08-phenotype_outcomes_ukb.Rmd

# Phenotype outcome and control exclusion events using UKB assessment center data {#phenotype-outcomes-ukb}

In this chapter, we use the master UKB event table created in chapter \@ref(curate-master-ukb-events-table) to generate different outcome event tables. We phenotype the following outcomes:

  - Diabetes (DM)
  - Myocardial Infarction (MI)
  - Unstable Angina (UA)
  - Ischemic Stroke (IS)
  - Hemorrhagic Stroke (HS)
  - Stroke (ST)
  - PCI
  - Composite CVD
  - Diabetic retinopathy (DR)
  - Diabetic kidney disease (DKD)
  
The following event tables will also be created to exclude certain controls from time-to-event tables in later chapters:

  - Diabetic eye disease control exclusion events
  - Diabetic kidney disease control exclusion events
  - Cardiovascular control exclusion events
  - Cerebrovascular control exclusion events
  - Non-coronary revascularization control exclusion events

The general phenotyping procedure for complication outcomes and control exclusion outcomes are the same. All we have to do is to define patterns we want to search and match these patterns from the master UKB event table.

The pattern searching is abstracted away in a function `get_phenotype_tab()` in `functions.R`. The function searches certain regular expression patterns corresponding to outcomes from the master UKB event table and outputs the table containing all of the events matching the patterns. The following are the types of patterns the function accepts:

- UKB-defined field patterns 
- custom-defined field patterns
- ICD10 code patterns (also accepts more specific ICD10 codes for primary death and secondary death events)
- OPCS4 code patterns
- self-reported condition code patterns
- self-reported operation code patterns

The UKB-defined field patterns match fields that are associated with first occurrence fields and algorithimically defined fields as defined by UKB study. The custom field pattern matches a custom field defined in chapter \@ref(curate-master-ukb-events-table). As a reminder, the only custom field we have defined is `dr_self` used in phenotyping diabetes related eye disease. The code patterns including ICD10, OPCS4, self-reported condition and self-reported operation match the codes that represent some clinical events. Thus, to phenotype an outcome, one should first identify the fields and codes associated with an outcome, define the patterns which are one of the inputs to `get_phenotype_tab()`. Internally, `get_phenotype_tab` function uses `grepl` function in `dplyr` package to filter the master event table.

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

Load packages and functions.
```{r}
library(tidyverse)
library(data.table)
source("functions.R")
```

Import the master event table.
```{r}
event_tab <- readRDS("generated_data/all_ukb_events_tab.RDS")
```

## Phenotype outcomes

### DM

Outcome fields:

- `f.130706.0.0`: the date of first occurrence of Type 1 diabetes
- `f.130708.0.0`: the date of first occurrence of Type 2 diabetes
- `f.130714.0.0`: the date of first occurrence of diabetes of unspecified type

```{r}
dm_field_patterns <- "130706|130708|130714"
```
  
Get DM event table and first occurrence DM event table.
```{r}
dm_ukb <- get_phenotype_tab(field_patterns = dm_field_patterns,event_tab = event_tab,firstoccur = F)
dm_firstoccur_ukb <- get_phenotype_tab(field_patterns = dm_field_patterns,event_tab = event_tab,firstoccur = T)
```

Categorize diabetes event to either Type1, Type2 or Uncertain in a new field `dm_type`.
```{r}
dm_ukb <- dm_ukb %>% mutate(dm_type = ifelse(grepl("130706", key), "Type1", 
                                             ifelse(grepl("130708", key), "Type2","Uncertain")))
```

```{r}
saveRDS(dm_ukb,"generated_data/dm_ukb.RDS")
saveRDS(dm_firstoccur_ukb,"generated_data/dm_firstoccur_ukb.RDS")
```

### MI

Outcome fields:

- f.131298.0.0: the date of first occurrence of MI
- f.131300.0.0: the date of subsequent occurrence of MI
- f.42000.0.0: the date of algorithmically defined MI outcome

ICD10 code prefixes: 

- I21
- I22
- I23

Define patterns to search.
```{r}
mi_field_patterns <- "131298|131300|42000"
mi_icd10_patterns <- "^I21|^I22|^I23"
```

