R/ImportFunctionsModular.R
In ISARICDataPlatform/CovidClinicalDataProcessor: Data Import and Processing for ISARIC COVID-19 Hospital Data
Defines functions
process.outcome.data
process.laboratory.data
process.vital.sign.data
process.treatment.icu.data
process.common.treatment.data
process.treatment.data
process.ICU.data
process.pregnancy.data
process.symptom.data
process.comorbidity.data
import.symptom.and.comorbidity.data
import.microb.data
import.demographic.data
Documented in
import.demographic.data
import.microb.data
import.symptom.and.comorbidity.data
process.common.treatment.data
process.comorbidity.data
process.ICU.data
process.laboratory.data
process.outcome.data
process.pregnancy.data
process.symptom.data
process.treatment.data
process.treatment.icu.data
process.vital.sign.data
#' Import demographic data
#' @param file.name Path of the demographics data file (CDISC format)
#' @param dtplyr.step Return the output as \code{dtplyr_step} to avoid unnecessary future calls to \code{as_tibble} or \code{as.data.table}
#' @import dplyr tibble
#' @return Formatted demographic data as a tibble or \code{dtplyr_step}
#' @export import.demographic.data
import.demographic.data <- function(file.name, dtplyr.step = FALSE){
wdi_dat <- WDI(indicator = c("NY.GDP.PCAP.KD", "SP.DYN.LE00.IN", "SP.DYN.IMRT.IN"),
start = 2020, end = 2020, extra = TRUE)%>%
filter(region != "Aggregates")%>%
select("Alpha_3"=iso3c,income,region)
country.lookup <- ISOcodes::ISO_3166_1 %>% as_tibble%>%
mutate(Name=case_when(!is.na(Common_name)~Common_name,
Name=="Lao People's Democratic Republic"~"Lao PDR",
TRUE~Name))%>%select(Alpha_3, Name)%>%left_join(wdi_dat)
#regexp <- "[[:digit:]]+" # process string
out <- dm %>%
###delete patients duplicates
group_by(usubjid) %>%
mutate(count=1)%>%
mutate(n = sum(count)) %>%
filter(n == 1) %>%
ungroup()%>%
rename(date_admit=rfstdtc)%>%
as.data.frame()%>%
mutate(age_d=case_when(ageu=="MONTHS"~12,
ageu=="YEARS" ~ 1,
ageu=="DAYS" ~ 365.25,
TRUE~ NA_real_))%>%
mutate(age2=age/age_d)%>%
select(-(age))%>%
rename(age=age2)%>%
mutate(age=replace(age,age<0,NA))%>%
mutate(ethnic = iconv(ethnic, to ="ASCII//TRANSLIT") %>% tolower()) %>%
mutate(ethnic = str_remove_all(ethnic, "\\s*\\([^)]*\\)")) %>%
mutate(ethnic = str_replace_all(ethnic, " - ", "_")) %>%
mutate(ethnic = str_replace_all(ethnic, "-", "_")) %>%
mutate(ethnic = str_replace_all(ethnic, "/| / ", "_")) %>%
mutate(ethnic = str_replace_all(ethnic, " ", "_")) %>%
mutate(ethnic = str_replace_all(ethnic, ",", "_")) %>%
mutate(ethnic = replace(ethnic, ethnic == "n_a" | ethnic == "na" | ethnic == "", NA))%>%
#mutate(studyid=substr(usubjid,1, 7))%>%
mutate(siteid_final=invid)%>%
mutate(siteid_final=case_when(invid=="00741cca_network"~ substr(subjid,1, 12),
invid=="227inserm"~ sub("\\-.*", "",subjid),
invid=="00689us_nhlbi_peta"~ sub("\\-.*", "",subjid),
invid==""~studyid,
studyid=="CVPRQTA"~"CVPRQTA",
TRUE~invid))%>%
mutate(sex = case_when(sex == "M" ~ "Male",
sex == "F" ~ "Female",
TRUE ~ NA_character_))%>%
mutate(date_admit=substr(date_admit,1, 10))%>%
mutate(date_admit=as_date(date_admit))%>%
mutate(date_admit=replace(date_admit,date_admit >date_pull,NA))%>%
select(usubjid, studyid, siteid_final, date_admit, age, sex, ethnic, country)
site_id_country<-out%>%
mutate(country=ifelse(siteid_final=="321cub_erasme__bru","BEL",country))%>%
mutate(country=ifelse(siteid_final=="435civil_hospital","BEL",country))%>%
mutate(country=ifelse(siteid_final=="00657hospital_de_c","PRT",country))%>%
mutate(country=ifelse(siteid_final==" 00727clinica_unive","COL",country))%>%
mutate(country=ifelse(siteid_final=="00580netcare_unita","ITA",country))%>%
mutate(country=ifelse(siteid_final=="00835consortium_im","POL",country))%>%
mutate(country=ifelse(siteid_final=="00831nicvd_dhaka","BGD",country))%>%
mutate(country = replace(country, country == "", NA)) %>%
left_join(country.lookup, by = c("country" = "Alpha_3")) %>%
select(-country) %>%
rename(country = Name) %>%
filter(!is.na(country))%>%
arrange(desc(country, income, siteid_final))%>%
distinct(siteid_final, .keep_all =T)%>%
select(siteid_final, 'country_2'=country,income, region)
out<-out%>%
left_join(site_id_country)%>%
mutate(country=country_2)%>%select(-c(country_2))
if(dtplyr.step){
return(out)
} else {
return(out %>% as_tibble())
}
}
#' Import microb data
#' @param file.name Path of the microbio data file (CDISC format)
#' @param dtplyr.step Return the output as \code{dtplyr_step} to avoid unnecessary future calls to \code{as_tibble} or \code{as.data.table}
#' @import dplyr tibble
#' @return Formatted demographic data as a tibble or \code{dtplyr_step}
#' @export import.microb.data
import.microb.data <- function(file.name, dtplyr.step = FALSE){
#mb<-shared.data.import(file.name, dtplyr.step = TRUE)
mb<-mb
detection<- mb%>%
#select(usubjid,mbtestcd,mbtest,mbtstdtl,mbcat,mbstresc,mbspec,mbloc,mbmethod)%>%
filter(mbtstdtl=="DETECTION")%>%
filter(mbtestcd=="CRONAVIR"|mbtestcd=="SARSCOV2")%>%
mutate(mbstresc = case_when(mbstresc == "NO" ~ "NEGATIVE",
mbstresc == "NEGATIVE" ~ "NEGATIVE",
mbstresc == "POSITIVE" ~ "POSITIVE",
TRUE ~ NA_character_)) %>%
mutate(mbtestcd = paste0("cov_det_",mbtestcd)%>% tolower%>%str_replace_all(" ", "_")) %>%
arrange(desc(mbstresc))%>%
distinct(usubjid, mbtestcd, .keep_all =T)%>%
as.data.table() %>%
dt_pivot_wider(id_cols = usubjid, names_from = mbtestcd, values_from = mbstresc) %>%
as.data.frame()
identification<-mb%>%
filter(mbtstdtl=="IDENTIFICATION")%>%
distinct(usubjid, mbstresc, .keep_all =T)%>%
filter(mbstresc=="SEVERE ACUTE RESPIRATORY SYNDROME CORONAVIRUS 2"|
mbstresc=="CORONAVIRIDAE")%>%
mutate(mbstresc=replace(mbstresc,mbstresc=="SEVERE ACUTE RESPIRATORY SYNDROME CORONAVIRUS 2","SARSCOV2"))%>%
mutate(mbstresc=replace(mbstresc,mbstresc=="SEVERE ACUTE RESPIRATORY SYNDROME-RELATED CORONAVIRUS","SARSCOV2"))%>%
mutate(mbstresc=replace(mbstresc,mbstresc=="CORONAVIRIDAE","CRONAVIR"))%>%
mutate(result="POSITIVE")%>%
mutate(mbstresc = paste0("cov_id_",mbstresc)%>%
tolower%>%
str_replace_all(" ", "_")) %>%
as.data.table() %>%
dt_pivot_wider(id_cols = usubjid, names_from = mbstresc, values_from = result) %>%
as.data.frame()
out<-full_join(detection,identification)%>%
mutate(cov_det_id="NEGATIVE")%>%
mutate(cov_det_id=case_when(cov_det_cronavir=="POSITIVE"|
cov_det_sarscov2=="POSITIVE"|
cov_id_cronavir=="POSITIVE"|
cov_id_sarscov2=="POSITIVE"~
"POSITIVE",
is.na(cov_det_cronavir)&
is.na(cov_det_sarscov2)&
is.na(cov_id_cronavir)&
is.na(cov_id_sarscov2)~
NA_character_,
TRUE~cov_det_id))
if(dtplyr.step){
return(out)
} else {
return(out %>% as_tibble())
}
}
#' Import data on symptoms and comorbidities
#' @param file.name Path of the symptoms data file (CDISC format)
#' @param dtplyr.step Return the output as \code{dtplyr_step} to avoid unnecessary future calls to \code{as_tibble} or \code{as.data.table}
#' @import dplyr tibble stringr
#' @return Formatted comorbidity and symptom data as a tibble or \code{dtplyr_step}
#' @export import.symptom.and.comorbidity.data
import.symptom.and.comorbidity.data <- function(file.name, dtplyr.step = TRUE){
#out <- shared.data.import(file.name,
# dtplyr.step = TRUE,
# immutable = TRUE) %>% # this will often by used twice, so should be immutable
out<-sa%>%
#imp_sa<-sa%>%
select(usubjid, saterm, sacat, samodify, sapresp, saoccur, sastdtc) %>%
mutate(sacat=case_when(
saterm=="CLINICALLY-DIAGNOSED COVID-19"~"CLINICALLY-DIAGNOSED COVID-19",
TRUE~sacat))%>%
filter(
sacat=="MEDICAL HISTORY"|
sacat=="SIGNS AND SYMPTOMS AT HOSPITAL ADMISSION"
|sacat=="CLINICALLY-DIAGNOSED COVID-19"
)%>%
#mutate(sacat=replace(sacat,saterm=="MALNUTRITION","MEDICAL HISTORY"))%>%#temporary correction
#filter( sapresp=="Y") %>%
mutate(saoccur = case_when(saoccur == "Y" ~ TRUE,
saoccur == "N" ~ FALSE,
TRUE ~ NA)) %>%
filter(!is.na(saoccur)) %>%
mutate(saterm=toupper(saterm))%>%
mutate(saterm=case_when(samodify!=""|is.na(samodify)~samodify,
TRUE ~ saterm))%>%#to add
mutate(saterm=case_when(saterm%like%'CARDIAC ARRHYTHMIA'~'CHRONIC CARDIAC DISEASE',
saterm%like%'CARDIAC DISEASE'~'CHRONIC CARDIAC DISEASE',
saterm%like%'CHORNIC CARDIAC DISEASE'~'CHRONIC CARDIAC DISEASE',
saterm%like%'CHRONIC HEART DISEASE'~'CHRONIC CARDIAC DISEASE',
saterm%like%'CONGENITAL CA'~'CHRONIC CARDIAC DISEASE',
saterm%like%'CONGENTIAL CARDIOPATHY'~'CHRONIC CARDIAC DISEASE',
saterm=='CORONARY DISEASE'~'CHRONIC CARDIAC DISEASE',
saterm=='HEART FAILURE'~'CHRONIC CARDIAC DISEASE',
saterm=='OROVALVA DISEASE'~'CHRONIC CARDIAC DISEASE',
saterm=='RHEUMATIC HEART DISEASE'~'CHRONIC CARDIAC DISEASE',
saterm=='VALVULAR HEART DISEASE'~'CHRONIC CARDIAC DISEASE',
saterm=='CONGESTIVE HEART FAILURE'~'CHRONIC CARDIAC DISEASE',
saterm=='CORONARY ARTERY DISEASE'~'CHRONIC CARDIAC DISEASE',
TRUE~saterm))%>%
mutate(saterm=case_when(saterm=='CHRONIC DIALYSIS'~'CHRONIC KIDNEY DISEASE',
saterm%like%'DEPRESSION'~'PSYCHIATRIC CONDITION',
saterm%like%'PSYCHOSIS'~'PSYCHIATRIC CONDITION',
saterm%like%'DYSLIPIDEMIA'~'CHRONIC METABOLIC DISORDER',
saterm%like%'HYPOTHYROIDISM'~'CHRONIC ENDOCRINE DISORDER NON DIABITES',
saterm%like%'HEPATITIS'~'LIVER DISEASE',
saterm%like%'MARASUMAS'~'MALNUTRITION',
saterm=='SAM UNDEFINED'~'MALNUTRITION',
saterm=='MIXED MARASMIC-KWASH'~'MALNUTRITION',
saterm=='OSA/ HOME CPAP/BI-PAP USE'~'OBESITY',
saterm=='PAOD'~'OTHER COMORBIDITIES',
saterm=='PEPTIC ULCER DISEASE EXCLUDING BLEEDING'~'OTHER COMORBIDITIES',
saterm=='PARALYSIS'~'CHRONIC NEUROLOGICAL DISORDER',
saterm=='STROKE OR OTHER NEUROLOGICAL DISORDERS'~'CHRONIC NEUROLOGICAL DISORDER',
saterm=='PULMONARY CIRCULATION DISORDER'~'CHRONIC CARDIAC DISEASE',
saterm%like%'ARRHYTHMIA'~'CHRONIC CARDIAC DISEASE',
saterm=='SUBSTANCE USE DISORDER'~'SUBSTANCE MISUSE',
saterm=='VENOUS THROMBOEMBOLISM- DVT/PE'~'THROMBOLIC DISORDERS',
saterm=='CHILLS/RIGORS'~'RIGOR OR SWEATING',
saterm=='NIGHT SWEAT'~'RIGOR OR SWEATING',
saterm=='CONGESTION/RHINORRHEA'~'RUNNY NOSE',
saterm=='CONJUNCTIVAL CONGESTION'~'UPPER RESPIRATORY TRACT SYMPTOMS',
saterm=='SNEEZING'~'UPPER RESPIRATORY TRACT SYMPTOMS',
saterm=='DELIRIUM / ENCEPHALOPATHY'~'ALTERED CONSCIOUSNESS CONFUSION',
saterm=='DIZZINESS/LIGHTHEADEDNESS'~'OTHER SIGNS AND SYMPTOMS',
saterm=='GASTROGASTROINTESTINAL HEMORRHAGE'~'OTHER SIGNS AND SYMPTOMS',
TRUE~saterm))%>%
mutate(saterm=case_when(saterm%like%'TUBERCULOSIS'~'TUBERCULOSIS',
saterm%like%'MALIGNANCY'~'MALIGNANT NEOPLASM',
saterm%like%'SPECIFIC CANCERS'~'MALIGNANT NEOPLASM',
saterm%like%'SOLID TUMOR'~'MALIGNANT NEOPLASM',
saterm%like%'METASTATIC CANCER'~'MALIGNANT NEOPLASM',
saterm=='SORE THROAT/THROAT PAIN'~'SORE THROAT',
saterm=='COAGULOPATHY'~'CHRONIC HEMATOLOGIC DISEASE',
saterm=='DYSLIPIDEMIA/HYPERLIPIDEMIA'~'CHRONIC HEMATOLOGIC DISEASE',
saterm=='IRON DEFICIENCY ANEMIA'~'CHRONIC HEMATOLOGIC DISEASE',
saterm=='BLOOD LOSS ANEMIA'~'CHRONIC HEMATOLOGIC DISEASE',
saterm=='CHRONIC HEMATOLOGICAL DISEASE'~'CHRONIC HEMATOLOGIC DISEASE',
saterm=='CHRONIC LIVER DISEASE'~'LIVER DISEASE',
saterm%like%'ACUTE LIVER'~'LIVER DISEASE',
saterm%like%'CHRONIC RENAL FAILURE'~'CHRONIC KIDNEY DISEASE',
saterm%like%'CHRONIC LUNG DISEASE'~'CHRONIC PULMONARY DISEASE',
saterm%like%'CHROMIC PULMONARY DISEASE'~'CHRONIC PULMONARY DISEASE',
TRUE~saterm))%>%
mutate(saterm=case_when(saterm%like%'RHEMATOLOGICAL DISORDER'~'rheumatologic disorder',
saterm%like%'CHRONIC NEUROLOGICAL'~'CHRONIC NEUROLOGICAL DISORDER',
saterm%like%'CURRENT SMOK'~'SMOKING',
saterm%like%'DIABETES'~'DIABETES',
saterm=='HISTORY OF PERIPHERAL OR CARDIAC REVASCULARIZATION'~'HISTORY OF PERIPHERAL OR CARDIAC REVASCULARIZATION',
saterm=='HISTORY OF SMOKING'~'SMOKING',
saterm%like%'SMOKING'~'SMOKING',
saterm%like%'HIV'~'AIDS/HIV',
saterm%like%'LIVER DISEASE'~'LIVER DISEASE',
saterm%like%'OTHER RELEVANT RISK'~'OTHER COMORBIDITIES',
saterm=='OTHER RISK FACTOR'~'OTHER COMORBIDITIES',
