healthpopR helps researchers and analysts classify populations into exposure and response groups, visualize ICD-10-based health profiles, and perform survival and Cox regression analysis using tidy workflows.
The package is designed for analyzing how exposure diagnoses affect response diagnoses in population health data research. Package uses regex code for identifying ICD-10/9/8 codes. Check Quick Example for motivation.
Dataset Limitations
Currently functions are developed so that they fit our specific dataset correctly. Which means that analyses are limited to only one gender and follow up starts at when population is 50 years old. Theses limitations can be removed later, when we use same functions to another datasets.
Installation
Install the development version from GitHub:
# Install healthpopR
remotes::install_github("janikmiet/healthpopR")Quick Example
Package uses regex to search diagnoses. Here’s how to classify a population to exposure and response groups, where
- Exposure is ICD-10 code E11 (Type 2 diabetes mellitus)
- Response is ICD-10 codes I20-I25 (Ischemic heart diseases)
In diagnoses dataset we use only 2 registry sources local and hilmo.
library(healthpopR)
## Load Datasets
population <- read.csv("your_population_data.csv")
diagnoses <- read.csv("your_diagnoses_data.csv")
## Classify population
dpop <- classify_population(diagnoses,
exposure_icd10 = "^E11",
exposure_src = c("local", "hilmo"),
response_icd10 = "^I2[0-5]",
response_src = c("local", "hilmo"),
data_population = population,
data_diagnoses = diagnoses)
head(dpop)Check data model -vignette to view how you should arrange your population and diagnoses -datasets.
Key Features
- 📊 Population classification:
classify_population(),classify_icd10_profile() - 🧠 ICD-10 diagnostics:
plot_diagnoses_src(),plot_health_icd10_profile() - ⏳ Survival analysis:
plot_survival_km(),cox_create_data(),cox_plot_spline() - 📈 Comparison tools:
summary_exp_resp_order(),tbl_icd10_diff_by_exposure() - 📊 Incidence modeling tools:
pirr_results()