05-prep_pcp_data.Rmd

# Prepare UKB Primary Care Data {#prep-pcp-data}

The primary care data contain clinical information of approximately 500,000 patients on their visits to primary care physicians. A reason for a visit is recorded as a code. The date of visit associated with the code is also available. We accomplish the following in this chapter:

1. Convert special dates to their mapping
2. Generate subject ID file (used later for excluding control subjects when phenotyping DR and CKD)
3. Exclude subjects with sex mismatch (although the number of unique subjects remain the same after applying sex mismatch filtering)

We use the following mapping which is identical to the one we used for preparing the master event table in chapter \@ref(curate-master-ukb-events-table):

|Special date|Map|
|:------------|:---|
|1900-01-01|Missing|
|1901-01-01|Missing|
|2037-07-07|Missing||
|1902-02-02|DOB of a subject|
|1903-03-03|DOB of a subject|

The primary care data is `entire_gp_clinical_30March2021_formatted.txt` in `raw_data` folder. The data can be downloaded by approved researchers from the UKB Record Repository if you have requested field `42040`.

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE, eval=F)
```

Load packages and a file containing functions.
```{r, message=F}
library(data.table)
library(tidyverse)
source("functions.R")
```

Import primary care data.
```{r}
gp_clinical <- fread("raw_data/entire_gp_clinical_30March2021.txt")
```

Fix the date format so that other scripts can use it more easily.
```{r}
gp_clinical$event_dt <- as.Date(gp_clinical$event_dt, tryFormats = "%d/%m/%Y")
```

Create code and terminology fields in place of read_2 and read_3 fields.
```{r}
gp_clinical <- gp_clinical %>% mutate(code = ifelse(read_2 != "", read_2, read_3)) %>%
  mutate(terminology = ifelse(read_2 != "", "read2", "read3")) %>%
  select(-read_2, -read_3) %>%
  distinct()
```

Rename fid (subject id) field.
```{r}
gp_clinical <- gp_clinical %>% rename(f.eid = eid)
```

Load demographic data.
```{r}
demog <- readRDS("generated_data/demog_selected.RDS")
```

Add a field indicating whether the date of event is special.
```{r}
gp_clinical$special_dt <- is_special_date(gp_clinical$event_dt)
```

Convert the special dates.
```{r}
demog_dob <- demog %>% select(f.eid, DOB)
gp_clinical <- gp_clinical %>% left_join(demog_dob, by = "f.eid")
gp_clinical$event_dt <- cleandates(gp_clinical$event_dt,gp_clinical$DOB)
gp_clinical <- gp_clinical %>% select(-DOB)
```

Save subject IDs represented in the primary care data.
```{r}
gp_subject_ids <- gp_clinical$f.eid %>% unique()
saveRDS(gp_subject_ids,"generated_data/gp_subject_ids.RDS")
```

Remove any events with missing dates.
```{r}
gp_clinical <- gp_clinical %>% filter(!is.na(event_dt))
```

Filter out subjects with sex mismatch.
```{r}
sex_mismatch_subject_ids <- readRDS("generated_data/sex_mismatch_subject_ids.RDS")
gp_clinical <- gp_clinical %>% filter(!(f.eid %in% sex_mismatch_subject_ids))
```

Save the primary care data.
```{r}
fwrite(gp_clinical, "generated_data/entire_gp_clinical_30March2021_formatted.txt", sep="\t", row.names=F, quote=T)
```