Get MI event table and first occurrence MI event table.
```{r}
mi_ukb <- 
  get_phenotype_tab(field_patterns = mi_field_patterns, icd10_patterns_any = mi_icd10_patterns,event_tab = event_tab,firstoccur = F)
mi_firstoccur_ukb <- 
  get_phenotype_tab(field_patterns = mi_field_patterns, icd10_patterns_any = mi_icd10_patterns,event_tab = event_tab,firstoccur = T)
```

```{r}
saveRDS(mi_ukb,"generated_data/mi_ukb.RDS")
saveRDS(mi_firstoccur_ukb,"generated_data/mi_firstoccur_ukb.RDS")
```

### Unstable Angina

ICD10 code prefixes: 

- I200

```{r}
unstable_angina_icd10_patterns <- "I200"
```

Get unstable angina table and first occurrence unstable angina event table.
```{r}
unstable_angina_ukb <- get_phenotype_tab(icd10_patterns_any = unstable_angina_icd10_patterns, event_tab = event_tab,firstoccur = F)
unstable_angina_firstoccur_ukb <- get_phenotype_tab(icd10_patterns_any = unstable_angina_icd10_patterns, event_tab = event_tab,firstoccur = T)
```

```{r}
saveRDS(unstable_angina_ukb,"generated_data/unstable_angina_ukb.RDS")
saveRDS(unstable_angina_firstoccur_ukb,"generated_data/unstable_angina_firstoccur_ukb.RDS")
```

### Ischemic Stroke

Outcome fields:
- 131366: Date I63 first reported (cerebral infarction)
- 42008: algorithmically defined outcome for ischemic stroke 

ICD10 code prefixes: 
- I63

Self-reported condition codes:
- 1583

Define patterns to search.
```{r}
stroke_infarct_field_patterns <- "131366|42008"
stroke_infarct_icd10_patterns <- "^I63"
stroke_infarct_selfrep_patterns <- "^1583$"
```

Get IS event table and first occurrence IS event table.
```{r}
stroke_infacrt_ukb <- get_phenotype_tab(field_patterns = stroke_infarct_field_patterns,
                                        icd10_patterns_any = stroke_infarct_icd10_patterns,
                                        selfrep_patterns = stroke_infarct_selfrep_patterns,
                                        event_tab = event_tab,firstoccur = F)

stroke_infacrt_firstoccur_ukb <- get_phenotype_tab(field_patterns = stroke_infarct_field_patterns,
                                        icd10_patterns_any = stroke_infarct_icd10_patterns,
                                        selfrep_patterns = stroke_infarct_selfrep_patterns,
                                        event_tab = event_tab,firstoccur = T)
```

```{r}
saveRDS(stroke_infacrt_ukb,"generated_data/stroke_infarct_ukb.RDS")
saveRDS(stroke_infacrt_firstoccur_ukb,"generated_data/stroke_infarct_firstoccur_ukb.RDS")
```

### Hemorrhagic Stroke

Outcome fields:

- 131360: Subarachnoid Hemorrhage
- 131362: Intracerebral Hemorrhage
- 131364: Other non-traumatic Hemorrhage, 
- 42010: Intracerebral Hemorrhage
- 42012: Subarachnoid Hemorrhage

ICD10 code prefixes: 
- I60
- I61
- I62

Self-reported condition codes:
- 1086

Define patterns to search.
```{r}
stroke_hem_field_patterns <- "131360|131362|131364|42010|42012"
stroke_hem_icd10_patterns <- "^I60|^I61|^I62"
stroke_hem_selfrep_patterns <- "^1086$"
```

Get HS event table and first occurrence HS table.
```{r}
stroke_hem_ukb <- get_phenotype_tab(field_patterns = stroke_hem_field_patterns,
                                            icd10_patterns_any = stroke_hem_icd10_patterns,
                                            selfrep_patterns = stroke_hem_selfrep_patterns,
                                            event_tab = event_tab,firstoccur = F)

stroke_hem_firstoccur_ukb <- get_phenotype_tab(field_patterns = stroke_hem_field_patterns,
                                            icd10_patterns_any = stroke_hem_icd10_patterns,
                                            selfrep_patterns = stroke_hem_selfrep_patterns,
                                            event_tab = event_tab,firstoccur = T)
```

```{r}
saveRDS(stroke_hem_ukb,"generated_data/stroke_hem_ukb.RDS")
saveRDS(stroke_hem_firstoccur_ukb,"generated_data/stroke_hem_firstoccur_ukb.RDS")
```