saterm%like%'RHEUMATOLOGICAL DISORD'~'RHEUMATOLOGIC DISORDER',
saterm=='SMOKER'~'SMOKING',
saterm=='SMOKER - CURRENT'~'SMOKING',
saterm=='SMOKER - FORMER'~'SMOKING - FORMER',
saterm=='FEEDING INTOLERANCE (PAEDIATRICS)'~'ANOREXIA',
saterm=='REFUSING TO EAT OR DRINK/HISTORY OF POOR ORAL INTAKE'~'ANOREXIA',
saterm%like%'ANOREXIA'~'ANOREXIA',
saterm=='ANOREXIA - LOSS OF APPETITE'~'ANOREXIA',
saterm=='CHEST PAIN/TIGHTNESS'~'CHEST PAIN',
saterm=='SWOLLEN NECK GLANDS/LYMPHADENOPATHY'~'LYMPHADENOPATHY',
TRUE~saterm))%>%
mutate(saterm=case_when(saterm%like%'COUGH'~'COUGH',
saterm%like%'COUTH'~'COUGH',
saterm=='HEMOPTYSIS'~'COUGH',
saterm=='DIARRHEA'~'DIARRHOEA',
saterm=='CONJUNCTIVAL CONGESTION '~'CONJUNCTIVITIS',
saterm%like%'FEVER'~'HISTORY OF FEVER',
saterm=='SEIZURE'~'SEIZURES',
saterm%like%'TRANSPLANT'~'TRANSPLANTATION',
saterm%like%'ANOSMIA'~'LOSS OF SMELL',
saterm%like%'AGEUSIA'~'LOSS OF TASTE',
saterm=="LOSS OF TASTE OR LOSS OF SMELL"~'LOSS OF SMELL/TASTE',
saterm=="NAUSEA/VOMITING"~'VOMITING/NAUSEA',
saterm%like%'MYALGIA OR FATIGUE'~'MUSCLE ACHES/JOINT PAIN',
saterm%like%'JOINT PAIN'~'MUSCLE ACHES/JOINT PAIN',
saterm%like%'MUSCLE ACHES'~'MUSCLE ACHES/JOINT PAIN',
saterm=='OTHER SIGN OR SYMPTOM'~'OTHER SIGNS AND SYMPTOMS',
saterm=='LOWER CHEST WALL INDRAWING'~'SHORTNESS OF BREATH',
saterm%like%'DEHYDRATION'~'SEVERE DEHYDRATION',
saterm%like%'RASH'~'SKIN RASH',
saterm=='EARPAIN'~'EAR PAIN',
TRUE ~ saterm ))%>%
mutate(saterm = iconv(saterm, to ="ASCII//TRANSLIT") %>% tolower()) %>%
mutate(saterm = str_remove_all(saterm, "\\s*\\([^)]*\\)")) %>%
mutate(saterm = str_replace_all(saterm, " - ", "_")) %>%
mutate(saterm = str_replace_all(saterm, "/| / ", "_")) %>%
mutate(saterm = str_replace_all(saterm, " ", "_")) %>%
arrange(desc(saoccur))%>%
distinct(usubjid,saterm, .keep_all =T)
if(dtplyr.step){
return(out)
} else {
return(out %>% as_tibble())
}
}
#' Process data on comorbidities
#' @param input Either the path of the symptoms/comorbidities data file (CDISC format) or output of \code{import.symptom.and.comorbidity.data}
#' @param dtplyr.step Return the output as \code{dtplyr_step} to avoid unnecessary future calls to \code{as_tibble} or \code{as.data.table}
#' @import dplyr tibble stringr tidyfast
#' @importFrom data.table as.data.table
#' @importFrom glue glue
#' @return Formatted comorbidity data as a tibble or \code{dtplyr_step}
#' @export process.comorbidity.data
process.comorbidity.data <- function(input, minimum=100, dtplyr.step = FALSE){
comorbid <- imp_sa_com %>%
filter(sacat=="MEDICAL HISTORY") %>%
filter(!is.na(sacat))%>%
filter(!is.na(saterm))%>%
arrange(desc(saoccur))%>%
group_by(saterm) %>%
arrange(desc(saoccur))%>%
mutate(n = sum(!is.na(saoccur))) %>%
filter(n >= eval(!!minimum))%>%
ungroup()%>%
mutate(saterm = paste0("comorbid_",saterm)) %>%
as.data.table() %>%
dt_pivot_wider(id_cols = usubjid, names_from = saterm, values_from = saoccur)
if(dtplyr.step){
return(comorbid %>% lazy_dt(immutable = FALSE))
} else {
return(comorbid %>% as_tibble())
}
}
#' Process data on symptoms
#' @param input Either the path of the symptoms/comorbidities data file (CDISC format) or output of \code{import.symptom.and.comorbidity.data}
#' @param dtplyr.step Return the output as \code{dtplyr_step} to avoid unnecessary future calls to \code{as_tibble} or \code{as.data.table}
#' @import dplyr tibble tidyfast dtplyr
#' @importFrom data.table as.data.table
#' @importFrom glue glue
#' @return Formatted symptom data as a tibble or \code{dtplyr_step}
#' @export process.symptom.data
process.symptom.data <- function(input, minimum=100, dtplyr.step = FALSE){
symptom_w <- imp_sa%>%mutate(studyid=substr(usubjid,1, 7))%>%filter(studyid!="CVZXZMV")%>%
filter(sacat=="SIGNS AND SYMPTOMS AT HOSPITAL ADMISSION") %>%
filter(saterm!="covid-19_symptoms")%>%
arrange(desc(saoccur))%>%
group_by(saterm) %>%
arrange(desc(saoccur))%>%
mutate(n = sum(!is.na(saoccur))) %>%
filter(n >= eval(!!minimum))%>%
ungroup()%>%
mutate(saterm = paste0("symptoms_",saterm)) %>%
#mutate(saterm = glue("symptoms_{saterm}", .envir = .SD)) %>%
as.data.table() %>%
dt_pivot_wider(id_cols = usubjid, names_from = saterm, values_from = saoccur) %>%
as.data.frame()
date_onset<-imp_sa%>%
ungroup()%>%
filter(sacat=="SIGNS AND SYMPTOMS AT HOSPITAL ADMISSION" & saoccur==TRUE) %>%
mutate(sastdtc=as.character(sastdtc))%>%
mutate(sastdtc = replace(sastdtc, sastdtc =="" , NA))%>%
mutate(sastdtc=substr(sastdtc,1, 10))%>%
mutate(sastdtc=as_date(sastdtc))%>%
filter(sastdtc >= "2020-01-01")%>%
filter(sastdtc < date_pull)%>%
arrange(sastdtc)%>%
distinct(usubjid, .keep_all =T)%>%
select(usubjid, "date_onset"=sastdtc)
covid_clinic_diagn<- imp_sa%>%
filter(sacat=="CLINICALLY-DIAGNOSED COVID-19")%>%
mutate(saoccur=case_when(is.na(sapresp)~TRUE,
TRUE~saoccur))%>%
arrange(desc(saoccur))%>%
distinct(usubjid, .keep_all =T)%>%
rename("clin_diag_covid_19"=saoccur)%>%
select(usubjid,clin_diag_covid_19)
symptomatic<-imp_sa%>%mutate(studyid=substr(usubjid,1, 7))%>%filter(studyid!="CVZXZMV")%>%
ungroup()%>%
filter(sacat=="SIGNS AND SYMPTOMS AT HOSPITAL ADMISSION")%>%
mutate(symptomatic=case_when(saterm=="asymptomatic" & saoccur==TRUE~FALSE,
saterm=="asymptomatic" & saoccur==FALSE~TRUE,
TRUE~saoccur
))%>%
arrange(desc(symptomatic))%>%
distinct(usubjid, .keep_all =T)%>%
select(usubjid, symptomatic)
symptom<- date_onset%>%
full_join(covid_clinic_diagn, by=c("usubjid"))%>%
full_join(symptomatic, by = c("usubjid"))%>%
full_join(symptom_w, by = c("usubjid"))
if(dtplyr.step){
return(symptom %>% lazy_dt(immutable = FALSE))
} else {
return(symptom %>% as_tibble())
}
}
#' Process data on pregnancy (as comorbidity)
#' @param file.name Path of the dispositions data file (CDISC format)
#' @param dtplyr.step Return the output as \code{dtplyr_step} to avoid unnecessary future calls to \code{as_tibble} or \code{as.data.table}
#' @import dplyr tibble stringr
#' @return Formatted pregnancy data as a tibble or \code{dtplyr_step}
#' @export process.pregnancy.data
process.pregnancy.data <- function(file.name, dtplyr.step = FALSE){
comorbid_pregnancy <- rp%>%
#comorbid_pregnancy <- shared.data.import(pregnancy.file.name, dtplyr.step = FALSE)%>%
#comorbid_pregnancy <- rp_open%>%
filter(rptestcd=="PREGIND") %>%
mutate(comorbid_pregnancy=rpstresc) %>%
mutate(comorbid_pregnancy = case_when(comorbid_pregnancy == "Y" ~ TRUE,
comorbid_pregnancy == "N" ~ FALSE,
TRUE ~ NA)) %>%
select(usubjid,comorbid_pregnancy)
if(dtplyr.step){
return(comorbid_pregnancy %>% lazy_dt(immutable = FALSE))
} else {
return(comorbid_pregnancy %>% as_tibble())
}
}
#' Process data on ICU admission
#' @param file.name Path of the healthcare encounters data file (CDISC format)
#' @param dtplyr.step Return the output as \code{dtplyr_step} to avoid unnecessary future calls to \code{as_tibble} or \code{as.data.table}
#' @import dplyr tibble tidyfast dtplyr
#' @importFrom data.table as.data.table
#' @return Formatted symptom data as a tibble or \code{dtplyr_step}
#' @export process.ICU.data
process.ICU.data <- function(file.name, dtplyr.step = FALSE){
icu <- ho%>%
mutate(hooccur = case_when(hooccur == "Y" ~ TRUE,
hooccur == "N" ~ FALSE,
TRUE ~ NA)) %>%
filter(!is.na (hooccur))%>%
select(usubjid, hodecod, hostdtc, hoendtc, hooccur, hostdy,hoendy)%>%
mutate(hostdtc=substr(hostdtc,1, 10))%>%
mutate(hostdtc=as_date(hostdtc))%>%
mutate(hoendtc=substr(hoendtc,1, 10))%>%
mutate(hoendtc=as_date(hoendtc))
last_ho_datea<-icu%>%
filter(hooccur==TRUE)%>%
filter(hostdtc >= "2020-01-01"|hostdtc<date_pull )%>%
arrange(desc(hostdtc))%>%
distinct(usubjid, .keep_all =T)%>%
select(usubjid,hostdtc)
last_ho_dates<-icu%>%
filter(hooccur==TRUE)%>%
filter(hoendtc>= "2020-01-01"|hoendtc<date_pull)%>%
arrange(desc(hoendtc))%>%
distinct(usubjid, .keep_all =T)%>%
select(usubjid,hoendtc)%>%
left_join(last_ho_datea, by = c("usubjid"))%>%
mutate(date_ho_last=case_when(is.na(hoendtc) ~ hostdtc,
is.na(hostdtc) ~ hoendtc,
hostdtc>hoendtc ~ hostdtc,
hostdtc<=hoendtc ~ hoendtc))%>%
select(usubjid,date_ho_last)
icu <-icu%>%
filter(hodecod=="INTENSIVE CARE UNIT")%>%
arrange(desc(hostdtc))%>%
distinct(usubjid, .keep_all =T)%>%
rename(ever_icu=hooccur)%>%
rename(icu_in=hostdtc)%>%
mutate(icu_in=as_date(icu_in))%>%
mutate(icu_in=replace(icu_in,icu_in < "2020-01-01" | icu_in >date_pull,NA))%>%
rename(icu_out=hoendtc)%>%
mutate(icu_out=as_date(icu_out))%>%
mutate(icu_out=replace(icu_out,icu_out < "2020-01-01" | icu_out>date_pull,NA))%>%
select(-c(hodecod))%>%
full_join(last_ho_dates, by = c("usubjid"))
if(dtplyr.step){
return(icu)
} else {
return(icu %>% as_tibble())
}
}
#' Process data on treatments
#' @param file.name Path of the intervention data file (CDISC format)
#' @param dtplyr.step Return the output as \code{dtplyr_step} to avoid unnecessary future calls to \code{as_tibble} or \code{as.data.table}
#' @import dplyr tibble stringr
#' @return Formatted treatment data (long format) as a tibble or \code{dtplyr_step}
#' @export process.treatment.data
process.treatment.data <- function(file.name, dtplyr.step = FALSE){
#int<-int%>%filter(studyid!="CVZXZMV")
treatment<-int%>%
filter(inpresp =="Y") %>%
filter(inevintx!="BEFORE HOSPITAL ADMISSION")%>%
mutate(inoccur = case_when(inoccur == "Y" ~ TRUE,
inoccur == "N" ~ FALSE,
TRUE ~ NA))%>%
filter(!is.na(inoccur))%>%
filter(incat!="MEDICAL HISTORY" | is.na (incat))%>%
mutate(intrt_original=intrt)%>%
mutate(intrt=toupper(intrt))%>%
mutate(intrt=as.character(intrt))%>%
mutate(inmodify=as.character(inmodify))%>%
mutate(incat=as.character(incat))%>%
mutate(intrt=case_when(inmodify!=""~inmodify,
TRUE ~ intrt))%>%
mutate(intrt=case_when(incat=="EXTRACORPOREAL"~'EXTRACORPOREAL',
incat=="INVASIVE VENTILATION"~'INVASIVE VENTILATION',
incat=="FACE MASK"~'NASAL OR MASK OXYGEN THERAPY',
incat=="NASAL / MASK OXYGEN THERAPY"~'NASAL OR MASK OXYGEN THERAPY',
incat=="INVASIVE VENTILATION"~'INVASIVE VENTILATION',
incat=="NON-INVASIVE VENTILATION "~'NON-INVASIVE VENTILATION ',
incat=="OTHER INTEVENTIONS"~'OTHER INTERVENTIONS',
incat=="PRONE POSITIONING"~'PRONE POSITIONING',
incat=="PRONE VENTILATION"~'PRONE VENTILATION',
incat=="ANTIBIOTIC AGENTS"~ "ANTIBIOTIC AGENTS",
incat=="ANTIFUNGAL AGENTS"~ "ANTIFUNGAL AGENTS",
incat=="ANTIVIRAL AGENTS"~ "ANTIVIRAL AGENTS",
incat=="CORTICOSTEROIDS"~ "CORTICOSTEROIDS",
incat=="ANTIMALARIAL AGENTS"~ "ANTIMALARIAL AGENTS",
incat=="NSAIDS"~"NON-STEROIDAL ANTI-INFLAMMATORY (NSAIDS)",
TRUE~intrt)) %>%
mutate(intrt=case_when(intrt%like%'ECMO'~'EXTRACORPOREAL',
intrt=='EXTRA CORPOREAL LIFE SUPPORT'~'EXTRACORPOREAL',
intrt=='EXTRACORPOREAL SUPPORT'~'EXTRACORPOREAL',
intrt=='PRONE POSITIONING WITH UNKNOWN VENTILATION'~'PRONE POSITION VENTILATION',
intrt=='PRONE VENTILATION'~'PRONE POSITION VENTILATION',
intrt=='CONTINUOUS RENAL REPLACEMENT THERAPIES (CRRT)'~'RENAL REPLACEMENT THERAPIES',
intrt%like%'RENAL REPLACEMENT THERAPY' |
intrt%like% 'DIALYSIS'~ 'RENAL REPLACEMENT THERAPIES',
intrt%like% 'HEMOFILTRATION'~ 'RENAL REPLACEMENT THERAPIES',
intrt=='ERP CVVH'~ 'RENAL REPLACEMENT THERAPIES',
###IMV
intrt=='INVASIVE MECHANICAL LUNG VENTILATION'~'INVASIVE VENTILATION',
intrt=='INVASIVE MECHANICAL VENTILATION'~'INVASIVE VENTILATION',
intrt=='MECHANICAL VENTILATION'~'INVASIVE VENTILATION',
intrt=='RE-INTUBATION'~'INVASIVE VENTILATION',
intrt=='INVASIVE VENTILATION'~'INVASIVE VENTILATION',
intrt%like%'APRV'~'INVASIVE VENTILATION',
intrt=='INTUBATION AND MECHANICAL VENTILATION'~'INVASIVE VENTILATION',
intrt=='MECHANICAL SUPPORT'~'INVASIVE VENTILATION',
intrt%like%'EXTUBATION'~'INVASIVE VENTILATION',
intrt=="VENTILATED"~'INVASIVE VENTILATION',
###NIV
intrt%like%'CPAP'~'NON-INVASIVE VENTILATION',
intrt%like%'BIPAP'~'NON-INVASIVE VENTILATION',
intrt%like%'NON-INVASIVE MECHANICAL VENTILATION (BIPAP, CPAP, OCNAF (OPTIFLOW) ...)'~'NON-INVASIVE VENTILATION',
