epidict
Epidemiology Data Dictionaries and Random Data Generators
The 'R4EPIs' project < https://r4epi.github.io/sitrep/> seeks to provide a set of standardized tools for analysis of outbreak and survey data in humanitarian aid settings. This package currently provides standardized data dictionaries from Medecins Sans Frontieres Operational Centre Amsterdam for outbreak scenarios (Acute Jaundice Syndrome, Cholera, Diphtheria, Measles, Meningitis) and surveys (Retrospective mortality and access to care, Malnutrition, Vaccination coverage and Event Based Surveillance) - as described in the following < https://scienceportal.msf.org/assets/standardised-mortality-surveys?utm_source=chatgpt.com>. In addition, a data generator from these dictionaries is provided. It is also possible to read in any Open Data Kit format data dictionary.
epidict
The goal of {epidict} is to provide standardized data dictionaries for use in epidemiological data analysis templates. Currently it supports standardised dictionaries from MSF OCA. This is a product of the R4EPIs project; learn more at https://r4epi.github.io/sitrep/
Installation
You can install {epidict} from CRAN:
install.packages("epidict")
Click here for alternative installation options
If there is a bugfix or feature that is not yet on CRAN, you can install it via the {drat} package:
You can also install the in-development version from GitHub using the {remotes} package (but there’s no guarantee that it will be stable):
# install.packages("remotes")
remotes::install_github("R4EPI/epidict")
Accessing dictionaries
The dictionaries can be obtained via the msf_dict() function, which
specifies a variable and their possible options (if categorical).
There are MSF intersectional outbreak dictionaries available in {epidict} based on ODK exports.
There are MSF OCA outbreak dictionaries available in {epidict} based on DHIS2 exports. > You can read more about the outbreak dictionaries at https://r4epi.github.io/epidict/articles/Outbreaks.html
In addition, there are MSF survey dictionaries available based on ODK exports. > You can read more about the survey dictionaries at https://r4epi.github.io/epidict/articles/Surveys.html
You can also read in your own ODK dictionaries using read_dict().
Generating data
The {epidict} package has a function for generating data that’s called
gen_data(), which takes three arguments: The dictionary, which column
describes the variable names, and how many rows are needed in the
output.
Click here for code examples
library("epidict")
gen_data("Measles", varnames = "data_element_shortname", numcases = 100, org = "MSF")
#> # A tibble: 100 × 52
#> case_number date_of_consultation_admis…¹ patient_facility_type patient_origin
#> <chr> <date> <fct> <chr>
#> 1 A1 2018-04-23 OP Village D
#> 2 A2 2018-02-23 OP Village C
#> 3 A3 2018-04-15 OP Village C
#> 4 A4 2018-04-30 OP Village A
#> 5 A5 2018-01-09 IP Village D
#> 6 A6 2018-03-14 OP Village D
#> 7 A7 2018-03-20 OP Village D
#> 8 A8 2018-03-23 OP Village A
#> 9 A9 2018-01-23 IP Village A
#> 10 A10 2018-03-19 OP Village D
#> # ℹ 90 more rows
#> # ℹ abbreviated name: ¹date_of_consultation_admission
#> # ℹ 48 more variables: age_years <int>, age_months <int>, age_days <int>,
#> # sex <fct>, pregnant <fct>, trimester <fct>,
#> # foetus_alive_at_admission <fct>, exit_status <fct>, date_of_exit <date>,
#> # time_to_death <fct>, pregnancy_outcome_at_exit <fct>,
#> # baby_born_with_complications <fct>, previously_vaccinated <fct>, …
gen_data("Vaccination_long", varnames = "name", numcases = 100, org = "MSF")
#> # A tibble: 100 × 120
#> start end today deviceid date team_number village_name village_other
#> <lgl> <lgl> <lgl> <lgl> <date> <lgl> <fct> <lgl>
#> 1 NA NA NA NA 2018-02-11 NA village_4 NA
#> 2 NA NA NA NA 2018-03-24 NA village_2 NA
#> 3 NA NA NA NA 2018-02-02 NA village_9 NA
#> 4 NA NA NA NA 2018-02-20 NA village_3 NA
#> 5 NA NA NA NA 2018-04-09 NA village_5 NA
#> 6 NA NA NA NA 2018-01-27 NA village_7 NA
#> 7 NA NA NA NA 2018-03-12 NA village_6 NA
#> 8 NA NA NA NA 2018-04-11 NA village_3 NA
#> 9 NA NA NA NA 2018-04-28 NA village_3 NA
#> 10 NA NA NA NA 2018-03-09 NA other NA
#> # ℹ 90 more rows
#> # ℹ 112 more variables: cluster_number <dbl>, household_number <int>,
#> # households_building <int>, random_hh <int>, consent <chr>,
#> # no_consent_reason <fct>, no_consent_other <lgl>, caretaker_relation <fct>,