### Stroke

Outcome fields:
- 42006: unspecified stroke
- 131368: unspecified stroke

ICD10 code prefix
- I64

Self-reported condition codes:
- 1081

Outcome tables:
- recurrent ischemic stroke event table
- recurrent hemorrhagic stroke event table

Define patterns to search.
```{r}
stroke_unspec_field_patterns <- "42006|131368"
stroke_unspec_icd10_patterns <- "^I64"
stroke_other_selfrep_patterns <- "^1081$"
```

Get stroke event table
```{r}
stroke_infarct_hem_ukb <- 
  full_join(stroke_infacrt_ukb,stroke_hem_ukb) %>% arrange(f.eid,event_dt)

stroke_unspec_ukb <- get_phenotype_tab(field_patterns = stroke_unspec_field_patterns,
                                   icd10_patterns_any = stroke_unspec_icd10_patterns,
                                   event_tab = event_tab,firstoccur = F)

stroke_other_ukb <- get_phenotype_tab(selfrep_patterns = stroke_other_selfrep_patterns,
                                   event_tab = event_tab,firstoccur = F)

stroke_ukb <- 
  stroke_infarct_hem_ukb %>% 
  full_join(stroke_unspec_ukb, by = c("f.eid","event_dt")) %>%
  full_join(stroke_other_ukb, by = c("f.eid","event_dt")) %>% 
  arrange(f.eid,event_dt)
```

Get first occurrence stroke event table
```{r}
stroke_firstoccur_ukb <- stroke_ukb %>% group_by(f.eid) %>% arrange(event_dt) %>% slice(1)
```

```{r}
saveRDS(stroke_ukb,"generated_data/stroke_ukb.RDS")
saveRDS(stroke_firstoccur_ukb,"generated_data/stroke_firstoccur_ukb.RDS")
```

### PCI

OPCS4 codes:
- K40, K41, K42, K43, K44, K45, K46, K483, K49, K501, K75, K76

Self-reported operation codes:
- 1070 (Coronary Angioplasty)
- 1095 (Coronary bypass grafts)

Define patterns to search.
```{r}
# OPCS patterns start with 'K' so do not have to start the pattern with '^'
pci_opcs_patterns <- "K40|K41|K42|K43|K44|K45|K46|K483|K49|K501|K75|K76"
pci_selfrep_op_patterns <- "^1070$|^1095$"
```

Get PCI event table and first occurrence PCI event table.
```{r}
pci_ukb <- get_phenotype_tab(opcs_patterns = pci_opcs_patterns,
                                     selfrep_op_patterns = pci_selfrep_op_patterns,
                                     event_tab = event_tab,firstoccur = F)

pci_firstoccur_ukb <- get_phenotype_tab(opcs_patterns = pci_opcs_patterns,
                                     selfrep_op_patterns = pci_selfrep_op_patterns,
                                     event_tab = event_tab,firstoccur = T)
```

```{r}
saveRDS(pci_ukb,"generated_data/pci_ukb.RDS")
saveRDS(pci_firstoccur_ukb,"generated_data/pci_firstoccur_ukb.RDS")
```

### CVD death

ICD10 code prefixes (primary death):

-I*

Define patterns to search.
```{r}
cvd_death_icd10_death_primary_patterns <- "^I"
```

Get CVD death event table
```{r}
cvd_death_ukb <- get_phenotype_tab(icd10_patterns_pd = cvd_death_icd10_death_primary_patterns,
                                     event_tab = event_tab,firstoccur = F)
```

```{r}
saveRDS(cvd_death_ukb,"generated_data/cvd_death_ukb.RDS")
```

### Composite CVD

Composite CVD was defined to be any of the following events:

