard errors of predictions

Description

Computes necessary information to calculate standard errors and confidence intervals in shiny app. This is adapted from parts of survival::survfit.coxph(). This function is meant to be used in conjunction with predict_se().

Usage

surv_pred_info(model, ctype, individual = FALSE, id, se.fit = TRUE, stype = 2)

Arguments

model	a coxph object
ctype	whether the cumulative hazard computation should have a correction for ties, 1=no, 2=yes.
individual	deprecated argument, replaced by id
id	optional variable name of subject identifiers. Not supported in app
se.fit	a logical value indicating whether standard errors should be computed. Default is TRUE for standard models, FALSE for multi-state (code not yet present for that case.)
stype	computation of the survival curve, 1=direct, 2=exponential of the cumulative hazard. Default is 2.

Value

A list of information needed for computing predicted standard errors.

Examples

library(survival)

colondeaths <- colon[colon$etype == 2, ]
split_colon <- split(colondeaths, colondeaths$rx)
colon_arm1 <- split_colon$Obs

colon1ph <- coxph(Surv(time, status) ~ factor(extent) + nodes + strata(surg)
                   + factor(differ),
                   colon_arm1,
                   x = TRUE, model = TRUE)
surv_pred_info(colon1ph)

---