#> # caretaker_other <lgl>, number_children <dbl>, child_number <chr>,
#> # sex <fct>, date_birth <date>, age_years <int>, age_months <int>,
#> # any_vaccine <fct>, vaccine_card <fct>, hf_records <fct>, …
Cleaning data with the dictionaries
You can use the dictionaries to clean the data via the {matchmaker} package:
Click here for code examples
library("matchmaker")
library("dplyr")
dat <- gen_data(dictionary = "Cholera",
varnames = "data_element_shortname",
numcases = 20,
org = "MSF"
)
print(dat)
#> # A tibble: 20 × 45
#> case_number date_of_consultation_admiss…¹ patient_origin age_years age_months
#> <chr> <date> <chr> <int> <int>
#> 1 A1 2018-03-23 Village D 18 NA
#> 2 A2 2018-01-29 Village B 27 NA
#> 3 A3 2018-01-23 Village D 40 NA
#> 4 A4 2018-02-02 Village D 15 NA
#> 5 A5 2018-01-23 Village B 28 NA
#> 6 A6 2018-01-15 Village A 40 NA
#> 7 A7 2018-01-04 Village B 34 NA
#> 8 A8 2018-04-25 Village D 29 NA
#> 9 A9 2018-04-18 Village D 45 NA
#> 10 A10 2018-01-22 Village C 10 NA
#> 11 A11 2018-02-06 Village A 19 NA
#> 12 A12 2018-03-03 Village D 61 NA
#> 13 A13 2018-01-08 Village B 20 NA
#> 14 A14 2018-03-08 Village C 73 NA
#> 15 A15 2018-03-08 Village C 66 NA
#> 16 A16 2018-03-18 Village A 52 NA
#> 17 A17 2018-04-03 Village A 59 NA
#> 18 A18 2018-01-12 Village B 24 NA
#> 19 A19 2018-01-19 Village D 53 NA
#> 20 A20 2018-03-07 Village A 7 NA
#> # ℹ abbreviated name: ¹date_of_consultation_admission
#> # ℹ 40 more variables: age_days <int>, sex <fct>, pregnant <fct>,
#> # trimester <fct>, foetus_alive_at_admission <fct>, exit_status <fct>,
#> # date_of_exit <date>, time_to_death <fct>, pregnancy_outcome_at_exit <fct>,
#> # previously_vaccinated <fct>, previous_vaccine_doses_received <fct>,
#> # readmission <fct>, msf_involvement <fct>,
#> # cholera_treatment_facility_type <fct>, residential_status_brief <fct>, …
# We want the expanded dictionary, so we will select `compact = FALSE`
dict <- msf_dict(dictionary = "Cholera",
long = TRUE,
compact = FALSE,
tibble = TRUE
)
print(dict)
#> # A tibble: 182 × 11
#> data_element_uid data_element_name data_element_shortname
#> <chr> <chr> <chr>
#> 1 AafTlSwliVQ egen_001_patient_case_number case_number
#> 2 OTGOtWBz39J egen_004_date_of_consultation_admiss… date_of_consultation_…
#> 3 wnmMr2V3T3u egen_006_patient_origin patient_origin
#> 4 sbgqjeVwtb8 egen_008_age_years age_years
#> 5 eXYhovYyl61 egen_009_age_months age_months
#> 6 UrYJSk2Wp46 egen_010_age_days age_days
#> 7 D1Ky5K7pFN6 egen_011_sex sex
#> 8 D1Ky5K7pFN6 egen_011_sex sex
#> 9 D1Ky5K7pFN6 egen_011_sex sex
#> 10 dTm5R53YYXC egen_012_pregnancy_status pregnant
#> # ℹ 172 more rows
#> # ℹ 8 more variables: data_element_description <chr>,
#> # data_element_valuetype <chr>, data_element_formname <chr>,
#> # used_optionset_uid <chr>, option_code <chr>, option_name <chr>,
#> # option_uid <chr>, option_order_in_set <dbl>
# Now we can use matchmaker to filter the data
dat_clean <- matchmaker::match_df(dat, dict,
from = "option_code",
to = "option_name",
by = "data_element_shortname",
order = "option_order_in_set"
)
print(dat_clean)
#> # A tibble: 20 × 45
#> case_number date_of_consultation_admiss…¹ patient_origin age_years age_months
#> <chr> <date> <chr> <int> <int>
#> 1 A1 2018-03-23 Village D 18 NA
#> 2 A2 2018-01-29 Village B 27 NA
#> 3 A3 2018-01-23 Village D 40 NA
#> 4 A4 2018-02-02 Village D 15 NA
#> 5 A5 2018-01-23 Village B 28 NA
#> 6 A6 2018-01-15 Village A 40 NA
#> 7 A7 2018-01-04 Village B 34 NA
#> 8 A8 2018-04-25 Village D 29 NA
#> 9 A9 2018-04-18 Village D 45 NA
#> 10 A10 2018-01-22 Village C 10 NA
#> 11 A11 2018-02-06 Village A 19 NA
#> 12 A12 2018-03-03 Village D 61 NA
#> 13 A13 2018-01-08 Village B 20 NA
#> 14 A14 2018-03-08 Village C 73 NA
#> 15 A15 2018-03-08 Village C 66 NA
#> 16 A16 2018-03-18 Village A 52 NA
#> 17 A17 2018-04-03 Village A 59 NA
#> 18 A18 2018-01-12 Village B 24 NA
#> 19 A19 2018-01-19 Village D 53 NA
#> 20 A20 2018-03-07 Village A 7 NA
#> # ℹ abbreviated name: ¹date_of_consultation_admission
#> # ℹ 40 more variables: age_days <int>, sex <fct>, pregnant <fct>,
#> # trimester <fct>, foetus_alive_at_admission <fct>, exit_status <fct>,
#> # date_of_exit <date>, time_to_death <fct>, pregnancy_outcome_at_exit <fct>,
#> # previously_vaccinated <fct>, previous_vaccine_doses_received <fct>,
#> # readmission <fct>, msf_involvement <fct>,
#> # cholera_treatment_facility_type <fct>, residential_status_brief <fct>, …