- MI
- IS
- UA
- PCI
- CVD death

Get compositie CVD event table and first occurrence composite CVD event table.
```{r}
cvd_ukb <- 
  mi_ukb %>%
  full_join(stroke_infacrt_ukb) %>%
  full_join(unstable_angina_ukb) %>%
  full_join(pci_ukb) %>% 
  full_join(cvd_death_ukb) %>% distinct() %>% arrange(f.eid,event_dt)

cvd_firstoccur_ukb <- cvd_ukb %>% group_by(f.eid) %>% arrange(event_dt) %>% slice(1)
```

```{r}
saveRDS(cvd_ukb,"generated_data/cvd_ukb.RDS")
saveRDS(cvd_firstoccur_ukb,"generated_data/cvd_firstoccur_ukb.RDS")
```

### DR

ICD10 codes:

- E1*.3: Diabetes Mellitus with Ophthalmic Complications
- H36.0: Diabetic Retinopathy
- H28.0: Diabetic Cataract

ICD9 codes:

- 2504: Diabetes with Ophthalmic manifestations
- 3620: Diabetic retinopathy

Self-reported condition codes:

- 1276

Custom defined fields:

- dr_self: the date of DR diagnosis. As a reminder `dr_self` was produced by combining information from the fields `f.5901.0.0`, `f.5901.1.0`, `f.5901.2.0` and `f.5901.3.0` which record age of subjects at which DR was diagnosed at four different time points. Please see chapter \@ref(curate-master-ukb-events-table) for details.

Define patterns.
```{r}
dr_icd10 <- "E103|E113|E133|E143|H360|H280"
dr_icd9 <- "^2504|^3620"
dr_self <- "1276"
dr_custom <- "dr_self"
```

Get first occurrence DR event table.
```{r}
dr_firstoccur_ukb <- get_phenotype_tab(icd10_patterns_any = dr_icd10,
                                       icd9_patterns_any = dr_icd9,
                                       selfrep_patterns = dr_self,
                                       custom_field_patterns = dr_custom,
                                    event_tab = event_tab,firstoccur = T)
```

```{r}
saveRDS(dr_firstoccur_ukb,"generated_data/dr_firstoccur_ukb.RDS")
```

### DKD

Outcome fields:

- 42026: algorithmically defined End-Stage Renal Disease (ESRD)

ICD10 codes:

- E1.2: Diabetes Mellitus with renal complication
- N18[0345]: CKD Stage 3-5, end stage
- N083: Glomerular disorders in Diabetes Mellitus

ICD9 codes:
- 2503: Diabetes with renal manifestations
- 5859: Renal failure

Self-reported condition codes:

- 1607

Define patterns.
```{r}
dkd_field_patterns <- "42026"
dkd_icd10 <- "E102|E112|E132|E142|N180|N183|N184|N185|N083"
dkd_icd9 <- "^2503|^5859"
dkd_self <- "1607"
```

Get DKD event table and first occurrence DKD event table.
```{r}
dkd_ukb <- get_phenotype_tab(field_patterns = dkd_field_patterns,
                             icd10_patterns_any = dkd_icd10,
                             icd9_patterns_any = dkd_icd9,
                             selfrep_patterns = dkd_self,
                  event_tab = event_tab,firstoccur = F)

dkd_firstoccur_ukb <- get_phenotype_tab(field_patterns = dkd_field_patterns,
                                        icd10_patterns_any = dkd_icd10,
                                        icd9_patterns_any = dkd_icd9,
                                        selfrep_patterns = dkd_self,
                  event_tab = event_tab,firstoccur = T)
```

```{r}
saveRDS(dkd_ukb,"generated_data/dkd_ukb.RDS")
saveRDS(dkd_firstoccur_ukb,"generated_data/dkd_firstoccur_ukb.RDS")
```

## Phenotype control exclusion events

### Cardiovascular control exclusion events

Define patterns.
```{r}
exclude_ctrl_cardio_icd10 <- "I2|I3|I5|I6|I7"
exclude_ctrl_cardio_self <- "1074|1076|1077|1078|1079|1080|1471|1489|1490|1492|1584|1585|1586|1587|1588|1589|1590|1591|1592"
exclude_ctrl_cardio_selfop <- "1069|1096|1097|1098|1099|1100|1101|1104|1523|1553|1554"
```