intrt%like%'NON-INVASIVE VENTILATION'~'NON-INVASIVE VENTILATION',
intrt=='NON-INVASIVE MECHANICAL VENTILATION'~'NON-INVASIVE VENTILATION',
intrt=='NON-INVASIVE POSITIVE PRESSURE VENTILATION'~'NON-INVASIVE VENTILATION',
intrt=='NON-INVASIVE RESPIRATORY SUPPORT'~'NON-INVASIVE VENTILATION',
TRUE ~ intrt))%>%
mutate(intrt=case_when(intrt%like%'OTHER INTERVENTION'~'OTHER INTERVENTIONS',
intrt%like%'CHEMOTHERAPY'| intrt%like%'ANTI-DIABETIC MEDICATIONS'|intrt%like%'BRONCHOSCOPY'|
intrt%like%'PROTON PUMP INHIBITORS'|intrt%like%'STATINS'|intrt%like%'MORPHINE'|
intrt%like%'HALOPERIDOL'|intrt%like%'OLANZAPINE'~'OTHER INTERVENTIONS',
intrt=='OTHER TARGETED COVID-19 MEDICATIONS'~'OTHER INTERVENTIONS',
intrt=='OTHER TREATMENTS FOR COVID19'~'OTHER INTERVENTIONS',
intrt%like%"NON-STEROIDAL"~"NON-STEROIDAL ANTI-INFLAMMATORY",
intrt%like%"NON STEROIDAL"~"NON-STEROIDAL ANTI-INFLAMMATORY",
TRUE ~ intrt))%>%
mutate(intrt=case_when(intrt=='NASAL CANULA'|intrt=='NASAL CANNULA'~'NASAL OXYGEN THERAPY',
intrt%like%'SURGICAL FEEDING TUBE'~'TOTAL PARENTERAL NUTRITION',
intrt=='FACE MASK'~'MASK OXYGEN THERAPY',
####HFNC
intrt=='OXYGEN THERAPY WITH HIGH FLOW NASAL CANULA'~'HIGH-FLOW NASAL CANULA OXYGEN THERAPY',
intrt=='HIGH-FLOW NASAL CANNULA OXYGEN THERAPY'~'HIGH-FLOW NASAL CANULA OXYGEN THERAPY',
###Prone positioning
intrt%like%'PRONACI'~'PRONE POSITIONING',
intrt=='PRONE POSITIONING'~'PRONE POSITIONING',
intrt%like%'TRACHEOSTOMY'~'TRACHEOSTOMY',
intrt%like%'NITRIC OXIDE'~'INHALED NITRIC OXIDE',
###Corticosteroids
intrt=="CORTICOSTEROID"~ "CORTICOSTEROIDS",
intrt=="DEXAMETHASONE"~ "CORTICOSTEROIDS",
intrt=="BETAMETHASONE"~ "CORTICOSTEROIDS",
intrt%like%"PREDNISOLONE"~ "CORTICOSTEROIDS",
intrt=="ORAL STEROIDS"~ "CORTICOSTEROIDS",
intrt=="STEROIDS"~ "CORTICOSTEROIDS",
intrt=="STEROIDS"~ "convalescent_plasma",
intrt%like%"HYDROCORTISONE"~ "CORTICOSTEROIDS",
intrt%like%"BLOOD TRANSFUSION OR BLOOD PRODUCT"~ "BLOOD TRANSFUSION OR BLOOD PRODUCT",
TRUE ~ intrt))%>%
mutate(intrt=case_when(intrt%like%"ANTIVIRAL" ~ "ANTIVIRAL AGENTS",
intrt%like%"ARV" ~ "ANTIVIRAL AGENTS",
intrt%like%"ANTIRETROVIRAL" ~ "ANTIVIRAL AGENTS",
intrt%like%"RIBAVIRIN" ~ "ANTIVIRAL AGENTS",
intrt%like%"LOPINAVIR AND RITONAVIR" ~ "ANTIVIRAL AGENTS",
intrt%like%"LOPINAVIR" ~ "ANTIVIRAL AGENTS",
intrt%like%"OSELTAMIVIR" ~ "ANTIVIRAL AGENTS",
intrt%like%"REMDESIVIR" ~ "ANTIVIRAL AGENTS",
intrt%like%"REMDESIVIR" ~ "ANTIVIRAL AGENTS",
intrt%like%"NEURAMINIDASE INHIBITORS" ~ "ANTIVIRAL AGENTS",
intrt%like%"ZANAMIVIR" ~ "ANTIVIRAL AGENTS",
intrt%like%"RIBAVARIN" ~ "ANTIVIRAL AGENTS",
intrt%like%"FLUCLOXACILLIN"~ "ANTIBIOTIC AGENTS",
intrt%like%"ANTIBIOTIC"~ "ANTIBIOTIC AGENTS",
intrt%like%"AMIKACIN"~ "ANTIBIOTIC AGENTS",
intrt%like%"AMOX"~ "ANTIBIOTIC AGENTS",
intrt%like%"AUGUMENTIN"~ "ANTIBIOTIC AGENTS",
intrt%like%"AZITHROMYCIN"~ "ANTIBIOTIC AGENTS",
intrt%like%"AZITHRYOMYCIN"~ "ANTIBIOTIC AGENTS",
intrt%like%"BENZY"~ "ANTIBIOTIC AGENTS",
intrt%like%"AUGUMENTIN"~ "ANTIBIOTIC AGENTS",
TRUE ~ intrt))%>%
mutate(intrt=case_when(intrt%like%"AZITHRYOMYCIN"~ "ANTIBIOTIC AGENTS",
intrt%like%"CEFTR"~ "ANTIBIOTIC AGENTS",
intrt%like%"CEFR"~ "ANTIBIOTIC AGENTS",
intrt%like%"DOXYCYCLINE"~ "ANTIBIOTIC AGENTS",
intrt%like%"CHLORAMPHENICOL"~ "ANTIBIOTIC AGENTS",
intrt%like%"CIPROFLOXACIN"~ "ANTIBIOTIC AGENTS",
intrt%like%"GENTAMICIN"~ "ANTIBIOTIC AGENTS",
intrt%like%"MEROPENEM"~ "ANTIBIOTIC AGENTS",
intrt%like%"METRONIDAZOLE"~ "ANTIBIOTIC AGENTS",
intrt%like%"ANTIMALARIAL" | intrt%like%"CHLOROQUINE" ~ "ANTIMALARIAL AGENTS",
intrt%like%"ANTIFUNGAL" ~ "ANTIFUNGAL AGENTS",
intrt %like% "OROGASTRIC"~"NASO/ NASOGASTRIC ORAL/OROGASTRIC FLUIDS",
intrt %like% "NGT OR OGT REQUIRED FOR NUTRITION"~"NASO/ NASOGASTRIC ORAL/OROGASTRIC FLUIDS",
intrt%like%'DOBUTAMINE' | intrt%like%'DOPAMINE' | intrt%like%'MILRINONE'
| intrt%like%'LEVOSIMENDAN' | intrt%like%'EPINEPHRINE' | intrt%like%'NOREPINEPRINE'
| intrt%like%'INOTROPES' |intrt%like%'VASOPRESS' |intrt%like%'NORADRENALINE' |
intrt%like%'ADRENALINE' |intrt%like%'BETA BLOCKER' ~'INOTROPES / VASOPRESSORS',
TRUE ~ intrt))%>%
mutate(intrt=case_when(intrt%like%'IMMUNOGLOBULI' ~ "convalescent_plasma",
intrt%like%'IMMUNOSUPPRES' ~ "IMMUNOSUPPRESSANTS",
intrt%like%'IMMUNOSTIMULANTS' ~ "IMMUNOSUPPRESSANTS",
intrt%like%'IMMUNOTHERAPY' ~ "IMMUNOSUPPRESSANTS",
intrt=="IL6 INHIBITOR" ~ "IMMUNOSUPPRESSANTS",
intrt=="TOCILIZUMAB" ~ "IMMUNOSUPPRESSANTS",
intrt%like%"INTERFERON" ~ "IMMUNOSTIMULANTS",
intrt%like%"HEPARIN" ~ "THERAPEUTIC ANTICOAGULANT",
intrt%like%"NOXAPARIN" ~ "THERAPEUTIC ANTICOAGULANT",
intrt=="ENOXAPARIN" ~ "THERAPEUTIC ANTICOAGULANT",
intrt%like%"SPIRONOLACTONE" ~ "DIURETICS",
intrt%like%"DIURETIC" ~ "DIURETICS",
intrt%like%"NITROUS OXIDE" ~ "inhaled_nitric_oxide",
intrt=="CPR" ~ "Cardiopulmonary resuscitation",
intrt%like%"EXPERIMENTAL AGENT" ~ "EXPERIMENTAL AGENTS",
intrt%like%"SARILUMAB" ~ "EXPERIMENTAL AGENTS",
intrt%like%"IV FLUID" ~ "INTRAVENOUS FLUIDS",
intrt%like%"I.V. SOLUTIONS" ~ "INTRAVENOUS FLUIDS",
intrt %like% "ANGIOTENSIN" | intrt %like% "ACE"~ "AGENTS ACTING ON THE RENIN-ANGIOTENSIN SYSTEM",
intrt%like%"ANTIINFLAMMATORY" ~ "ANTIINFLAMMATORY",
TRUE ~ intrt))%>% as.data.frame()%>%
select(studyid,usubjid,'treatment'=intrt,inoccur,intrt_original,inmodify,incat, inevintx,
indur,indtc,instdtc,inendtc,indy)%>%
mutate(treatment = iconv(treatment, to ="ASCII//TRANSLIT") %>% tolower()) %>%
mutate(treatment = str_remove_all(treatment, "\\s*\\([^)]*\\)")) %>%
mutate(treatment = str_replace_all(treatment, " - ", "_")) %>%
mutate(treatment = str_replace_all(treatment, "-", "_")) %>%
mutate(treatment = str_replace_all(treatment, "/| / ", "_")) %>%
mutate(treatment = str_replace_all(treatment, " ", "_"))
if(dtplyr.step){
return(treatment)
} else {
return(treatment %>% as_tibble())
}
}
#' Process data on the most common treatments
#' @param input Either the path of the interventions data file (CDISC format) or output of \code{process.treatment.data}
#' @param minimum The minimum number of times a treatment need appear to be considered "common"; default 1000.
#' @param dtplyr.step Return the output as \code{dtplyr_step} to avoid unnecessary future calls to \code{as_tibble} or \code{as.data.table}
#' @import dplyr tibble dtplyr tidyfast
#' @importFrom data.table as.data.table
#' @importFrom glue glue
#' @return Formatted common treatment data (wide format) as a tibble or \code{dtplyr_step}
#' @export process.common.treatment.data
process.common.treatment.data <- function(file.name, minimum=10, dtplyr.step = FALSE){
oxy_within_d1<-imp_int%>%
mutate(treatment=case_when(treatment=="extracorporeal" |
treatment=="inhaled_nitric_oxide" |
treatment=="prone_position_ventilation" |
treatment=="respiratory_support" |
treatment=="tracheostomy" |
treatment=="high_flow_nasal_cannula" |
treatment=="invasive_ventilation" |
treatment=="mask_oxygen_therapy" |
treatment=="nasal_oxygen_therapy" |
treatment=="non_invasive_ventilation"~"oxygen_therapy",
TRUE~treatment))%>%
filter(treatment=="oxygen_therapy")%>%
filter(inevintx=="AT HOSPITAL ADMISSION"|indy==1)%>%
mutate(oxytreat_when=case_when(inevintx=="AT HOSPITAL ADMISSION"~"at_admi",
indy==1~"within_24h"))%>%
arrange(desc(inoccur))%>%
distinct(usubjid,treatment, .keep_all =T)%>%
select(usubjid,"d1_oxygen_therapy"=inoccur)
date_in_last <- imp_int %>%
filter(inoccur==TRUE)%>%
mutate(date_in_last=substr(indtc,1, 10))%>%
mutate(date_in_last=as_date(date_in_last))%>%
filter(date_in_last >= "2020-01-01"| date_in_last<date_pull)%>%
arrange(desc(inoccur))%>%
distinct(usubjid, .keep_all =T)%>%
select(usubjid, date_in_last )
treatment <- imp_int%>%
#filter(!is.na(indtc))%>%
group_by(treatment)%>%
arrange(desc(inoccur))%>%
mutate(n = sum(!is.na(inoccur)))%>%
filter(n >= eval(!!minimum))%>%
ungroup()%>%
filter(treatment!="extracorporeal" &
treatment!="inhaled_nitric_oxide" &
treatment!="oxygen_therapy" &
treatment!="prone_position_ventilation" &
treatment!="prone_ventilation" &
treatment!="respiratory_support" &
treatment!="tracheostomy" &
treatment!="prone_positioning")%>%
#mutate(treatment=replace(treatment,treatment=="cpr","cardiopulmonary_resuscitation"))%>%
filter(treatment!="covid_19_vaccination")%>%
filter(treatment!="supplemental_oxygen_fio2")%>%
arrange(desc(inoccur))%>%
distinct(usubjid, treatment, .keep_all =T)%>%
#mutate(treatment = paste0("treat_",treatment)) %>%
mutate(treatment = glue("treat_{treatment}", treatment = treatment))%>%
as.data.table()%>%
dt_pivot_wider(id_cols = usubjid, names_from = treatment, values_from = inoccur)%>%
as.data.frame()%>%
full_join(date_in_last)
####calculating oxygen therapy overall
treat_oxy <- imp_int%>%
mutate(treatment=case_when(treatment=="extracorporeal" |
treatment=="inhaled_nitric_oxide" |
treatment=="prone_position_ventilation" |
treatment=="respiratory_support" |
treatment=="tracheostomy" |
treatment=="high_flow_nasal_cannula" |
treatment=="invasive_ventilation" |
treatment=="mask_oxygen_therapy" |
treatment=="nasal_oxygen_therapy" |
treatment=="oxygen_therapy"|
treatment=="non_invasive_ventilation"~"treat_oxygen_therapy",
TRUE~treatment))%>%
filter(treatment=="treat_oxygen_therapy")%>%
arrange(desc(inoccur))%>%
distinct(usubjid, .keep_all =T)%>%
select(usubjid,"treat_oxygen_therapy"=inoccur)
###adding duration for inasive_ventilation and non_invasive_ventilation
indur<-imp_int%>%select(usubjid,treatment, inoccur,indur,indtc,instdtc,inendtc,indy)%>%
filter(treatment=="invasive_ventilation"|treatment=="non_invasive_ventilation")%>%
mutate(treatment=case_when(treatment=='non_invasive_ventilation'~'dur_niv',
treatment=='invasive_ventilation'~'dur_imv',
TRUE~treatment))%>%
mutate(indur_clean=as.numeric(gsub("[^0-9.]", "",indur)))%>%
filter(!is.na(indur_clean) | indur_clean!="")%>%
distinct(usubjid,treatment, .keep_all =T)%>%
dt_pivot_wider(id_cols = usubjid, names_from = treatment, values_from = indur_clean)%>%
as_tibble()
vent_st_instdtc<-imp_int%>%select(usubjid,treatment, inoccur,indur,indtc,instdtc)%>%
filter(treatment=="invasive_ventilation"|treatment=="non_invasive_ventilation")%>%
mutate(treatment=case_when(treatment=='non_invasive_ventilation'~'date_niv_st',
treatment=='invasive_ventilation'~'date_imv_st',
TRUE~treatment))%>%
filter(inoccur==TRUE)%>%
mutate(instdtc=substr(instdtc,1, 10))%>%
mutate(instdtc=as_date(instdtc))%>%
filter(!is.na(instdtc))%>%
arrange(instdtc)%>%
distinct(usubjid,treatment, .keep_all =T)%>%
dt_pivot_wider(id_cols = usubjid, names_from = treatment, values_from = instdtc)%>%
as_tibble()
vent_st_indtc<-imp_int%>%
filter(inevintx=="00:00-24:00 ON DAY OF ASSESSMENT")%>%
filter(treatment=="invasive_ventilation"|treatment=="non_invasive_ventilation")%>%
mutate(treatment=case_when(treatment=='non_invasive_ventilation'~'date_niv_indtc_st',
treatment=='invasive_ventilation'~'date_imv_indtc_st',
TRUE~treatment))%>%
filter(inoccur==TRUE)%>%
mutate(indtc=substr(indtc,1, 10))%>%
mutate(indtc=as_date(indtc))%>%
filter(!is.na(indtc))%>%
arrange(instdtc)%>%
distinct(usubjid,treatment, .keep_all =T)%>%
dt_pivot_wider(id_cols = usubjid, names_from = treatment, values_from = indtc)%>%
as_tibble()
treatment <-treatment%>%
full_join(treat_oxy)%>%
full_join(oxy_within_d1)%>%
full_join(indur)%>%
full_join(vent_st_instdtc)%>%
full_join(vent_st_indtc)%>%
#full_join(vent_at_adm)%>%
mutate(date_imv_st=case_when(is.na(date_imv_st)~date_imv_indtc_st,
TRUE~date_imv_st))%>%
select(-c(date_imv_indtc_st))%>%
rename("date_niv_st"=date_niv_indtc_st)
if(dtplyr.step){
return(treatment) %>% lazy_dt(immutable = FALSE)
} else {
return(treatment %>% as_tibble())
}
}
#' Process data on the most common icu treatments
#' @param input Either the path of the interventions data file (CDISC format) or output of \code{process.treatment.data}
#' @param minimum The minimum number of times a treatment need appear to be considered "common"; default 1000.
