Make Lookup Table

library(conflicted)
suppressMessages(conflict_prefer("filter", "dplyr"))

library(DOPE)
library(tibble)
library(dplyr)
library(stringr)
library(usethis)

The lookup table has information on drug category, class and synonyms. It is made from the DEA and “No Slang” tables. The core of the table is the dea_factsheets table. It is tweaked to simplify complex names

Fix dea_brands

The dea brands file has:

dea_brands <- DOPE::dea_brands %>%
  rename("brand" = brands) %>%
  mutate(across(where(is.character), tolower)) %>%
  mutate(brand = str_remove_all(brand, "®")) %>%
  mutate(brand = case_when(brand == "kadianms-contin" ~ "kadian", # two names
                           TRUE ~ brand)) %>%
  bind_rows(c(category = "morphine", brand = "ms contin"))

Fix DEA category and class

The fact sheet file has:

These two do not fit well

dea_class_cat <- DOPE::dea_factsheets %>%
  mutate(across(where(is.character), tolower)) %>%
  mutate(category =
           case_when(category == "peyote and mescaline" ~ "mescaline",
                     category == "ghb - gamma-hydroxybutyric acid" ~ "ghb",
                     category == "ecstasy or mdma (also known as molly)" ~ "mdma",
                     category == "flakka (alpha-pvp)" ~ "flakka",

                     TRUE ~ category))

The DEA street name file has:

dea_street_names <- DOPE::dea_street_names %>%
  mutate(across(where(is.character), tolower)) %>%
  mutate(category =
           case_when(category == "fentanyl and fentanyl derivatives"    ~ "fentanyl",
                     TRUE ~ category))
# the these drugs do not have fact sheets but they appear on the DEA

# DEA website (in slang)
missing <- tribble(
  ~category, ~class, ~fact_path,
  "acetaminophen and oxycodone", "narcotics (opioids)", "DEA slang",
  "alprazolam", "benzodiazepine", "DEA slang" ,
  "clonazepam", "benzodiazepine", "DEA slang"  ,
  "crack cocaine", "stimulants", "DEA slang",
  "hydrocodone", "narcotics (opioids)", "DEA slang",
  "marijuana concentrates", "cannabis", "DEA slang",
  "mushrooms", "hallucinogen", "DEA slang",
  "PCP", "hallucinogen", "DEA slang",
  "promethazine with codeine", "narcotics (opioids)", "DEA slang",
  "ritalin", "stimulants", "DEA slang",
  "synthetic cannabinoids", "designer drugs", "DEA slang",
  "synthetic cathinones", "stimulants", "DEA slang")

# extras from NoSlang
extraNewSlang <- tribble(
  ~category,  ~class, ~fact_path,
  "2cb", "hallucinogen", "noSlang Term",
  "alpha-ethyltryptamine", "hallucinogen", "noSlang Term",
  "alpha-methyltryptamine", "hallucinogen", "noSlang Term",
  "amobarbital", "depressant", "noSlang Term",
  "amyl nitrite", "inhalant", "noSlang Term",
  "dextromethorphan", "depressant", "noSlang Term",
  "diazepam", "benzodiazepine", "noSlang Term",
  "dimethyltryptamine", "hallucinogen", "noSlang Term",
  "gbl", "depressant", "noSlang Term",
  "isobutyl nitrite", "inhalant", "noSlang Term",
  "methcathinone", "stimulant", "noSlang Term",
  "methaqualone", "depressant", "noSlang Term",
  "nitrous oxide", "inhalant", "noSlang Term"
)
categories <- data.frame(allCat = c(dea_brands$category,
                                   dea_street_names$category,
                                   dea_class_cat$category,
                                   missing$category,
                                   extraNewSlang$category)) %>%
  distinct() %>%
  arrange()
library(sqldf)
#> Loading required package: gsubfn
#> Loading required package: proto
#> Loading required package: RSQLite
lookup_df <- sqldf("select cc.class, a.category, a.syn synonym from
                (select b.category, b.brand as syn from dea_brands as b
                 union
                 select s1.category, s1.slang as syn from dea_street_names as s1
                 union
                 select s2.category, s2.brand as syn from dea_street_names as s2 where s2.brand <> NULL
                 union
                 select ns.drug, ns.street_name as syn from noslang_street_names as ns) as a
              left join dea_class_cat as cc on a.category = cc.category")

#use_data(lookup_df, overwrite = TRUE)

Several records do not have a class.

