-
Notifications
You must be signed in to change notification settings - Fork 0
/
01-get_table.r
170 lines (144 loc) · 6.31 KB
/
01-get_table.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
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
# intro -------------------------------------------------------------------
source("libraries.R")
source("functions.R")
# connection --------------------------------------------------------------
source("00-connection.r")
# input tables or views ---------------------------------------------------
# sacar "fv" de "vitalsfirstdayfv" si no se usan trends que puse
t_firstday = c(
"icustays"
,"gcsfirstday", "heightfirstday", "labsfirstday", "rrtfirstday"
,"uofirstday", "ventfirstday", "vitalsfirstdayfv", "weightfirstday"
,"icustay_detail"
,"sofa","sapsii","oasis"
,"martin_sepsis"
)
for (t in t_firstday) assign(t, tbl(mimic,t))
# make query ---------------------------------------------------------------------
library(dbplyr)
stays = icustays %>%
# remove stays with transfers (seguro?)
dplyr::filter((first_wardid == last_wardid) & (first_careunit == last_careunit)) %>%
# only admissions to MICU (seguro? CAMBIA NROW POR MUCHO 30K vs 10K)
dplyr::filter(first_careunit=='MICU') %>%
# only metavision (or carevue) dbsource (seguro? usarlo si hace falta bajar el n) %>%
# dplyr::filter(dbsource=='carevue') %>%
# get stays details (data from admits and pats)
select(subject_id,hadm_id,icustay_id) %>%
inner_join(icustay_detail, by=c('subject_id', 'hadm_id', 'icustay_id')) %>%
# only first hospital stays AND first icu stays
dplyr::filter(first_hosp_stay==T & first_icu_stay==T) %>%
# only records without outlier ages (6% tienen age aprox 300!!!)
dplyr::filter(admission_age<110)
# remove multiple stays per admit
keep_admits = stays %>%
group_by(hadm_id) %>%
summarise(n=count(icustay_id)) %>%
dplyr::filter(n>=1 & n<=1) %>%
select(hadm_id)
# join stays-admits
stays = stays %>%
inner_join(keep_admits, by='hadm_id') %>%
mutate(
# create mortality in hospital
mortality_inhospital = ( (!is.na(dod)) & (admittime <= dod) & (dischtime >= dod) ),
# create mortality in unit
mortality_inunit = ( (!is.na(dod)) & (intime <= dod) & (outtime >= dod) ),
# create 30-day mortality
mortality_30 = ( (!is.na(dod)) & (DATE_PART('day',dod-admittime) < 30) ),
# create 7-day mortality
mortality_7 = ( (!is.na(dod)) & (DATE_PART('day',dod-admittime) < 7) )
)
# get data from first 24hs, scores and sepsis flag
base = stays %>%
inner_join(gcsfirstday, by=c('subject_id', 'hadm_id', 'icustay_id')) %>%
inner_join(heightfirstday, by=c('icustay_id')) %>%
inner_join(rrtfirstday, by=c('subject_id', 'hadm_id', 'icustay_id')) %>%
inner_join(uofirstday, by=c('subject_id', 'hadm_id', 'icustay_id')) %>%
inner_join(ventfirstday, by=c('subject_id', 'hadm_id', 'icustay_id')) %>%
inner_join(labsfirstday, by=c('subject_id', 'hadm_id', 'icustay_id')) %>%
# aca usa fv, version modificada que agrega trend
# glucose_min y glucose_max are also in labs -> agrega suffix
inner_join(vitalsfirstdayfv, by=c('subject_id', 'hadm_id', 'icustay_id'),
suffix=c("_labs","_vitals")) %>%
inner_join(weightfirstday, by=c('icustay_id')) %>%
inner_join(sofa, by=c('subject_id', 'hadm_id', 'icustay_id')) %>%
inner_join(sapsii, by=c('subject_id', 'hadm_id', 'icustay_id')) %>%
inner_join(oasis %>% select(subject_id, hadm_id, icustay_id
,mechvent ,electivesurgery
, oasis,oasis_prob)
, by=c('subject_id', 'hadm_id', 'icustay_id')) %>%
inner_join(martin_sepsis, by=c('subject_id', 'hadm_id'))
# collect query -----------------------------------------------------------
tabla = collect(base)
# mutate vars -------------------------------------------------------------
# not useful vars
k_remove = c(
"dod","admittime","dischtime","intime", "outtime","hospital_expire_flag"
,"hospstay_seq", "icustay_seq","first_hosp_stay", "first_icu_stay"
# usamos weight a secas que combina todas
,"weight_admit","weight_daily","weight_echoinhosp","weight_echoprehosp"
#vars de tabla sofa
,"respiration","coagulation","liver","cardiovascular.x","cns","renal.x"
#vars de tabla sapsii
,"age_score","hr_score","sysbp_score","temp_score","pao2fio2_score","uo_score"
,"bun_score","wbc_score","potassium_score","sodium_score","bicarbonate_score"
,"bilirubin_score","gcs_score","comorbidity_score","admissiontype_score"
#vars de tabla martin_sepsis
,"organ_failure","respiratory","cardiovascular.y","renal.y","hepatic"
,"hematologic","metabolic","neurologic"
)
tabla = tabla %>%
# remove not useful vars
select(-k_remove) %>%
# rename targets as y_*
rename(y_morh = mortality_inhospital,
y_moru = mortality_inunit,
y_mor30 = mortality_30,
y_mor7 = mortality_7,
y_seps = sepsis,
y_losh = los_hospital,
y_losu = los_icu) %>%
# rename scores as score_*
rename(
score_sofa = sofa
,score_sapsii = sapsii
,score_sapsiiprob = sapsii_prob
,score_oasis = oasis
,score_oasisprob = oasis_prob
) %>%
# rename ids as id_*
rename(id_subject = subject_id,
id_hadm = hadm_id,
id_icustay = icustay_id) %>%
# create id unico as 1:nrow
mutate(id_tot = 1:nrow(.)) %>%
# binarias as logical
mutate(
endotrachflag = as.logical(endotrachflag)
,vent = as.logical(vent)
,rrt = as.logical(rrt)
,mechvent = as.logical(mechvent)
,electivesurgery = as.logical(electivesurgery)
) %>%
mutate_at(vars(y_morh, y_moru, y_mor30, y_mor7, y_seps), as.logical)
# transform ethnicity
tabla$ethnicity = tabla$ethnicity %>%
stringr::str_replace_all(' OR ','/') %>%
stringr::str_split(" - ", simplify=T) %>% "["(,1) %>%
stringr::str_split("/", simplify=T) %>% "["(,1) %>%
{case_when(
. %in% c("UNKNOWN","PATIENT DECLINED TO ANSWER","UNABLE TO OBTAIN") ~ NA_character_,
. %in% c("SOUTH AMERICAN") ~ "HISPANIC",
. %in% c("PORTUGUESE") ~ "WHITE",
. %in% c("AMERICAN INDIAN","CARIBBEAN ISLAND","NATIVE HAWAIIAN","OTHER",
"MULTI RACE ETHNICITY","MIDDLE EASTERN") ~ "OTHER",
T ~ .)}
# glasgow coma scale (gcs) as ordered factor
tabla = tabla %>%
mutate_at(vars(gcsmotor, gcsverbal, gcseyes, mingcs),
function(x) forcats::fct_inseq(factor(x), ordered=T))
# save tabla --------------------------------------------------------------
saveRDS(tabla, "data/working/tabla.rds")
# end connection --------------------------------------------------------------
dbDisconnect(mimic)