Create Pairing values with later values to follow progression

Author: HeatherARobinson

Pairing values with later values to follow progression

@@ -0,0 +1,103 @@
+#It can be useful to follow patient progression over time. The following code calls the first recorded metric after a 6 month interval
+#where this data is available, and appends this to each row.
+
+###################################################################################
+#FIRST LOG CREATININE AFTER 6M
+
+first<-crea.rep[,c("PatientID","event.date","log_CREA")]
+first$event.date2<-first$event.date+180
+indx1<-neardate(first$PatientID,crea.rep$PatientID,first$event.date2,crea.rep$event.date,best="after",nomatch=NA_integer_)
+first$LogCrea6M<-crea.rep[indx1,"log_CREA"]
+
+#THE ABOVE CODE PREFERABLY MATCHES A RESULT FROM AFTER 6 MONTHS, BUT OTHERWISE SELECTS ONE BEFORE IT. 
+#WE NEED TO REMOVE DATA COLLECTED FROM BEFORE 6 MONTHS:
+
+maxes<-first %>%
+group_by(PatientID) %>%
+slice(which.max(event.date)) %>%
+as.data.frame
+maxes<-maxes[,c(1,2)]
+colnames(maxes)<-c("PatientID","MaxDate")
+first<-merge(first,maxes)
+first<-first[first$event.date2<=first$MaxDate,c(1,2,5)]
+ first$PatientID<-as.character(first$PatientID)
+crea.rep$PatientID<-as.character(crea.rep$PatientID)
+crea.rep<-merge(crea.rep,first,all.x=TRUE,all.y=FALSE)
+crea.rep$LogCrea6M<-unlist(crea.rep$LogCrea6M)
+
+###################################################################################
+#FIRST EGFR AFTER 6M-MDRD
+
+first<-crea.rep[,c("PatientID","event.date","MDRDeGFR")]
+first$event.date2<-first$event.date+180
+
+indx1<-neardate(first$PatientID,crea.rep$PatientID,first$event.date2,crea.rep$event.date,best="prior",nomatch=NA_integer_)
+first$MDRDeGFR6M<-crea.rep[indx1,"MDRDeGFR"]
+
+#THE ABOVE CODE PREFERABLY MATCHES A RESULT FROM AFTER 6 MONTHS, BUT OTHERWISE SELECTS ONE BEFORE IT. 
+#WE NEED TO REMOVE DATA COLLECTED FROM BEFORE 6 MONTHS:
+
+maxes<-first %>%
+group_by(PatientID) %>%
+slice(which.max(event.date)) %>%
+as.data.frame
+maxes<-maxes[,c(1,2)]
+colnames(maxes)<-c("PatientID","MaxDate")
+
+first$PatientID<-as.character(first$PatientID)
+maxes$PatientID<-as.character(maxes$PatientID)
+first<-merge(first,maxes)
+first<-first[first$event.date2<=first$MaxDate,c(1,2,5)]
+
+crea.rep<-merge(crea.rep,first,all.x=TRUE)
+crea.rep$MDRDeGFR6M<-unlist(crea.rep$MDRDeGFR6M)
+
+################################################################################### CHECKED
+#FIRST CKDEPI AFTER 6M
+
+first<-crea.rep[,c("PatientID","event.date","CKDEPIeGFR")]
+first$event.date2<-first$event.date+180
+
+indx1<-neardate(first$PatientID,crea.rep$PatientID,first$event.date2,crea.rep$event.date,best="prior",nomatch=NA_integer_)
+first$CKDEPIeGFR6M<-crea.rep[indx1,"CKDEPIeGFR"]
+
+#THE ABOVE CODE PREFERABLY MATCHES A RESULT FROM AFTER 6 MONTHS, BUT OTHERWISE SELECTS ONE BEFORE IT. 
+#WE NEED TO REMOVE DATA COLLECTED FROM BEFORE 6 MONTHS:
+
+maxes<-first %>%
+group_by(PatientID) %>%
+slice(which.max(event.date)) %>%
+as.data.frame
+maxes<-maxes[,c(1,2)]
+colnames(maxes)<-c("PatientID","MaxDate")
+first<-merge(first,maxes)
+first<-first[first$event.date2<=first$MaxDate,c(1,2,5)]
+
+crea.rep<-merge(crea.rep,first,all.x=TRUE)
+crea.rep$CKDEPIeGFR6M<-unlist(crea.rep$CKDEPIeGFR6M)
+
+################################################################################### CHECKED
+
+#FIRST CREATININE AFTER 6M
+first<-crea.rep[,c("PatientID","event.date","Creatinine")]
+first$event.date2<-first$event.date+180
+
+indx1<-neardate(first$PatientID,crea.rep$PatientID,first$event.date2,crea.rep$event.date,best="prior",nomatch=NA_integer_)
+first$Creatinine6M<-crea.rep[indx1,"Creatinine"]
+
+#THE ABOVE CODE PREFERABLY MATCHES A RESULT FROM AFTER 6 MONTHS, BUT OTHERWISE SELECTS ONE BEFORE IT. 
+#WE NEED TO REMOVE DATA COLLECTED FROM BEFORE 6 MONTHS:
+
+maxes<-first %>%
+group_by(PatientID) %>%
+slice(which.max(event.date)) %>%
+as.data.frame
+maxes<-maxes[,c(1,2)]
+colnames(maxes)<-c("PatientID","MaxDate")
+first<-merge(first,maxes)
+first<-first[first$event.date2<=first$MaxDate,c(1,2,5)]
+
+crea.rep<-merge(crea.rep,first,all.x=TRUE)
+crea.rep$Creatinine6M<-unlist(crea.rep$Creatinine6M)
+
+################################################################################