README.Rmd

---
title: "REIMB"
output:
  github_document:
    toc: true
    df_print: kable
editor_options: 
  chunk_output_type: inline
---

```{r setup, include=FALSE}
rm(list = ls())

knitr::opts_chunk$set(echo = FALSE)

library(tidyverse)
library(scales)

source("../lib/vocabulary_reporting_fuctions.r")


# Document variables 
N_TOPS <- 10

# output status
db_statuses <- tibble()

```

```{r load, message=FALSE, warning=FALSE}
reimb_concept <- read_csv("./1_source_vocabulary/CONCEPT_REIMB.csv") %>% 
  filter(domain_id!="Metadata")

reimb_mapsto <- read_csv("../ATC/2_mapping_to_standard/CONCEPT_RELATIONSHIP_ATC.csv") %>% 
  filter(FALSE) ## read an other table to keep same colums

reimb_freqs <- read_csv("./3_freq_of_source_codes/freq_total.csv",
  col_types = cols(
    code = col_character(),
    freq_finngen = col_double(),
    freq_total = col_double()
  )
)
```


# Intro
TODO:


# Formating source vocabulary to OMOP
The reimb codes were downloaded from [nose](nose) (3.4.2020). Missing codes, translations and validity dates were manually colledted from kela's web page into table [./1_source_vocabulary/Kela_reimbursement_codes.xlsx](./1_source_vocabulary/Kela_reimbursement_codes.xlsx) by [\@CoderMikolaj](github.com/CoderMikolaj). 

Resulting table contains the codes, names, and validity periods. Some codes (e.g 307) have a different names during different periods, and therefore codes are repeated on the table for name and time period. 

TODO: we got the name changes over time only for the top 21 codes. The rest should be checked. (top 21 bcs were the ones with no name at all kela's )

Missing English translations were made using Google-tralator

```{r}
reimb_concept %>% count(tmp_name_en_source, sort = T)
```

Source table with added translation has been formatted to be similar to OMOP in  [1_source_vocabulary/reimb_for_source_vocabulary.csv](1_source_vocabulary/reimb_for_source_vocabulary.csv). 



## Progess in number of codes
```{r}
n_codes <- reimb_concept %>% nrow()
n_codes_accepted <- reimb_mapsto %>% distinct(concept_id_1) %>% nrow()
```


From `r scales::number(n_codes)` codes `r scales::number(n_codes_accepted)` have been approved. 

This makes `r scales::percent(n_codes_accepted/n_codes)` of codes approved. 


```{r}
db_statuses <- bind_rows(
  tibble(
    status = "mapped",
    n_codes = n_codes_accepted,
    per_codes = percent(n_codes_accepted/!!n_codes),
    n_events = as.double(NA),
    per_events = as.character(NA), 
    db_name = "source"
  ),
 tibble(
    status = "not_mapped",
    n_codes = !!n_codes-n_codes_accepted,
    per_codes = percent((!!n_codes-n_codes_accepted)/!!n_codes),
    n_events =  as.double(NA),
    per_events = as.character(NA), 
    db_name = "source"
  )
)
```





# Assessing coverage of databases


```{r  results='asis'}

#write_csv(db_statuses, "status_table.csv" )
 database_coverage_md("REIMB", reimb_concept, reimb_mapsto, reimb_freqs) 

```