Skip to content

A Python implementation of Hierarchical Condition Categories

License

Notifications You must be signed in to change notification settings

yubin-park/hccpy

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

96 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

NOTICE - March 31, 2024

I am in the process of rewriting the entire hccpy incorporating all the feedback and issues in the past. If you want to provide mode thoughts while I am in the process, please reach out to me on LinkedIn (https://www.linkedin.com/in/yubin-park-phd/) - the fastest way to reach me.

The new project will be hosted by https://github.com/mimilabs - for more details, please see: https://www.mimilabs.ai/projects/65e908d2972dde0f134d21db

Best,

Yubin

hccpy

Hierachical Condition Categories Python Package.

This module implements the Hierachical Condition Categories that are used for adjusting risks for the Medicare population. The original SAS implementation can be found here.

The latest version is 0.1.10 which was released on 04/10/2024.

Currently, hccpy supports:

  • CMS-HCC V22
  • CMS-HCC V23
  • CMS-HCC V24
  • CMS-HCC V28
  • CMS-HCC ESRD
  • HHS-HCC 2019 (V05)
  • HHS-HCC 2022 (V07)

Note that hccpy does not have support for ICD-9.

Installing

Installing from the source:

$ git clone git@github.com:yubin-park/hccpy.git
$ cd hccpy
$ python setup.py develop

Or, simply using pip:

$ pip install hccpy

File Structure

  • hccpy/ : The package source code is located here.
    • data/: The raw data files directly downloaded from the National Burequ of Economics Research
      • Here, you see the original SAS scripts and data files for the CMS HCC models.
    • _AGESEXV2.py: a Python re-write of the AGESEXV2.TXT SAS script.
    • _V2218O1M.py: a Python re-write of the V2218O1M.TXT SAS script.
    • _V2218O1P.py: a Python re-write of the V2219O1P.TXT SAS script.
    • _V22I0ED2.py: a Python re-write of the V22I0ED2.TXT SAS script.
    • _V2318P1M.py: a Python re-write of the V2318P1M.TXT SAS script.
    • _V2419P1M.py: a Python re-write of the V2419P1M.TXT SAS script.
    • hcc.py: the main module that combines the various logical components for CMS-HCC
    • hhshcc.py: the main module for HHS-HCC
    • utils.py: utility functions for reading data files
  • tests/: test scripts to check the validity of the outputs.
  • LICENSE.txt: Apache 2.0.
  • README.md: This README file.
  • setup.py: a set-up script.

Code Examples

hccpy is really simple to use. Please see some examples below:

Importing

To import the HCCEngine class from hccpy:

>>> import json
>>> from hccpy.hcc import HCCEngine
>>> he = HCCEngine()
>>> print(he.profile.__doc__)
Returns the HCC risk profile of a given patient information.

        Parameters
        ----------
        dx_lst : list of str
                 A list of ICD10 codes for the measurement year.
        age : int or float
              The age of the patient.
        sex : str 
              The sex of the patient; {"M", "F"}
        elig : str
               The eligibility segment of the patient.
               Allowed values are as follows:
               - "CFA": Community Full Benefit Dual Aged
               - "CFD": Community Full Benefit Dual Disabled
               - "CNA": Community NonDual Aged
               - "CND": Community NonDual Disabled
               - "CPA": Community Partial Benefit Dual Aged
               - "CPD": Community Partial Benefit Dual Disabled
               - "INS": Long Term Institutional
               - "NE": New Enrollee
               - "SNPNE": SNP NE
        orec: str
              Original reason for entitlement code.
              - "0": Old age and survivor's insurance
              - "1": Disability insurance benefits
              - "2": End-stage renal disease 
              - "3": Both DIB and ESRD
        medicaid: bool
                  If the patient is in Medicaid or not.
>>>

HCC-Profiling a Member with Diagnosis Codes

To get a HCC profile from a list of diagnosis codes (in ICD-10):

>>> rp = he.profile(["E1169", "I5030", "I509", "I211", "I209", "R05"])
>>> print(json.dumps(rp, indent=2))
{
  "risk_score": 1.3139999999999998,
  "details": {
    "CNA_M70_74": 0.379,
    "CNA_HCC85": 0.323,
    "CNA_HCC88": 0.14,
    "CNA_HCC18": 0.318,
    "CNA_HCC85_gDiabetesMellit": 0.154,
    "CNA_DIABETES_CHF": 0.0
  },
  "hcc_lst": [
    "HCC85",
    "HCC88",
    "HCC18"
  ],
  "hcc_map": {
    "I5030": "HCC85",
    "I209": "HCC88",
    "E1169": "HCC18",
    "I509": "HCC85"
  },
  "parameters": {
    "age": 70,
    "sex": "M",
    "elig": "CNA",
    "medicaid": false,
    "disabled": 0,
    "origds": 0
  }
}
>>>

HCC for the new model, V28

Please use "V28" when initializing the engine.

