-
Notifications
You must be signed in to change notification settings - Fork 1
/
PostSequelae_Dec31.R
200 lines (144 loc) · 10.5 KB
/
PostSequelae_Dec31.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
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
setwd("~/GitHub/longhauler")
rm(list = ls())
library(dplyr)
library(ggplot2)
library(reshape2)
library(stringr)
library(Hmisc)
#====
#==== Read Phase 2.1. Data
#====
#======!!! Note >> I have added the information about sequence of encounters !!!
#======!!! It is necessary to have the information about encounters in the PatientClinicalCourse file as in the simulated data
#== added the simulated data files from Arianna
#PatientSummary<-read.csv("./SimulatedData/PatientSummarySim.csv")
#PatientClinicalCourse<-read.csv("./SimulatedData/PatientClinicalCourseSim.csv")
#PatientObservations<-read.csv("./SimulatedData/PatientObservationsSim.csv")
#== added the new data files from i2b2syn (synthetic data from OMOP project) andth UTF to read Windows files
PatientSummary<-read.csv("./SimulatedData/synPatientSummary.csv",fileEncoding="UTF-8-BOM")
PatientClinicalCourse<-read.csv("./SimulatedData/synPatientClinicalCourse.csv",fileEncoding="UTF-8-BOM")
PatientObservations<-read.csv("./SimulatedData/synPatientObservations.csv",fileEncoding="UTF-8-BOM")
#== added the new data files from ACT_Stage (real data from ACT project) andth UTF to read Windows files
#PatientSummary<-read.csv("./ACTData/actPatientSummary.csv",fileEncoding="UTF-8-BOM")
#PatientClinicalCourse<-read.csv("./ACTData/actPatientClinicalCourse.csv",fileEncoding="UTF-8-BOM")
#PatientObservations<-read.csv("./ACTData/actPatientObservations.csv",fileEncoding="UTF-8-BOM")
PostSequelaeList<-list()
#====
#==== Number of Encounters per patient
#====
NumberOfEncounters<-PatientClinicalCourse[PatientClinicalCourse$in_hospital==1,]%>%
group_by(patient_num) %>%
summarise(NumbEncounters = as.factor(n_distinct(encounter_seq)))
NumberOfEncountersTbl<-as.data.frame(table(NumberOfEncounters$NumbEncounters))
PostSequelaeList$NumberofEncounters<-ggplot(data=NumberOfEncountersTbl, aes(x=Var1, y=Freq)) +
geom_bar(stat="identity")+xlab("Number of Encounters")+ylab("Count of Patients")
#==== #====
#==== Duration of admissions, Days btw admissions
#==== #====
#====
#==== Duration of first Admission
#====
TotalDaysFirstEncounter<-PatientClinicalCourse[PatientClinicalCourse$encounter_seq==1,] %>%
group_by(patient_num) %>%
summarise(Duration = max(days_since_admission))
PostSequelaeList$Durationofthefirstadmission<-ggplot(data=TotalDaysFirstEncounter, aes(Duration)) +
geom_histogram(binwidth = 1) + xlab("Duration of the first admission")+ylab("")
PostSequelaeList$DurationofthefirstadmissionSummary<-summary(TotalDaysFirstEncounter$Duration)
#====
#==== Duration of first Admission for patients with/without a second readmission
#====
PtsReadmitted<-unique(PatientClinicalCourse[PatientClinicalCourse$encounter_seq>1,c("patient_num")])
TotalDaysFirstEncounter$Readmitted<-ifelse(TotalDaysFirstEncounter$patient_num %in% PtsReadmitted, "YES","NO")
PostSequelaeList$DurationofthefirstadmissionGroups<-ggplot(data=TotalDaysFirstEncounter, aes(x=Duration, fill=as.factor(Readmitted))) +
geom_histogram(binwidth = 1, alpha=0.5) + xlab("Duration of the first admission")+ylab("")
wilcox.test(Duration ~ Readmitted, data = TotalDaysFirstEncounter)
#====
#==== DIAGNOSIS AND PROCEDURES
#===