#' @param dtplyr.step Return the output as \code{dtplyr_step} to avoid unnecessary future calls to \code{as_tibble} or \code{as.data.table}
#' @import dplyr tibble dtplyr tidyfast
#' @importFrom data.table as.data.table
#' @importFrom glue glue
#' @return Formatted common treatment data (wide format) as a tibble or \code{dtplyr_step}
#' @export process.treatment.icu.data
process.treatment.icu.data <- function(file.name,imp_icu,imp_dm,imp_ds, minimum=10, dtplyr.step = FALSE){
adm_date<-imp_dm%>%
select(usubjid,date_admit)
#tabindy_icu<-tabyl(imp_treat_icu$indy)
icu_ever<-imp_icu%>%
filter(ever_icu==TRUE)%>%
filter(!is.na(icu_in))%>%
left_join(adm_date)%>%
left_join(imp_ds)%>%
mutate(icu_dy_in=icu_in-date_admit+1)%>%
mutate(icu_dy_in=as.double(icu_dy_in))%>%
mutate(icu_dy_in=case_when(icu_dy_in<0~NA_real_,
TRUE~icu_dy_in))%>%
mutate(icu_dy_out=case_when(!is.na(icu_out)~icu_dy_in+(icu_out-icu_in),
is.na(icu_out)&!is.na(date_ho_last)~icu_dy_in+(date_ho_last-icu_in),
is.na(icu_out)&is.na(date_ho_last)~icu_dy_in+(date_outcome-icu_in),
#is.na(icu_out)&is.na(date_ho_last)~icu_dy_in+(date_outcome-icu_in),
TRUE~NA_real_))%>%
mutate(icu_dy_out=as.integer(icu_dy_out))%>%
mutate(hoendy=case_when(is.na(hoendy)~icu_dy_out,
TRUE~hoendy))%>%
select(usubjid,hostdy,hoendy)#%>%
#filter(!is.na(hoendy)& hoendy>-1 &!is.na(hostdy)& hostdy>-1)%>%
#filter(!is.na(hoendy) & !is.na(hostdy))%>%
#left_join(imp_int)%>%
treat_oxy_icu <- imp_int%>%
mutate(treatment=case_when(treatment=="extracorporeal" |
treatment=="inhaled_nitric_oxide" |
treatment=="prone_position_ventilation" |
treatment=="respiratory_support" |
treatment=="tracheostomy" |
treatment=="high_flow_nasal_cannula" |
treatment=="invasive_ventilation" |
treatment=="mask_oxygen_therapy" |
treatment=="nasal_oxygen_therapy" |
treatment=="oxygen_therapy"|
treatment=="non_invasive_ventilation"~"treat_oxygen_therapy",
TRUE~treatment))%>%
filter(treatment=="treat_oxygen_therapy")%>%
arrange(desc(inoccur))%>%
left_join(icu_ever,by = c("usubjid"))%>%
mutate(indy=as.numeric(indy))%>%
mutate(hostdy=as.numeric(hostdy))%>%
mutate(hoendy=as.numeric(hoendy))%>%
mutate(int_icu=case_when((indy>=hostdy)~ TRUE,
TRUE ~ FALSE))%>%
filter(int_icu==TRUE)%>%
arrange(desc(inoccur))%>%
distinct(usubjid, treatment, .keep_all =T)%>%
select(usubjid,"icu_treat_oxygen_therapy"=inoccur)
imp_treat_icu<-imp_int%>%
filter(!is.na(indy))%>%
group_by(treatment) %>%
arrange(desc(inoccur))%>%
mutate(n = sum(!is.na(inoccur))) %>%
filter(n >= eval(!!minimum)) %>%
ungroup()%>%
mutate(treatment=replace(treatment,treatment=="cpr","cardiopulmonary_resuscitation"))%>%
filter(treatment!="covid_19_vaccination")%>%
filter(treatment!="supplemental_oxygen_fio2")%>%
filter(treatment!="extracorporeal" &
treatment!="inhaled_nitric_oxide" &
treatment!="oxygen_therapy" &
treatment!="prone_position_ventilation" &
treatment!="prone_ventilation" &
treatment!="respiratory_support" &
treatment!="tracheostomy" &
treatment!="prone_positioning")%>%
arrange(desc(inoccur))%>%
#mutate(indtc=as.Date(indtc))%>%
#filter(indtc>= "2020-01-01"|indtc<date_pull)%>%
left_join(icu_ever,by = c("usubjid"))%>%
#filter((indy>=hostdy)&(indy<=hoendy))%>%
mutate(indy=as.numeric(indy))%>%
mutate(hostdy=as.numeric(hostdy))%>%
mutate(hoendy=as.numeric(hoendy))%>%
#mutate(int_icu=case_when((indy>=hostdy)&(indy<=hoendy) ~ TRUE,
# TRUE ~ FALSE))%>%
mutate(int_icu=case_when((indy>=hostdy)~ TRUE,
TRUE ~ FALSE))%>%
#mutate(int_icu=case_when(indtc>=icu_in ~ TRUE,
# TRUE ~ FALSE))%>%
filter(int_icu==TRUE)%>%
arrange(desc(inoccur))%>%
distinct(usubjid, treatment, .keep_all =T)%>%
mutate(treatment = glue("icu_treat_{treatment}", treatment = treatment)) %>%
as.data.table() %>%
dt_pivot_wider(id_cols = usubjid, names_from = treatment, values_from = inoccur)%>%
full_join(treat_oxy)
if(dtplyr.step){
return(imp_treat_icu) %>% lazy_dt(immutable = FALSE)
} else {
return(imp_treat_icu %>% as_tibble())
}
}
#' Process data on vital sign
#' @param file.name Path of the dispositions data file (CDISC format)
#' @param dtplyr.step Return the output as \code{dtplyr_step} to avoid unnecessary future calls to \code{as_tibble} or \code{as.data.table}
#' @import dplyr tibble dtplyr tidyfast
#' @importFrom data.table as.data.table
#' @importFrom glue glue
#' @return Formatted vital sign (wide format) as a tibble or \code{dtplyr_step}
#' @export process.vital.sign.data
process.vital.sign.data <- function(file.name, dtplyr.step = FALSE){
vital_sign <- vs %>%
select(usubjid, vstestcd, vscat,vsstresn,vsstresu, vsdtc, vso2src) %>%
filter(vscat=="SIGNS AND SYMPTOMS AT HOSPITAL ADMISSION" | vscat=="SIGNS AND SYMPTOMS AT ADMISSION")%>%
mutate(vsstresn=as.numeric(vsstresn))%>%
mutate(vsstresn=case_when(vstestcd=="OXYSAT"& vsstresn< 1~ NA_real_,
vstestcd=="OXYSAT"& vsstresn> 100~ NA_real_,
vstestcd=="BMI"& vsstresn< 0~ NA_real_,
vstestcd=="BMI"& vsstresn> 100~ NA_real_,
vstestcd=="DIABP"& vsstresn< 0~ NA_real_,
vstestcd=="DIABP"& vsstresn> 300~ NA_real_,
vstestcd=="HEIGHT"& vsstresn< 0~ NA_real_,
vstestcd=="HEIGHT"& vsstresn> 250~ NA_real_,
vstestcd=="HR"& vsstresn< 0~ NA_real_,
vstestcd=="HR"& vsstresn> 250~ NA_real_,
vstestcd=="MAP"& vsstresn< 0~ NA_real_,
vstestcd=="MAP"& vsstresn> 250~ NA_real_,
vstestcd=="MUARMCIR"& vsstresn< 0~ NA_real_,
vstestcd=="MUARMCIR"& vsstresn> 100~ NA_real_,
vstestcd=="PULSE"& vsstresn< 0~ NA_real_,
vstestcd=="PULSE"& vsstresn> 250~ NA_real_,
vstestcd=="RESP"& vsstresn< 0~ NA_real_,
vstestcd=="RESP"& vsstresn> 60~ NA_real_,
vstestcd=="SYSBP"& vsstresn< 0~ NA_real_,
vstestcd=="SYSBP"& vsstresn> 250~ NA_real_,
vstestcd=="TEMP"& vsstresn< 30~ NA_real_,
vstestcd=="TEMP"& vsstresn> 44~ NA_real_,
vstestcd=="WEIGHT"& vsstresn< 0~ NA_real_,
vstestcd=="WEIGHT"& vsstresn> 300~ NA_real_,
TRUE~vsstresn))%>%
filter(!is.na(vsstresn))%>%
arrange(desc(vsdtc))%>%
distinct(usubjid,vstestcd, .keep_all =T)%>%
mutate(vso2src=case_when(vso2src==""&vstestcd=="OXYSAT"~'UNKNOWN',
TRUE~vso2src))%>%
mutate(vso2src= str_replace_all(vso2src, " ", "_"))%>%
mutate(vstestcd=case_when(vstestcd=="OXYSAT"~paste0(vstestcd,"_",vso2src),
TRUE~vstestcd))%>%
mutate(vstestcd = paste0("vs_",vstestcd)) %>%
#mutate(vstestcd = glue("vs_{vstestcd}", vstestcd = vstestcd))%>%
mutate(vstestcd = iconv(vstestcd, to ="ASCII//TRANSLIT") %>% tolower()) %>%
as.data.table() %>%
dt_pivot_wider(id_cols = usubjid, names_from = vstestcd, values_from = vsstresn)%>%
as.data.frame() %>%
mutate(vs_oxysat=case_when(!is.na(vs_oxysat_oxygen_therapy)~vs_oxysat_oxygen_therapy,
!is.na(vs_oxysat_room_air)~vs_oxysat_room_air,
TRUE~vs_oxysat_unknown))
if(dtplyr.step){
return(vital_sign)
} else {
return(vital_sign %>% as_tibble())
}
}
#' Process data on laboratory
#' @param file.name Path of the dispositions data file (CDISC format)
#' @param dtplyr.step Return the output as \code{dtplyr_step} to avoid unnecessary future calls to \code{as_tibble} or \code{as.data.table}
#' @import dplyr tibble dtplyr tidyfast
#' @importFrom data.table as.data.table
#' @importFrom glue glue
#' @return Formatted laboratory (wide format) as a tibble or \code{dtplyr_step}
#' @export process.laboratory.data
process.laboratory.data <- function(file.name, dtplyr.step = FALSE){
laboratory <- lb%>%
select(usubjid, lbdy, lbtestcd, lbcat,lborres,lbdtc) %>%
mutate(lborres=replace(lborres,lborres=="",NA))%>%
mutate(studyid=substr(usubjid,1, 7))%>%
mutate(lbcat=case_when(lbdy==1 & (studyid=="CVCCPUK"|
studyid=="CVMEWUS" |
studyid=="CORE"|
studyid=="CVTDWXD"|
studyid=="CVTTYLU"|
studyid=="CVZXZMV"|
studyid=="CVKBQEI") ~"LABORATORY RESULTS ON ADMISSION",
#lbdy==1 & studyid=="CVMEWUS"~"LABORATORY RESULTS ON ADMISSION",
TRUE~as.character(lbcat)))%>%
filter(lbcat=="LABORATORY RESULTS ON ADMISSION")%>%
filter(lbtestcd=="ALT"|
lbtestcd=="APTT"|
lbtestcd=="CRP"|
lbtestcd=="LYM"|
lbtestcd=="NEUT"|
lbtestcd=="PT"|
lbtestcd=="WBC"|
lbtestcd=="BILI"|
lbtestcd=="AST"|
lbtestcd=="UREAN")%>%
mutate(lborres=as.numeric(lborres))%>%
filter(!is.na(lborres))%>%
arrange(desc(lbdtc))%>%
distinct(usubjid,lbtestcd, .keep_all =T)%>%
mutate(lborres=case_when(lbtestcd=="NEUT" & lborres>100 ~ lborres/1000,
lbtestcd=="LYM" & lborres>100 ~ lborres/1000,
lbtestcd=="WBC" & lborres>100 ~ lborres/1000,
lbtestcd=="ALT" & lborres>2000 ~ NA_real_,
lbtestcd=="ALT" & lborres<0 ~ NA_real_,
lbtestcd=="AST" & lborres>2000 ~ NA_real_,
lbtestcd=="AST" & lborres<0 ~ NA_real_,
lbtestcd=="BILI" & lborres>2000 ~ NA_real_,
lbtestcd=="BILI" & lborres<0 ~ NA_real_,
lbtestcd=="CRP" & lborres>500 ~ NA_real_,
lbtestcd=="CRP" & lborres<0 ~ NA_real_,
lbtestcd=="PT" & lborres>105 ~ NA_real_,
lbtestcd=="PT" & lborres<0 ~ NA_real_,
lbtestcd=="UREAN" & lborres>100 ~ NA_real_,
lbtestcd=="UREAN" & lborres<0 ~ NA_real_,
lbtestcd=="APTT" & lborres>2000 ~ NA_real_,
lbtestcd=="APTT" & lborres<0 ~ NA_real_,
TRUE ~ lborres ))%>%
mutate(lborres=case_when(lbtestcd=="NEUT" & lborres>100 ~ NA_real_,
lbtestcd=="NEUT" & lborres<0 ~ NA_real_,
lbtestcd=="LYM" & lborres>100 ~ NA_real_,
lbtestcd=="LYM" & lborres<0 ~ NA_real_,
lbtestcd=="WBC" & lborres>100 ~ NA_real_,
lbtestcd=="WBC" & lborres<0 ~ NA_real_,
TRUE ~ lborres ))%>%
mutate(lbtestcd = paste0("lab_",lbtestcd )) %>%
#mutate(lbtestcd = glue("lab_{lbtestcd}", lbtestcd = lbtestcd)) %>%
mutate(lbtestcd = iconv(lbtestcd, to ="ASCII//TRANSLIT") %>% tolower()) %>%
as.data.table() %>%
dt_pivot_wider(id_cols = usubjid, names_from = lbtestcd, values_from = lborres)
if(dtplyr.step){
return(laboratory)
} else {
return(laboratory%>% as_tibble())
}
}
#' Process data on outcomes
#' @param file.name Path of the dispositions data file (CDISC format)
#' @param dtplyr.step Return the output as \code{dtplyr_step} to avoid unnecessary future calls to \code{as_tibble} or \code{as.data.table}
#' @import dplyr tibble stringr
#' @return Formatted outcome data (long format) as a tibble or \code{dtplyr_step}
#' @export process.outcome.data
process.outcome.data <- function(file.name, dtplyr.step = FALSE){
outcome <- ds%>%
select(usubjid, dsterm, "date_outcome" = dsstdtc, dsmodify) %>%
mutate(date_outcome=substr(date_outcome,1, 10))%>%
mutate(date_outcome=as_date(date_outcome))%>%
mutate(date_outcome=replace(date_outcome,date_outcome< "2020-01-01",NA))%>%
mutate(date_outcome=replace(date_outcome,date_outcome>date_pull,NA))%>%
mutate(outcome=tolower(dsterm))%>%
mutate(outcome=case_when(outcome=="palliative"~"transferred",
outcome=="transferred to another unit"~"ongoing care",
outcome=="Ongoing health care needs NOT related to COVID episode"~"discharge",
outcome==""~NA_character_,
TRUE~outcome))%>%
mutate(outcome=case_when(outcome%like%"hospitalis"~"ongoing care",
outcome%like%"hospitaliz"~"ongoing care",
outcome%like%"ongoing"~"ongoing care",
outcome=="in hospital"~"ongoing care",
outcome%like%"death"~"death",
outcome=="died"~"death",
outcome=="deceased"~"death",
outcome=="died (non-covid)"~"death",
#outcome=="Death In Hospital"~"Death",
outcome=="alive"~"discharge",
outcome%like%"discharge"~"discharge",
outcome%like%"transfer"~"transferred",
outcome=="long term care facility"~"transferred",
outcome=="quarantine center"~"transferred",
outcome=="missing in database"~"unknown outcome",
outcome=="unknown"~"unknown outcome",
outcome=="not recorded"~"unknown outcome",
TRUE ~ outcome))%>%
group_by(usubjid) %>%
mutate(count=1)%>%
mutate(n = sum(count)) %>%
filter(n == 1)%>%
select(-c(dsterm,dsmodify,n,count))
if(dtplyr.step){
return(outcome)
} else {
return(outcome %>% as_tibble())
}
}
ISARICDataPlatform/CovidClinicalDataProcessor documentation built on March 22, 2022, 7:51 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Defines functions process.outcome.data process.laboratory.data process.vital.sign.data process.treatment.icu.data process.common.treatment.data process.treatment.data process.ICU.data process.pregnancy.data process.symptom.data process.comorbidity.data import.symptom.and.comorbidity.data import.microb.data import.demographic.data
