validatehealth/cmshcc

This R package calculates CMS-HCC risk scores as an alternative to using SAS

Based on presentation from: http://ase.uva.nl/binaries/content/assets/subsites/amsterdam-school-of-economics/r-in-insurance/webster-risk-adjustment-in-r.pdf

1. install.packages("devtools") if not already installed
2. install cmshcc package using: devtools::install_github("healthactuary/cmshcc")
3. load the cmshcc package using library(cmshcc)
4. generate a sample PERSON data frame or use your own with same format: PERSON <- generate_PERSON()
5. load the CMS-HCC diagnosis code to HCC mapping: cmshcc_map <- load_cmshcc_map()
6. create a diagnosis data frame using mapping or use your own data with same format: DIAG <- generate_DIAG(cmshcc_map)
7. Evaluate the 2017 V22 CMS-HCC models: Community_NonDual_Aged = evaluate_v22_2017(PERSON, DIAG, "Community_NonDual_Aged") Community_NonDual_Disabled = evaluate_v22_2017(PERSON, DIAG, "Community_NonDual_Disabled") Community_FBDual_Aged = evaluate_v22_2017(PERSON, DIAG, "Community_FBDual_Aged") Community_FBDual_Disabled = evaluate_v22_2017(PERSON, DIAG, "Community_FBDual_Disabled") Community_PBDual_Aged = evaluate_v22_2017(PERSON, DIAG, "Community_PBDual_Aged") Community_PBDual_Disabled = evaluate_v22_2017(PERSON, DIAG, "Community_PBDual_Disabled") Institutional = evaluate_v22_2017(PERSON, DIAG, "Institutional")

Bind together results in one dataframe if preferable:

results <- cbind(Community_NonDual_Aged, Community_NonDual_Disabled, Community_FBDual_Aged, Community_FBDual_Disabled, Community_PBDual_Aged, Community_PBDual_Disabled, Institutional)

cmshcc's Issues

Performance Improvement - Remove Unnecessary cbinds

The calculation columns do not necessarily need to be joined with each other or with the input PERSON data frame. Undo this to avoid unnecessary calculations.

Make Age as of date a parameter with default the current date

object 'factor_type' not found

I am following the guide from the README file and received an error below:

Community_NonDual_Aged = evaluate_v22_2017(PERSON, DIAG, "Community_NonDual_Aged")
Error in eval(e, x, parent.frame()) : object 'factor_type' not found

Code:

library(cmshcc)
PERSON <- generate_PERSON()
cmshcc_map <- load_cmshcc_map()
DIAG <- generate_DIAG(cmshcc_map)
Community_NonDual_Aged = evaluate_v22_2017(PERSON, DIAG, "Community_NonDual_Aged")

Documentation - Add robust commenting using Roxygen

Hard Coded Values in Code

Move the hard coded HCC factor scores to a list argument to the evaluation function. Load the factors as an .RData file during package loading to make available in R environment.