Natcourse/natcourse_1_data.R at master · jerudolph13/Natcourse · GitHub

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

###############################################################################################
#
# Project: Natural course paper
#
# Purpose: Read in EAGeR data and prep for use in analyses
#
# Author: Jacqueline Rudolph
#
# Last Update: 11 Nov 2020
#
###############################################################################################


packages <- c("tidyverse")
for (package in packages) {
  update.packages(package, ask=F)
  library(package, character.only=T)
}


# Read in data ------------------------------------------------------------

eager <- read.csv(file="../data/2020_11_06-eager_weekly.csv", header=TRUE) %>%
  select(id, week, GA, treatment, age, BMI, smoke, eligibility, studymed_b_imputed, bleed_d_imputed,
         nausea_d_imputed, outcome) %>%
  #Remove ID 417199 due to missing information
  filter(id != 417199)


# Manipulate data ---------------------------------------------------------

#Define exposure and covariates
eager <- eager %>%
    mutate(
      #baseline variables
      eligibility = as.numeric(eligibility == "new"),

      #t-varying compliance exposure
      compliance = as.numeric(studymed_b_imputed >= 5/7),
      compliance_last = if_else(week == 1, 0, as.double(lag(compliance))),

      #t-varying confounders
      bleed = as.numeric(bleed_d_imputed > 0),
      bleed_last = if_else(week == 1, 0, as.double(lag(bleed))),
      nausea = as.numeric(nausea_d_imputed > 0),
      nausea_last = if_else(week == 1, 0, as.double(lag(nausea))),

      #flag last record
      last = as.numeric(!duplicated(id, fromLast=T)))

#Define the outcome
  #This treats women with unknown GA but a post-pregnancy outcome as censored when we last see them
  #We could instead impute GA for these women
eager_last <- eager %>%
  filter(last==1) %>%
  mutate(c_week = week - GA) %>%
  select(id, c_week)
eager <- left_join(eager, eager_last, by="id") %>%
  mutate(conception = ifelse(GA==99, 0, as.numeric((week - c_week) >= 0)))

#Define drop out
eager$drop <- ifelse(eager$last==1 & eager$outcome=="withdrawal", 1,
              ifelse(eager$last==1 & eager$conception==0 & eager$outcome %in% c("live birth", "pregnancy loss"), 1, 0))

#Subset to records through event or censoring
eager <- filter(eager, week <= 26)

#Remove extra records after conception
eager <- eager %>%
  group_by(id) %>%
  mutate(cum_concep = cumsum(conception)) %>%
  filter(cum_concep <= 1) %>%
  ungroup(id) %>%
  select(-c(cum_concep, last))

#For t-fixed analysis, obtain each woman's average compliance through 26 weeks of follow-up
  #Compare average compliance to the 5/7 rule
  comp <- eager %>%
    group_by(id) %>%
    summarize(avg.comp=mean(studymed_b_imputed)) %>%
    mutate(compliance=as.numeric(avg.comp>=5/7)) %>%
    select(-avg.comp)


# Select final variables --------------------------------------------------

#For time-fixed analysis, keep relevant baseline variables, outcome, and drop out
eager_base <- eager %>%
    select(id, week, treatment, age, BMI, smoke, eligibility, conception, drop) %>%
    inner_join(comp, by="id")  #Merge average compliance

#For time-varying analysis, keep relevant baseline variables, compliance, bleeding, nausea, outcome, and drop out
eager_tvar <- select(eager, id, week, treatment, age, BMI, smoke, eligibility, compliance, compliance_last,
                     bleed, bleed_last, nausea, nausea_last, conception, drop)


###################################################################################################################################
# Output data

write.table(eager_base, file="../data/eager_base.txt", sep="\t", row.names=FALSE)
write.table(eager_tvar, file="../data/eager_tvar.txt", sep="\t", row.names=FALSE)