#> [1] 30

Add data from IQVIA

iqvia <- DOPE::iqvia %>%
  mutate(across(where(is.character), tolower)) 

lookup_df <- lookup_df %>% 
  bind_rows(iqvia)

Fix pluralization

# get classes & categories
unique(lookup_df$class)
#>  [1] "hallucinogen"        "narcotics (opioids)" "depressants"        
#>  [4] "stimulants"          "antitussives"        "heroin"             
#>  [7] "inhalant"            "cannabis"            "drugs of concern"   
#> [10] "steroids"            "analgesic"           "diuretic"           
#> [13] "treatment drug"      "reversal agent"
unique(lookup_df$category)
#>  [1] "2cb"                                 
#>  [2] "acetaminophen and oxycodone"         
#>  [3] "alpha-ethyltryptamine"               
#>  [4] "alprazolam"                          
#>  [5] "amobarbital"                         
#>  [6] "amphetamine"                         
#>  [7] "amphetamines"                        
#>  [8] "amt"                                 
#>  [9] "amyl nitrite"                        
#> [10] "barbiturates"                        
#> [11] "benzodiazepines"                     
#> [12] "clonazepam"                          
#> [13] "cocaine"                             
#> [14] "codeine"                             
#> [15] "crack"                               
#> [16] "crack cocaine"                       
#> [17] "dextromethorphan"                    
#> [18] "diazepam"                            
#> [19] "dmt"                                 
#> [20] "fentanyl"                            
#> [21] "gbl"                                 
#> [22] "ghb"                                 
#> [23] "heroin"                              
#> [24] "hydrocodone"                         
#> [25] "hydromorphone"                       
#> [26] "inhalants"                           
#> [27] "isobutyl nitrite"                    
#> [28] "ketamine"                            
#> [29] "khat"                                
#> [30] "lsd"                                 
#> [31] "marijuana"                           
#> [32] "marijuana concentrates"              
#> [33] "mdma"                                
#> [34] "mescaline"                           
#> [35] "methadone"                           
#> [36] "methamphetamine"                     
#> [37] "methaqualone"                        
#> [38] "methcathinone"                       
#> [39] "methylphenidate"                     
#> [40] "morphine"                            
#> [41] "mushrooms"                           
#> [42] "nitrous oxide"                       
#> [43] "opium"                               
#> [44] "oxycodone"                           
#> [45] "pcp"                                 
#> [46] "peyote"                              
#> [47] "promethazine with codeine"           
#> [48] "psilocybin"                          
#> [49] "ritalin"                             
#> [50] "rohypnol"                            
#> [51] "salvia divinorum"                    
#> [52] "steroids"                            
#> [53] "synthetic cannabinoids"              
#> [54] "synthetic cathinones"                
#> [55] "pain relief"                         
#> [56] "codeine combinations, non-injectable"
#> [57] "caffeine"                            
#> [58] "treatment drug"                      
#> [59] "morphine-opium, injectable"          
#> [60] "morphine-opium, non-injectable"      
#> [61] "reversal agent"                      
#> [62] "synthetic narcotic, injectable"      
#> [63] "synthetic narcotic, non-injectable"

lookup_df3 <- lookup_df %>% 
  mutate(class = if_else(class != "cannabis", str_replace(class, "s$", ""), class),
         class = case_when(class == "narcotics (opioids)" ~ "narcotic (opioid)",
                           TRUE ~ class),
         category = if_else(category != "mushrooms", str_replace(category, "s$", ""), category)
         )
orange_book <-
  tibble(synonym = c(dea_controlled$substance, dea_controlled$synonym)) %>%
  mutate(synonym = tolower(synonym)) %>% 
  distinct() %>% 
  mutate(
    difficult = str_count(synonym, "[(]") > 0 | 
      str_detect(synonym, ",(?=\\S)") |
      str_detect(synonym, '[-|/|&|,|"| ]')
  ) %>% 
  filter(!difficult) %>% 
  anti_join(lookup_df, by = "synonym") %>% 
  anti_join(lookup_df, by = c("synonym" = "class")) %>% 
  anti_join(lookup_df, by = c("synonym" = "category")) %>% 
  mutate(category = "Unknown",  class = "Unknown") %>% 
  select(class, category, synonym) 
  
    
lookup_df <- bind_rows(lookup_df3 , orange_book)
use_data(lookup_df, overwrite = TRUE)
#> v Setting active project to 'C:/Users/godom/AppData/Local/Temp/RtmpWA2jey/Rbuild1540c1cf043da/DOPE'
#> v Saving 'lookup_df' to 'data/lookup_df.rda'
#> * Document your data (see 'https://r-pkgs.org/data.html')

mirror server hosted at Truenetwork, Russian Federation.