Documented in import.demographic.data import.microb.data import.symptom.and.comorbidity.data process.common.treatment.data process.comorbidity.data process.ICU.data process.laboratory.data process.outcome.data process.pregnancy.data process.symptom.data process.treatment.data process.treatment.icu.data process.vital.sign.data
#' Import demographic data
#' @param file.name Path of the demographics data file (CDISC format)
#' @param dtplyr.step Return the output as \code{dtplyr_step} to avoid unnecessary future calls to \code{as_tibble} or \code{as.data.table}
#' @import dplyr tibble
#' @return Formatted demographic data as a tibble or \code{dtplyr_step}
#' @export import.demographic.data
import.demographic.data <- function(file.name, dtplyr.step = FALSE){
wdi_dat <- WDI(indicator = c("NY.GDP.PCAP.KD", "SP.DYN.LE00.IN", "SP.DYN.IMRT.IN"),
start = 2020, end = 2020, extra = TRUE)%>%
filter(region != "Aggregates")%>%
select("Alpha_3"=iso3c,income,region)
country.lookup <- ISOcodes::ISO_3166_1 %>% as_tibble%>%
mutate(Name=case_when(!is.na(Common_name)~Common_name,
Name=="Lao People's Democratic Republic"~"Lao PDR",
TRUE~Name))%>%select(Alpha_3, Name)%>%left_join(wdi_dat)
#regexp <- "[[:digit:]]+" # process string
out <- dm %>%
###delete patients duplicates
group_by(usubjid) %>%
mutate(count=1)%>%
mutate(n = sum(count)) %>%
filter(n == 1) %>%
ungroup()%>%
rename(date_admit=rfstdtc)%>%
as.data.frame()%>%
mutate(age_d=case_when(ageu=="MONTHS"~12,
ageu=="YEARS" ~ 1,
ageu=="DAYS" ~ 365.25,
TRUE~ NA_real_))%>%
mutate(age2=age/age_d)%>%
select(-(age))%>%
rename(age=age2)%>%
mutate(age=replace(age,age<0,NA))%>%
mutate(ethnic = iconv(ethnic, to ="ASCII//TRANSLIT") %>% tolower()) %>%
mutate(ethnic = str_remove_all(ethnic, "\\s*\\([^)]*\\)")) %>%
mutate(ethnic = str_replace_all(ethnic, " - ", "_")) %>%
mutate(ethnic = str_replace_all(ethnic, "-", "_")) %>%
mutate(ethnic = str_replace_all(ethnic, "/| / ", "_")) %>%
mutate(ethnic = str_replace_all(ethnic, " ", "_")) %>%
mutate(ethnic = str_replace_all(ethnic, ",", "_")) %>%
mutate(ethnic = replace(ethnic, ethnic == "n_a" | ethnic == "na" | ethnic == "", NA))%>%
#mutate(studyid=substr(usubjid,1, 7))%>%
mutate(siteid_final=invid)%>%
mutate(siteid_final=case_when(invid=="00741cca_network"~ substr(subjid,1, 12),
invid=="227inserm"~ sub("\\-.*", "",subjid),
invid=="00689us_nhlbi_peta"~ sub("\\-.*", "",subjid),
invid==""~studyid,
studyid=="CVPRQTA"~"CVPRQTA",
TRUE~invid))%>%
mutate(sex = case_when(sex == "M" ~ "Male",
sex == "F" ~ "Female",
TRUE ~ NA_character_))%>%
mutate(date_admit=substr(date_admit,1, 10))%>%
mutate(date_admit=as_date(date_admit))%>%
mutate(date_admit=replace(date_admit,date_admit >date_pull,NA))%>%
select(usubjid, studyid, siteid_final, date_admit, age, sex, ethnic, country)
site_id_country<-out%>%
mutate(country=ifelse(siteid_final=="321cub_erasme__bru","BEL",country))%>%
mutate(country=ifelse(siteid_final=="435civil_hospital","BEL",country))%>%
mutate(country=ifelse(siteid_final=="00657hospital_de_c","PRT",country))%>%
mutate(country=ifelse(siteid_final==" 00727clinica_unive","COL",country))%>%
mutate(country=ifelse(siteid_final=="00580netcare_unita","ITA",country))%>%
mutate(country=ifelse(siteid_final=="00835consortium_im","POL",country))%>%
mutate(country=ifelse(siteid_final=="00831nicvd_dhaka","BGD",country))%>%
mutate(country = replace(country, country == "", NA)) %>%
left_join(country.lookup, by = c("country" = "Alpha_3")) %>%
select(-country) %>%
rename(country = Name) %>%
filter(!is.na(country))%>%
arrange(desc(country, income, siteid_final))%>%
distinct(siteid_final, .keep_all =T)%>%
select(siteid_final, 'country_2'=country,income, region)
out<-out%>%
left_join(site_id_country)%>%
mutate(country=country_2)%>%select(-c(country_2))
if(dtplyr.step){
return(out)
} else {
return(out %>% as_tibble())
}
}
#' Import microb data
#' @param file.name Path of the microbio data file (CDISC format)
#' @param dtplyr.step Return the output as \code{dtplyr_step} to avoid unnecessary future calls to \code{as_tibble} or \code{as.data.table}
#' @import dplyr tibble
#' @return Formatted demographic data as a tibble or \code{dtplyr_step}
#' @export import.microb.data
import.microb.data <- function(file.name, dtplyr.step = FALSE){
#mb<-shared.data.import(file.name, dtplyr.step = TRUE)
mb<-mb
detection<- mb%>%
#select(usubjid,mbtestcd,mbtest,mbtstdtl,mbcat,mbstresc,mbspec,mbloc,mbmethod)%>%
filter(mbtstdtl=="DETECTION")%>%
filter(mbtestcd=="CRONAVIR"|mbtestcd=="SARSCOV2")%>%
mutate(mbstresc = case_when(mbstresc == "NO" ~ "NEGATIVE",
mbstresc == "NEGATIVE" ~ "NEGATIVE",
mbstresc == "POSITIVE" ~ "POSITIVE",
TRUE ~ NA_character_)) %>%
mutate(mbtestcd = paste0("cov_det_",mbtestcd)%>% tolower%>%str_replace_all(" ", "_")) %>%
arrange(desc(mbstresc))%>%
distinct(usubjid, mbtestcd, .keep_all =T)%>%
as.data.table() %>%
dt_pivot_wider(id_cols = usubjid, names_from = mbtestcd, values_from = mbstresc) %>%
as.data.frame()
identification<-mb%>%
filter(mbtstdtl=="IDENTIFICATION")%>%
distinct(usubjid, mbstresc, .keep_all =T)%>%
filter(mbstresc=="SEVERE ACUTE RESPIRATORY SYNDROME CORONAVIRUS 2"|
mbstresc=="CORONAVIRIDAE")%>%
mutate(mbstresc=replace(mbstresc,mbstresc=="SEVERE ACUTE RESPIRATORY SYNDROME CORONAVIRUS 2","SARSCOV2"))%>%
mutate(mbstresc=replace(mbstresc,mbstresc=="SEVERE ACUTE RESPIRATORY SYNDROME-RELATED CORONAVIRUS","SARSCOV2"))%>%
mutate(mbstresc=replace(mbstresc,mbstresc=="CORONAVIRIDAE","CRONAVIR"))%>%
mutate(result="POSITIVE")%>%
mutate(mbstresc = paste0("cov_id_",mbstresc)%>%
tolower%>%
str_replace_all(" ", "_")) %>%
as.data.table() %>%
dt_pivot_wider(id_cols = usubjid, names_from = mbstresc, values_from = result) %>%
as.data.frame()
out<-full_join(detection,identification)%>%
mutate(cov_det_id="NEGATIVE")%>%
mutate(cov_det_id=case_when(cov_det_cronavir=="POSITIVE"|
cov_det_sarscov2=="POSITIVE"|
cov_id_cronavir=="POSITIVE"|
cov_id_sarscov2=="POSITIVE"~
"POSITIVE",
is.na(cov_det_cronavir)&
is.na(cov_det_sarscov2)&
is.na(cov_id_cronavir)&
is.na(cov_id_sarscov2)~
NA_character_,
TRUE~cov_det_id))
if(dtplyr.step){
return(out)
} else {
return(out %>% as_tibble())
}
}
#' Import data on symptoms and comorbidities
#' @param file.name Path of the symptoms data file (CDISC format)
#' @param dtplyr.step Return the output as \code{dtplyr_step} to avoid unnecessary future calls to \code{as_tibble} or \code{as.data.table}
#' @import dplyr tibble stringr
#' @return Formatted comorbidity and symptom data as a tibble or \code{dtplyr_step}
#' @export import.symptom.and.comorbidity.data
import.symptom.and.comorbidity.data <- function(file.name, dtplyr.step = TRUE){
#out <- shared.data.import(file.name,
# dtplyr.step = TRUE,
# immutable = TRUE) %>% # this will often by used twice, so should be immutable
out<-sa%>%
#imp_sa<-sa%>%
select(usubjid, saterm, sacat, samodify, sapresp, saoccur, sastdtc) %>%
mutate(sacat=case_when(
saterm=="CLINICALLY-DIAGNOSED COVID-19"~"CLINICALLY-DIAGNOSED COVID-19",
TRUE~sacat))%>%
filter(
sacat=="MEDICAL HISTORY"|
sacat=="SIGNS AND SYMPTOMS AT HOSPITAL ADMISSION"
|sacat=="CLINICALLY-DIAGNOSED COVID-19"
)%>%
#mutate(sacat=replace(sacat,saterm=="MALNUTRITION","MEDICAL HISTORY"))%>%#temporary correction
#filter( sapresp=="Y") %>%
mutate(saoccur = case_when(saoccur == "Y" ~ TRUE,
saoccur == "N" ~ FALSE,
TRUE ~ NA)) %>%
filter(!is.na(saoccur)) %>%
mutate(saterm=toupper(saterm))%>%
mutate(saterm=case_when(samodify!=""|is.na(samodify)~samodify,
TRUE ~ saterm))%>%#to add
mutate(saterm=case_when(saterm%like%'CARDIAC ARRHYTHMIA'~'CHRONIC CARDIAC DISEASE',
saterm%like%'CARDIAC DISEASE'~'CHRONIC CARDIAC DISEASE',
saterm%like%'CHORNIC CARDIAC DISEASE'~'CHRONIC CARDIAC DISEASE',
saterm%like%'CHRONIC HEART DISEASE'~'CHRONIC CARDIAC DISEASE',
saterm%like%'CONGENITAL CA'~'CHRONIC CARDIAC DISEASE',
saterm%like%'CONGENTIAL CARDIOPATHY'~'CHRONIC CARDIAC DISEASE',
saterm=='CORONARY DISEASE'~'CHRONIC CARDIAC DISEASE',
saterm=='HEART FAILURE'~'CHRONIC CARDIAC DISEASE',
saterm=='OROVALVA DISEASE'~'CHRONIC CARDIAC DISEASE',
saterm=='RHEUMATIC HEART DISEASE'~'CHRONIC CARDIAC DISEASE',
saterm=='VALVULAR HEART DISEASE'~'CHRONIC CARDIAC DISEASE',
saterm=='CONGESTIVE HEART FAILURE'~'CHRONIC CARDIAC DISEASE',
saterm=='CORONARY ARTERY DISEASE'~'CHRONIC CARDIAC DISEASE',
TRUE~saterm))%>%
mutate(saterm=case_when(saterm=='CHRONIC DIALYSIS'~'CHRONIC KIDNEY DISEASE',
saterm%like%'DEPRESSION'~'PSYCHIATRIC CONDITION',
saterm%like%'PSYCHOSIS'~'PSYCHIATRIC CONDITION',
saterm%like%'DYSLIPIDEMIA'~'CHRONIC METABOLIC DISORDER',
saterm%like%'HYPOTHYROIDISM'~'CHRONIC ENDOCRINE DISORDER NON DIABITES',
saterm%like%'HEPATITIS'~'LIVER DISEASE',
saterm%like%'MARASUMAS'~'MALNUTRITION',
saterm=='SAM UNDEFINED'~'MALNUTRITION',
saterm=='MIXED MARASMIC-KWASH'~'MALNUTRITION',
saterm=='OSA/ HOME CPAP/BI-PAP USE'~'OBESITY',
saterm=='PAOD'~'OTHER COMORBIDITIES',
saterm=='PEPTIC ULCER DISEASE EXCLUDING BLEEDING'~'OTHER COMORBIDITIES',
saterm=='PARALYSIS'~'CHRONIC NEUROLOGICAL DISORDER',
saterm=='STROKE OR OTHER NEUROLOGICAL DISORDERS'~'CHRONIC NEUROLOGICAL DISORDER',
saterm=='PULMONARY CIRCULATION DISORDER'~'CHRONIC CARDIAC DISEASE',
saterm%like%'ARRHYTHMIA'~'CHRONIC CARDIAC DISEASE',
saterm=='SUBSTANCE USE DISORDER'~'SUBSTANCE MISUSE',
saterm=='VENOUS THROMBOEMBOLISM- DVT/PE'~'THROMBOLIC DISORDERS',
saterm=='CHILLS/RIGORS'~'RIGOR OR SWEATING',
saterm=='NIGHT SWEAT'~'RIGOR OR SWEATING',
saterm=='CONGESTION/RHINORRHEA'~'RUNNY NOSE',
saterm=='CONJUNCTIVAL CONGESTION'~'UPPER RESPIRATORY TRACT SYMPTOMS',
saterm=='SNEEZING'~'UPPER RESPIRATORY TRACT SYMPTOMS',
saterm=='DELIRIUM / ENCEPHALOPATHY'~'ALTERED CONSCIOUSNESS CONFUSION',
saterm=='DIZZINESS/LIGHTHEADEDNESS'~'OTHER SIGNS AND SYMPTOMS',
saterm=='GASTROGASTROINTESTINAL HEMORRHAGE'~'OTHER SIGNS AND SYMPTOMS',
TRUE~saterm))%>%
mutate(saterm=case_when(saterm%like%'TUBERCULOSIS'~'TUBERCULOSIS',
saterm%like%'MALIGNANCY'~'MALIGNANT NEOPLASM',
saterm%like%'SPECIFIC CANCERS'~'MALIGNANT NEOPLASM',
saterm%like%'SOLID TUMOR'~'MALIGNANT NEOPLASM',
saterm%like%'METASTATIC CANCER'~'MALIGNANT NEOPLASM',
saterm=='SORE THROAT/THROAT PAIN'~'SORE THROAT',
saterm=='COAGULOPATHY'~'CHRONIC HEMATOLOGIC DISEASE',
saterm=='DYSLIPIDEMIA/HYPERLIPIDEMIA'~'CHRONIC HEMATOLOGIC DISEASE',
saterm=='IRON DEFICIENCY ANEMIA'~'CHRONIC HEMATOLOGIC DISEASE',
saterm=='BLOOD LOSS ANEMIA'~'CHRONIC HEMATOLOGIC DISEASE',
saterm=='CHRONIC HEMATOLOGICAL DISEASE'~'CHRONIC HEMATOLOGIC DISEASE',
saterm=='CHRONIC LIVER DISEASE'~'LIVER DISEASE',
saterm%like%'ACUTE LIVER'~'LIVER DISEASE',
saterm%like%'CHRONIC RENAL FAILURE'~'CHRONIC KIDNEY DISEASE',
saterm%like%'CHRONIC LUNG DISEASE'~'CHRONIC PULMONARY DISEASE',
saterm%like%'CHROMIC PULMONARY DISEASE'~'CHRONIC PULMONARY DISEASE',
TRUE~saterm))%>%
mutate(saterm=case_when(saterm%like%'RHEMATOLOGICAL DISORDER'~'rheumatologic disorder',
saterm%like%'CHRONIC NEUROLOGICAL'~'CHRONIC NEUROLOGICAL DISORDER',
saterm%like%'CURRENT SMOK'~'SMOKING',
saterm%like%'DIABETES'~'DIABETES',
saterm=='HISTORY OF PERIPHERAL OR CARDIAC REVASCULARIZATION'~'HISTORY OF PERIPHERAL OR CARDIAC REVASCULARIZATION',
saterm=='HISTORY OF SMOKING'~'SMOKING',
saterm%like%'SMOKING'~'SMOKING',
saterm%like%'HIV'~'AIDS/HIV',
saterm%like%'LIVER DISEASE'~'LIVER DISEASE',
saterm%like%'OTHER RELEVANT RISK'~'OTHER COMORBIDITIES',
saterm=='OTHER RISK FACTOR'~'OTHER COMORBIDITIES',
saterm%like%'RHEUMATOLOGICAL DISORD'~'RHEUMATOLOGIC DISORDER',
saterm=='SMOKER'~'SMOKING',
saterm=='SMOKER - CURRENT'~'SMOKING',
saterm=='SMOKER - FORMER'~'SMOKING - FORMER',