>>> from hccpy.hcc import HCCEngine
>>> he = HCCEngine("28")

Also, see the test_v23() examples in tests/hcc_tests.py.

Coding Intensity Factor and Normalization Factors

You can add normalization factors and coding intensity factors to directly calculate the adjusted risk score.

By default, these two parameters are set as:

cif = 0.059, # coding intensity factor.
norm_params={ 
  "C": 1.015, # community/institution models
  "D": 1.022, # ESRD Dialysis
  "G": 1.028 # ESRD Graft
}

You can overwrite these parameters. For example, this setting below would not adjust the raw risk score.

HCCEngine(version="28", cif = 0, norm_params={"C": 1})

To see the adjusted risk scores,

>>> from hccpy.hcc import HCCEngine
>>> he = HCCEngine("28")
>>> rp = he.profile(["E1169", "I5030", "I509", "I211", "I209", "R05"],
                    age=70, sex="M", elig="CNA") 
>>> rp["risk_score_adj"]

Also, see the test_norm_factors() examples in tests/hcc_tests.py.

HCC-Profiling a New Member

If a member is new, then provide the elig="NE" in the input:

>>> rp = he.profile([], elig="NE", age=65)
>>> print(json.dumps(rp, indent=2))
{
  "risk_score": 0.514,
  "details": {
    "NE_NMCAID_NORIGDIS_NEM65": 0.514
  },
  "hcc_lst": [],
  "hcc_map": {},
  "parameters": {
    "age": 65,
    "sex": "M",
    "elig": "NE_NMCAID_NORIGDIS_NE",
    "medicaid": false,
    "disabled": 0,
    "origds": 0
  }
}
>>>

HCC-Profiling a Intitutionalized Member

If a member has a different eligibility status, change the eligibility as follows (e.g. institutionalized member):

>>> rp = he.profile(["E1169", "I5030", "I509", "I209"], elig="INS")
>>> print(json.dumps(rp, indent=2))
{
  "risk_score": 2.6059999999999994,
  "details": {
    "INS_M70_74": 1.323,
    "INS_HCC88": 0.497,
    "INS_HCC18": 0.441,
    "INS_HCC85": 0.191,
    "INS_HCC85_gDiabetesMellit": 0.0,
    "INS_DIABETES_CHF": 0.154
  },
  "hcc_lst": [
    "HCC88",
    "HCC18",
    "HCC85"
  ],
  "hcc_map": {
    "I209": "HCC88",
    "E1169": "HCC18",
    "I509": "HCC85",
    "I5030": "HCC85"
  },
  "parameters": {
    "age": 70,
    "sex": "M",
    "elig": "INS",
    "medicaid": false,
    "disabled": 0,
    "origds": 0
  }
}

HCC-Describing a Hierachical Condition Category

To get the description, hierarchy parents and children of a HCC:

>>> hcc_doc = he.describe_hcc("HCC19")  # either "HCC19", "hcc19" or "19"
>>> print(json.dumps(hcc_doc, indent=2))
{
  "description": "Diabetes without Complication",
  "children": [],
  "parents": [
    "HCC17",
    "HCC18"
  ]
}

Eligible Risk Adjustment Codes

Not all claims are eligible for risk adjustment. For professional claims, a certain set of CPT codes is required to be eligible, while for institutional claims, a certain set of bill types is needed. This module provides an easy interface for determining if a certain claim is eligible for risk adjustment or not.

NOTE: This function uses CPT codes, and this requires AMA CPT license. Once you carefully review the license, you need to download a data file.

>>> from hccpy.raeligible import RAEligible
>>> rae = RAEligible()
>>> rae.load(fn="CY2019Q2_CPTHCPCS_CMS_20190425.csv")
>>> rae.is_eligible(pr_lst=["C5271"])
True
>>> rae.is_eligible(pr_lst=["C5270"])
False
>>>

NOTE: The data file (CY2019Q2_CPTHCPCS_CMS_20190425.csv) should be located in the same folder.

NOTE

python -m build
twine upload dist/*

License

Apache 2.0

Authors/Maintainers

  • Yubin Park @yubin-park
  • Thomas Chen @t-kychen
  • Matt Walker @mwalker14
  • David Roberts @dr00b
  • Kevin Buchan Jr. @kevinbuchanjr

References

About

A Python implementation of Hierarchical Condition Categories

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages