generated from opensafely/research-template
/
cumulative_incidence.R
143 lines (121 loc) · 6 KB
/
cumulative_incidence.R
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
######################################
# This script
# - produces a cumulative incidence plot of follow-up-time
# - saves plot as svg
######################################
# Preliminaries ----
## Import libraries
library('tidyverse')
library('lubridate')
library('reshape2')
library('here')
library('survival')
library('survminer')
## Custom function
fct_case_when <- function(...) {
# uses dplyr::case_when but converts the output to a factor,
# with factors ordered as they appear in the case_when's ... argument
args <- as.list(match.call())
levels <- sapply(args[-1], function(f) f[[3]]) # extract RHS of formula
levels <- levels[!is.na(levels)]
factor(dplyr::case_when(...), levels=levels)
}
## Import data
data_processed <- read_rds(here::here("output", "data", "data_processed.rds"))
data_trial <- read_rds(here::here("output", "data", "data_processed_trial.rds")) %>%
filter(time_to_positive_test >= 0)
## Format groups
data_processed <- data_processed %>%
mutate(group = ifelse(care_home_65plus == 1, 1, NA),
group = ifelse(is.na(group) & ageband == 3, 2, group),
group = ifelse(is.na(group) & hscworker == 1, 3, group),
group = ifelse(is.na(group) & ageband == 2, 4, group),
group = ifelse(is.na(group) & shielded == 1, 5, group),
group = ifelse(is.na(group) & age >=50 & age <70, 6, group),
group = ifelse(is.na(group), 7, group),
group = fct_case_when(
group == "1" ~ "Care home (priority group 1)",
group == "2" ~ "80+ (priority group 2)",
group == "3" ~ "Health/care workers (priority groups 1-2)",
group == "4" ~ "70-79 (priority groups 3-4)",
group == "5" ~ "Shielding (age 16-69) (priority group 4)",
group == "6" ~ "50-69 (priority groups 5-9)",
group == "7" ~ "Others not in the above groups (under 50)",
#TRUE ~ "Unknown"
TRUE ~ NA_character_)) %>%
filter(time_to_positive_test >= 0)
# Plot ----
threshold <- 7
## Data
surv_data_all <- survfit(Surv(time = time_to_positive_test, event = covid_positive_test) ~ 1,
data = data_processed) %>%
broom::tidy() %>%
filter(estimate > 0) %>%
mutate(
estimate = pmin(1,plyr::round_any(estimate, threshold/max(n.risk)), na.rm=TRUE),
conf.low = pmin(1, plyr::round_any(conf.low, threshold/max(n.risk)), na.rm=TRUE),
conf.high = pmin(1, plyr::round_any(conf.high, threshold/max(n.risk)), na.rm=TRUE),
cum.in = 1 - estimate,
lci = 1- conf.high,
uci = 1 - conf.low,
group = "All"
) %>%
select(group, time, cum.in, lci, uci)
surv_data_groups <- survfit(Surv(time = time_to_positive_test, event = covid_positive_test) ~ group,
data = data_processed) %>%
broom::tidy() %>%
filter(estimate > 0) %>%
mutate(
estimate = pmin(1,plyr::round_any(estimate, threshold/max(n.risk)), na.rm=TRUE),
conf.low = pmin(1, plyr::round_any(conf.low, threshold/max(n.risk)), na.rm=TRUE),
conf.high = pmin(1, plyr::round_any(conf.high, threshold/max(n.risk)), na.rm=TRUE),
cum.in = 1 - estimate,
lci = 1- conf.high,
uci = 1 - conf.low
) %>%
mutate(group = gsub(".*=","", strata),
group = factor(group, levels = c("All",
"Care home (priority group 1)",
"80+ (priority group 2)",
"Health/care workers (priority groups 1-2)",
"70-79 (priority groups 3-4)",
"Shielding (age 16-69) (priority group 4)",
"50-69 (priority groups 5-9)",
"Others not in the above groups (under 50)"))) %>%
select(group, time, cum.in, lci, uci)
surv_data_trial <- survfit(Surv(time = time_to_positive_test, event = covid_positive_test) ~ 1,
data = data_trial) %>%
broom::tidy() %>%
filter(estimate > 0) %>%
mutate(
estimate = pmin(1,plyr::round_any(estimate, threshold/max(n.risk)), na.rm=TRUE),
conf.low = pmin(1, plyr::round_any(conf.low, threshold/max(n.risk)), na.rm=TRUE),
conf.high = pmin(1, plyr::round_any(conf.high, threshold/max(n.risk)), na.rm=TRUE),
cum.in = 1 - estimate,
lci = 1- conf.high,
uci = 1 - conf.low,
group = "Trial"
) %>%
select(group, time, cum.in, lci, uci)
surv_data_risk_table <- ggsurvplot(survfit(Surv(time = time_to_positive_test, event = covid_positive_test) ~ group,
data = data_processed), risk.table = TRUE)$data.survtable %>%
select(group, time, n.risk) %>%
mutate(`n.risk` = ifelse(`n.risk` < 8, "<8", `n.risk`),
group = factor(group, levels = c("Care home (priority group 1)",
"80+ (priority group 2)",
"Health/care workers (priority groups 1-2)",
"70-79 (priority groups 3-4)",
"Shielding (age 16-69) (priority group 4)",
"50-69 (priority groups 5-9)",
"Others not in the above groups (under 50)")))
surv_data_risk_table_trial <- ggsurvplot(survfit(Surv(time = time_to_positive_test, event = covid_positive_test) ~ 1,
data = data_trial), risk.table = TRUE)$data.survtable %>%
select(time, n.risk) %>%
mutate(`n.risk` = ifelse(`n.risk` < 8, "<8", `n.risk`),
group = "Trial")
## Save data behind plots
write_csv(surv_data_all, here::here("output", "data", "surv_data_all.csv"))
write_csv(surv_data_groups, here::here("output", "data", "surv_data_groups.csv"))
write_csv(surv_data_trial, here::here("output", "data", "surv_data_trial.csv"))
write_csv(surv_data_risk_table, here::here("output", "data", "surv_data_risk_table.csv"))
write_csv(surv_data_risk_table_trial, here::here("output", "data", "surv_data_risk_table_trial.csv"))