saterm=='FEEDING INTOLERANCE (PAEDIATRICS)'~'ANOREXIA',
saterm=='REFUSING TO EAT OR DRINK/HISTORY OF POOR ORAL INTAKE'~'ANOREXIA',
saterm%like%'ANOREXIA'~'ANOREXIA',
saterm=='ANOREXIA - LOSS OF APPETITE'~'ANOREXIA',
saterm=='CHEST PAIN/TIGHTNESS'~'CHEST PAIN',
saterm=='SWOLLEN NECK GLANDS/LYMPHADENOPATHY'~'LYMPHADENOPATHY',
TRUE~saterm))%>%
mutate(saterm=case_when(saterm%like%'COUGH'~'COUGH',
saterm%like%'COUTH'~'COUGH',
saterm=='HEMOPTYSIS'~'COUGH',
saterm=='DIARRHEA'~'DIARRHOEA',
saterm=='CONJUNCTIVAL CONGESTION '~'CONJUNCTIVITIS',
saterm%like%'FEVER'~'HISTORY OF FEVER',
saterm=='SEIZURE'~'SEIZURES',
saterm%like%'TRANSPLANT'~'TRANSPLANTATION',
saterm%like%'ANOSMIA'~'LOSS OF SMELL',
saterm%like%'AGEUSIA'~'LOSS OF TASTE',
saterm=="LOSS OF TASTE OR LOSS OF SMELL"~'LOSS OF SMELL/TASTE',
saterm=="NAUSEA/VOMITING"~'VOMITING/NAUSEA',
saterm%like%'MYALGIA OR FATIGUE'~'MUSCLE ACHES/JOINT PAIN',
saterm%like%'JOINT PAIN'~'MUSCLE ACHES/JOINT PAIN',
saterm%like%'MUSCLE ACHES'~'MUSCLE ACHES/JOINT PAIN',
saterm=='OTHER SIGN OR SYMPTOM'~'OTHER SIGNS AND SYMPTOMS',
saterm=='LOWER CHEST WALL INDRAWING'~'SHORTNESS OF BREATH',
saterm%like%'DEHYDRATION'~'SEVERE DEHYDRATION',
saterm%like%'RASH'~'SKIN RASH',
saterm=='EARPAIN'~'EAR PAIN',
TRUE ~ saterm ))%>%
mutate(saterm = iconv(saterm, to ="ASCII//TRANSLIT") %>% tolower()) %>%
mutate(saterm = str_remove_all(saterm, "\\s*\\([^)]*\\)")) %>%
mutate(saterm = str_replace_all(saterm, " - ", "_")) %>%
mutate(saterm = str_replace_all(saterm, "/| / ", "_")) %>%
mutate(saterm = str_replace_all(saterm, " ", "_")) %>%
arrange(desc(saoccur))%>%
distinct(usubjid,saterm, .keep_all =T)
if(dtplyr.step){
return(out)
} else {
return(out %>% as_tibble())
}
}
#' Process data on comorbidities
#' @param input Either the path of the symptoms/comorbidities data file (CDISC format) or output of \code{import.symptom.and.comorbidity.data}
#' @param dtplyr.step Return the output as \code{dtplyr_step} to avoid unnecessary future calls to \code{as_tibble} or \code{as.data.table}
#' @import dplyr tibble stringr tidyfast
#' @importFrom data.table as.data.table
#' @importFrom glue glue
#' @return Formatted comorbidity data as a tibble or \code{dtplyr_step}
#' @export process.comorbidity.data
process.comorbidity.data <- function(input, minimum=100, dtplyr.step = FALSE){
comorbid <- imp_sa_com %>%
filter(sacat=="MEDICAL HISTORY") %>%
filter(!is.na(sacat))%>%
filter(!is.na(saterm))%>%
arrange(desc(saoccur))%>%
group_by(saterm) %>%
arrange(desc(saoccur))%>%
mutate(n = sum(!is.na(saoccur))) %>%
filter(n >= eval(!!minimum))%>%
ungroup()%>%
mutate(saterm = paste0("comorbid_",saterm)) %>%
as.data.table() %>%
dt_pivot_wider(id_cols = usubjid, names_from = saterm, values_from = saoccur)
if(dtplyr.step){
return(comorbid %>% lazy_dt(immutable = FALSE))
} else {
return(comorbid %>% as_tibble())
}
}
#' Process data on symptoms
#' @param input Either the path of the symptoms/comorbidities data file (CDISC format) or output of \code{import.symptom.and.comorbidity.data}
#' @param dtplyr.step Return the output as \code{dtplyr_step} to avoid unnecessary future calls to \code{as_tibble} or \code{as.data.table}
#' @import dplyr tibble tidyfast dtplyr
#' @importFrom data.table as.data.table
#' @importFrom glue glue
#' @return Formatted symptom data as a tibble or \code{dtplyr_step}
#' @export process.symptom.data
process.symptom.data <- function(input, minimum=100, dtplyr.step = FALSE){
symptom_w <- imp_sa%>%mutate(studyid=substr(usubjid,1, 7))%>%filter(studyid!="CVZXZMV")%>%
filter(sacat=="SIGNS AND SYMPTOMS AT HOSPITAL ADMISSION") %>%
filter(saterm!="covid-19_symptoms")%>%
arrange(desc(saoccur))%>%
group_by(saterm) %>%
arrange(desc(saoccur))%>%
mutate(n = sum(!is.na(saoccur))) %>%
filter(n >= eval(!!minimum))%>%
ungroup()%>%
mutate(saterm = paste0("symptoms_",saterm)) %>%
#mutate(saterm = glue("symptoms_{saterm}", .envir = .SD)) %>%
as.data.table() %>%
dt_pivot_wider(id_cols = usubjid, names_from = saterm, values_from = saoccur) %>%
as.data.frame()
date_onset<-imp_sa%>%
ungroup()%>%
filter(sacat=="SIGNS AND SYMPTOMS AT HOSPITAL ADMISSION" & saoccur==TRUE) %>%
mutate(sastdtc=as.character(sastdtc))%>%
mutate(sastdtc = replace(sastdtc, sastdtc =="" , NA))%>%
mutate(sastdtc=substr(sastdtc,1, 10))%>%
mutate(sastdtc=as_date(sastdtc))%>%
filter(sastdtc >= "2020-01-01")%>%
filter(sastdtc < date_pull)%>%
arrange(sastdtc)%>%
distinct(usubjid, .keep_all =T)%>%
select(usubjid, "date_onset"=sastdtc)
covid_clinic_diagn<- imp_sa%>%
filter(sacat=="CLINICALLY-DIAGNOSED COVID-19")%>%
mutate(saoccur=case_when(is.na(sapresp)~TRUE,
TRUE~saoccur))%>%
arrange(desc(saoccur))%>%
distinct(usubjid, .keep_all =T)%>%
rename("clin_diag_covid_19"=saoccur)%>%
select(usubjid,clin_diag_covid_19)
symptomatic<-imp_sa%>%mutate(studyid=substr(usubjid,1, 7))%>%filter(studyid!="CVZXZMV")%>%
ungroup()%>%
filter(sacat=="SIGNS AND SYMPTOMS AT HOSPITAL ADMISSION")%>%
mutate(symptomatic=case_when(saterm=="asymptomatic" & saoccur==TRUE~FALSE,
saterm=="asymptomatic" & saoccur==FALSE~TRUE,
TRUE~saoccur
))%>%
arrange(desc(symptomatic))%>%
distinct(usubjid, .keep_all =T)%>%
select(usubjid, symptomatic)
symptom<- date_onset%>%
full_join(covid_clinic_diagn, by=c("usubjid"))%>%
full_join(symptomatic, by = c("usubjid"))%>%
full_join(symptom_w, by = c("usubjid"))
if(dtplyr.step){
return(symptom %>% lazy_dt(immutable = FALSE))
} else {
return(symptom %>% as_tibble())
}
}
#' Process data on pregnancy (as comorbidity)
#' @param file.name Path of the dispositions data file (CDISC format)
#' @param dtplyr.step Return the output as \code{dtplyr_step} to avoid unnecessary future calls to \code{as_tibble} or \code{as.data.table}
#' @import dplyr tibble stringr
#' @return Formatted pregnancy data as a tibble or \code{dtplyr_step}
#' @export process.pregnancy.data
process.pregnancy.data <- function(file.name, dtplyr.step = FALSE){
comorbid_pregnancy <- rp%>%
#comorbid_pregnancy <- shared.data.import(pregnancy.file.name, dtplyr.step = FALSE)%>%
#comorbid_pregnancy <- rp_open%>%
filter(rptestcd=="PREGIND") %>%
mutate(comorbid_pregnancy=rpstresc) %>%
mutate(comorbid_pregnancy = case_when(comorbid_pregnancy == "Y" ~ TRUE,
comorbid_pregnancy == "N" ~ FALSE,
TRUE ~ NA)) %>%
select(usubjid,comorbid_pregnancy)
if(dtplyr.step){
return(comorbid_pregnancy %>% lazy_dt(immutable = FALSE))
} else {
return(comorbid_pregnancy %>% as_tibble())
}
}
#' Process data on ICU admission
#' @param file.name Path of the healthcare encounters data file (CDISC format)
#' @param dtplyr.step Return the output as \code{dtplyr_step} to avoid unnecessary future calls to \code{as_tibble} or \code{as.data.table}
#' @import dplyr tibble tidyfast dtplyr
#' @importFrom data.table as.data.table
#' @return Formatted symptom data as a tibble or \code{dtplyr_step}
#' @export process.ICU.data
process.ICU.data <- function(file.name, dtplyr.step = FALSE){
icu <- ho%>%
mutate(hooccur = case_when(hooccur == "Y" ~ TRUE,
hooccur == "N" ~ FALSE,
TRUE ~ NA)) %>%
filter(!is.na (hooccur))%>%
select(usubjid, hodecod, hostdtc, hoendtc, hooccur, hostdy,hoendy)%>%
mutate(hostdtc=substr(hostdtc,1, 10))%>%
mutate(hostdtc=as_date(hostdtc))%>%
mutate(hoendtc=substr(hoendtc,1, 10))%>%
mutate(hoendtc=as_date(hoendtc))
last_ho_datea<-icu%>%
filter(hooccur==TRUE)%>%
filter(hostdtc >= "2020-01-01"|hostdtc<date_pull )%>%
arrange(desc(hostdtc))%>%
distinct(usubjid, .keep_all =T)%>%
select(usubjid,hostdtc)
last_ho_dates<-icu%>%
filter(hooccur==TRUE)%>%
filter(hoendtc>= "2020-01-01"|hoendtc<date_pull)%>%
arrange(desc(hoendtc))%>%
distinct(usubjid, .keep_all =T)%>%
select(usubjid,hoendtc)%>%
left_join(last_ho_datea, by = c("usubjid"))%>%
mutate(date_ho_last=case_when(is.na(hoendtc) ~ hostdtc,
is.na(hostdtc) ~ hoendtc,
hostdtc>hoendtc ~ hostdtc,
hostdtc<=hoendtc ~ hoendtc))%>%
select(usubjid,date_ho_last)
icu <-icu%>%
filter(hodecod=="INTENSIVE CARE UNIT")%>%
arrange(desc(hostdtc))%>%
distinct(usubjid, .keep_all =T)%>%
rename(ever_icu=hooccur)%>%
rename(icu_in=hostdtc)%>%
mutate(icu_in=as_date(icu_in))%>%
mutate(icu_in=replace(icu_in,icu_in < "2020-01-01" | icu_in >date_pull,NA))%>%
rename(icu_out=hoendtc)%>%
mutate(icu_out=as_date(icu_out))%>%
mutate(icu_out=replace(icu_out,icu_out < "2020-01-01" | icu_out>date_pull,NA))%>%
select(-c(hodecod))%>%
full_join(last_ho_dates, by = c("usubjid"))
if(dtplyr.step){
return(icu)
} else {
return(icu %>% as_tibble())
}
}
#' Process data on treatments
#' @param file.name Path of the intervention data file (CDISC format)
#' @param dtplyr.step Return the output as \code{dtplyr_step} to avoid unnecessary future calls to \code{as_tibble} or \code{as.data.table}
#' @import dplyr tibble stringr
#' @return Formatted treatment data (long format) as a tibble or \code{dtplyr_step}
#' @export process.treatment.data
process.treatment.data <- function(file.name, dtplyr.step = FALSE){
#int<-int%>%filter(studyid!="CVZXZMV")
treatment<-int%>%
filter(inpresp =="Y") %>%
filter(inevintx!="BEFORE HOSPITAL ADMISSION")%>%
mutate(inoccur = case_when(inoccur == "Y" ~ TRUE,
inoccur == "N" ~ FALSE,
TRUE ~ NA))%>%
filter(!is.na(inoccur))%>%
filter(incat!="MEDICAL HISTORY" | is.na (incat))%>%
mutate(intrt_original=intrt)%>%
mutate(intrt=toupper(intrt))%>%
mutate(intrt=as.character(intrt))%>%
mutate(inmodify=as.character(inmodify))%>%
mutate(incat=as.character(incat))%>%
mutate(intrt=case_when(inmodify!=""~inmodify,
TRUE ~ intrt))%>%
mutate(intrt=case_when(incat=="EXTRACORPOREAL"~'EXTRACORPOREAL',
incat=="INVASIVE VENTILATION"~'INVASIVE VENTILATION',
incat=="FACE MASK"~'NASAL OR MASK OXYGEN THERAPY',
incat=="NASAL / MASK OXYGEN THERAPY"~'NASAL OR MASK OXYGEN THERAPY',
incat=="INVASIVE VENTILATION"~'INVASIVE VENTILATION',
incat=="NON-INVASIVE VENTILATION "~'NON-INVASIVE VENTILATION ',
incat=="OTHER INTEVENTIONS"~'OTHER INTERVENTIONS',
incat=="PRONE POSITIONING"~'PRONE POSITIONING',
incat=="PRONE VENTILATION"~'PRONE VENTILATION',
incat=="ANTIBIOTIC AGENTS"~ "ANTIBIOTIC AGENTS",
incat=="ANTIFUNGAL AGENTS"~ "ANTIFUNGAL AGENTS",
incat=="ANTIVIRAL AGENTS"~ "ANTIVIRAL AGENTS",
incat=="CORTICOSTEROIDS"~ "CORTICOSTEROIDS",
incat=="ANTIMALARIAL AGENTS"~ "ANTIMALARIAL AGENTS",
incat=="NSAIDS"~"NON-STEROIDAL ANTI-INFLAMMATORY (NSAIDS)",
TRUE~intrt)) %>%
mutate(intrt=case_when(intrt%like%'ECMO'~'EXTRACORPOREAL',
intrt=='EXTRA CORPOREAL LIFE SUPPORT'~'EXTRACORPOREAL',
intrt=='EXTRACORPOREAL SUPPORT'~'EXTRACORPOREAL',
intrt=='PRONE POSITIONING WITH UNKNOWN VENTILATION'~'PRONE POSITION VENTILATION',
intrt=='PRONE VENTILATION'~'PRONE POSITION VENTILATION',
intrt=='CONTINUOUS RENAL REPLACEMENT THERAPIES (CRRT)'~'RENAL REPLACEMENT THERAPIES',
intrt%like%'RENAL REPLACEMENT THERAPY' |
intrt%like% 'DIALYSIS'~ 'RENAL REPLACEMENT THERAPIES',
intrt%like% 'HEMOFILTRATION'~ 'RENAL REPLACEMENT THERAPIES',
intrt=='ERP CVVH'~ 'RENAL REPLACEMENT THERAPIES',
###IMV
intrt=='INVASIVE MECHANICAL LUNG VENTILATION'~'INVASIVE VENTILATION',
intrt=='INVASIVE MECHANICAL VENTILATION'~'INVASIVE VENTILATION',
intrt=='MECHANICAL VENTILATION'~'INVASIVE VENTILATION',
intrt=='RE-INTUBATION'~'INVASIVE VENTILATION',
intrt=='INVASIVE VENTILATION'~'INVASIVE VENTILATION',
intrt%like%'APRV'~'INVASIVE VENTILATION',
intrt=='INTUBATION AND MECHANICAL VENTILATION'~'INVASIVE VENTILATION',
intrt=='MECHANICAL SUPPORT'~'INVASIVE VENTILATION',
intrt%like%'EXTUBATION'~'INVASIVE VENTILATION',
intrt=="VENTILATED"~'INVASIVE VENTILATION',
###NIV
intrt%like%'CPAP'~'NON-INVASIVE VENTILATION',
intrt%like%'BIPAP'~'NON-INVASIVE VENTILATION',
intrt%like%'NON-INVASIVE MECHANICAL VENTILATION (BIPAP, CPAP, OCNAF (OPTIFLOW) ...)'~'NON-INVASIVE VENTILATION',
intrt%like%'NON-INVASIVE VENTILATION'~'NON-INVASIVE VENTILATION',
intrt=='NON-INVASIVE MECHANICAL VENTILATION'~'NON-INVASIVE VENTILATION',
