Skip to contents

Survival Difference (Life-Gained) Plot

Source: R/hazard-plot.R
survival_difference_plot.Rd

Superseded.

survival_difference_plot() has been superseded by the S3 constructor hv_survival_difference() plus plot.hv_survival_difference().

Plots the difference in survival between two groups over time, with an optional confidence band. Covers tp.hp.dead.life-gained.sas and the survival-difference component of tp.hp.numtreat.survdiff.matched.sas.

SAS context: The HAZDIFL macro in tp.hp.dead.life-gained.sas bootstraps the difference S_2(t) - S_1(t) (or S_1(t) - S_2(t)) and stores the result in a diffout dataset. Export that dataset and pass it here with the appropriate column names.

Usage

survival_difference_plot(
  diff_data,
  x_col = "time",
  estimate_col = "difference",
  lower_col = NULL,
  upper_col = NULL,
  group_col = NULL,
  ci_alpha = 0.2,
  line_width = 1
)

Arguments

diff_data

Data frame of pre-computed survival differences. See sample_survival_difference_data().

x_col

Name of the time column. Default "time".

estimate_col

Name of the difference column. Default "difference".

lower_col

Name of the lower CI column, or NULL. Default NULL.

upper_col

Name of the upper CI column, or NULL. Default NULL.

group_col

Name of a grouping column for multiple comparisons, or NULL. Default NULL.

ci_alpha

Transparency of the CI ribbon. Default 0.20.

line_width

Line width. Default 1.0.

Value

A ggplot2::ggplot() object.

References

SAS templates: tp.hp.dead.life-gained.sas, tp.hp.numtreat.survdiff.matched.sas.

See also

sample_survival_difference_data(), nnt_plot(), hazard_plot()

Examples

library(ggplot2)

diff_dat <- sample_survival_difference_data(
  n = 500, time_max = 10,
  groups = c("Control" = 1.0, "Treatment" = 0.70)
)

# --- (1) Single comparison: TF-TAVR vs TA-TAVR (tp.hp.dead.life-gained) --
survival_difference_plot(
  diff_dat,
  lower_col = "diff_lower",
  upper_col = "diff_upper"
) +
  scale_colour_manual(values = c("steelblue"), guide = "none") +
  scale_fill_manual(values = c("steelblue"), guide = "none") +
  geom_hline(yintercept = 0, linetype = "dashed", colour = "grey50") +
  scale_x_continuous(limits = c(0, 10), breaks = 0:10) +
  scale_y_continuous(limits = c(-5, 30),
                     labels = function(x) paste0(x, "%")) +
  labs(x = "Years", y = "Survival Difference (%)") +
  hv_theme("poster")


# --- (2) Multiple treatment comparisons ----------------------------------
# Simulate three comparisons and combine (each row = one comparison)
d1 <- sample_survival_difference_data(
  groups = c("Medical Mgmt" = 1.0, "TF-TAVR" = 0.70), seed = 1
)
d1$comparison <- "TF-TAVR vs Medical Mgmt"

d2 <- sample_survival_difference_data(
  groups = c("TA-TAVR" = 0.90, "TF-TAVR" = 0.70), seed = 2
)
d2$comparison <- "TF-TAVR vs TA-TAVR"

d3 <- sample_survival_difference_data(
  groups = c("AVR" = 0.80, "TF-TAVR" = 0.70), seed = 3
)
d3$comparison <- "TF-TAVR vs AVR"

dall <- rbind(d1, d2, d3)

survival_difference_plot(dall, group_col = "comparison") +
  scale_colour_brewer(palette = "Set1", name = NULL) +
  scale_fill_brewer(palette = "Set1", guide = "none") +
  geom_hline(yintercept = 0, linetype = "dashed", colour = "grey50") +
  scale_x_continuous(limits = c(0, 10), breaks = 0:10) +
  labs(x = "Years", y = "Survival Difference (%)") +
  hv_theme("poster")