Get cardiovascular control exclusion event table.
```{r}
cardio_control_exclusion_events <- 
  get_phenotype_tab(icd10_patterns_any = exclude_ctrl_cardio_icd10,
                  selfrep_patterns = exclude_ctrl_cardio_self,
                  selfrep_op_patterns = exclude_ctrl_cardio_selfop,event_tab = event_tab,firstoccur = F)
```

```{r}
saveRDS(cardio_control_exclusion_events,"generated_data/cardio_control_exclusion_events_ukb.RDS")
```

### Cerebrovascular control exclusion events

Define patterns.
```{r}
exclude_ctrl_cereb_icd10 <- "G45|G46|I65|I66|I67|I68|I69"
exclude_ctrl_cereb_self <- "1082"
```

Get cerebrovascular control exclusion event table.
```{r}
cerebro_control_exclusion_events <- 
  get_phenotype_tab(icd10_patterns_any = exclude_ctrl_cereb_icd10,
                    selfrep_patterns = exclude_ctrl_cereb_self,event_tab = event_tab,firstoccur = F)
```

```{r}
saveRDS(cerebro_control_exclusion_events,"generated_data/cerebro_control_exclusion_events_ukb.RDS")
```

### Non-coronary revascularization procedure control exclusion events

Define patterns.
```{r}
other_revasc_selfop <- "1071|1102|1105|1107|1108|1109|1110"
```

Get non-coronary revascularization procedure control exclusion event table.
```{r}
other_revas_control_exclusion_events <-
  get_phenotype_tab(selfrep_op_patterns = other_revasc_selfop, event_tab = event_tab, firstoccur = F)
```

```{r}
saveRDS(other_revas_control_exclusion_events,"generated_data/other_revas_control_exclusion_events_ukb.RDS")
```

### Renal disease control exclusion events

Define patterns.
```{r}
exclude_ctrl_renal_icd10 <- "N0|N1|N2|Z49|Z992"
exclude_ctrl_renal_icd9 <- "^58[0-9]|^59[1-4]|^590[23]|^V420|^V454|^V560|^V568"
exclude_ctrl_renal_self <- "1192|1193|1194|1519|1520|1608|1609"
exclude_ctrl_renal_selfop <- "1195|1487"
exclude_ctrl_renal_opcs <- "M01|M02|M03|X40|X41|X42"
```

Get renal diseasae control exclusion event table.
```{r}
dkd_control_exclusion_events <-
  get_phenotype_tab(icd10_patterns_any = exclude_ctrl_renal_icd10,
                    icd9_patterns_any = exclude_ctrl_renal_icd9,
                    selfrep_patterns = exclude_ctrl_renal_self,
                    selfrep_op_patterns = exclude_ctrl_renal_selfop,
                    opcs_patterns = exclude_ctrl_renal_opcs,event_tab = event_tab,firstoccur = F)
```

```{r}
saveRDS(dkd_control_exclusion_events,"generated_data/dkd_control_exclusion_events_ukb.RDS")
```

### Eye disease control exclusion events

Define patterns.
```{r}
exclude_eye_icd10 <- "H25|H26|H28|H34|H35|H36|H40|H42"
exclude_eye_icd9 <- "^36[256]"
exclude_eye_self <- "1275|1277|1278|1281|1282|1527|1538|1530"
exclude_eye_selfop <- "1434|1435|1436|1437"
```

Get eye disease control exclusion event table.
```{r}
dr_control_exclusion_events <- 
  get_phenotype_tab(icd10_patterns_any = exclude_eye_icd10,
                    icd9_patterns_any = exclude_eye_icd9,
                    selfrep_patterns = exclude_eye_self,
                    selfrep_op_patterns = exclude_eye_selfop,event_tab = event_tab,firstoccur = F)
```

```{r}
saveRDS(dr_control_exclusion_events,"generated_data/dr_control_exclusion_events_ukb.RDS")
```