PatientClinicalCourseGrouped<-PatientClinicalCourse %>%
group_by(siteid,patient_num, encounter_seq) %>%
summarise(FirstDay = min(days_since_admission), LastDay=max(days_since_admission))
PatientObservations<-na.omit(PatientObservations)
# Add encounter information (observation admission day > 0)
PatientObservationsMerge<-merge(PatientObservations[PatientObservations$days_since_admission>=0,],PatientClinicalCourseGrouped, by=c("siteid","patient_num") , all = TRUE)
PatientObservationsMergeSelect<-PatientObservationsMerge[PatientObservationsMerge$days_since_admission>= PatientObservationsMerge$FirstDay &
PatientObservationsMerge$days_since_admission<= PatientObservationsMerge$LastDay ,]
# Create encounters (observation admission day < 0 )
PatientObservationsMergePreviousHosp<-PatientObservations[PatientObservations$days_since_admission<0,]
if (nrow(PatientObservationsMergePreviousHosp)>0) {
PatientObservationsMergePreviousHosp$encounter_seq<--1
#PatientObservationsMergePreviousHosp$ward<-"Pre-Covid" #== not sure why i had to comment this to work, but columns did not match in merge otherwise
PatientObservationsMergePreviousHosp$FirstDay<-PatientObservationsMergePreviousHosp$days_since_admission
PatientObservationsMergePreviousHosp$LastDay<--1
}
PatientObservationsEnctrs<-rbind(PatientObservationsMergeSelect,PatientObservationsMergePreviousHosp)
PatientObservationsEnctrs$FirstDay<-NULL
PatientObservationsEnctrs$LastDay<-NULL
#=========
#=========DIAGNOSIS - PheCode
#=========
#PheCodes<-read.csv("./PheCodesMapping/map_file_icd9_all.csv")
#=== added the new phecode mapping file provided from Victor
PheCodes<-read.csv("./PheCodesMapping/phecode_map_ICD910CM.csv")
load(file = "./PheCodesMapping/phecode_map_file.rda")
PheCodes<-merge(PheCodes, phecode_description, by=c("Phecode"),all.x = TRUE)
PatientObservationsEnctrs$concept_code<-str_trim(PatientObservationsEnctrs$concept_code, side = c("both"))
#PatientObservationsEnctrsDiagPheCodes<-merge(PatientObservationsEnctrs[PatientObservationsEnctrs$concept_type=="DIAG-ICD9",],
# PheCodes[,c("code","Description","Phecode")], by.x=c("concept_code"), by.y = ("code"), all.x = TRUE )
#== changed this to ICD10 concept type from the i2b2 files
PatientObservationsEnctrsDiagPheCodes<-merge(PatientObservationsEnctrs[PatientObservationsEnctrs$concept_type=="DIAG-ICD10",],
PheCodes[,c("code","Description","Phecode")], by.x=c("concept_code"), by.y = ("code"), all.x = TRUE )
PatientObservationsEnctrsDiagPheCodes$PhecodeLength<-nchar(PatientObservationsEnctrsDiagPheCodes$Phecode)
#=========Time Window (use days form admission or days from the first discharge)
PatientObservationsEnctrsDiagPheCodes<-na.omit(PatientObservationsEnctrsDiagPheCodes)
PatientObservationsEnctrsDiagPheCodes<-PatientObservationsEnctrsDiagPheCodes[PatientObservationsEnctrsDiagPheCodes$encounter_seq>0,] ## changed from 1
#== uncommented the fixed windows
#PatientObservationsEnctrsDiagPheCodes$timewindw<-cut(PatientObservationsEnctrsDiagPheCodes$days_since_admission,
# breaks=c(-1000,0, 30, 60, 90, 120,1000),
# right = FALSE, labels = c( "<0" , "1-30" , "31-60" , "61-90", "91-120" ,">120" ))
#PatientObservationsEnctrsDiagPheCodes$timewindw<-cut(PatientObservationsEnctrsDiagPheCodes$days_since_admission,
# breaks=c(-Inf,0, 30, 60, 90, 120,Inf),
# right = FALSE, labels = c( "<0" , "0-29" , "30-59" , "60-89", "90-119" ,"120-inf" ))
#PatientObservationsEnctrsDiagPheCodes$timewindw<-cut(PatientObservationsEnctrsDiagPheCodes$days_since_admission,