intrt=='NON-INVASIVE POSITIVE PRESSURE VENTILATION'~'NON-INVASIVE VENTILATION',
intrt=='NON-INVASIVE RESPIRATORY SUPPORT'~'NON-INVASIVE VENTILATION',
TRUE ~ intrt))%>%
mutate(intrt=case_when(intrt%like%'OTHER INTERVENTION'~'OTHER INTERVENTIONS',
intrt%like%'CHEMOTHERAPY'| intrt%like%'ANTI-DIABETIC MEDICATIONS'|intrt%like%'BRONCHOSCOPY'|
intrt%like%'PROTON PUMP INHIBITORS'|intrt%like%'STATINS'|intrt%like%'MORPHINE'|
intrt%like%'HALOPERIDOL'|intrt%like%'OLANZAPINE'~'OTHER INTERVENTIONS',
intrt=='OTHER TARGETED COVID-19 MEDICATIONS'~'OTHER INTERVENTIONS',
intrt=='OTHER TREATMENTS FOR COVID19'~'OTHER INTERVENTIONS',
intrt%like%"NON-STEROIDAL"~"NON-STEROIDAL ANTI-INFLAMMATORY",
intrt%like%"NON STEROIDAL"~"NON-STEROIDAL ANTI-INFLAMMATORY",
TRUE ~ intrt))%>%
mutate(intrt=case_when(intrt=='NASAL CANULA'|intrt=='NASAL CANNULA'~'NASAL OXYGEN THERAPY',
intrt%like%'SURGICAL FEEDING TUBE'~'TOTAL PARENTERAL NUTRITION',
intrt=='FACE MASK'~'MASK OXYGEN THERAPY',
####HFNC
intrt=='OXYGEN THERAPY WITH HIGH FLOW NASAL CANULA'~'HIGH-FLOW NASAL CANULA OXYGEN THERAPY',
intrt=='HIGH-FLOW NASAL CANNULA OXYGEN THERAPY'~'HIGH-FLOW NASAL CANULA OXYGEN THERAPY',
###Prone positioning
intrt%like%'PRONACI'~'PRONE POSITIONING',
intrt=='PRONE POSITIONING'~'PRONE POSITIONING',
intrt%like%'TRACHEOSTOMY'~'TRACHEOSTOMY',
intrt%like%'NITRIC OXIDE'~'INHALED NITRIC OXIDE',
###Corticosteroids
intrt=="CORTICOSTEROID"~ "CORTICOSTEROIDS",
intrt=="DEXAMETHASONE"~ "CORTICOSTEROIDS",
intrt=="BETAMETHASONE"~ "CORTICOSTEROIDS",
intrt%like%"PREDNISOLONE"~ "CORTICOSTEROIDS",
intrt=="ORAL STEROIDS"~ "CORTICOSTEROIDS",
intrt=="STEROIDS"~ "CORTICOSTEROIDS",
intrt=="STEROIDS"~ "convalescent_plasma",
intrt%like%"HYDROCORTISONE"~ "CORTICOSTEROIDS",
intrt%like%"BLOOD TRANSFUSION OR BLOOD PRODUCT"~ "BLOOD TRANSFUSION OR BLOOD PRODUCT",
TRUE ~ intrt))%>%
mutate(intrt=case_when(intrt%like%"ANTIVIRAL" ~ "ANTIVIRAL AGENTS",
intrt%like%"ARV" ~ "ANTIVIRAL AGENTS",
intrt%like%"ANTIRETROVIRAL" ~ "ANTIVIRAL AGENTS",
intrt%like%"RIBAVIRIN" ~ "ANTIVIRAL AGENTS",
intrt%like%"LOPINAVIR AND RITONAVIR" ~ "ANTIVIRAL AGENTS",
intrt%like%"LOPINAVIR" ~ "ANTIVIRAL AGENTS",
intrt%like%"OSELTAMIVIR" ~ "ANTIVIRAL AGENTS",
intrt%like%"REMDESIVIR" ~ "ANTIVIRAL AGENTS",
intrt%like%"REMDESIVIR" ~ "ANTIVIRAL AGENTS",
intrt%like%"NEURAMINIDASE INHIBITORS" ~ "ANTIVIRAL AGENTS",
intrt%like%"ZANAMIVIR" ~ "ANTIVIRAL AGENTS",
intrt%like%"RIBAVARIN" ~ "ANTIVIRAL AGENTS",
intrt%like%"FLUCLOXACILLIN"~ "ANTIBIOTIC AGENTS",
intrt%like%"ANTIBIOTIC"~ "ANTIBIOTIC AGENTS",
intrt%like%"AMIKACIN"~ "ANTIBIOTIC AGENTS",
intrt%like%"AMOX"~ "ANTIBIOTIC AGENTS",
intrt%like%"AUGUMENTIN"~ "ANTIBIOTIC AGENTS",
intrt%like%"AZITHROMYCIN"~ "ANTIBIOTIC AGENTS",
intrt%like%"AZITHRYOMYCIN"~ "ANTIBIOTIC AGENTS",
intrt%like%"BENZY"~ "ANTIBIOTIC AGENTS",
intrt%like%"AUGUMENTIN"~ "ANTIBIOTIC AGENTS",
TRUE ~ intrt))%>%
mutate(intrt=case_when(intrt%like%"AZITHRYOMYCIN"~ "ANTIBIOTIC AGENTS",
intrt%like%"CEFTR"~ "ANTIBIOTIC AGENTS",
intrt%like%"CEFR"~ "ANTIBIOTIC AGENTS",
intrt%like%"DOXYCYCLINE"~ "ANTIBIOTIC AGENTS",
intrt%like%"CHLORAMPHENICOL"~ "ANTIBIOTIC AGENTS",
intrt%like%"CIPROFLOXACIN"~ "ANTIBIOTIC AGENTS",
intrt%like%"GENTAMICIN"~ "ANTIBIOTIC AGENTS",
intrt%like%"MEROPENEM"~ "ANTIBIOTIC AGENTS",
intrt%like%"METRONIDAZOLE"~ "ANTIBIOTIC AGENTS",
intrt%like%"ANTIMALARIAL" | intrt%like%"CHLOROQUINE" ~ "ANTIMALARIAL AGENTS",
intrt%like%"ANTIFUNGAL" ~ "ANTIFUNGAL AGENTS",
intrt %like% "OROGASTRIC"~"NASO/ NASOGASTRIC ORAL/OROGASTRIC FLUIDS",
intrt %like% "NGT OR OGT REQUIRED FOR NUTRITION"~"NASO/ NASOGASTRIC ORAL/OROGASTRIC FLUIDS",
intrt%like%'DOBUTAMINE' | intrt%like%'DOPAMINE' | intrt%like%'MILRINONE'
| intrt%like%'LEVOSIMENDAN' | intrt%like%'EPINEPHRINE' | intrt%like%'NOREPINEPRINE'
| intrt%like%'INOTROPES' |intrt%like%'VASOPRESS' |intrt%like%'NORADRENALINE' |
intrt%like%'ADRENALINE' |intrt%like%'BETA BLOCKER' ~'INOTROPES / VASOPRESSORS',
TRUE ~ intrt))%>%
mutate(intrt=case_when(intrt%like%'IMMUNOGLOBULI' ~ "convalescent_plasma",
intrt%like%'IMMUNOSUPPRES' ~ "IMMUNOSUPPRESSANTS",
intrt%like%'IMMUNOSTIMULANTS' ~ "IMMUNOSUPPRESSANTS",
intrt%like%'IMMUNOTHERAPY' ~ "IMMUNOSUPPRESSANTS",
intrt=="IL6 INHIBITOR" ~ "IMMUNOSUPPRESSANTS",
intrt=="TOCILIZUMAB" ~ "IMMUNOSUPPRESSANTS",
intrt%like%"INTERFERON" ~ "IMMUNOSTIMULANTS",
intrt%like%"HEPARIN" ~ "THERAPEUTIC ANTICOAGULANT",
intrt%like%"NOXAPARIN" ~ "THERAPEUTIC ANTICOAGULANT",
intrt=="ENOXAPARIN" ~ "THERAPEUTIC ANTICOAGULANT",
intrt%like%"SPIRONOLACTONE" ~ "DIURETICS",
intrt%like%"DIURETIC" ~ "DIURETICS",
intrt%like%"NITROUS OXIDE" ~ "inhaled_nitric_oxide",
intrt=="CPR" ~ "Cardiopulmonary resuscitation",
intrt%like%"EXPERIMENTAL AGENT" ~ "EXPERIMENTAL AGENTS",
intrt%like%"SARILUMAB" ~ "EXPERIMENTAL AGENTS",
intrt%like%"IV FLUID" ~ "INTRAVENOUS FLUIDS",
intrt%like%"I.V. SOLUTIONS" ~ "INTRAVENOUS FLUIDS",
intrt %like% "ANGIOTENSIN" | intrt %like% "ACE"~ "AGENTS ACTING ON THE RENIN-ANGIOTENSIN SYSTEM",
intrt%like%"ANTIINFLAMMATORY" ~ "ANTIINFLAMMATORY",
TRUE ~ intrt))%>% as.data.frame()%>%
select(studyid,usubjid,'treatment'=intrt,inoccur,intrt_original,inmodify,incat, inevintx,
indur,indtc,instdtc,inendtc,indy)%>%
mutate(treatment = iconv(treatment, to ="ASCII//TRANSLIT") %>% tolower()) %>%
mutate(treatment = str_remove_all(treatment, "\\s*\\([^)]*\\)")) %>%
mutate(treatment = str_replace_all(treatment, " - ", "_")) %>%
mutate(treatment = str_replace_all(treatment, "-", "_")) %>%
mutate(treatment = str_replace_all(treatment, "/| / ", "_")) %>%
mutate(treatment = str_replace_all(treatment, " ", "_"))
if(dtplyr.step){
return(treatment)
} else {
return(treatment %>% as_tibble())
}
}
#' Process data on the most common treatments
#' @param input Either the path of the interventions data file (CDISC format) or output of \code{process.treatment.data}
#' @param minimum The minimum number of times a treatment need appear to be considered "common"; default 1000.
#' @param dtplyr.step Return the output as \code{dtplyr_step} to avoid unnecessary future calls to \code{as_tibble} or \code{as.data.table}
#' @import dplyr tibble dtplyr tidyfast
#' @importFrom data.table as.data.table
#' @importFrom glue glue
#' @return Formatted common treatment data (wide format) as a tibble or \code{dtplyr_step}
#' @export process.common.treatment.data
process.common.treatment.data <- function(file.name, minimum=10, dtplyr.step = FALSE){
oxy_within_d1<-imp_int%>%
mutate(treatment=case_when(treatment=="extracorporeal" |
treatment=="inhaled_nitric_oxide" |
treatment=="prone_position_ventilation" |
treatment=="respiratory_support" |
treatment=="tracheostomy" |
treatment=="high_flow_nasal_cannula" |
treatment=="invasive_ventilation" |
treatment=="mask_oxygen_therapy" |
treatment=="nasal_oxygen_therapy" |
treatment=="non_invasive_ventilation"~"oxygen_therapy",
TRUE~treatment))%>%
filter(treatment=="oxygen_therapy")%>%
filter(inevintx=="AT HOSPITAL ADMISSION"|indy==1)%>%
mutate(oxytreat_when=case_when(inevintx=="AT HOSPITAL ADMISSION"~"at_admi",
indy==1~"within_24h"))%>%
arrange(desc(inoccur))%>%
distinct(usubjid,treatment, .keep_all =T)%>%
select(usubjid,"d1_oxygen_therapy"=inoccur)
date_in_last <- imp_int %>%
filter(inoccur==TRUE)%>%
mutate(date_in_last=substr(indtc,1, 10))%>%
mutate(date_in_last=as_date(date_in_last))%>%
filter(date_in_last >= "2020-01-01"| date_in_last<date_pull)%>%
arrange(desc(inoccur))%>%
distinct(usubjid, .keep_all =T)%>%
select(usubjid, date_in_last )
treatment <- imp_int%>%
#filter(!is.na(indtc))%>%
group_by(treatment)%>%
arrange(desc(inoccur))%>%
mutate(n = sum(!is.na(inoccur)))%>%
filter(n >= eval(!!minimum))%>%
ungroup()%>%
filter(treatment!="extracorporeal" &
treatment!="inhaled_nitric_oxide" &
treatment!="oxygen_therapy" &
treatment!="prone_position_ventilation" &
treatment!="prone_ventilation" &
treatment!="respiratory_support" &
treatment!="tracheostomy" &
treatment!="prone_positioning")%>%
#mutate(treatment=replace(treatment,treatment=="cpr","cardiopulmonary_resuscitation"))%>%
filter(treatment!="covid_19_vaccination")%>%
filter(treatment!="supplemental_oxygen_fio2")%>%
arrange(desc(inoccur))%>%
distinct(usubjid, treatment, .keep_all =T)%>%
#mutate(treatment = paste0("treat_",treatment)) %>%
mutate(treatment = glue("treat_{treatment}", treatment = treatment))%>%
as.data.table()%>%
dt_pivot_wider(id_cols = usubjid, names_from = treatment, values_from = inoccur)%>%
as.data.frame()%>%
full_join(date_in_last)
####calculating oxygen therapy overall
treat_oxy <- imp_int%>%
mutate(treatment=case_when(treatment=="extracorporeal" |
treatment=="inhaled_nitric_oxide" |
treatment=="prone_position_ventilation" |
treatment=="respiratory_support" |
treatment=="tracheostomy" |
treatment=="high_flow_nasal_cannula" |
treatment=="invasive_ventilation" |
treatment=="mask_oxygen_therapy" |
treatment=="nasal_oxygen_therapy" |
treatment=="oxygen_therapy"|
treatment=="non_invasive_ventilation"~"treat_oxygen_therapy",
TRUE~treatment))%>%
filter(treatment=="treat_oxygen_therapy")%>%
arrange(desc(inoccur))%>%
distinct(usubjid, .keep_all =T)%>%
select(usubjid,"treat_oxygen_therapy"=inoccur)
###adding duration for inasive_ventilation and non_invasive_ventilation
indur<-imp_int%>%select(usubjid,treatment, inoccur,indur,indtc,instdtc,inendtc,indy)%>%
filter(treatment=="invasive_ventilation"|treatment=="non_invasive_ventilation")%>%
mutate(treatment=case_when(treatment=='non_invasive_ventilation'~'dur_niv',
treatment=='invasive_ventilation'~'dur_imv',
TRUE~treatment))%>%
mutate(indur_clean=as.numeric(gsub("[^0-9.]", "",indur)))%>%
filter(!is.na(indur_clean) | indur_clean!="")%>%
distinct(usubjid,treatment, .keep_all =T)%>%
dt_pivot_wider(id_cols = usubjid, names_from = treatment, values_from = indur_clean)%>%
as_tibble()
vent_st_instdtc<-imp_int%>%select(usubjid,treatment, inoccur,indur,indtc,instdtc)%>%
filter(treatment=="invasive_ventilation"|treatment=="non_invasive_ventilation")%>%
mutate(treatment=case_when(treatment=='non_invasive_ventilation'~'date_niv_st',
treatment=='invasive_ventilation'~'date_imv_st',
TRUE~treatment))%>%
filter(inoccur==TRUE)%>%
mutate(instdtc=substr(instdtc,1, 10))%>%
mutate(instdtc=as_date(instdtc))%>%
filter(!is.na(instdtc))%>%
arrange(instdtc)%>%
distinct(usubjid,treatment, .keep_all =T)%>%
dt_pivot_wider(id_cols = usubjid, names_from = treatment, values_from = instdtc)%>%
as_tibble()
vent_st_indtc<-imp_int%>%
filter(inevintx=="00:00-24:00 ON DAY OF ASSESSMENT")%>%
filter(treatment=="invasive_ventilation"|treatment=="non_invasive_ventilation")%>%
mutate(treatment=case_when(treatment=='non_invasive_ventilation'~'date_niv_indtc_st',
treatment=='invasive_ventilation'~'date_imv_indtc_st',
TRUE~treatment))%>%
filter(inoccur==TRUE)%>%
mutate(indtc=substr(indtc,1, 10))%>%
mutate(indtc=as_date(indtc))%>%
filter(!is.na(indtc))%>%
arrange(instdtc)%>%
distinct(usubjid,treatment, .keep_all =T)%>%
dt_pivot_wider(id_cols = usubjid, names_from = treatment, values_from = indtc)%>%
as_tibble()
treatment <-treatment%>%
full_join(treat_oxy)%>%
full_join(oxy_within_d1)%>%
full_join(indur)%>%
full_join(vent_st_instdtc)%>%
full_join(vent_st_indtc)%>%
#full_join(vent_at_adm)%>%
mutate(date_imv_st=case_when(is.na(date_imv_st)~date_imv_indtc_st,
TRUE~date_imv_st))%>%
select(-c(date_imv_indtc_st))%>%
rename("date_niv_st"=date_niv_indtc_st)
if(dtplyr.step){
return(treatment) %>% lazy_dt(immutable = FALSE)
} else {
return(treatment %>% as_tibble())
}
}
#' Process data on the most common icu treatments
#' @param input Either the path of the interventions data file (CDISC format) or output of \code{process.treatment.data}
#' @param minimum The minimum number of times a treatment need appear to be considered "common"; default 1000.
