Reputation: 23
I'd like to distinguish between statistical significance (OR = 1.0) and clinical significance (OR = 1.5) in my forest plot. I created this plot using the forestplot package, sample code below. Is adding a vertical line possible (while maintaining the line of no difference)?
library(forestplot)
test_data <- structure(list(
mean = c(NA, NA, 1, 0.5, 2),
lower = c(NA, NA, .5, .25, 1.5),
upper = c(NA, NA, 1.5, .75, 2.5)),
.Names = c("mean", "lower", "upper"),
row.names = c(NA, -5L),
class = "data.frame")
tabletext <- cbind(
c("", "Outcome", "Outcome 1", "Outcome 2", "Outcome 3"),
c("", "OR", "1 (0.5 - 1.5)", "0.5 (0.25 - 0.75)", "2.0 (1.5 - 2.5)"))
forestplot(tabletext,
test_data,
new_page = TRUE,
xlog = TRUE,
boxsize = .25
)
Upvotes: 2
Views: 4360
Reputation: 83
Is this what you were looking for?
forestplot(tabletext,
test_data,
new_page = TRUE,
xlog = TRUE,
grid = structure(c(log(1.5)),
gp = gpar(lty = 2, col = "#CCCCFF")),
zero = 1,
boxsize = .25)
Upvotes: 3
Reputation: 3514
A suboptimal (and not very elegant) solution could be: 1- creating an empty plot with no axis or labels, 2- then plot a vertical line (abline(v=1.5)
) and 3- call your forestplot
with new_page = F
.
Upvotes: 0