# breaks=c(-Inf, 0, 15, 30, 45, 60, 75, 90, 105, 120, Inf),
# right = FALSE, labels = c( "<0","0-14","15-29","30-44","45-59","60-74","75-89","90-104","105-119","120-inf"))
PatientObservationsEnctrsDiagPheCodes$timewindw<-cut(PatientObservationsEnctrsDiagPheCodes$days_since_admission,
breaks=c(seq(0,120, by = 5)),
right = FALSE)
#== commented the computed windows
#PatientObservationsEnctrsDiagPheCodes$timewindw <- cut2(PatientObservationsEnctrsDiagPheCodes$days_since_firstdischarge, g =5)
#=========Diagnosis in time Window - Bubble Graph
CountDiagnosisTW<-PatientObservationsEnctrsDiagPheCodes[PatientObservationsEnctrsDiagPheCodes$PhecodeLength>=3,] %>%
group_by(Description,timewindw) %>%
#summarise(Freq=n()) ## this was giving duplicates
summarise(Freq=n_distinct(patient_num))
CountDiagnosisNPts<-PatientObservationsEnctrsDiagPheCodes %>%
group_by(timewindw) %>%
summarise(Pts=n_distinct(patient_num))
CountDiagnosisDescr<-PatientObservationsEnctrsDiagPheCodes %>%
group_by(Description) %>%
summarise(Pts=n_distinct(patient_num))
CountDiagnosisDescr$perc<-(CountDiagnosisDescr$Pts/length(unique(CountDiagnosisDescr$Pts)))*100
CountDiagnosisTW$perc<-(CountDiagnosisTW$Freq/length(unique(CountDiagnosisTW$Freq)))*100 ################################# HERE
#For the Bubble Chart > select only Frequent Phecodes
#keep<-data.frame(Description=unique(CountDiagnosisDescr[CountDiagnosisDescr$perc>200,c("Description")])) #specify bubble cutoff here
#keep<-data.frame(Description=unique(CountDiagnosisTW[ (CountDiagnosisTW$Freq>24) & (CountDiagnosisTW$timewindw==c("60-89")), c("Description") ])) #specify bubble cutoff here
#keep<-data.frame(Description=unique(CountDiagnosisTW[ (CountDiagnosisTW$perc>20) & (CountDiagnosisTW$timewindw==c("60-74")), c("Description") ])) #specify bubble cutoff here ################################# HERE
#keep<-data.frame(Description=unique(CountDiagnosisTW[ (CountDiagnosisTW$perc>10) & (CountDiagnosisTW$timewindw==c("[20,25)")), c("Description") ])) #specify bubble cutoff here ################################# HERE
keep<-data.frame(Description=unique(CountDiagnosisDescr[CountDiagnosisDescr$perc>20,c("Description")])) #specify bubble cutoff here
CountDiagnosis<-merge(CountDiagnosisTW,CountDiagnosisNPts, by=c("timewindw"), all.x = TRUE)
CountDiagnosis$perc<-(CountDiagnosis$Freq/CountDiagnosis$Pts)*100
PostSequelaeList$CountDiagnosisTot<-CountDiagnosis
CountDiagnosis<-CountDiagnosis[CountDiagnosis$Description %in% keep$Description,]
CountDiagnosis$Description <- factor(CountDiagnosis$Description, levels = CountDiagnosisDescr$Description[order(CountDiagnosisDescr$perc)])
PostSequelaeList$DiagnosisBubblePlot<-ggplot(CountDiagnosis, aes(x=timewindw, y=Description, size = perc,color=Freq)) +
geom_point(alpha=0.7)+ scale_size(range = c(.1, 15), name="% Patients")+
scale_colour_gradient(low = "#4895ef", high = "#3a0ca3", name="# Patients")+
theme(text = element_text(size=15))+
xlab("Day since admission")+ylab("PheCode")
#xlab("Day since First Discharge")+ylab("Murphy PheCode")
#save(PostSequelaeList, file="./SimulatedData/PostSequelaeListSIMULATED.RData")
save(PostSequelaeList, file="./SimulatedData/synPostSequelaeList.RData")
#save(PostSequelaeList, file="./ACTData/actPostSequelaeList.RData")
# 371 61-90 Viral infection 370 342 108.1871345
PostSequelaeList