#' @param dtplyr.step Return the output as \code{dtplyr_step} to avoid unnecessary future calls to \code{as_tibble} or \code{as.data.table}
#' @import dplyr tibble dtplyr tidyfast
#' @importFrom data.table as.data.table
#' @importFrom glue glue
#' @return Formatted common treatment data (wide format) as a tibble or \code{dtplyr_step}
#' @export process.treatment.icu.data
process.treatment.icu.data <- function(file.name,imp_icu,imp_dm,imp_ds, minimum=10, dtplyr.step = FALSE){
adm_date<-imp_dm%>%
select(usubjid,date_admit)
#tabindy_icu<-tabyl(imp_treat_icu$indy)
icu_ever<-imp_icu%>%
filter(ever_icu==TRUE)%>%
filter(!is.na(icu_in))%>%
left_join(adm_date)%>%
left_join(imp_ds)%>%
mutate(icu_dy_in=icu_in-date_admit+1)%>%
mutate(icu_dy_in=as.double(icu_dy_in))%>%
mutate(icu_dy_in=case_when(icu_dy_in<0~NA_real_,
TRUE~icu_dy_in))%>%
mutate(icu_dy_out=case_when(!is.na(icu_out)~icu_dy_in+(icu_out-icu_in),
is.na(icu_out)&!is.na(date_ho_last)~icu_dy_in+(date_ho_last-icu_in),
is.na(icu_out)&is.na(date_ho_last)~icu_dy_in+(date_outcome-icu_in),
#is.na(icu_out)&is.na(date_ho_last)~icu_dy_in+(date_outcome-icu_in),
TRUE~NA_real_))%>%
mutate(icu_dy_out=as.integer(icu_dy_out))%>%
mutate(hoendy=case_when(is.na(hoendy)~icu_dy_out,
TRUE~hoendy))%>%
select(usubjid,hostdy,hoendy)#%>%
#filter(!is.na(hoendy)& hoendy>-1 &!is.na(hostdy)& hostdy>-1)%>%
#filter(!is.na(hoendy) & !is.na(hostdy))%>%
#left_join(imp_int)%>%
treat_oxy_icu <- imp_int%>%
mutate(treatment=case_when(treatment=="extracorporeal" |
treatment=="inhaled_nitric_oxide" |
treatment=="prone_position_ventilation" |
treatment=="respiratory_support" |
treatment=="tracheostomy" |
treatment=="high_flow_nasal_cannula" |
treatment=="invasive_ventilation" |
treatment=="mask_oxygen_therapy" |
treatment=="nasal_oxygen_therapy" |
treatment=="oxygen_therapy"|
treatment=="non_invasive_ventilation"~"treat_oxygen_therapy",
TRUE~treatment))%>%
filter(treatment=="treat_oxygen_therapy")%>%
arrange(desc(inoccur))%>%
left_join(icu_ever,by = c("usubjid"))%>%
mutate(indy=as.numeric(indy))%>%
mutate(hostdy=as.numeric(hostdy))%>%
mutate(hoendy=as.numeric(hoendy))%>%
mutate(int_icu=case_when((indy>=hostdy)~ TRUE,
TRUE ~ FALSE))%>%
filter(int_icu==TRUE)%>%
arrange(desc(inoccur))%>%
distinct(usubjid, treatment, .keep_all =T)%>%
select(usubjid,"icu_treat_oxygen_therapy"=inoccur)
imp_treat_icu<-imp_int%>%
filter(!is.na(indy))%>%
group_by(treatment) %>%
arrange(desc(inoccur))%>%
mutate(n = sum(!is.na(inoccur))) %>%
filter(n >= eval(!!minimum)) %>%
ungroup()%>%
mutate(treatment=replace(treatment,treatment=="cpr","cardiopulmonary_resuscitation"))%>%
filter(treatment!="covid_19_vaccination")%>%
filter(treatment!="supplemental_oxygen_fio2")%>%
filter(treatment!="extracorporeal" &
treatment!="inhaled_nitric_oxide" &
treatment!="oxygen_therapy" &
treatment!="prone_position_ventilation" &
treatment!="prone_ventilation" &
treatment!="respiratory_support" &
treatment!="tracheostomy" &
treatment!="prone_positioning")%>%
arrange(desc(inoccur))%>%
#mutate(indtc=as.Date(indtc))%>%
#filter(indtc>= "2020-01-01"|indtc<date_pull)%>%
left_join(icu_ever,by = c("usubjid"))%>%
#filter((indy>=hostdy)&(indy<=hoendy))%>%
mutate(indy=as.numeric(indy))%>%
mutate(hostdy=as.numeric(hostdy))%>%
mutate(hoendy=as.numeric(hoendy))%>%
#mutate(int_icu=case_when((indy>=hostdy)&(indy<=hoendy) ~ TRUE,
# TRUE ~ FALSE))%>%
mutate(int_icu=case_when((indy>=hostdy)~ TRUE,
TRUE ~ FALSE))%>%
#mutate(int_icu=case_when(indtc>=icu_in ~ TRUE,
# TRUE ~ FALSE))%>%
filter(int_icu==TRUE)%>%
arrange(desc(inoccur))%>%
distinct(usubjid, treatment, .keep_all =T)%>%
mutate(treatment = glue("icu_treat_{treatment}", treatment = treatment)) %>%
as.data.table() %>%
dt_pivot_wider(id_cols = usubjid, names_from = treatment, values_from = inoccur)%>%
full_join(treat_oxy)
if(dtplyr.step){
return(imp_treat_icu) %>% lazy_dt(immutable = FALSE)
} else {
return(imp_treat_icu %>% as_tibble())
}
}
#' Process data on vital sign
#' @param file.name Path of the dispositions data file (CDISC format)
#' @param dtplyr.step Return the output as \code{dtplyr_step} to avoid unnecessary future calls to \code{as_tibble} or \code{as.data.table}
#' @import dplyr tibble dtplyr tidyfast
#' @importFrom data.table as.data.table
#' @importFrom glue glue
#' @return Formatted vital sign (wide format) as a tibble or \code{dtplyr_step}
#' @export process.vital.sign.data
process.vital.sign.data <- function(file.name, dtplyr.step = FALSE){
vital_sign <- vs %>%
select(usubjid, vstestcd, vscat,vsstresn,vsstresu, vsdtc, vso2src) %>%
filter(vscat=="SIGNS AND SYMPTOMS AT HOSPITAL ADMISSION" | vscat=="SIGNS AND SYMPTOMS AT ADMISSION")%>%
mutate(vsstresn=as.numeric(vsstresn))%>%
mutate(vsstresn=case_when(vstestcd=="OXYSAT"& vsstresn< 1~ NA_real_,
vstestcd=="OXYSAT"& vsstresn> 100~ NA_real_,
vstestcd=="BMI"& vsstresn< 0~ NA_real_,
vstestcd=="BMI"& vsstresn> 100~ NA_real_,
vstestcd=="DIABP"& vsstresn< 0~ NA_real_,
vstestcd=="DIABP"& vsstresn> 300~ NA_real_,
vstestcd=="HEIGHT"& vsstresn< 0~ NA_real_,
vstestcd=="HEIGHT"& vsstresn> 250~ NA_real_,
vstestcd=="HR"& vsstresn< 0~ NA_real_,
vstestcd=="HR"& vsstresn> 250~ NA_real_,
vstestcd=="MAP"& vsstresn< 0~ NA_real_,
vstestcd=="MAP"& vsstresn> 250~ NA_real_,
vstestcd=="MUARMCIR"& vsstresn< 0~ NA_real_,
vstestcd=="MUARMCIR"& vsstresn> 100~ NA_real_,
vstestcd=="PULSE"& vsstresn< 0~ NA_real_,
vstestcd=="PULSE"& vsstresn> 250~ NA_real_,
vstestcd=="RESP"& vsstresn< 0~ NA_real_,
vstestcd=="RESP"& vsstresn> 60~ NA_real_,
vstestcd=="SYSBP"& vsstresn< 0~ NA_real_,
vstestcd=="SYSBP"& vsstresn> 250~ NA_real_,
vstestcd=="TEMP"& vsstresn< 30~ NA_real_,
vstestcd=="TEMP"& vsstresn> 44~ NA_real_,
vstestcd=="WEIGHT"& vsstresn< 0~ NA_real_,
vstestcd=="WEIGHT"& vsstresn> 300~ NA_real_,
TRUE~vsstresn))%>%
filter(!is.na(vsstresn))%>%
arrange(desc(vsdtc))%>%
distinct(usubjid,vstestcd, .keep_all =T)%>%
mutate(vso2src=case_when(vso2src==""&vstestcd=="OXYSAT"~'UNKNOWN',
TRUE~vso2src))%>%
mutate(vso2src= str_replace_all(vso2src, " ", "_"))%>%
mutate(vstestcd=case_when(vstestcd=="OXYSAT"~paste0(vstestcd,"_",vso2src),
TRUE~vstestcd))%>%
mutate(vstestcd = paste0("vs_",vstestcd)) %>%
#mutate(vstestcd = glue("vs_{vstestcd}", vstestcd = vstestcd))%>%
mutate(vstestcd = iconv(vstestcd, to ="ASCII//TRANSLIT") %>% tolower()) %>%
as.data.table() %>%
dt_pivot_wider(id_cols = usubjid, names_from = vstestcd, values_from = vsstresn)%>%
as.data.frame() %>%
mutate(vs_oxysat=case_when(!is.na(vs_oxysat_oxygen_therapy)~vs_oxysat_oxygen_therapy,
!is.na(vs_oxysat_room_air)~vs_oxysat_room_air,
TRUE~vs_oxysat_unknown))
if(dtplyr.step){
return(vital_sign)
} else {
return(vital_sign %>% as_tibble())
}
}
#' Process data on laboratory
#' @param file.name Path of the dispositions data file (CDISC format)
#' @param dtplyr.step Return the output as \code{dtplyr_step} to avoid unnecessary future calls to \code{as_tibble} or \code{as.data.table}
#' @import dplyr tibble dtplyr tidyfast
#' @importFrom data.table as.data.table
#' @importFrom glue glue
#' @return Formatted laboratory (wide format) as a tibble or \code{dtplyr_step}
#' @export process.laboratory.data
process.laboratory.data <- function(file.name, dtplyr.step = FALSE){
laboratory <- lb%>%
select(usubjid, lbdy, lbtestcd, lbcat,lborres,lbdtc) %>%
mutate(lborres=replace(lborres,lborres=="",NA))%>%
mutate(studyid=substr(usubjid,1, 7))%>%
mutate(lbcat=case_when(lbdy==1 & (studyid=="CVCCPUK"|
studyid=="CVMEWUS" |
studyid=="CORE"|
studyid=="CVTDWXD"|
studyid=="CVTTYLU"|
studyid=="CVZXZMV"|
studyid=="CVKBQEI") ~"LABORATORY RESULTS ON ADMISSION",
#lbdy==1 & studyid=="CVMEWUS"~"LABORATORY RESULTS ON ADMISSION",
TRUE~as.character(lbcat)))%>%
filter(lbcat=="LABORATORY RESULTS ON ADMISSION")%>%
filter(lbtestcd=="ALT"|
lbtestcd=="APTT"|
lbtestcd=="CRP"|
lbtestcd=="LYM"|
lbtestcd=="NEUT"|
lbtestcd=="PT"|
lbtestcd=="WBC"|
lbtestcd=="BILI"|
lbtestcd=="AST"|
lbtestcd=="UREAN")%>%
mutate(lborres=as.numeric(lborres))%>%
filter(!is.na(lborres))%>%
arrange(desc(lbdtc))%>%
distinct(usubjid,lbtestcd, .keep_all =T)%>%
mutate(lborres=case_when(lbtestcd=="NEUT" & lborres>100 ~ lborres/1000,
lbtestcd=="LYM" & lborres>100 ~ lborres/1000,
lbtestcd=="WBC" & lborres>100 ~ lborres/1000,
lbtestcd=="ALT" & lborres>2000 ~ NA_real_,
lbtestcd=="ALT" & lborres<0 ~ NA_real_,
lbtestcd=="AST" & lborres>2000 ~ NA_real_,
lbtestcd=="AST" & lborres<0 ~ NA_real_,
lbtestcd=="BILI" & lborres>2000 ~ NA_real_,
lbtestcd=="BILI" & lborres<0 ~ NA_real_,
lbtestcd=="CRP" & lborres>500 ~ NA_real_,
lbtestcd=="CRP" & lborres<0 ~ NA_real_,
lbtestcd=="PT" & lborres>105 ~ NA_real_,
lbtestcd=="PT" & lborres<0 ~ NA_real_,
lbtestcd=="UREAN" & lborres>100 ~ NA_real_,
lbtestcd=="UREAN" & lborres<0 ~ NA_real_,
lbtestcd=="APTT" & lborres>2000 ~ NA_real_,
lbtestcd=="APTT" & lborres<0 ~ NA_real_,
TRUE ~ lborres ))%>%
mutate(lborres=case_when(lbtestcd=="NEUT" & lborres>100 ~ NA_real_,
lbtestcd=="NEUT" & lborres<0 ~ NA_real_,
lbtestcd=="LYM" & lborres>100 ~ NA_real_,
lbtestcd=="LYM" & lborres<0 ~ NA_real_,
lbtestcd=="WBC" & lborres>100 ~ NA_real_,
lbtestcd=="WBC" & lborres<0 ~ NA_real_,
TRUE ~ lborres ))%>%
mutate(lbtestcd = paste0("lab_",lbtestcd )) %>%
#mutate(lbtestcd = glue("lab_{lbtestcd}", lbtestcd = lbtestcd)) %>%
mutate(lbtestcd = iconv(lbtestcd, to ="ASCII//TRANSLIT") %>% tolower()) %>%
as.data.table() %>%
dt_pivot_wider(id_cols = usubjid, names_from = lbtestcd, values_from = lborres)
if(dtplyr.step){
return(laboratory)
} else {
return(laboratory%>% as_tibble())
}
}
#' Process data on outcomes
#' @param file.name Path of the dispositions data file (CDISC format)
#' @param dtplyr.step Return the output as \code{dtplyr_step} to avoid unnecessary future calls to \code{as_tibble} or \code{as.data.table}
#' @import dplyr tibble stringr
#' @return Formatted outcome data (long format) as a tibble or \code{dtplyr_step}
#' @export process.outcome.data
process.outcome.data <- function(file.name, dtplyr.step = FALSE){
outcome <- ds%>%
select(usubjid, dsterm, "date_outcome" = dsstdtc, dsmodify) %>%
mutate(date_outcome=substr(date_outcome,1, 10))%>%
mutate(date_outcome=as_date(date_outcome))%>%
mutate(date_outcome=replace(date_outcome,date_outcome< "2020-01-01",NA))%>%
mutate(date_outcome=replace(date_outcome,date_outcome>date_pull,NA))%>%
mutate(outcome=tolower(dsterm))%>%
mutate(outcome=case_when(outcome=="palliative"~"transferred",
outcome=="transferred to another unit"~"ongoing care",
outcome=="Ongoing health care needs NOT related to COVID episode"~"discharge",
outcome==""~NA_character_,
TRUE~outcome))%>%
mutate(outcome=case_when(outcome%like%"hospitalis"~"ongoing care",
outcome%like%"hospitaliz"~"ongoing care",
outcome%like%"ongoing"~"ongoing care",
outcome=="in hospital"~"ongoing care",
outcome%like%"death"~"death",
outcome=="died"~"death",
outcome=="deceased"~"death",
outcome=="died (non-covid)"~"death",
#outcome=="Death In Hospital"~"Death",
outcome=="alive"~"discharge",
outcome%like%"discharge"~"discharge",
outcome%like%"transfer"~"transferred",
outcome=="long term care facility"~"transferred",
outcome=="quarantine center"~"transferred",
outcome=="missing in database"~"unknown outcome",
outcome=="unknown"~"unknown outcome",
outcome=="not recorded"~"unknown outcome",
TRUE ~ outcome))%>%
group_by(usubjid) %>%
mutate(count=1)%>%
mutate(n = sum(count)) %>%
filter(n == 1)%>%
select(-c(dsterm,dsmodify,n,count))
if(dtplyr.step){
return(outcome)
} else {
return(outcome %>% as_tibble())
}
}
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.